Neural network versus classical time series forecasting models

NASA Astrophysics Data System (ADS)

Nor, Maria Elena; Safuan, Hamizah Mohd; Shab, Noorzehan Fazahiyah Md; Asrul, Mohd; Abdullah, Affendi; Mohamad, Nurul Asmaa Izzati; Lee, Muhammad Hisyam

2017-05-01

Artificial neural network (ANN) has advantage in time series forecasting as it has potential to solve complex forecasting problems. This is because ANN is data driven approach which able to be trained to map past values of a time series. In this study the forecast performance between neural network and classical time series forecasting method namely seasonal autoregressive integrated moving average models was being compared by utilizing gold price data. Moreover, the effect of different data preprocessing on the forecast performance of neural network being examined. The forecast accuracy was evaluated using mean absolute deviation, root mean square error and mean absolute percentage error. It was found that ANN produced the most accurate forecast when Box-Cox transformation was used as data preprocessing.

Evaluating the performance of infectious disease forecasts: A comparison of climate-driven and seasonal dengue forecasts for Mexico.

PubMed

Johansson, Michael A; Reich, Nicholas G; Hota, Aditi; Brownstein, John S; Santillana, Mauricio

2016-09-26

Dengue viruses, which infect millions of people per year worldwide, cause large epidemics that strain healthcare systems. Despite diverse efforts to develop forecasting tools including autoregressive time series, climate-driven statistical, and mechanistic biological models, little work has been done to understand the contribution of different components to improved prediction. We developed a framework to assess and compare dengue forecasts produced from different types of models and evaluated the performance of seasonal autoregressive models with and without climate variables for forecasting dengue incidence in Mexico. Climate data did not significantly improve the predictive power of seasonal autoregressive models. Short-term and seasonal autocorrelation were key to improving short-term and long-term forecasts, respectively. Seasonal autoregressive models captured a substantial amount of dengue variability, but better models are needed to improve dengue forecasting. This framework contributes to the sparse literature of infectious disease prediction model evaluation, using state-of-the-art validation techniques such as out-of-sample testing and comparison to an appropriate reference model.

Evaluating the performance of infectious disease forecasts: A comparison of climate-driven and seasonal dengue forecasts for Mexico

PubMed Central

Johansson, Michael A.; Reich, Nicholas G.; Hota, Aditi; Brownstein, John S.; Santillana, Mauricio

2016-01-01

Dengue viruses, which infect millions of people per year worldwide, cause large epidemics that strain healthcare systems. Despite diverse efforts to develop forecasting tools including autoregressive time series, climate-driven statistical, and mechanistic biological models, little work has been done to understand the contribution of different components to improved prediction. We developed a framework to assess and compare dengue forecasts produced from different types of models and evaluated the performance of seasonal autoregressive models with and without climate variables for forecasting dengue incidence in Mexico. Climate data did not significantly improve the predictive power of seasonal autoregressive models. Short-term and seasonal autocorrelation were key to improving short-term and long-term forecasts, respectively. Seasonal autoregressive models captured a substantial amount of dengue variability, but better models are needed to improve dengue forecasting. This framework contributes to the sparse literature of infectious disease prediction model evaluation, using state-of-the-art validation techniques such as out-of-sample testing and comparison to an appropriate reference model. PMID:27665707

Nonlinear time series modeling and forecasting the seismic data of the Hindu Kush region

NASA Astrophysics Data System (ADS)

Khan, Muhammad Yousaf; Mittnik, Stefan

2018-01-01

In this study, we extended the application of linear and nonlinear time models in the field of earthquake seismology and examined the out-of-sample forecast accuracy of linear Autoregressive (AR), Autoregressive Conditional Duration (ACD), Self-Exciting Threshold Autoregressive (SETAR), Threshold Autoregressive (TAR), Logistic Smooth Transition Autoregressive (LSTAR), Additive Autoregressive (AAR), and Artificial Neural Network (ANN) models for seismic data of the Hindu Kush region. We also extended the previous studies by using Vector Autoregressive (VAR) and Threshold Vector Autoregressive (TVAR) models and compared their forecasting accuracy with linear AR model. Unlike previous studies that typically consider the threshold model specifications by using internal threshold variable, we specified these models with external transition variables and compared their out-of-sample forecasting performance with the linear benchmark AR model. The modeling results show that time series models used in the present study are capable of capturing the dynamic structure present in the seismic data. The point forecast results indicate that the AR model generally outperforms the nonlinear models. However, in some cases, threshold models with external threshold variables specification produce more accurate forecasts, indicating that specification of threshold time series models is of crucial importance. For raw seismic data, the ACD model does not show an improved out-of-sample forecasting performance over the linear AR model. The results indicate that the AR model is the best forecasting device to model and forecast the raw seismic data of the Hindu Kush region.

Evaluation of streamflow forecast for the National Water Model of U.S. National Weather Service

NASA Astrophysics Data System (ADS)

Rafieeinasab, A.; McCreight, J. L.; Dugger, A. L.; Gochis, D.; Karsten, L. R.; Zhang, Y.; Cosgrove, B.; Liu, Y.

2016-12-01

The National Water Model (NWM), an implementation of the community WRF-Hydro modeling system, is an operational hydrologic forecasting model for the contiguous United States. The model forecasts distributed hydrologic states and fluxes, including soil moisture, snowpack, ET, and ponded water. In particular, the NWM provides streamflow forecasts at more than 2.7 million river reaches for three forecast ranges: short (15 hr), medium (10 days), and long (30 days). In this study, we verify short and medium range streamflow forecasts in the context of the verification of their respective quantitative precipitation forecasts/forcing (QPF), the High Resolution Rapid Refresh (HRRR) and the Global Forecast System (GFS). The streamflow evaluation is performed for summer of 2016 at more than 6,000 USGS gauges. Both individual forecasts and forecast lead times are examined. Selected case studies of extreme events aim to provide insight into the quality of the NWM streamflow forecasts. A goal of this comparison is to address how much streamflow bias originates from precipitation forcing bias. To this end, precipitation verification is performed over the contributing areas above (and between assimilated) USGS gauge locations. Precipitation verification is based on the aggregated, blended StageIV/StageII data as the "reference truth". We summarize the skill of the streamflow forecasts, their skill relative to the QPF, and make recommendations for improving NWM forecast skill.

Hybrid Forecasting of Daily River Discharges Considering Autoregressive Heteroscedasticity

NASA Astrophysics Data System (ADS)

Szolgayová, Elena Peksová; Danačová, Michaela; Komorniková, Magda; Szolgay, Ján

2017-06-01

It is widely acknowledged that in the hydrological and meteorological communities, there is a continuing need to improve the quality of quantitative rainfall and river flow forecasts. A hybrid (combined deterministic-stochastic) modelling approach is proposed here that combines the advantages offered by modelling the system dynamics with a deterministic model and a deterministic forecasting error series with a data-driven model in parallel. Since the processes to be modelled are generally nonlinear and the model error series may exhibit nonstationarity and heteroscedasticity, GARCH-type nonlinear time series models are considered here. The fitting, forecasting and simulation performance of such models have to be explored on a case-by-case basis. The goal of this paper is to test and develop an appropriate methodology for model fitting and forecasting applicable for daily river discharge forecast error data from the GARCH family of time series models. We concentrated on verifying whether the use of a GARCH-type model is suitable for modelling and forecasting a hydrological model error time series on the Hron and Morava Rivers in Slovakia. For this purpose we verified the presence of heteroscedasticity in the simulation error series of the KLN multilinear flow routing model; then we fitted the GARCH-type models to the data and compared their fit with that of an ARMA - type model. We produced one-stepahead forecasts from the fitted models and again provided comparisons of the model's performance.

Evaluation of CMAQ and CAMx Ensemble Air Quality Forecasts during the 2015 MAPS-Seoul Field Campaign

NASA Astrophysics Data System (ADS)

Kim, E.; Kim, S.; Bae, C.; Kim, H. C.; Kim, B. U.

2015-12-01

The performance of Air quality forecasts during the 2015 MAPS-Seoul Field Campaign was evaluated. An forecast system has been operated to support the campaign's daily aircraft route decisions for airborne measurements to observe long-range transporting plume. We utilized two real-time ensemble systems based on the Weather Research and Forecasting (WRF)-Sparse Matrix Operator Kernel Emissions (SMOKE)-Comprehensive Air quality Model with extensions (CAMx) modeling framework and WRF-SMOKE- Community Multi_scale Air Quality (CMAQ) framework over northeastern Asia to simulate PM10 concentrations. Global Forecast System (GFS) from National Centers for Environmental Prediction (NCEP) was used to provide meteorological inputs for the forecasts. For an additional set of retrospective simulations, ERA Interim Reanalysis from European Centre for Medium-Range Weather Forecasts (ECMWF) was also utilized to access forecast uncertainties from the meteorological data used. Model Inter-Comparison Study for Asia (MICS-Asia) and National Institute of Environment Research (NIER) Clean Air Policy Support System (CAPSS) emission inventories are used for foreign and domestic emissions, respectively. In the study, we evaluate the CMAQ and CAMx model performance during the campaign by comparing the results to the airborne and surface measurements. Contributions of foreign and domestic emissions are estimated using a brute force method. Analyses on model performance and emissions will be utilized to improve air quality forecasts for the upcoming KORUS-AQ field campaign planned in 2016.

ENSURF: multi-model sea level forecast - implementation and validation results for the IBIROOS and Western Mediterranean regions

NASA Astrophysics Data System (ADS)

Pérez, B.; Brower, R.; Beckers, J.; Paradis, D.; Balseiro, C.; Lyons, K.; Cure, M.; Sotillo, M. G.; Hacket, B.; Verlaan, M.; Alvarez Fanjul, E.

2011-04-01

ENSURF (Ensemble SURge Forecast) is a multi-model application for sea level forecast that makes use of existing storm surge or circulation models today operational in Europe, as well as near-real time tide gauge data in the region, with the following main goals: - providing an easy access to existing forecasts, as well as to its performance and model validation, by means of an adequate visualization tool - generation of better forecasts of sea level, including confidence intervals, by means of the Bayesian Model Average Technique (BMA) The system was developed and implemented within ECOOP (C.No. 036355) European Project for the NOOS and the IBIROOS regions, based on MATROOS visualization tool developed by Deltares. Both systems are today operational at Deltares and Puertos del Estado respectively. The Bayesian Modelling Average technique generates an overall forecast probability density function (PDF) by making a weighted average of the individual forecasts PDF's; the weights represent the probability that a model will give the correct forecast PDF and are determined and updated operationally based on the performance of the models during a recent training period. This implies the technique needs the availability of sea level data from tide gauges in near-real time. Results of validation of the different models and BMA implementation for the main harbours will be presented for the IBIROOS and Western Mediterranean regions, where this kind of activity is performed for the first time. The work has proved to be useful to detect problems in some of the circulation models not previously well calibrated with sea level data, to identify the differences on baroclinic and barotropic models for sea level applications and to confirm the general improvement of the BMA forecasts.

A new scoring method for evaluating the performance of earthquake forecasts and predictions

NASA Astrophysics Data System (ADS)

Zhuang, J.

2009-12-01

This study presents a new method, namely the gambling score, for scoring the performance of earthquake forecasts or predictions. Unlike most other scoring procedures that require a regular scheme of forecast and treat each earthquake equally, regardless their magnitude, this new scoring method compensates the risk that the forecaster has taken. A fair scoring scheme should reward the success in a way that is compatible with the risk taken. Suppose that we have the reference model, usually the Poisson model for usual cases or Omori-Utsu formula for the case of forecasting aftershocks, which gives probability p0 that at least 1 event occurs in a given space-time-magnitude window. The forecaster, similar to a gambler, who starts with a certain number of reputation points, bets 1 reputation point on ``Yes'' or ``No'' according to his forecast, or bets nothing if he performs a NA-prediction. If the forecaster bets 1 reputation point of his reputations on ``Yes" and loses, the number of his reputation points is reduced by 1; if his forecasts is successful, he should be rewarded (1-p0)/p0 reputation points. The quantity (1-p0)/p0 is the return (reward/bet) ratio for bets on ``Yes''. In this way, if the reference model is correct, the expected return that he gains from this bet is 0. This rule also applies to probability forecasts. Suppose that p is the occurrence probability of an earthquake given by the forecaster. We can regard the forecaster as splitting 1 reputation point by betting p on ``Yes'' and 1-p on ``No''. In this way, the forecaster's expected pay-off based on the reference model is still 0. From the viewpoints of both the reference model and the forecaster, the rule for rewarding and punishment is fair. This method is also extended to the continuous case of point process models, where the reputation points bet by the forecaster become a continuous mass on the space-time-magnitude range of interest. We also calculate the upper bound of the gambling score when the true model is a renewal process, the stress release model or the ETAS model and when the reference model is the Poisson model.

Tourism forecasting using modified empirical mode decomposition and group method of data handling

NASA Astrophysics Data System (ADS)

Yahya, N. A.; Samsudin, R.; Shabri, A.

2017-09-01

In this study, a hybrid model using modified Empirical Mode Decomposition (EMD) and Group Method of Data Handling (GMDH) model is proposed for tourism forecasting. This approach reconstructs intrinsic mode functions (IMFs) produced by EMD using trial and error method. The new component and the remaining IMFs is then predicted respectively using GMDH model. Finally, the forecasted results for each component are aggregated to construct an ensemble forecast. The data used in this experiment are monthly time series data of tourist arrivals from China, Thailand and India to Malaysia from year 2000 to 2016. The performance of the model is evaluated using Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE) where conventional GMDH model and EMD-GMDH model are used as benchmark models. Empirical results proved that the proposed model performed better forecasts than the benchmarked models.

An application of ensemble/multi model approach for wind power production forecast.

NASA Astrophysics Data System (ADS)

Alessandrini, S.; Decimi, G.; Hagedorn, R.; Sperati, S.

2010-09-01

The wind power forecast of the 3 days ahead period are becoming always more useful and important in reducing the problem of grid integration and energy price trading due to the increasing wind power penetration. Therefore it's clear that the accuracy of this forecast is one of the most important requirements for a successful application. The wind power forecast is based on a mesoscale meteorological models that provides the 3 days ahead wind data. A Model Output Statistic correction is then performed to reduce systematic error caused, for instance, by a wrong representation of surface roughness or topography in the meteorological models. The corrected wind data are then used as input in the wind farm power curve to obtain the power forecast. These computations require historical time series of wind measured data (by an anemometer located in the wind farm or on the nacelle) and power data in order to be able to perform the statistical analysis on the past. For this purpose a Neural Network (NN) is trained on the past data and then applied in the forecast task. Considering that the anemometer measurements are not always available in a wind farm a different approach has also been adopted. A training of the NN to link directly the forecasted meteorological data and the power data has also been performed. The normalized RMSE forecast error seems to be lower in most cases by following the second approach. We have examined two wind farms, one located in Denmark on flat terrain and one located in a mountain area in the south of Italy (Sicily). In both cases we compare the performances of a prediction based on meteorological data coming from a single model with those obtained by using two or more models (RAMS, ECMWF deterministic, LAMI, HIRLAM). It is shown that the multi models approach reduces the day-ahead normalized RMSE forecast error of at least 1% compared to the singles models approach. Moreover the use of a deterministic global model, (e.g. ECMWF deterministic model) seems to reach similar level of accuracy of those of the mesocale models (LAMI and RAMS). Finally we have focused on the possibility of using the ensemble model (ECMWF) to estimate the hourly, three days ahead, power forecast accuracy. Contingency diagram between RMSE of the deterministic power forecast and the ensemble members spread of wind forecast have been produced. From this first analysis it seems that ensemble spread could be used as an indicator of the forecast's accuracy at least for the first day ahead period. In fact low spreads often correspond to low forecast error. For longer forecast horizon the correlation between RMSE and ensemble spread decrease becoming too low to be used for this purpose.

State updating and calibration period selection to improve dynamic monthly streamflow forecasts for an environmental flow management application

NASA Astrophysics Data System (ADS)

Gibbs, Matthew S.; McInerney, David; Humphrey, Greer; Thyer, Mark A.; Maier, Holger R.; Dandy, Graeme C.; Kavetski, Dmitri

2018-02-01

Monthly to seasonal streamflow forecasts provide useful information for a range of water resource management and planning applications. This work focuses on improving such forecasts by considering the following two aspects: (1) state updating to force the models to match observations from the start of the forecast period, and (2) selection of a shorter calibration period that is more representative of the forecast period, compared to a longer calibration period traditionally used. The analysis is undertaken in the context of using streamflow forecasts for environmental flow water management of an open channel drainage network in southern Australia. Forecasts of monthly streamflow are obtained using a conceptual rainfall-runoff model combined with a post-processor error model for uncertainty analysis. This model set-up is applied to two catchments, one with stronger evidence of non-stationarity than the other. A range of metrics are used to assess different aspects of predictive performance, including reliability, sharpness, bias and accuracy. The results indicate that, for most scenarios and metrics, state updating improves predictive performance for both observed rainfall and forecast rainfall sources. Using the shorter calibration period also improves predictive performance, particularly for the catchment with stronger evidence of non-stationarity. The results highlight that a traditional approach of using a long calibration period can degrade predictive performance when there is evidence of non-stationarity. The techniques presented can form the basis for operational monthly streamflow forecasting systems and provide support for environmental decision-making.

Comparisons of Three-Dimensional Variational Data Assimilation and Model Output Statistics in Improving Atmospheric Chemistry Forecasts

NASA Astrophysics Data System (ADS)

Ma, Chaoqun; Wang, Tijian; Zang, Zengliang; Li, Zhijin

2018-07-01

Atmospheric chemistry models usually perform badly in forecasting wintertime air pollution because of their uncertainties. Generally, such uncertainties can be decreased effectively by techniques such as data assimilation (DA) and model output statistics (MOS). However, the relative importance and combined effects of the two techniques have not been clarified. Here, a one-month air quality forecast with the Weather Research and Forecasting-Chemistry (WRF-Chem) model was carried out in a virtually operational setup focusing on Hebei Province, China. Meanwhile, three-dimensional variational (3DVar) DA and MOS based on one-dimensional Kalman filtering were implemented separately and simultaneously to investigate their performance in improving the model forecast. Comparison with observations shows that the chemistry forecast with MOS outperforms that with 3DVar DA, which could be seen in all the species tested over the whole 72 forecast hours. Combined use of both techniques does not guarantee a better forecast than MOS only, with the improvements and degradations being small and appearing rather randomly. Results indicate that the implementation of MOS is more suitable than 3DVar DA in improving the operational forecasting ability of WRF-Chem.

Value versus Accuracy: application of seasonal forecasts to a hydro-economic optimization model for the Sudanese Blue Nile

NASA Astrophysics Data System (ADS)

Satti, S.; Zaitchik, B. F.; Siddiqui, S.; Badr, H. S.; Shukla, S.; Peters-Lidard, C. D.

2015-12-01

The unpredictable nature of precipitation within the East African (EA) region makes it one of the most vulnerable, food insecure regions in the world. There is a vital need for forecasts to inform decision makers, both local and regional, and to help formulate the region's climate change adaptation strategies. Here, we present a suite of different seasonal forecast models, both statistical and dynamical, for the EA region. Objective regionalization is performed for EA on the basis of interannual variability in precipitation in both observations and models. This regionalization is applied as the basis for calculating a number of standard skill scores to evaluate each model's forecast accuracy. A dynamically linked Land Surface Model (LSM) is then applied to determine forecasted flows, which drive the Sudanese Hydroeconomic Optimization Model (SHOM). SHOM combines hydrologic, agronomic and economic inputs to determine the optimal decisions that maximize economic benefits along the Sudanese Blue Nile. This modeling sequence is designed to derive the potential added value of information of each forecasting model to agriculture and hydropower management. A rank of each model's forecasting skill score along with its added value of information is analyzed in order compare the performance of each forecast. This research aims to improve understanding of how characteristics of accuracy, lead time, and uncertainty of seasonal forecasts influence their utility to water resources decision makers who utilize them.

Precipitable water vapour forecasting: a tool for optimizing IR observations at Roque de los Muchachos Observatory

NASA Astrophysics Data System (ADS)

Pérez-Jordán, wG; Castro-Almazán, J. A.; Muñoz-Tuñón, C.

2018-07-01

We validate the Weather Research and Forecasting (WRF) model for precipitable water vapour (PWV) forecasting as a fully operational tool for optimizing astronomical infrared observations at Roque de los Muchachos Observatory (ORM). For the model validation, we used GNSS-based (Global Navigation Satellite System) data from the PWV monitor located at the ORM. We have run WRF every 24 h for near two months, with a horizon of 48 h (hourly forecasts), from 2016 January 11 to March 04. These runs represent 1296 hourly forecast points. The validation is carried out using different approaches: performance as a function of the forecast range, time horizon accuracy, performance as a function of the PWV value, and performance of the operational WRF time series with 24- and 48-h horizons. Excellent agreement was found between the model forecasts and observations, with R = 0.951 and 0.904 for the 24- and 48-h forecast time series, respectively. The 48-h forecast was further improved by correcting a time lag of 2 h found in the predictions. The final errors, taking into account all the uncertainties involved, are 1.75 mm for the 24-h forecasts and 1.99 mm for 48 h. We found linear trends in both the correlation and root-mean-square error of the residuals (measurements - forecasts) as a function of the forecast range within the horizons analysed (up to 48 h). In summary, the WRF performance is excellent and accurate, thus allowing it to be implemented as an operational tool at the ORM.

Precipitable water vapour forecasting: a tool for optimizing IR observations at Roque de los Muchachos Observatory.

NASA Astrophysics Data System (ADS)

Pérez-Jordán, G.; Castro-Almazán, J. A.; Muñoz-Tuñón, C.

2018-04-01

We validate the Weather Research and Forecasting (WRF) model for precipitable water vapour (PWV) forecasting as a fully operational tool for optimizing astronomical infrared (IR) observations at Roque de los Muchachos Observatory (ORM). For the model validation we used GNSS-based (Global Navigation Satellite System) data from the PWV monitor located at the ORM. We have run WRF every 24 h for near two months, with a horizon of 48 hours (hourly forecasts), from 2016 January 11 to 2016 March 4. These runs represent 1296 hourly forecast points. The validation is carried out using different approaches: performance as a function of the forecast range, time horizon accuracy, performance as a function of the PWV value, and performance of the operational WRF time series with 24- and 48-hour horizons. Excellent agreement was found between the model forecasts and observations, with R =0.951 and R =0.904 for the 24- and 48-h forecast time series respectively. The 48-h forecast was further improved by correcting a time lag of 2 h found in the predictions. The final errors, taking into account all the uncertainties involved, are 1.75 mm for the 24-h forecasts and 1.99 mm for 48 h. We found linear trends in both the correlation and RMSE of the residuals (measurements - forecasts) as a function of the forecast range within the horizons analysed (up to 48 h). In summary, the WRF performance is excellent and accurate, thus allowing it to be implemented as an operational tool at the ORM.

Leveraging organismal biology to forecast the effects of climate change.

PubMed

Buckley, Lauren B; Cannistra, Anthony F; John, Aji

2018-04-26

Despite the pressing need for accurate forecasts of ecological and evolutionary responses to environmental change, commonly used modelling approaches exhibit mixed performance because they omit many important aspects of how organisms respond to spatially and temporally variable environments. Integrating models based on organismal phenotypes at the physiological, performance and fitness levels can improve model performance. We summarize current limitations of environmental data and models and discuss potential remedies. The paper reviews emerging techniques for sensing environments at fine spatial and temporal scales, accounting for environmental extremes, and capturing how organisms experience the environment. Intertidal mussel data illustrate biologically important aspects of environmental variability. We then discuss key challenges in translating environmental conditions into organismal performance including accounting for the varied timescales of physiological processes, for responses to environmental fluctuations including the onset of stress and other thresholds, and for how environmental sensitivities vary across lifecycles. We call for the creation of phenotypic databases to parameterize forecasting models and advocate for improved sharing of model code and data for model testing. We conclude with challenges in organismal biology that must be solved to improve forecasts over the next decade.acclimation, biophysical models, ecological forecasting, extremes, microclimate, spatial and temporal variability.

Forecasting electricity usage using univariate time series models

NASA Astrophysics Data System (ADS)

Hock-Eam, Lim; Chee-Yin, Yip

2014-12-01

Electricity is one of the important energy sources. A sufficient supply of electricity is vital to support a country's development and growth. Due to the changing of socio-economic characteristics, increasing competition and deregulation of electricity supply industry, the electricity demand forecasting is even more important than before. It is imperative to evaluate and compare the predictive performance of various forecasting methods. This will provide further insights on the weakness and strengths of each method. In literature, there are mixed evidences on the best forecasting methods of electricity demand. This paper aims to compare the predictive performance of univariate time series models for forecasting the electricity demand using a monthly data of maximum electricity load in Malaysia from January 2003 to December 2013. Results reveal that the Box-Jenkins method produces the best out-of-sample predictive performance. On the other hand, Holt-Winters exponential smoothing method is a good forecasting method for in-sample predictive performance.

Water balance models in one-month-ahead streamflow forecasting

USGS Publications Warehouse

Alley, William M.

1985-01-01

Techniques are tested that incorporate information from water balance models in making 1-month-ahead streamflow forecasts in New Jersey. The results are compared to those based on simple autoregressive time series models. The relative performance of the models is dependent on the month of the year in question. The water balance models are most useful for forecasts of April and May flows. For the stations in northern New Jersey, the April and May forecasts were made in order of decreasing reliability using the water-balance-based approaches, using the historical monthly means, and using simple autoregressive models. The water balance models were useful to a lesser extent for forecasts during the fall months. For the rest of the year the improvements in forecasts over those obtained using the simpler autoregressive models were either very small or the simpler models provided better forecasts. When using the water balance models, monthly corrections for bias are found to improve minimum mean-square-error forecasts as well as to improve estimates of the forecast conditional distributions.

Optimization of autoregressive, exogenous inputs-based typhoon inundation forecasting models using a multi-objective genetic algorithm

NASA Astrophysics Data System (ADS)

Ouyang, Huei-Tau

2017-07-01

Three types of model for forecasting inundation levels during typhoons were optimized: the linear autoregressive model with exogenous inputs (LARX), the nonlinear autoregressive model with exogenous inputs with wavelet function (NLARX-W) and the nonlinear autoregressive model with exogenous inputs with sigmoid function (NLARX-S). The forecast performance was evaluated by three indices: coefficient of efficiency, error in peak water level and relative time shift. Historical typhoon data were used to establish water-level forecasting models that satisfy all three objectives. A multi-objective genetic algorithm was employed to search for the Pareto-optimal model set that satisfies all three objectives and select the ideal models for the three indices. Findings showed that the optimized nonlinear models (NLARX-W and NLARX-S) outperformed the linear model (LARX). Among the nonlinear models, the optimized NLARX-W model achieved a more balanced performance on the three indices than the NLARX-S models and is recommended for inundation forecasting during typhoons.

Application of Support Vector Machine to Forex Monitoring

NASA Astrophysics Data System (ADS)

Kamruzzaman, Joarder; Sarker, Ruhul A.

Previous studies have demonstrated superior performance of artificial neural network (ANN) based forex forecasting models over traditional regression models. This paper applies support vector machines to build a forecasting model from the historical data using six simple technical indicators and presents a comparison with an ANN based model trained by scaled conjugate gradient (SCG) learning algorithm. The models are evaluated and compared on the basis of five commonly used performance metrics that measure closeness of prediction as well as correctness in directional change. Forecasting results of six different currencies against Australian dollar reveal superior performance of SVM model using simple linear kernel over ANN-SCG model in terms of all the evaluation metrics. The effect of SVM parameter selection on prediction performance is also investigated and analyzed.

Wavelet regression model in forecasting crude oil price

NASA Astrophysics Data System (ADS)

Hamid, Mohd Helmie; Shabri, Ani

2017-05-01

This study presents the performance of wavelet multiple linear regression (WMLR) technique in daily crude oil forecasting. WMLR model was developed by integrating the discrete wavelet transform (DWT) and multiple linear regression (MLR) model. The original time series was decomposed to sub-time series with different scales by wavelet theory. Correlation analysis was conducted to assist in the selection of optimal decomposed components as inputs for the WMLR model. The daily WTI crude oil price series has been used in this study to test the prediction capability of the proposed model. The forecasting performance of WMLR model were also compared with regular multiple linear regression (MLR), Autoregressive Moving Average (ARIMA) and Generalized Autoregressive Conditional Heteroscedasticity (GARCH) using root mean square errors (RMSE) and mean absolute errors (MAE). Based on the experimental results, it appears that the WMLR model performs better than the other forecasting technique tested in this study.

The IDEA model: A single equation approach to the Ebola forecasting challenge.

PubMed

Tuite, Ashleigh R; Fisman, David N

2018-03-01

Mathematical modeling is increasingly accepted as a tool that can inform disease control policy in the face of emerging infectious diseases, such as the 2014-2015 West African Ebola epidemic, but little is known about the relative performance of alternate forecasting approaches. The RAPIDD Ebola Forecasting Challenge (REFC) tested the ability of eight mathematical models to generate useful forecasts in the face of simulated Ebola outbreaks. We used a simple, phenomenological single-equation model (the "IDEA" model), which relies only on case counts, in the REFC. Model fits were performed using a maximum likelihood approach. We found that the model performed reasonably well relative to other more complex approaches, with performance metrics ranked on average 4th or 5th among participating models. IDEA appeared better suited to long- than short-term forecasts, and could be fit using nothing but reported case counts. Several limitations were identified, including difficulty in identifying epidemic peak (even retrospectively), unrealistically precise confidence intervals, and difficulty interpolating daily case counts when using a model scaled to epidemic generation time. More realistic confidence intervals were generated when case counts were assumed to follow a negative binomial, rather than Poisson, distribution. Nonetheless, IDEA represents a simple phenomenological model, easily implemented in widely available software packages that could be used by frontline public health personnel to generate forecasts with accuracy that approximates that which is achieved using more complex methodologies. Copyright © 2016 The Author(s). Published by Elsevier B.V. All rights reserved.

Multi-RCM ensemble downscaling of global seasonal forecasts (MRED)

NASA Astrophysics Data System (ADS)

Arritt, R. W.

2008-12-01

The Multi-RCM Ensemble Downscaling (MRED) project was recently initiated to address the question, Can regional climate models provide additional useful information from global seasonal forecasts? MRED will use a suite of regional climate models to downscale seasonal forecasts produced by the new National Centers for Environmental Prediction (NCEP) Climate Forecast System (CFS) seasonal forecast system and the NASA GEOS5 system. The initial focus will be on wintertime forecasts in order to evaluate topographic forcing, snowmelt, and the potential usefulness of higher resolution, especially for near-surface fields influenced by high resolution orography. Each regional model will cover the conterminous US (CONUS) at approximately 32 km resolution, and will perform an ensemble of 15 runs for each year 1982-2003 for the forecast period 1 December - 30 April. MRED will compare individual regional and global forecasts as well as ensemble mean precipitation and temperature forecasts, which are currently being used to drive macroscale land surface models (LSMs), as well as wind, humidity, radiation, turbulent heat fluxes, which are important for more advanced coupled macro-scale hydrologic models. Metrics of ensemble spread will also be evaluated. Extensive analysis will be performed to link improvements in downscaled forecast skill to regional forcings and physical mechanisms. Our overarching goal is to determine what additional skill can be provided by a community ensemble of high resolution regional models, which we believe will eventually define a strategy for more skillful and useful regional seasonal climate forecasts.

Value-at-Risk forecasts by a spatiotemporal model in Chinese stock market

NASA Astrophysics Data System (ADS)

Gong, Pu; Weng, Yingliang

2016-01-01

This paper generalizes a recently proposed spatial autoregressive model and introduces a spatiotemporal model for forecasting stock returns. We support the view that stock returns are affected not only by the absolute values of factors such as firm size, book-to-market ratio and momentum but also by the relative values of factors like trading volume ranking and market capitalization ranking in each period. This article studies a new method for constructing stocks' reference groups; the method is called quartile method. Applying the method empirically to the Shanghai Stock Exchange 50 Index, we compare the daily volatility forecasting performance and the out-of-sample forecasting performance of Value-at-Risk (VaR) estimated by different models. The empirical results show that the spatiotemporal model performs surprisingly well in terms of capturing spatial dependences among individual stocks, and it produces more accurate VaR forecasts than the other three models introduced in the previous literature. Moreover, the findings indicate that both allowing for serial correlation in the disturbances and using time-varying spatial weight matrices can greatly improve the predictive accuracy of a spatial autoregressive model.

Synoptic scale forecast skill and systematic errors in the MASS 2.0 model. [Mesoscale Atmospheric Simulation System

NASA Technical Reports Server (NTRS)

Koch, S. E.; Skillman, W. C.; Kocin, P. J.; Wetzel, P. J.; Brill, K. F.

1985-01-01

The synoptic scale performance characteristics of MASS 2.0 are determined by comparing filtered 12-24 hr model forecasts to same-case forecasts made by the National Meteorological Center's synoptic-scale Limited-area Fine Mesh model. Characteristics of the two systems are contrasted, and the analysis methodology used to determine statistical skill scores and systematic errors is described. The overall relative performance of the two models in the sample is documented, and important systematic errors uncovered are presented.

Performance Comparison of the European Storm Surge Models and Chaotic Model in Forecasting Extreme Storm Surges

NASA Astrophysics Data System (ADS)

Siek, M. B.; Solomatine, D. P.

2009-04-01

Storm surge modeling has rapidly developed considerably over the past 30 years. A number of significant advances on operational storm surge models have been implemented and tested, consisting of: refining computational grids, calibrating the model, using a better numerical scheme (i.e. more realistic model physics for air-sea interaction), implementing data assimilation and ensemble model forecasts. This paper addresses the performance comparison between the existing European storm surge models and the recently developed methods of nonlinear dynamics and chaos theory in forecasting storm surge dynamics. The chaotic model is built using adaptive local models based on the dynamical neighbours in the reconstructed phase space of observed time series data. The comparison focused on the model accuracy in forecasting a recently extreme storm surge in the North Sea on November 9th, 2007 that hit the coastlines of several European countries. The combination of a high tide, north-westerly winds exceeding 50 mph and low pressure produced an exceptional storm tide. The tidal level was exceeded 3 meters above normal sea levels. Flood warnings were issued for the east coast of Britain and the entire Dutch coast. The Maeslant barrier's two arc-shaped steel doors in the Europe's biggest port of Rotterdam was closed for the first time since its construction in 1997 due to this storm surge. In comparison to the chaotic model performance, the forecast data from several European physically-based storm surge models were provided from: BSH Germany, DMI Denmark, DNMI Norway, KNMI Netherlands and MUMM Belgium. The performance comparison was made over testing datasets for two periods/conditions: non-stormy period (1-Sep-2007 till 14-Oct-2007) and stormy period (15-Oct-2007 till 20-Nov-2007). A scalar chaotic model with optimized parameters was developed by utilizing an hourly training dataset of observations (11-Sep-2005 till 31-Aug-2007). The comparison results indicated the chaotic model yields better forecasts than the existing European storm surge models. The best performance of European storm surge models for non-storm and storm conditions was achieved by KNMI (with Kalman filter data assimilation) and BSH with errors of 8.95cm and 10.92cm, respectively. Whereas the chaotic model can provide 6 and 48 hours forecasts with errors of 3.10cm and 8.55cm for non-storm condition and 5.04cm and 15.21cm for storm condition, respectively. The chaotic model can provide better forecasts primarily due to the fact that the chaotic model forecasting are estimated by local models which model and identify the similar development of storm surges in the past. In practice, the chaotic model can serve as a reliable and accurate model to support decision-makers in operational ship navigation and flood forecasting.

How can we deal with ANN in flood forecasting? As a simulation model or updating kernel!

NASA Astrophysics Data System (ADS)

Hassan Saddagh, Mohammad; Javad Abedini, Mohammad

2010-05-01

Flood forecasting and early warning, as a non-structural measure for flood control, is often considered to be the most effective and suitable alternative to mitigate the damage and human loss caused by flood. Forecast results which are output of hydrologic, hydraulic and/or black box models should secure accuracy of flood values and timing, especially for long lead time. The application of the artificial neural network (ANN) in flood forecasting has received extensive attentions in recent years due to its capability to capture the dynamics inherent in complex processes including flood. However, results obtained from executing plain ANN as simulation model demonstrate dramatic reduction in performance indices as lead time increases. This paper is intended to monitor the performance indices as it relates to flood forecasting and early warning using two different methodologies. While the first method employs a multilayer neural network trained using back-propagation scheme to forecast output hydrograph of a hypothetical river for various forecast lead time up to 6.0 hr, the second method uses 1D hydrodynamic MIKE11 model as forecasting model and multilayer neural network as updating kernel to monitor and assess the performance indices compared to ANN alone in light of increase in lead time. Results presented in both graphical and tabular format indicate superiority of MIKE11 coupled with ANN as updating kernel compared to ANN as simulation model alone. While plain ANN produces more accurate results for short lead time, the errors increase expeditiously for longer lead time. The second methodology provides more accurate and reliable results for longer forecast lead time.

Validation of WRF forecasts for the Chajnantor region

NASA Astrophysics Data System (ADS)

Pozo, Diana; Marín, J. C.; Illanes, L.; Curé, M.; Rabanus, D.

2016-06-01

This study assesses the performance of the Weather Research and Forecasting (WRF) model to represent the near-surface weather conditions and the precipitable water vapour (PWV) in the Chajnantor plateau, in the north of Chile, from 2007 April to December. The WRF model shows a very good performance forecasting the near-surface temperature and zonal wind component, although it overestimates the 2 m water vapour mixing ratio and underestimates the 10 m meridional wind component. The model represents very well the seasonal, intraseasonal and the diurnal variation of PWV. However, the PWV errors increase after the 12 h of simulation. Errors in the simulations are larger than 1.5 mm only during 10 per cent of the study period, they do not exceed 0.5 mm during 65 per cent of the time and they are below 0.25 mm more than 45 per cent of the time, which emphasizes the good performance of the model to forecast the PWV over the region. The misrepresentation of the near-surface humidity in the region by the WRF model may have a negative impact on the PWV forecasts. Thus, having accurate forecasts of humidity near the surface may result in more accurate PWV forecasts. Overall, results from this, as well as recent studies, supports the use of the WRF model to provide accurate weather forecasts for the region, particularly for the PWV, which can be of great benefit for astronomers in the planning of their scientific operations and observing time.

The meta-Gaussian Bayesian Processor of forecasts and associated preliminary experiments

NASA Astrophysics Data System (ADS)

Chen, Fajing; Jiao, Meiyan; Chen, Jing

2013-04-01

Public weather services are trending toward providing users with probabilistic weather forecasts, in place of traditional deterministic forecasts. Probabilistic forecasting techniques are continually being improved to optimize available forecasting information. The Bayesian Processor of Forecast (BPF), a new statistical method for probabilistic forecast, can transform a deterministic forecast into a probabilistic forecast according to the historical statistical relationship between observations and forecasts generated by that forecasting system. This technique accounts for the typical forecasting performance of a deterministic forecasting system in quantifying the forecast uncertainty. The meta-Gaussian likelihood model is suitable for a variety of stochastic dependence structures with monotone likelihood ratios. The meta-Gaussian BPF adopting this kind of likelihood model can therefore be applied across many fields, including meteorology and hydrology. The Bayes theorem with two continuous random variables and the normal-linear BPF are briefly introduced. The meta-Gaussian BPF for a continuous predictand using a single predictor is then presented and discussed. The performance of the meta-Gaussian BPF is tested in a preliminary experiment. Control forecasts of daily surface temperature at 0000 UTC at Changsha and Wuhan stations are used as the deterministic forecast data. These control forecasts are taken from ensemble predictions with a 96-h lead time generated by the National Meteorological Center of the China Meteorological Administration, the European Centre for Medium-Range Weather Forecasts, and the US National Centers for Environmental Prediction during January 2008. The results of the experiment show that the meta-Gaussian BPF can transform a deterministic control forecast of surface temperature from any one of the three ensemble predictions into a useful probabilistic forecast of surface temperature. These probabilistic forecasts quantify the uncertainty of the control forecast; accordingly, the performance of the probabilistic forecasts differs based on the source of the underlying deterministic control forecasts.

Improved Neural Networks with Random Weights for Short-Term Load Forecasting

PubMed Central

Lang, Kun; Zhang, Mingyuan; Yuan, Yongbo

2015-01-01

An effective forecasting model for short-term load plays a significant role in promoting the management efficiency of an electric power system. This paper proposes a new forecasting model based on the improved neural networks with random weights (INNRW). The key is to introduce a weighting technique to the inputs of the model and use a novel neural network to forecast the daily maximum load. Eight factors are selected as the inputs. A mutual information weighting algorithm is then used to allocate different weights to the inputs. The neural networks with random weights and kernels (KNNRW) is applied to approximate the nonlinear function between the selected inputs and the daily maximum load due to the fast learning speed and good generalization performance. In the application of the daily load in Dalian, the result of the proposed INNRW is compared with several previously developed forecasting models. The simulation experiment shows that the proposed model performs the best overall in short-term load forecasting. PMID:26629825

Improved Neural Networks with Random Weights for Short-Term Load Forecasting.

PubMed

Lang, Kun; Zhang, Mingyuan; Yuan, Yongbo

2015-01-01

An effective forecasting model for short-term load plays a significant role in promoting the management efficiency of an electric power system. This paper proposes a new forecasting model based on the improved neural networks with random weights (INNRW). The key is to introduce a weighting technique to the inputs of the model and use a novel neural network to forecast the daily maximum load. Eight factors are selected as the inputs. A mutual information weighting algorithm is then used to allocate different weights to the inputs. The neural networks with random weights and kernels (KNNRW) is applied to approximate the nonlinear function between the selected inputs and the daily maximum load due to the fast learning speed and good generalization performance. In the application of the daily load in Dalian, the result of the proposed INNRW is compared with several previously developed forecasting models. The simulation experiment shows that the proposed model performs the best overall in short-term load forecasting.

Adaptive time-variant models for fuzzy-time-series forecasting.

PubMed

Wong, Wai-Keung; Bai, Enjian; Chu, Alice Wai-Ching

2010-12-01

A fuzzy time series has been applied to the prediction of enrollment, temperature, stock indices, and other domains. Related studies mainly focus on three factors, namely, the partition of discourse, the content of forecasting rules, and the methods of defuzzification, all of which greatly influence the prediction accuracy of forecasting models. These studies use fixed analysis window sizes for forecasting. In this paper, an adaptive time-variant fuzzy-time-series forecasting model (ATVF) is proposed to improve forecasting accuracy. The proposed model automatically adapts the analysis window size of fuzzy time series based on the prediction accuracy in the training phase and uses heuristic rules to generate forecasting values in the testing phase. The performance of the ATVF model is tested using both simulated and actual time series including the enrollments at the University of Alabama, Tuscaloosa, and the Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX). The experiment results show that the proposed ATVF model achieves a significant improvement in forecasting accuracy as compared to other fuzzy-time-series forecasting models.

Using Meteorological Analogues for Reordering Postprocessed Precipitation Ensembles in Hydrological Forecasting

NASA Astrophysics Data System (ADS)

Bellier, Joseph; Bontron, Guillaume; Zin, Isabella

2017-12-01

Meteorological ensemble forecasts are nowadays widely used as input of hydrological models for probabilistic streamflow forecasting. These forcings are frequently biased and have to be statistically postprocessed, using most of the time univariate techniques that apply independently to individual locations, lead times and weather variables. Postprocessed ensemble forecasts therefore need to be reordered so as to reconstruct suitable multivariate dependence structures. The Schaake shuffle and ensemble copula coupling are the two most popular methods for this purpose. This paper proposes two adaptations of them that make use of meteorological analogues for reconstructing spatiotemporal dependence structures of precipitation forecasts. Performances of the original and adapted techniques are compared through a multistep verification experiment using real forecasts from the European Centre for Medium-Range Weather Forecasts. This experiment evaluates not only multivariate precipitation forecasts but also the corresponding streamflow forecasts that derive from hydrological modeling. Results show that the relative performances of the different reordering methods vary depending on the verification step. In particular, the standard Schaake shuffle is found to perform poorly when evaluated on streamflow. This emphasizes the crucial role of the precipitation spatiotemporal dependence structure in hydrological ensemble forecasting.

Comparison of Two Hybrid Models for Forecasting the Incidence of Hemorrhagic Fever with Renal Syndrome in Jiangsu Province, China.

PubMed

Wu, Wei; Guo, Junqiao; An, Shuyi; Guan, Peng; Ren, Yangwu; Xia, Linzi; Zhou, Baosen

2015-01-01

Cases of hemorrhagic fever with renal syndrome (HFRS) are widely distributed in eastern Asia, especially in China, Russia, and Korea. It is proved to be a difficult task to eliminate HFRS completely because of the diverse animal reservoirs and effects of global warming. Reliable forecasting is useful for the prevention and control of HFRS. Two hybrid models, one composed of nonlinear autoregressive neural network (NARNN) and autoregressive integrated moving average (ARIMA) the other composed of generalized regression neural network (GRNN) and ARIMA were constructed to predict the incidence of HFRS in the future one year. Performances of the two hybrid models were compared with ARIMA model. The ARIMA, ARIMA-NARNN ARIMA-GRNN model fitted and predicted the seasonal fluctuation well. Among the three models, the mean square error (MSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) of ARIMA-NARNN hybrid model was the lowest both in modeling stage and forecasting stage. As for the ARIMA-GRNN hybrid model, the MSE, MAE and MAPE of modeling performance and the MSE and MAE of forecasting performance were less than the ARIMA model, but the MAPE of forecasting performance did not improve. Developing and applying the ARIMA-NARNN hybrid model is an effective method to make us better understand the epidemic characteristics of HFRS and could be helpful to the prevention and control of HFRS.

Evaluation and Applications of the Prediction of Intensity Model Error (PRIME) Model

NASA Astrophysics Data System (ADS)

Bhatia, K. T.; Nolan, D. S.; Demaria, M.; Schumacher, A.

2015-12-01

Forecasters and end users of tropical cyclone (TC) intensity forecasts would greatly benefit from a reliable expectation of model error to counteract the lack of consistency in TC intensity forecast performance. As a first step towards producing error predictions to accompany each TC intensity forecast, Bhatia and Nolan (2013) studied the relationship between synoptic parameters, TC attributes, and forecast errors. In this study, we build on previous results of Bhatia and Nolan (2013) by testing the ability of the Prediction of Intensity Model Error (PRIME) model to forecast the absolute error and bias of four leading intensity models available for guidance in the Atlantic basin. PRIME forecasts are independently evaluated at each 12-hour interval from 12 to 120 hours during the 2007-2014 Atlantic hurricane seasons. The absolute error and bias predictions of PRIME are compared to their respective climatologies to determine their skill. In addition to these results, we will present the performance of the operational version of PRIME run during the 2015 hurricane season. PRIME verification results show that it can reliably anticipate situations where particular models excel, and therefore could lead to a more informed protocol for hurricane evacuations and storm preparations. These positive conclusions suggest that PRIME forecasts also have the potential to lower the error in the original intensity forecasts of each model. As a result, two techniques are proposed to develop a post-processing procedure for a multimodel ensemble based on PRIME. The first approach is to inverse-weight models using PRIME absolute error predictions (higher predicted absolute error corresponds to lower weights). The second multimodel ensemble applies PRIME bias predictions to each model's intensity forecast and the mean of the corrected models is evaluated. The forecasts of both of these experimental ensembles are compared to those of the equal-weight ICON ensemble, which currently provides the most reliable forecasts in the Atlantic basin.

Performance evaluation of ionospheric time delay forecasting models using GPS observations at a low-latitude station

NASA Astrophysics Data System (ADS)

Sivavaraprasad, G.; Venkata Ratnam, D.

2017-07-01

Ionospheric delay is one of the major atmospheric effects on the performance of satellite-based radio navigation systems. It limits the accuracy and availability of Global Positioning System (GPS) measurements, related to critical societal and safety applications. The temporal and spatial gradients of ionospheric total electron content (TEC) are driven by several unknown priori geophysical conditions and solar-terrestrial phenomena. Thereby, the prediction of ionospheric delay is challenging especially over Indian sub-continent. Therefore, an appropriate short/long-term ionospheric delay forecasting model is necessary. Hence, the intent of this paper is to forecast ionospheric delays by considering day to day, monthly and seasonal ionospheric TEC variations. GPS-TEC data (January 2013-December 2013) is extracted from a multi frequency GPS receiver established at K L University, Vaddeswaram, Guntur station (geographic: 16.37°N, 80.37°E; geomagnetic: 7.44°N, 153.75°E), India. An evaluation, in terms of forecasting capabilities, of three ionospheric time delay models - an Auto Regressive Moving Average (ARMA) model, Auto Regressive Integrated Moving Average (ARIMA) model, and a Holt-Winter's model is presented. The performances of these models are evaluated through error measurement analysis during both geomagnetic quiet and disturbed days. It is found that, ARMA model is effectively forecasting the ionospheric delay with an accuracy of 82-94%, which is 10% more superior to ARIMA and Holt-Winter's models. Moreover, the modeled VTEC derived from International Reference Ionosphere, IRI (IRI-2012) model and new global TEC model, Neustrelitz TEC Model (NTCM-GL) have compared with forecasted VTEC values of ARMA, ARIMA and Holt-Winter's models during geomagnetic quiet days. The forecast results are indicating that ARMA model would be useful to set up an early warning system for ionospheric disturbances at low latitude regions.

Forecasting Space Weather-Induced GPS Performance Degradation Using Random Forest

NASA Astrophysics Data System (ADS)

Filjar, R.; Filic, M.; Milinkovic, F.

2017-12-01

Space weather and ionospheric dynamics have a profound effect on positioning performance of the Global Satellite Navigation System (GNSS). However, the quantification of that effect is still the subject of scientific activities around the world. In the latest contribution to the understanding of the space weather and ionospheric effects on satellite-based positioning performance, we conducted a study of several candidates for forecasting method for space weather-induced GPS positioning performance deterioration. First, a 5-days set of experimentally collected data was established, encompassing the space weather and ionospheric activity indices (including: the readings of the Sudden Ionospheric Disturbance (SID) monitors, components of geomagnetic field strength, global Kp index, Dst index, GPS-derived Total Electron Content (TEC) samples, standard deviation of TEC samples, and sunspot number) and observations of GPS positioning error components (northing, easting, and height positioning error) derived from the Adriatic Sea IGS reference stations' RINEX raw pseudorange files in quiet space weather periods. This data set was split into the training and test sub-sets. Then, a selected set of supervised machine learning methods based on Random Forest was applied to the experimentally collected data set in order to establish the appropriate regional (the Adriatic Sea) forecasting models for space weather-induced GPS positioning performance deterioration. The forecasting models were developed in the R/rattle statistical programming environment. The forecasting quality of the regional forecasting models developed was assessed, and the conclusions drawn on the advantages and shortcomings of the regional forecasting models for space weather-caused GNSS positioning performance deterioration.

Forecasting of dissolved oxygen in the Guanting reservoir using an optimized NGBM (1,1) model.

PubMed

An, Yan; Zou, Zhihong; Zhao, Yanfei

2015-03-01

An optimized nonlinear grey Bernoulli model was proposed by using a particle swarm optimization algorithm to solve the parameter optimization problem. In addition, each item in the first-order accumulated generating sequence was set in turn as an initial condition to determine which alternative would yield the highest forecasting accuracy. To test the forecasting performance, the optimized models with different initial conditions were then used to simulate dissolved oxygen concentrations in the Guanting reservoir inlet and outlet (China). The empirical results show that the optimized model can remarkably improve forecasting accuracy, and the particle swarm optimization technique is a good tool to solve parameter optimization problems. What's more, the optimized model with an initial condition that performs well in in-sample simulation may not do as well as in out-of-sample forecasting. Copyright © 2015. Published by Elsevier B.V.

An application of a multi model approach for solar energy prediction in Southern Italy

NASA Astrophysics Data System (ADS)

Avolio, Elenio; Lo Feudo, Teresa; Calidonna, Claudia Roberta; Contini, Daniele; Torcasio, Rosa Claudia; Tiriolo, Luca; Montesanti, Stefania; Transerici, Claudio; Federico, Stefano

2015-04-01

The accuracy of the short and medium range forecast of solar irradiance is very important for solar energy integration into the grid. This issue is particularly important for Southern Italy where a significant availability of solar energy is associated with a poor development of the grid. In this work we analyse the performance of two deterministic models for the prediction of surface temperature and short-wavelength radiance for two sites in southern Italy. Both parameters are needed to forecast the power production from solar power plants, so the performance of the forecast for these meteorological parameters is of paramount importance. The models considered in this work are the RAMS (Regional Atmospheric Modeling System) and the WRF (Weather Research and Forecasting Model) and they were run for the summer 2013 at 4 km horizontal resolution over Italy. The forecast lasts three days. Initial and dynamic boundary conditions are given by the 12 UTC deterministic forecast of the ECMWF-IFS (European Centre for Medium Weather Range Forecast - Integrated Forecasting System) model, and were available every 6 hours. Verification is given against two surface stations located in Southern Italy, Lamezia Terme and Lecce, and are based on hourly output of models forecast. Results for the whole period for temperature show a positive bias for the RAMS model and a negative bias for the WRF model. RMSE is between 1 and 2 °C for both models. Results for the whole period for the short-wavelength radiance show a positive bias for both models (about 30 W/m2 for both models) and a RMSE of 100 W/m2. To reduce the model errors, a statistical post-processing technique, i.e the multi-model, is adopted. In this approach the two model's outputs are weighted with an adequate set of weights computed for a training period. In general, the performance is improved by the application of the technique, and the RMSE is reduced by a sizeable fraction (i.e. larger than 10% of the initial RMSE) depending on the forecasting time and parameter. The performance of the multi model is discussed as a function of the length of the training period and is compared with the performance of the MOS (Model Output Statistics) approach. ACKNOWLEDGMENTS This work is partially supported by projects PON04a2E Sinergreen-ResNovae - "Smart Energy Master for the energetic government of the territory" and PONa3_00363 "High Technology Infrastructure for Climate and Environment Monitoring" (I-AMICA) founded by Italian Ministry of University and Research (MIUR) PON 2007-2013. The ECMWF and CNMCA (Centro Nazionale di Meteorologia e Climatologia Aeronautica) are acknowledged for the use of the MARS (Meteorological Archive and Retrieval System).

Predictive Skill of Meteorological Drought Based on Multi-Model Ensemble Forecasts: A Real-Time Assessment

NASA Astrophysics Data System (ADS)

Chen, L. C.; Mo, K. C.; Zhang, Q.; Huang, J.

2014-12-01

Drought prediction from monthly to seasonal time scales is of critical importance to disaster mitigation, agricultural planning, and multi-purpose reservoir management. Starting in December 2012, NOAA Climate Prediction Center (CPC) has been providing operational Standardized Precipitation Index (SPI) Outlooks using the North American Multi-Model Ensemble (NMME) forecasts, to support CPC's monthly drought outlooks and briefing activities. The current NMME system consists of six model forecasts from U.S. and Canada modeling centers, including the CFSv2, CM2.1, GEOS-5, CCSM3.0, CanCM3, and CanCM4 models. In this study, we conduct an assessment of the predictive skill of meteorological drought using real-time NMME forecasts for the period from May 2012 to May 2014. The ensemble SPI forecasts are the equally weighted mean of the six model forecasts. Two performance measures, the anomaly correlation coefficient and root-mean-square errors against the observations, are used to evaluate forecast skill.Similar to the assessment based on NMME retrospective forecasts, predictive skill of monthly-mean precipitation (P) forecasts is generally low after the second month and errors vary among models. Although P forecast skill is not large, SPI predictive skill is high and the differences among models are small. The skill mainly comes from the P observations appended to the model forecasts. This factor also contributes to the similarity of SPI prediction among the six models. Still, NMME SPI ensemble forecasts have higher skill than those based on individual models or persistence, and the 6-month SPI forecasts are skillful out to four months. The three major drought events occurred during the 2012-2014 period, the 2012 Central Great Plains drought, the 2013 Upper Midwest flash drought, and 2013-2014 California drought, are used as examples to illustrate the system's strength and limitations. For precipitation-driven drought events, such as the 2012 Central Great Plains drought, NMME SPI forecasts perform well in predicting drought severity and spatial patterns. For fast-developing drought events, such as the 2013 Upper Midwest flash drought, the system failed to capture the onset of the drought.

Hourly runoff forecasting for flood risk management: Application of various computational intelligence models

NASA Astrophysics Data System (ADS)

Badrzadeh, Honey; Sarukkalige, Ranjan; Jayawardena, A. W.

2015-10-01

Reliable river flow forecasts play a key role in flood risk mitigation. Among different approaches of river flow forecasting, data driven approaches have become increasingly popular in recent years due to their minimum information requirements and ability to simulate nonlinear and non-stationary characteristics of hydrological processes. In this study, attempts are made to apply four different types of data driven approaches, namely traditional artificial neural networks (ANN), adaptive neuro-fuzzy inference systems (ANFIS), wavelet neural networks (WNN), and, hybrid ANFIS with multi resolution analysis using wavelets (WNF). Developed models applied for real time flood forecasting at Casino station on Richmond River, Australia which is highly prone to flooding. Hourly rainfall and runoff data were used to drive the models which have been used for forecasting with 1, 6, 12, 24, 36 and 48 h lead-time. The performance of models further improved by adding an upstream river flow data (Wiangaree station), as another effective input. All models perform satisfactorily up to 12 h lead-time. However, the hybrid wavelet-based models significantly outperforming the ANFIS and ANN models in the longer lead-time forecasting. The results confirm the robustness of the proposed structure of the hybrid models for real time runoff forecasting in the study area.

Assessing a 3D smoothed seismicity model of induced earthquakes

NASA Astrophysics Data System (ADS)

Zechar, Jeremy; Király, Eszter; Gischig, Valentin; Wiemer, Stefan

2016-04-01

As more energy exploration and extraction efforts cause earthquakes, it becomes increasingly important to control induced seismicity. Risk management schemes must be improved and should ultimately be based on near-real-time forecasting systems. With this goal in mind, we propose a test bench to evaluate models of induced seismicity based on metrics developed by the CSEP community. To illustrate the test bench, we consider a model based on the so-called seismogenic index and a rate decay; to produce three-dimensional forecasts, we smooth past earthquakes in space and time. We explore four variants of this model using the Basel 2006 and Soultz-sous-Forêts 2004 datasets to make short-term forecasts, test their consistency, and rank the model variants. Our results suggest that such a smoothed seismicity model is useful for forecasting induced seismicity within three days, and giving more weight to recent events improves forecast performance. Moreover, the location of the largest induced earthquake is forecast well by this model. Despite the good spatial performance, the model does not estimate the seismicity rate well: it frequently overestimates during stimulation and during the early post-stimulation period, and it systematically underestimates around shut-in. In this presentation, we also describe a robust estimate of information gain, a modification that can also benefit forecast experiments involving tectonic earthquakes.

Cloud Computing Applications in Support of Earth Science Activities at Marshall Space Flight Center

NASA Technical Reports Server (NTRS)

Molthan, Andrew L.; Limaye, Ashutosh S.; Srikishen, Jayanthi

2011-01-01

Currently, the NASA Nebula Cloud Computing Platform is available to Agency personnel in a pre-release status as the system undergoes a formal operational readiness review. Over the past year, two projects within the Earth Science Office at NASA Marshall Space Flight Center have been investigating the performance and value of Nebula s "Infrastructure as a Service", or "IaaS" concept and applying cloud computing concepts to advance their respective mission goals. The Short-term Prediction Research and Transition (SPoRT) Center focuses on the transition of unique NASA satellite observations and weather forecasting capabilities for use within the operational forecasting community through partnerships with NOAA s National Weather Service (NWS). SPoRT has evaluated the performance of the Weather Research and Forecasting (WRF) model on virtual machines deployed within Nebula and used Nebula instances to simulate local forecasts in support of regional forecast studies of interest to select NWS forecast offices. In addition to weather forecasting applications, rapidly deployable Nebula virtual machines have supported the processing of high resolution NASA satellite imagery to support disaster assessment following the historic severe weather and tornado outbreak of April 27, 2011. Other modeling and satellite analysis activities are underway in support of NASA s SERVIR program, which integrates satellite observations, ground-based data and forecast models to monitor environmental change and improve disaster response in Central America, the Caribbean, Africa, and the Himalayas. Leveraging SPoRT s experience, SERVIR is working to establish a real-time weather forecasting model for Central America. Other modeling efforts include hydrologic forecasts for Kenya, driven by NASA satellite observations and reanalysis data sets provided by the broader meteorological community. Forecast modeling efforts are supplemented by short-term forecasts of convective initiation, determined by geostationary satellite observations processed on virtual machines powered by Nebula.

The RAPIDD ebola forecasting challenge: Synthesis and lessons learnt.

PubMed

Viboud, Cécile; Sun, Kaiyuan; Gaffey, Robert; Ajelli, Marco; Fumanelli, Laura; Merler, Stefano; Zhang, Qian; Chowell, Gerardo; Simonsen, Lone; Vespignani, Alessandro

2018-03-01

Infectious disease forecasting is gaining traction in the public health community; however, limited systematic comparisons of model performance exist. Here we present the results of a synthetic forecasting challenge inspired by the West African Ebola crisis in 2014-2015 and involving 16 international academic teams and US government agencies, and compare the predictive performance of 8 independent modeling approaches. Challenge participants were invited to predict 140 epidemiological targets across 5 different time points of 4 synthetic Ebola outbreaks, each involving different levels of interventions and "fog of war" in outbreak data made available for predictions. Prediction targets included 1-4 week-ahead case incidences, outbreak size, peak timing, and several natural history parameters. With respect to weekly case incidence targets, ensemble predictions based on a Bayesian average of the 8 participating models outperformed any individual model and did substantially better than a null auto-regressive model. There was no relationship between model complexity and prediction accuracy; however, the top performing models for short-term weekly incidence were reactive models with few parameters, fitted to a short and recent part of the outbreak. Individual model outputs and ensemble predictions improved with data accuracy and availability; by the second time point, just before the peak of the epidemic, estimates of final size were within 20% of the target. The 4th challenge scenario - mirroring an uncontrolled Ebola outbreak with substantial data reporting noise - was poorly predicted by all modeling teams. Overall, this synthetic forecasting challenge provided a deep understanding of model performance under controlled data and epidemiological conditions. We recommend such "peace time" forecasting challenges as key elements to improve coordination and inspire collaboration between modeling groups ahead of the next pandemic threat, and to assess model forecasting accuracy for a variety of known and hypothetical pathogens. Published by Elsevier B.V.

Validation of the Kp Geomagnetic Index Forecast at CCMC

NASA Astrophysics Data System (ADS)

Frechette, B. P.; Mays, M. L.

2017-12-01

The Community Coordinated Modeling Center (CCMC) Space Weather Research Center (SWRC) sub-team provides space weather services to NASA robotic mission operators and science campaigns and prototypes new models, forecasting techniques, and procedures. The Kp index is a measure of geomagnetic disturbances for space weather in the magnetosphere such as geomagnetic storms and substorms. In this study, we performed validation on the Newell et al. (2007) Kp prediction equation from December 2010 to July 2017. The purpose of this research is to understand the Kp forecast performance because it's critical for NASA missions to have confidence in the space weather forecast. This research was done by computing the Kp error for each forecast (average, minimum, maximum) and each synoptic period. Then to quantify forecast performance we computed the mean error, mean absolute error, root mean square error, multiplicative bias and correlation coefficient. A contingency table was made for each forecast and skill scores were computed. The results are compared to the perfect score and reference forecast skill score. In conclusion, the skill score and error results show that the minimum of the predicted Kp over each synoptic period from the Newell et al. (2007) Kp prediction equation performed better than the maximum or average of the prediction. However, persistence (reference forecast) outperformed all of the Kp forecasts (minimum, maximum, and average). Overall, the Newell Kp prediction still predicts within a range of 1, even though persistence beats it.

Nowcasting and Forecasting the Monthly Food Stamps Data in the US Using Online Search Data

PubMed Central

Fantazzini, Dean

2014-01-01

We propose the use of Google online search data for nowcasting and forecasting the number of food stamps recipients. We perform a large out-of-sample forecasting exercise with almost 3000 competing models with forecast horizons up to 2 years ahead, and we show that models including Google search data statistically outperform the competing models at all considered horizons. These results hold also with several robustness checks, considering alternative keywords, a falsification test, different out-of-samples, directional accuracy and forecasts at the state-level. PMID:25369315

Weather Research and Forecasting Model Wind Sensitivity Study at Edwards Air Force Base, CA

NASA Technical Reports Server (NTRS)

Watson, Leela R.; Bauman, William H., III

2008-01-01

NASA prefers to land the space shuttle at Kennedy Space Center (KSC). When weather conditions violate Flight Rules at KSC, NASA will usually divert the shuttle landing to Edwards Air Force Base (EAFB) in Southern California. But forecasting surface winds at EAFB is a challenge for the Spaceflight Meteorology Group (SMG) forecasters due to the complex terrain that surrounds EAFB, One particular phenomena identified by SMG is that makes it difficult to forecast the EAFB surface winds is called "wind cycling". This occurs when wind speeds and directions oscillate among towers near the EAFB runway leading to a challenging deorbit bum forecast for shuttle landings. The large-scale numerical weather prediction models cannot properly resolve the wind field due to their coarse horizontal resolutions, so a properly tuned high-resolution mesoscale model is needed. The Weather Research and Forecasting (WRF) model meets this requirement. The AMU assessed the different WRF model options to determine which configuration best predicted surface wind speed and direction at EAFB, To do so, the AMU compared the WRF model performance using two hot start initializations with the Advanced Research WRF and Non-hydrostatic Mesoscale Model dynamical cores and compared model performance while varying the physics options.

Assessing the performance of eight real-time updating models and procedures for the Brosna River

NASA Astrophysics Data System (ADS)

Goswami, M.; O'Connor, K. M.; Bhattarai, K. P.; Shamseldin, A. Y.

2005-10-01

The flow forecasting performance of eight updating models, incorporated in the Galway River Flow Modelling and Forecasting System (GFMFS), was assessed using daily data (rainfall, evaporation and discharge) of the Irish Brosna catchment (1207 km2), considering their one to six days lead-time discharge forecasts. The Perfect Forecast of Input over the Forecast Lead-time scenario was adopted, where required, in place of actual rainfall forecasts. The eight updating models were: (i) the standard linear Auto-Regressive (AR) model, applied to the forecast errors (residuals) of a simulation (non-updating) rainfall-runoff model; (ii) the Neural Network Updating (NNU) model, also using such residuals as input; (iii) the Linear Transfer Function (LTF) model, applied to the simulated and the recently observed discharges; (iv) the Non-linear Auto-Regressive eXogenous-Input Model (NARXM), also a neural network-type structure, but having wide options of using recently observed values of one or more of the three data series, together with non-updated simulated outflows, as inputs; (v) the Parametric Simple Linear Model (PSLM), of LTF-type, using recent rainfall and observed discharge data; (vi) the Parametric Linear perturbation Model (PLPM), also of LTF-type, using recent rainfall and observed discharge data, (vii) n-AR, an AR model applied to the observed discharge series only, as a naïve updating model; and (viii) n-NARXM, a naive form of the NARXM, using only the observed discharge data, excluding exogenous inputs. The five GFMFS simulation (non-updating) models used were the non-parametric and parametric forms of the Simple Linear Model and of the Linear Perturbation Model, the Linearly-Varying Gain Factor Model, the Artificial Neural Network Model, and the conceptual Soil Moisture Accounting and Routing (SMAR) model. As the SMAR model performance was found to be the best among these models, in terms of the Nash-Sutcliffe R2 value, both in calibration and in verification, the simulated outflows of this model only were selected for the subsequent exercise of producing updated discharge forecasts. All the eight forms of updating models for producing lead-time discharge forecasts were found to be capable of producing relatively good lead-1 (1-day ahead) forecasts, with R2 values almost 90% or above. However, for higher lead time forecasts, only three updating models, viz., NARXM, LTF, and NNU, were found to be suitable, with lead-6 values of R2 about 90% or higher. Graphical comparisons were made of the lead-time forecasts for the two largest floods, one in the calibration period and the other in the verification period.

Quantitative precipitation forecasts in the Alps - an assessment from the Forecast Demonstration Project MAP D-PHASE

NASA Astrophysics Data System (ADS)

Ament, F.; Weusthoff, T.; Arpagaus, M.; Rotach, M.

2009-04-01

The main aim of the WWRP Forecast Demonstration Project MAP D-PHASE is to demonstrate the performance of today's models to forecast heavy precipitation and flood events in the Alpine region. Therefore an end-to-end, real-time forecasting system was installed and operated during the D PHASE Operations Period from June to November 2007. Part of this system are 30 numerical weather prediction models (deterministic as well as ensemble systems) operated by weather services and research institutes, which issue alerts if predicted precipitation accumulations exceed critical thresholds. Additionally to the real-time alerts, all relevant model fields of these simulations are stored in a central data archive. This comprehensive data set allows a detailed assessment of today's quantitative precipitation forecast (QPF) performance in the Alpine region. We will present results of QPF verifications against Swiss radar and rain gauge data both from a qualitative point of view, in terms of alerts, as well as from a quantitative perspective, in terms of precipitation rate. Various influencing factors like lead time, accumulation time, selection of warning thresholds, or bias corrections will be discussed. Additional to traditional verifications of area average precipitation amounts, the performance of the models to predict the correct precipitation statistics without requiring a point-to-point match will be described by using modern Fuzzy verification techniques. Both analyses reveal significant advantages of deep convection resolving models compared to coarser models with parameterized convection. An intercomparison of the model forecasts themselves reveals a remarkably high variability between different models, and makes it worthwhile to evaluate the potential of a multi-model ensemble. Various multi-model ensemble strategies will be tested by combining D-PHASE models to virtual ensemble systems.

Performance of an Advanced MOS System in the 1996-97 National Collegiate Weather Forecasting Contest.

NASA Astrophysics Data System (ADS)

Vislocky, Robert L.; Fritsch, J. Michael

1997-12-01

A prototype advanced model output statistics (MOS) forecast system that was entered in the 1996-97 National Collegiate Weather Forecast Contest is described and its performance compared to that of widely available objective guidance and to contest participants. The prototype system uses an optimal blend of aviation (AVN) and nested grid model (NGM) MOS forecasts, explicit output from the NGM and Eta guidance, and the latest surface weather observations from the forecast site. The forecasts are totally objective and can be generated quickly on a personal computer. Other "objective" forms of guidance tracked in the contest are 1) the consensus forecast (i.e., the average of the forecasts from all of the human participants), 2) the combination of NGM raw output (for precipitation forecasts) and NGM MOS guidance (for temperature forecasts), and 3) the combination of Eta Model raw output (for precipitation forecasts) and AVN MOS guidance (for temperature forecasts).Results show that the advanced MOS system finished in 20th place out of 737 original entrants, or better than approximately 97% of the human forecasters who entered the contest. Moreover, the advanced MOS system was slightly better than consensus (23d place). The fact that an objective forecast system finished ahead of consensus is a significant accomplishment since consensus is traditionally a very formidable "opponent" in forecast competitions. Equally significant is that the advanced MOS system was superior to the traditional guidance products available from the National Centers for Environmental Prediction (NCEP). Specifically, the combination of NGM raw output and NGM MOS guidance finished in 175th place, and the combination of Eta Model raw output and AVN MOS guidance finished in 266th place. The latter result is most intriguing since the proposed elimination of all NGM products would likely result in a serious degradation of objective products disseminated by NCEP, unless they are replaced with equal or better substitutes. On the other hand, the positive performance of the prototype advanced MOS system shows that it is possible to create a single objective product that is not only superior to currently available objective guidance products, but is also on par with some of the better human forecasters.

Performance assessment of a Bayesian Forecasting System (BFS) for real-time flood forecasting

NASA Astrophysics Data System (ADS)

Biondi, D.; De Luca, D. L.

2013-02-01

SummaryThe paper evaluates, for a number of flood events, the performance of a Bayesian Forecasting System (BFS), with the aim of evaluating total uncertainty in real-time flood forecasting. The predictive uncertainty of future streamflow is estimated through the Bayesian integration of two separate processors. The former evaluates the propagation of input uncertainty on simulated river discharge, the latter computes the hydrological uncertainty of actual river discharge associated with all other possible sources of error. A stochastic model and a distributed rainfall-runoff model were assumed, respectively, for rainfall and hydrological response simulations. A case study was carried out for a small basin in the Calabria region (southern Italy). The performance assessment of the BFS was performed with adequate verification tools suited for probabilistic forecasts of continuous variables such as streamflow. Graphical tools and scalar metrics were used to evaluate several attributes of the forecast quality of the entire time-varying predictive distributions: calibration, sharpness, accuracy, and continuous ranked probability score (CRPS). Besides the overall system, which incorporates both sources of uncertainty, other hypotheses resulting from the BFS properties were examined, corresponding to (i) a perfect hydrological model; (ii) a non-informative rainfall forecast for predicting streamflow; and (iii) a perfect input forecast. The results emphasize the importance of using different diagnostic approaches to perform comprehensive analyses of predictive distributions, to arrive at a multifaceted view of the attributes of the prediction. For the case study, the selected criteria revealed the interaction of the different sources of error, in particular the crucial role of the hydrological uncertainty processor when compensating, at the cost of wider forecast intervals, for the unreliable and biased predictive distribution resulting from the Precipitation Uncertainty Processor.

Applications of the gambling score in evaluating earthquake predictions and forecasts

NASA Astrophysics Data System (ADS)

Zhuang, Jiancang; Zechar, Jeremy D.; Jiang, Changsheng; Console, Rodolfo; Murru, Maura; Falcone, Giuseppe

2010-05-01

This study presents a new method, namely the gambling score, for scoring the performance earthquake forecasts or predictions. Unlike most other scoring procedures that require a regular scheme of forecast and treat each earthquake equally, regardless their magnitude, this new scoring method compensates the risk that the forecaster has taken. Starting with a certain number of reputation points, once a forecaster makes a prediction or forecast, he is assumed to have betted some points of his reputation. The reference model, which plays the role of the house, determines how many reputation points the forecaster can gain if he succeeds, according to a fair rule, and also takes away the reputation points bet by the forecaster if he loses. This method is also extended to the continuous case of point process models, where the reputation points betted by the forecaster become a continuous mass on the space-time-magnitude range of interest. For discrete predictions, we apply this method to evaluate performance of Shebalin's predictions made by using the Reverse Tracing of Precursors (RTP) algorithm and of the outputs of the predictions from the Annual Consultation Meeting on Earthquake Tendency held by China Earthquake Administration. For the continuous case, we use it to compare the probability forecasts of seismicity in the Abruzzo region before and after the L'aquila earthquake based on the ETAS model and the PPE model.

Markov chain-incorporated and synthetic data-supported conditional artificial neural network models for forecasting monthly precipitation in arid regions

NASA Astrophysics Data System (ADS)

Aksoy, Hafzullah; Dahamsheh, Ahmad

2018-07-01

For forecasting monthly precipitation in an arid region, the feed forward back-propagation, radial basis function and generalized regression artificial neural networks (ANNs) are used in this study. The ANN models are improved after incorporation of a Markov chain-based algorithm (MC-ANNs) with which the percentage of dry months is forecasted perfectly, thus generation of any non-physical negative precipitation is eliminated. Due to the fact that recorded precipitation time series are usually shorter than the length needed for a proper calibration of ANN models, synthetic monthly precipitation data are generated by Thomas-Fiering model to further improve the performance of forecasting. For case studies from Jordan, it is seen that only a slightly better performance is achieved with the use of MC and synthetic data. A conditional statement is, therefore, established and imbedded into the ANN models after the incorporation of MC and support of synthetic data, to substantially improve the ability of the models for forecasting monthly precipitation in arid regions.

Purposes and methods of scoring earthquake forecasts

NASA Astrophysics Data System (ADS)

Zhuang, J.

2010-12-01

There are two kinds of purposes in the studies on earthquake prediction or forecasts: one is to give a systematic estimation of earthquake risks in some particular region and period in order to give advice to governments and enterprises for the use of reducing disasters, the other one is to search for reliable precursors that can be used to improve earthquake prediction or forecasts. For the first case, a complete score is necessary, while for the latter case, a partial score, which can be used to evaluate whether the forecasts or predictions have some advantages than a well know model, is necessary. This study reviews different scoring methods for evaluating the performance of earthquake prediction and forecasts. Especially, the gambling scoring method, which is developed recently, shows its capacity in finding good points in an earthquake prediction algorithm or model that are not in a reference model, even if its overall performance is no better than the reference model.

Gambling scores for earthquake predictions and forecasts

NASA Astrophysics Data System (ADS)

Zhuang, Jiancang

2010-04-01

This paper presents a new method, namely the gambling score, for scoring the performance earthquake forecasts or predictions. Unlike most other scoring procedures that require a regular scheme of forecast and treat each earthquake equally, regardless their magnitude, this new scoring method compensates the risk that the forecaster has taken. Starting with a certain number of reputation points, once a forecaster makes a prediction or forecast, he is assumed to have betted some points of his reputation. The reference model, which plays the role of the house, determines how many reputation points the forecaster can gain if he succeeds, according to a fair rule, and also takes away the reputation points betted by the forecaster if he loses. This method is also extended to the continuous case of point process models, where the reputation points betted by the forecaster become a continuous mass on the space-time-magnitude range of interest. We also calculate the upper bound of the gambling score when the true model is a renewal process, the stress release model or the ETAS model and when the reference model is the Poisson model.

Comparison of Two Hybrid Models for Forecasting the Incidence of Hemorrhagic Fever with Renal Syndrome in Jiangsu Province, China

PubMed Central

Wu, Wei; Guo, Junqiao; An, Shuyi; Guan, Peng; Ren, Yangwu; Xia, Linzi; Zhou, Baosen

2015-01-01

Background Cases of hemorrhagic fever with renal syndrome (HFRS) are widely distributed in eastern Asia, especially in China, Russia, and Korea. It is proved to be a difficult task to eliminate HFRS completely because of the diverse animal reservoirs and effects of global warming. Reliable forecasting is useful for the prevention and control of HFRS. Methods Two hybrid models, one composed of nonlinear autoregressive neural network (NARNN) and autoregressive integrated moving average (ARIMA) the other composed of generalized regression neural network (GRNN) and ARIMA were constructed to predict the incidence of HFRS in the future one year. Performances of the two hybrid models were compared with ARIMA model. Results The ARIMA, ARIMA-NARNN ARIMA-GRNN model fitted and predicted the seasonal fluctuation well. Among the three models, the mean square error (MSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) of ARIMA-NARNN hybrid model was the lowest both in modeling stage and forecasting stage. As for the ARIMA-GRNN hybrid model, the MSE, MAE and MAPE of modeling performance and the MSE and MAE of forecasting performance were less than the ARIMA model, but the MAPE of forecasting performance did not improve. Conclusion Developing and applying the ARIMA-NARNN hybrid model is an effective method to make us better understand the epidemic characteristics of HFRS and could be helpful to the prevention and control of HFRS. PMID:26270814

Large-watershed flood simulation and forecasting based on different-resolution distributed hydrological model

NASA Astrophysics Data System (ADS)

Li, J.

2017-12-01

Large-watershed flood simulation and forecasting is very important for a distributed hydrological model in the application. There are some challenges including the model's spatial resolution effect, model performance and accuracy and so on. To cope with the challenge of the model's spatial resolution effect, different model resolution including 1000m*1000m, 600m*600m, 500m*500m, 400m*400m, 200m*200m were used to build the distributed hydrological model—Liuxihe model respectively. The purpose is to find which one is the best resolution for Liuxihe model in Large-watershed flood simulation and forecasting. This study sets up a physically based distributed hydrological model for flood forecasting of the Liujiang River basin in south China. Terrain data digital elevation model (DEM), soil type and land use type are downloaded from the website freely. The model parameters are optimized by using an improved Particle Swarm Optimization(PSO) algorithm; And parameter optimization could reduce the parameter uncertainty that exists for physically deriving model parameters. The different model resolution (200m*200m—1000m*1000m ) are proposed for modeling the Liujiang River basin flood with the Liuxihe model in this study. The best model's spatial resolution effect for flood simulation and forecasting is 200m*200m.And with the model's spatial resolution reduction, the model performance and accuracy also become worse and worse. When the model resolution is 1000m*1000m, the flood simulation and forecasting result is the worst, also the river channel divided based on this resolution is differs from the actual one. To keep the model with an acceptable performance, minimum model spatial resolution is needed. The suggested threshold model spatial resolution for modeling the Liujiang River basin flood is a 500m*500m grid cell, but the model spatial resolution with a 200m*200m grid cell is recommended in this study to keep the model at a best performance.

A multi-source data assimilation framework for flood forecasting: Accounting for runoff routing lags

NASA Astrophysics Data System (ADS)

Meng, S.; Xie, X.

2015-12-01

In the flood forecasting practice, model performance is usually degraded due to various sources of uncertainties, including the uncertainties from input data, model parameters, model structures and output observations. Data assimilation is a useful methodology to reduce uncertainties in flood forecasting. For the short-term flood forecasting, an accurate estimation of initial soil moisture condition will improve the forecasting performance. Considering the time delay of runoff routing is another important effect for the forecasting performance. Moreover, the observation data of hydrological variables (including ground observations and satellite observations) are becoming easily available. The reliability of the short-term flood forecasting could be improved by assimilating multi-source data. The objective of this study is to develop a multi-source data assimilation framework for real-time flood forecasting. In this data assimilation framework, the first step is assimilating the up-layer soil moisture observations to update model state and generated runoff based on the ensemble Kalman filter (EnKF) method, and the second step is assimilating discharge observations to update model state and runoff within a fixed time window based on the ensemble Kalman smoother (EnKS) method. This smoothing technique is adopted to account for the runoff routing lag. Using such assimilation framework of the soil moisture and discharge observations is expected to improve the flood forecasting. In order to distinguish the effectiveness of this dual-step assimilation framework, we designed a dual-EnKF algorithm in which the observed soil moisture and discharge are assimilated separately without accounting for the runoff routing lag. The results show that the multi-source data assimilation framework can effectively improve flood forecasting, especially when the runoff routing has a distinct time lag. Thus, this new data assimilation framework holds a great potential in operational flood forecasting by merging observations from ground measurement and remote sensing retrivals.

A multifactor approach to forecasting Romanian gross domestic product (GDP) in the short run.

PubMed

Armeanu, Daniel; Andrei, Jean Vasile; Lache, Leonard; Panait, Mirela

2017-01-01

The purpose of this paper is to investigate the application of a generalized dynamic factor model (GDFM) based on dynamic principal components analysis to forecasting short-term economic growth in Romania. We have used a generalized principal components approach to estimate a dynamic model based on a dataset comprising 86 economic and non-economic variables that are linked to economic output. The model exploits the dynamic correlations between these variables and uses three common components that account for roughly 72% of the information contained in the original space. We show that it is possible to generate reliable forecasts of quarterly real gross domestic product (GDP) using just the common components while also assessing the contribution of the individual variables to the dynamics of real GDP. In order to assess the relative performance of the GDFM to standard models based on principal components analysis, we have also estimated two Stock-Watson (SW) models that were used to perform the same out-of-sample forecasts as the GDFM. The results indicate significantly better performance of the GDFM compared with the competing SW models, which empirically confirms our expectations that the GDFM produces more accurate forecasts when dealing with large datasets.

A multifactor approach to forecasting Romanian gross domestic product (GDP) in the short run

PubMed Central

Armeanu, Daniel; Lache, Leonard; Panait, Mirela

2017-01-01

The purpose of this paper is to investigate the application of a generalized dynamic factor model (GDFM) based on dynamic principal components analysis to forecasting short-term economic growth in Romania. We have used a generalized principal components approach to estimate a dynamic model based on a dataset comprising 86 economic and non-economic variables that are linked to economic output. The model exploits the dynamic correlations between these variables and uses three common components that account for roughly 72% of the information contained in the original space. We show that it is possible to generate reliable forecasts of quarterly real gross domestic product (GDP) using just the common components while also assessing the contribution of the individual variables to the dynamics of real GDP. In order to assess the relative performance of the GDFM to standard models based on principal components analysis, we have also estimated two Stock-Watson (SW) models that were used to perform the same out-of-sample forecasts as the GDFM. The results indicate significantly better performance of the GDFM compared with the competing SW models, which empirically confirms our expectations that the GDFM produces more accurate forecasts when dealing with large datasets. PMID:28742100

A Case Study on a Combination NDVI Forecasting Model Based on the Entropy Weight Method

DOE Office of Scientific and Technical Information (OSTI.GOV)

Huang, Shengzhi; Ming, Bo; Huang, Qiang

It is critically meaningful to accurately predict NDVI (Normalized Difference Vegetation Index), which helps guide regional ecological remediation and environmental managements. In this study, a combination forecasting model (CFM) was proposed to improve the performance of NDVI predictions in the Yellow River Basin (YRB) based on three individual forecasting models, i.e., the Multiple Linear Regression (MLR), Artificial Neural Network (ANN), and Support Vector Machine (SVM) models. The entropy weight method was employed to determine the weight coefficient for each individual model depending on its predictive performance. Results showed that: (1) ANN exhibits the highest fitting capability among the four orecastingmore » models in the calibration period, whilst its generalization ability becomes weak in the validation period; MLR has a poor performance in both calibration and validation periods; the predicted results of CFM in the calibration period have the highest stability; (2) CFM generally outperforms all individual models in the validation period, and can improve the reliability and stability of predicted results through combining the strengths while reducing the weaknesses of individual models; (3) the performances of all forecasting models are better in dense vegetation areas than in sparse vegetation areas.« less

A study comparison of two system model performance in estimated lifted index over Indonesia.

NASA Astrophysics Data System (ADS)

lestari, Juliana tri; Wandala, Agie

2018-05-01

Lifted index (LI) is one of atmospheric stability indices that used for thunderstorm forecasting. Numerical weather Prediction Models are essential for accurate weather forecast these day. This study has completed the attempt to compare the two NWP models these are Weather Research Forecasting (WRF) model and Global Forecasting System (GFS) model in estimates LI at 20 locations over Indonesia and verified the result with observation. Taylor diagram was used to comparing the models skill with shown the value of standard deviation, coefficient correlation and Root mean square error (RMSE). This study using the dataset on 00.00 UTC and 12.00 UTC during mid-March to Mid-April 2017. From the sample of LI distributions, both models have a tendency to overestimated LI value in almost all region in Indonesia while the WRF models has the better ability to catch the LI pattern distribution with observation than GFS model has. The verification result shows how both WRF and GFS model have such a weak relationship with observation except Eltari meteorologi station that its coefficient correlation reach almost 0.6 with the low RMSE value. Mean while WRF model have a better performance than GFS model. This study suggest that estimated LI of WRF model can provide the good performance for Thunderstorm forecasting over Indonesia in the future. However unsufficient relation between output models and observation in the certain location need a further investigation.

Short-term forecasting of individual household electricity loads with investigating impact of data resolution and forecast horizon

NASA Astrophysics Data System (ADS)

Yildiz, Baran; Bilbao, Jose I.; Dore, Jonathon; Sproul, Alistair B.

2018-05-01

Smart grid components such as smart home and battery energy management systems, high penetration of renewable energy systems, and demand response activities, require accurate electricity demand forecasts for the successful operation of the electricity distribution networks. For example, in order to optimize residential PV generation and electricity consumption and plan battery charge-discharge regimes by scheduling household appliances, forecasts need to target and be tailored to individual household electricity loads. The recent uptake of smart meters allows easier access to electricity readings at very fine resolutions; hence, it is possible to utilize this source of available data to create forecast models. In this paper, models which predominantly use smart meter data alongside with weather variables, or smart meter based models (SMBM), are implemented to forecast individual household loads. Well-known machine learning models such as artificial neural networks (ANN), support vector machines (SVM) and Least-Square SVM are implemented within the SMBM framework and their performance is compared. The analysed household stock consists of 14 households from the state of New South Wales, Australia, with at least a year worth of 5 min. resolution data. In order for the results to be comparable between different households, our study first investigates household load profiles according to their volatility and reveals the relationship between load standard deviation and forecast performance. The analysis extends previous research by evaluating forecasts over four different data resolution; 5, 15, 30 and 60 min, each resolution analysed for four different horizons; 1, 6, 12 and 24 h ahead. Both, data resolution and forecast horizon, proved to have significant impact on the forecast performance and the obtained results provide important insights for the operation of various smart grid applications. Finally, it is shown that the load profile of some households vary significantly across different days; as a result, providing a single model for the entire period may result in limited performance. By the use of a pre-clustering step, similar daily load profiles are grouped together according to their standard deviation, and instead of applying one SMBM for the entire data-set of a particular household, separate SMBMs are applied to each one of the clusters. This preliminary clustering step increases the complexity of the analysis however it results in significant improvements in forecast performance.

Development of visibility forecasting modeling framework for the Lower Fraser Valley of British Columbia using Canada's Regional Air Quality Deterministic Prediction System.

PubMed

So, Rita; Teakles, Andrew; Baik, Jonathan; Vingarzan, Roxanne; Jones, Keith

2018-05-01

Visibility degradation, one of the most noticeable indicators of poor air quality, can occur despite relatively low levels of particulate matter when the risk to human health is low. The availability of timely and reliable visibility forecasts can provide a more comprehensive understanding of the anticipated air quality conditions to better inform local jurisdictions and the public. This paper describes the development of a visibility forecasting modeling framework, which leverages the existing air quality and meteorological forecasts from Canada's operational Regional Air Quality Deterministic Prediction System (RAQDPS) for the Lower Fraser Valley of British Columbia. A baseline model (GM-IMPROVE) was constructed using the revised IMPROVE algorithm based on unprocessed forecasts from the RAQDPS. Three additional prototypes (UMOS-HYB, GM-MLR, GM-RF) were also developed and assessed for forecast performance of up to 48 hr lead time during various air quality and meteorological conditions. Forecast performance was assessed by examining their ability to provide both numerical and categorical forecasts in the form of 1-hr total extinction and Visual Air Quality Ratings (VAQR), respectively. While GM-IMPROVE generally overestimated extinction more than twofold, it had skill in forecasting the relative species contribution to visibility impairment, including ammonium sulfate and ammonium nitrate. Both statistical prototypes, GM-MLR and GM-RF, performed well in forecasting 1-hr extinction during daylight hours, with correlation coefficients (R) ranging from 0.59 to 0.77. UMOS-HYB, a prototype based on postprocessed air quality forecasts without additional statistical modeling, provided reasonable forecasts during most daylight hours. In terms of categorical forecasts, the best prototype was approximately 75 to 87% correct, when forecasting for a condensed three-category VAQR. A case study, focusing on a poor visual air quality yet low Air Quality Health Index episode, illustrated that the statistical prototypes were able to provide timely and skillful visibility forecasts with lead time up to 48 hr. This study describes the development of a visibility forecasting modeling framework, which leverages the existing air quality and meteorological forecasts from Canada's operational Regional Air Quality Deterministic Prediction System. The main applications include tourism and recreation planning, input into air quality management programs, and educational outreach. Visibility forecasts, when supplemented with the existing air quality and health based forecasts, can assist jurisdictions to anticipate the visual air quality impacts as perceived by the public, which can potentially assist in formulating the appropriate air quality bulletins and recommendations.

A Time-Series Water Level Forecasting Model Based on Imputation and Variable Selection Method.

PubMed

Yang, Jun-He; Cheng, Ching-Hsue; Chan, Chia-Pan

2017-01-01

Reservoirs are important for households and impact the national economy. This paper proposed a time-series forecasting model based on estimating a missing value followed by variable selection to forecast the reservoir's water level. This study collected data from the Taiwan Shimen Reservoir as well as daily atmospheric data from 2008 to 2015. The two datasets are concatenated into an integrated dataset based on ordering of the data as a research dataset. The proposed time-series forecasting model summarily has three foci. First, this study uses five imputation methods to directly delete the missing value. Second, we identified the key variable via factor analysis and then deleted the unimportant variables sequentially via the variable selection method. Finally, the proposed model uses a Random Forest to build the forecasting model of the reservoir's water level. This was done to compare with the listing method under the forecasting error. These experimental results indicate that the Random Forest forecasting model when applied to variable selection with full variables has better forecasting performance than the listing model. In addition, this experiment shows that the proposed variable selection can help determine five forecast methods used here to improve the forecasting capability.

A quantitative comparison of precipitation forecasts between the storm-scale numerical weather prediction model and auto-nowcast system in Jiangsu, China

NASA Astrophysics Data System (ADS)

Wang, Gaili; Yang, Ji; Wang, Dan; Liu, Liping

2016-11-01

Extrapolation techniques and storm-scale Numerical Weather Prediction (NWP) models are two primary approaches for short-term precipitation forecasts. The primary objective of this study is to verify precipitation forecasts and compare the performances of two nowcasting schemes: a Beijing Auto-Nowcast system (BJ-ANC) based on extrapolation techniques and a storm-scale NWP model called the Advanced Regional Prediction System (ARPS). The verification and comparison takes into account six heavy precipitation events that occurred in the summer of 2014 and 2015 in Jiangsu, China. The forecast performances of the two schemes were evaluated for the next 6 h at 1-h intervals using gridpoint-based measures of critical success index, bias, index of agreement, root mean square error, and using an object-based verification method called Structure-Amplitude-Location (SAL) score. Regarding gridpoint-based measures, BJ-ANC outperforms ARPS at first, but then the forecast accuracy decreases rapidly with lead time and performs worse than ARPS after 4-5 h of the initial forecast. Regarding the object-based verification method, most forecasts produced by BJ-ANC focus on the center of the diagram at the 1-h lead time and indicate high-quality forecasts. As the lead time increases, BJ-ANC overestimates precipitation amount and produces widespread precipitation, especially at a 6-h lead time. The ARPS model overestimates precipitation at all lead times, particularly at first.

An Objective Verification of the North American Mesoscale Model for Kennedy Space Center and Cape Canaveral Air Force Station

NASA Technical Reports Server (NTRS)

Bauman, William H., III

2010-01-01

The 45th Weather Squadron (45 WS) Launch Weather Officers use the 12-km resolution North American Mesoscale (NAM) model (MesoNAM) text and graphical product forecasts extensively to support launch weather operations. However, the actual performance of the model at Kennedy Space Center (KSC) and Cape Canaveral Air Force Station (CCAFS) has not been measured objectively. In order to have tangible evidence of model performance, the 45 WS tasked the Applied Meteorology Unit to conduct a detailed statistical analysis of model output compared to observed values. The model products are provided to the 45 WS by ACTA, Inc. and include hourly forecasts from 0 to 84 hours based on model initialization times of 00, 06, 12 and 18 UTC. The objective analysis compared the MesoNAM forecast winds, temperature and dew point, as well as the changes in these parameters over time, to the observed values from the sensors in the KSC/CCAFS wind tower network. Objective statistics will give the forecasters knowledge of the model's strength and weaknesses, which will result in improved forecasts for operations.

Monthly monsoon rainfall forecasting using artificial neural networks

NASA Astrophysics Data System (ADS)

Ganti, Ravikumar

2014-10-01

Indian agriculture sector heavily depends on monsoon rainfall for successful harvesting. In the past, prediction of rainfall was mainly performed using regression models, which provide reasonable accuracy in the modelling and forecasting of complex physical systems. Recently, Artificial Neural Networks (ANNs) have been proposed as efficient tools for modelling and forecasting. A feed-forward multi-layer perceptron type of ANN architecture trained using the popular back-propagation algorithm was employed in this study. Other techniques investigated for modeling monthly monsoon rainfall include linear and non-linear regression models for comparison purposes. The data employed in this study include monthly rainfall and monthly average of the daily maximum temperature in the North Central region in India. Specifically, four regression models and two ANN model's were developed. The performance of various models was evaluated using a wide variety of standard statistical parameters and scatter plots. The results obtained in this study for forecasting monsoon rainfalls using ANNs have been encouraging. India's economy and agricultural activities can be effectively managed with the help of the availability of the accurate monsoon rainfall forecasts.

A hybrid group method of data handling with discrete wavelet transform for GDP forecasting

NASA Astrophysics Data System (ADS)

Isa, Nadira Mohamed; Shabri, Ani

2013-09-01

This study is proposed the application of hybridization model using Group Method of Data Handling (GMDH) and Discrete Wavelet Transform (DWT) in time series forecasting. The objective of this paper is to examine the flexibility of the hybridization GMDH in time series forecasting by using Gross Domestic Product (GDP). A time series data set is used in this study to demonstrate the effectiveness of the forecasting model. This data are utilized to forecast through an application aimed to handle real life time series. This experiment compares the performances of a hybrid model and a single model of Wavelet-Linear Regression (WR), Artificial Neural Network (ANN), and conventional GMDH. It is shown that the proposed model can provide a promising alternative technique in GDP forecasting.

Model Forecast Skill and Sensitivity to Initial Conditions in the Seasonal Sea Ice Outlook

NASA Technical Reports Server (NTRS)

Blanchard-Wrigglesworth, E.; Cullather, R. I.; Wang, W.; Zhang, J.; Bitz, C. M.

2015-01-01

We explore the skill of predictions of September Arctic sea ice extent from dynamical models participating in the Sea Ice Outlook (SIO). Forecasts submitted in August, at roughly 2 month lead times, are skillful. However, skill is lower in forecasts submitted to SIO, which began in 2008, than in hindcasts (retrospective forecasts) of the last few decades. The multimodel mean SIO predictions offer slightly higher skill than the single-model SIO predictions, but neither beats a damped persistence forecast at longer than 2 month lead times. The models are largely unsuccessful at predicting each other, indicating a large difference in model physics and/or initial conditions. Motivated by this, we perform an initial condition sensitivity experiment with four SIO models, applying a fixed -1 m perturbation to the initial sea ice thickness. The significant range of the response among the models suggests that different model physics make a significant contribution to forecast uncertainty.

Assessing the value of post-processed state-of-the-art long-term weather forecast ensembles for agricultural water management mediated by farmers' behaviours

NASA Astrophysics Data System (ADS)

Li, Yu; Giuliani, Matteo; Castelletti, Andrea

2016-04-01

Recent advances in modelling of coupled ocean-atmosphere dynamics significantly improved skills of long-term climate forecast from global circulation models (GCMs). These more accurate weather predictions are supposed to be a valuable support to farmers in optimizing farming operations (e.g. crop choice, cropping and watering time) and for more effectively coping with the adverse impacts of climate variability. Yet, assessing how actually valuable this information can be to a farmer is not straightforward and farmers' response must be taken into consideration. Indeed, in the context of agricultural systems potentially useful forecast information should alter stakeholders' expectation, modify their decisions, and ultimately produce an impact on their performance. Nevertheless, long-term forecast are mostly evaluated in terms of accuracy (i.e., forecast quality) by comparing hindcast and observed values and only few studies investigated the operational value of forecast looking at the gain of utility within the decision-making context, e.g. by considering the derivative of forecast information, such as simulated crop yields or simulated soil moisture, which are essential to farmers' decision-making process. In this study, we contribute a step further in the assessment of the operational value of long-term weather forecasts products by embedding these latter into farmers' behavioral models. This allows a more critical assessment of the forecast value mediated by the end-users' perspective, including farmers' risk attitudes and behavioral patterns. Specifically, we evaluate the operational value of thirteen state-of-the-art long-range forecast products against climatology forecast and empirical prediction (i.e. past year climate and historical average) within an integrated agronomic modeling framework embedding an implicit model of the farmers' decision-making process. Raw ensemble datasets are bias-corrected and downscaled using a stochastic weather generator, in order to address the mismatch of the spatio-temporal scale between forecast data from GCMs and our model. For each product, the experiment is composed by two cascade simulations: 1) an ex-ante simulation using forecast data, and 2) an ex-post simulation with observations. Multi-year simulations are performed to account for climate variability, and the operational value of the different forecast products is evaluated against the perfect foresight on the basis of expected crop productivity as well as the final decisions under different decision-making criterions. Our results show that not all products generate beneficial effects to farmers' performance, and the forecast errors might be amplified due to farmers' decision-making process and risk attitudes, yielding little or even worse performance compared with the empirical approaches.

Box Office Forecasting considering Competitive Environment and Word-of-Mouth in Social Networks: A Case Study of Korean Film Market.

PubMed

Kim, Taegu; Hong, Jungsik; Kang, Pilsung

2017-01-01

Accurate box office forecasting models are developed by considering competition and word-of-mouth (WOM) effects in addition to screening-related information. Nationality, genre, ratings, and distributors of motion pictures running concurrently with the target motion picture are used to describe the competition, whereas the numbers of informative, positive, and negative mentions posted on social network services (SNS) are used to gauge the atmosphere spread by WOM. Among these candidate variables, only significant variables are selected by genetic algorithm (GA), based on which machine learning algorithms are trained to build forecasting models. The forecasts are combined to improve forecasting performance. Experimental results on the Korean film market show that the forecasting accuracy in early screening periods can be significantly improved by considering competition. In addition, WOM has a stronger influence on total box office forecasting. Considering both competition and WOM improves forecasting performance to a larger extent than when only one of them is considered.

Box Office Forecasting considering Competitive Environment and Word-of-Mouth in Social Networks: A Case Study of Korean Film Market

PubMed Central

Kim, Taegu; Hong, Jungsik

2017-01-01

Accurate box office forecasting models are developed by considering competition and word-of-mouth (WOM) effects in addition to screening-related information. Nationality, genre, ratings, and distributors of motion pictures running concurrently with the target motion picture are used to describe the competition, whereas the numbers of informative, positive, and negative mentions posted on social network services (SNS) are used to gauge the atmosphere spread by WOM. Among these candidate variables, only significant variables are selected by genetic algorithm (GA), based on which machine learning algorithms are trained to build forecasting models. The forecasts are combined to improve forecasting performance. Experimental results on the Korean film market show that the forecasting accuracy in early screening periods can be significantly improved by considering competition. In addition, WOM has a stronger influence on total box office forecasting. Considering both competition and WOM improves forecasting performance to a larger extent than when only one of them is considered. PMID:28819355

Improved ensemble-mean forecasting of ENSO events by a zero-mean stochastic error model of an intermediate coupled model

NASA Astrophysics Data System (ADS)

Zheng, Fei; Zhu, Jiang

2017-04-01

How to design a reliable ensemble prediction strategy with considering the major uncertainties of a forecasting system is a crucial issue for performing an ensemble forecast. In this study, a new stochastic perturbation technique is developed to improve the prediction skills of El Niño-Southern Oscillation (ENSO) through using an intermediate coupled model. We first estimate and analyze the model uncertainties from the ensemble Kalman filter analysis results through assimilating the observed sea surface temperatures. Then, based on the pre-analyzed properties of model errors, we develop a zero-mean stochastic model-error model to characterize the model uncertainties mainly induced by the missed physical processes of the original model (e.g., stochastic atmospheric forcing, extra-tropical effects, Indian Ocean Dipole). Finally, we perturb each member of an ensemble forecast at each step by the developed stochastic model-error model during the 12-month forecasting process, and add the zero-mean perturbations into the physical fields to mimic the presence of missing processes and high-frequency stochastic noises. The impacts of stochastic model-error perturbations on ENSO deterministic predictions are examined by performing two sets of 21-yr hindcast experiments, which are initialized from the same initial conditions and differentiated by whether they consider the stochastic perturbations. The comparison results show that the stochastic perturbations have a significant effect on improving the ensemble-mean prediction skills during the entire 12-month forecasting process. This improvement occurs mainly because the nonlinear terms in the model can form a positive ensemble-mean from a series of zero-mean perturbations, which reduces the forecasting biases and then corrects the forecast through this nonlinear heating mechanism.

Assessment of an ensemble seasonal streamflow forecasting system for Australia

NASA Astrophysics Data System (ADS)

Bennett, James C.; Wang, Quan J.; Robertson, David E.; Schepen, Andrew; Li, Ming; Michael, Kelvin

2017-11-01

Despite an increasing availability of skilful long-range streamflow forecasts, many water agencies still rely on simple resampled historical inflow sequences (stochastic scenarios) to plan operations over the coming year. We assess a recently developed forecasting system called forecast guided stochastic scenarios (FoGSS) as a skilful alternative to standard stochastic scenarios for the Australian continent. FoGSS uses climate forecasts from a coupled ocean-land-atmosphere prediction system, post-processed with the method of calibration, bridging and merging. Ensemble rainfall forecasts force a monthly rainfall-runoff model, while a staged hydrological error model quantifies and propagates hydrological forecast uncertainty through forecast lead times. FoGSS is able to generate ensemble streamflow forecasts in the form of monthly time series to a 12-month forecast horizon. FoGSS is tested on 63 Australian catchments that cover a wide range of climates, including 21 ephemeral rivers. In all perennial and many ephemeral catchments, FoGSS provides an effective alternative to resampled historical inflow sequences. FoGSS generally produces skilful forecasts at shorter lead times ( < 4 months), and transits to climatology-like forecasts at longer lead times. Forecasts are generally reliable and unbiased. However, FoGSS does not perform well in very dry catchments (catchments that experience zero flows more than half the time in some months), sometimes producing strongly negative forecast skill and poor reliability. We attempt to improve forecasts through the use of (i) ESP rainfall forcings, (ii) different rainfall-runoff models, and (iii) a Bayesian prior to encourage the error model to return climatology forecasts in months when the rainfall-runoff model performs poorly. Of these, the use of the prior offers the clearest benefit in very dry catchments, where it moderates strongly negative forecast skill and reduces bias in some instances. However, the prior does not remedy poor reliability in very dry catchments. Overall, FoGSS is an attractive alternative to historical inflow sequences in all but the driest catchments. We discuss ways in which forecast reliability in very dry catchments could be improved in future work.

Real-time validation of the Dst Predictor model

USGS Publications Warehouse

McCollough, James P.; Young, Shawn L.; Rigler, E. Joshua; Simpson, Hal A.

2015-01-01

The Dst Predictor model, which has been running real-time in the Space Weather Analysis and Forecast System (SWAFS), provides 1-hour and 4-hour forecasts of the Dst index. This is useful for awareness of impending geomagnetic activity, as well as driving other real-time models that use Dst as an input. In this report, we examine the performance of this forecast model in detail. When validating indices it should be noted that performance is only with respect to a reference index as they are derived quantities assumed to reflect a state of the magnetosphere that cannot be directly measured. In this case U.S. Geological Survey (USGS) Definitive Dst is the reference index (Section 3). Whether or not the model better reflects the actual activity level is nearly impossible to discern and is outside the scope of this report. We evaluate the performance of the model by computing continuous predictant skill scores against USGS Definitive Dst values as “observations” (Section 4.2). The two sets of data are not well-correlated for both 1-hour and 4-hour forecasts. The Dst Predictor Prediction Efficiency for both the 1- and 4-hour forecasts suggests poor performance versus the climatological mean. However, the skill score against a nowcast persistence model is positive, suggesting value added by the Dst Predictor model. We further examine statistics for storm times (Section 4.3) with similar results: nowcast persistence performs worse than Dst Predictor.  Dst Predictor is superior to the nowcast persistence model for the metric used in this study. We recommend continued use of the DstPredictor model for 1-and4-hour Dst predictions along with active study of other Dst forecast models that do not rely on nowcast inputs (Section 6). The lack of certified requirements makes further recommendations difficult. A study of how the error in Dst translates to error in models and a better understanding of operational needs for magnetic storm warning are needed to determine such requirements. Nowcast persistence is often hard to beat for short term forecasts and specification and Dst Predictor clearly performs well against that standard (with 1-hour and 4-hour skill-scores of 0.233 and 0.485 respectively), although poor in absolute terms (with1-hourand4-hour prediction efficiencies of-64.6and-43.1, respectively).

The value of model averaging and dynamical climate model predictions for improving statistical seasonal streamflow forecasts over Australia

NASA Astrophysics Data System (ADS)

Pokhrel, Prafulla; Wang, Q. J.; Robertson, David E.

2013-10-01

Seasonal streamflow forecasts are valuable for planning and allocation of water resources. In Australia, the Bureau of Meteorology employs a statistical method to forecast seasonal streamflows. The method uses predictors that are related to catchment wetness at the start of a forecast period and to climate during the forecast period. For the latter, a predictor is selected among a number of lagged climate indices as candidates to give the "best" model in terms of model performance in cross validation. This study investigates two strategies for further improvement in seasonal streamflow forecasts. The first is to combine, through Bayesian model averaging, multiple candidate models with different lagged climate indices as predictors, to take advantage of different predictive strengths of the multiple models. The second strategy is to introduce additional candidate models, using rainfall and sea surface temperature predictions from a global climate model as predictors. This is to take advantage of the direct simulations of various dynamic processes. The results show that combining forecasts from multiple statistical models generally yields more skillful forecasts than using only the best model and appears to moderate the worst forecast errors. The use of rainfall predictions from the dynamical climate model marginally improves the streamflow forecasts when viewed over all the study catchments and seasons, but the use of sea surface temperature predictions provide little additional benefit.

Daily water level forecasting using wavelet decomposition and artificial intelligence techniques

NASA Astrophysics Data System (ADS)

Seo, Youngmin; Kim, Sungwon; Kisi, Ozgur; Singh, Vijay P.

2015-01-01

Reliable water level forecasting for reservoir inflow is essential for reservoir operation. The objective of this paper is to develop and apply two hybrid models for daily water level forecasting and investigate their accuracy. These two hybrid models are wavelet-based artificial neural network (WANN) and wavelet-based adaptive neuro-fuzzy inference system (WANFIS). Wavelet decomposition is employed to decompose an input time series into approximation and detail components. The decomposed time series are used as inputs to artificial neural networks (ANN) and adaptive neuro-fuzzy inference system (ANFIS) for WANN and WANFIS models, respectively. Based on statistical performance indexes, the WANN and WANFIS models are found to produce better efficiency than the ANN and ANFIS models. WANFIS7-sym10 yields the best performance among all other models. It is found that wavelet decomposition improves the accuracy of ANN and ANFIS. This study evaluates the accuracy of the WANN and WANFIS models for different mother wavelets, including Daubechies, Symmlet and Coiflet wavelets. It is found that the model performance is dependent on input sets and mother wavelets, and the wavelet decomposition using mother wavelet, db10, can further improve the efficiency of ANN and ANFIS models. Results obtained from this study indicate that the conjunction of wavelet decomposition and artificial intelligence models can be a useful tool for accurate forecasting daily water level and can yield better efficiency than the conventional forecasting models.

Skill of ENSEMBLES seasonal re-forecasts for malaria prediction in West Africa

NASA Astrophysics Data System (ADS)

Jones, A. E.; Morse, A. P.

2012-12-01

This study examines the performance of malaria-relevant climate variables from the ENSEMBLES seasonal ensemble re-forecasts for sub-Saharan West Africa, using a dynamic malaria model to transform temperature and rainfall forecasts into simulated malaria incidence and verifying these forecasts against simulations obtained by driving the malaria model with General Circulation Model-derived reanalysis. Two subregions of forecast skill are identified: the highlands of Cameroon, where low temperatures limit simulated malaria during the forecast period and interannual variability in simulated malaria is closely linked to variability in temperature, and northern Nigeria/southern Niger, where simulated malaria variability is strongly associated with rainfall variability during the peak rain months.

Verification of space weather forecasts at the UK Met Office

NASA Astrophysics Data System (ADS)

Bingham, S.; Sharpe, M.; Jackson, D.; Murray, S.

2017-12-01

The UK Met Office Space Weather Operations Centre (MOSWOC) has produced space weather guidance twice a day since its official opening in 2014. Guidance includes 4-day probabilistic forecasts of X-ray flares, geomagnetic storms, high-energy electron events and high-energy proton events. Evaluation of such forecasts is important to forecasters, stakeholders, model developers and users to understand the performance of these forecasts and also strengths and weaknesses to enable further development. Met Office terrestrial near real-time verification systems have been adapted to provide verification of X-ray flare and geomagnetic storm forecasts. Verification is updated daily to produce Relative Operating Characteristic (ROC) curves and Reliability diagrams, and rolling Ranked Probability Skill Scores (RPSSs) thus providing understanding of forecast performance and skill. Results suggest that the MOSWOC issued X-ray flare forecasts are usually not statistically significantly better than a benchmark climatological forecast (where the climatology is based on observations from the previous few months). By contrast, the issued geomagnetic storm activity forecast typically performs better against this climatological benchmark.

Comparative study of four time series methods in forecasting typhoid fever incidence in China.

PubMed

Zhang, Xingyu; Liu, Yuanyuan; Yang, Min; Zhang, Tao; Young, Alistair A; Li, Xiaosong

2013-01-01

Accurate incidence forecasting of infectious disease is critical for early prevention and for better government strategic planning. In this paper, we present a comprehensive study of different forecasting methods based on the monthly incidence of typhoid fever. The seasonal autoregressive integrated moving average (SARIMA) model and three different models inspired by neural networks, namely, back propagation neural networks (BPNN), radial basis function neural networks (RBFNN), and Elman recurrent neural networks (ERNN) were compared. The differences as well as the advantages and disadvantages, among the SARIMA model and the neural networks were summarized and discussed. The data obtained for 2005 to 2009 and for 2010 from the Chinese Center for Disease Control and Prevention were used as modeling and forecasting samples, respectively. The performances were evaluated based on three metrics: mean absolute error (MAE), mean absolute percentage error (MAPE), and mean square error (MSE). The results showed that RBFNN obtained the smallest MAE, MAPE and MSE in both the modeling and forecasting processes. The performances of the four models ranked in descending order were: RBFNN, ERNN, BPNN and the SARIMA model.

Comparative Study of Four Time Series Methods in Forecasting Typhoid Fever Incidence in China

PubMed Central

Zhang, Xingyu; Liu, Yuanyuan; Yang, Min; Zhang, Tao; Young, Alistair A.; Li, Xiaosong

2013-01-01

Accurate incidence forecasting of infectious disease is critical for early prevention and for better government strategic planning. In this paper, we present a comprehensive study of different forecasting methods based on the monthly incidence of typhoid fever. The seasonal autoregressive integrated moving average (SARIMA) model and three different models inspired by neural networks, namely, back propagation neural networks (BPNN), radial basis function neural networks (RBFNN), and Elman recurrent neural networks (ERNN) were compared. The differences as well as the advantages and disadvantages, among the SARIMA model and the neural networks were summarized and discussed. The data obtained for 2005 to 2009 and for 2010 from the Chinese Center for Disease Control and Prevention were used as modeling and forecasting samples, respectively. The performances were evaluated based on three metrics: mean absolute error (MAE), mean absolute percentage error (MAPE), and mean square error (MSE). The results showed that RBFNN obtained the smallest MAE, MAPE and MSE in both the modeling and forecasting processes. The performances of the four models ranked in descending order were: RBFNN, ERNN, BPNN and the SARIMA model. PMID:23650546

Benchmark analysis of forecasted seasonal temperature over different climatic areas

NASA Astrophysics Data System (ADS)

Giunta, G.; Salerno, R.; Ceppi, A.; Ercolani, G.; Mancini, M.

2015-12-01

From a long-term perspective, an improvement of seasonal forecasting, which is often exclusively based on climatology, could provide a new capability for the management of energy resources in a time scale of just a few months. This paper regards a benchmark analysis in relation to long-term temperature forecasts over Italy in the year 2010, comparing the eni-kassandra meteo forecast (e-kmf®) model, the Climate Forecast System-National Centers for Environmental Prediction (CFS-NCEP) model, and the climatological reference (based on 25-year data) with observations. Statistical indexes are used to understand the reliability of the prediction of 2-m monthly air temperatures with a perspective of 12 weeks ahead. The results show how the best performance is achieved by the e-kmf® system which improves the reliability for long-term forecasts compared to climatology and the CFS-NCEP model. By using the reliable high-performance forecast system, it is possible to optimize the natural gas portfolio and management operations, thereby obtaining a competitive advantage in the European energy market.

PERFORMANCE AND DIAGNOSTIC EVALUATION OF OZONE PREDICTIONS BY THE ETA-COMMUNITY MULTISCALE AIR QUALITY FORECAST SYSTEM DURING THE 2002 NEW ENGLAND AIR QUALITY STUDY

EPA Science Inventory

A real-time air quality forecasting system (Eta-CMAQ model suite) has been developed by linking the NCEP Eta model to the U.S. EPA CMAQ model. This work presents results from the application of the Eta-CMAQ modeling system for forecasting O₃ over the northeastern U.S d...

Reservoir studies with geostatistics to forecast performance

DOE Office of Scientific and Technical Information (OSTI.GOV)

Tang, R.W.; Behrens, R.A.; Emanuel, A.S.

1991-05-01

In this paper example geostatistics and streamtube applications are presented for waterflood and CO{sub 2} flood in two low-permeability sandstone reservoirs. Thy hybrid approach of combining fine vertical resolution in cross-sectional models with streamtubes resulted in models that showed water channeling and provided realistic performance estimates. Results indicate that the combination of detailed geostatistical cross sections and fine-grid streamtube models offers a systematic approach for realistic performance forecasts.

Time-Series Forecast Modeling on High-Bandwidth Network Measurements

DOE Office of Scientific and Technical Information (OSTI.GOV)

Yoo, Wucherl; Sim, Alex

With the increasing number of geographically distributed scientific collaborations and the growing sizes of scientific data, it has become challenging for users to achieve the best possible network performance on a shared network. In this paper, we have developed a model to forecast expected bandwidth utilization on high-bandwidth wide area networks. The forecast model can improve the efficiency of the resource utilization and scheduling of data movements on high-bandwidth networks to accommodate ever increasing data volume for large-scale scientific data applications. A univariate time-series forecast model is developed with the Seasonal decomposition of Time series by Loess (STL) and themore » AutoRegressive Integrated Moving Average (ARIMA) on Simple Network Management Protocol (SNMP) path utilization measurement data. Compared with the traditional approach such as Box-Jenkins methodology to train the ARIMA model, our forecast model reduces computation time up to 92.6 %. It also shows resilience against abrupt network usage changes. Finally, our forecast model conducts the large number of multi-step forecast, and the forecast errors are within the mean absolute deviation (MAD) of the monitored measurements.« less

Time-Series Forecast Modeling on High-Bandwidth Network Measurements

DOE PAGES

Yoo, Wucherl; Sim, Alex

2016-06-24

With the increasing number of geographically distributed scientific collaborations and the growing sizes of scientific data, it has become challenging for users to achieve the best possible network performance on a shared network. In this paper, we have developed a model to forecast expected bandwidth utilization on high-bandwidth wide area networks. The forecast model can improve the efficiency of the resource utilization and scheduling of data movements on high-bandwidth networks to accommodate ever increasing data volume for large-scale scientific data applications. A univariate time-series forecast model is developed with the Seasonal decomposition of Time series by Loess (STL) and themore » AutoRegressive Integrated Moving Average (ARIMA) on Simple Network Management Protocol (SNMP) path utilization measurement data. Compared with the traditional approach such as Box-Jenkins methodology to train the ARIMA model, our forecast model reduces computation time up to 92.6 %. It also shows resilience against abrupt network usage changes. Finally, our forecast model conducts the large number of multi-step forecast, and the forecast errors are within the mean absolute deviation (MAD) of the monitored measurements.« less

A parimutuel gambling perspective to compare probabilistic seismicity forecasts

NASA Astrophysics Data System (ADS)

Zechar, J. Douglas; Zhuang, Jiancang

2014-10-01

Using analogies to gaming, we consider the problem of comparing multiple probabilistic seismicity forecasts. To measure relative model performance, we suggest a parimutuel gambling perspective which addresses shortcomings of other methods such as likelihood ratio, information gain and Molchan diagrams. We describe two variants of the parimutuel approach for a set of forecasts: head-to-head, in which forecasts are compared in pairs, and round table, in which all forecasts are compared simultaneously. For illustration, we compare the 5-yr forecasts of the Regional Earthquake Likelihood Models experiment for M4.95+ seismicity in California.

National Centers for Environmental Prediction

Science.gov Websites

Modeling Center continuously monitors its NWP model performance against different performance measures, and AIRCFT GFS SSI and forecast fits to RAOBS for last 7 days spatial bias maps for different regions different regions GFS SSI and forecast fits to RAOBS for calendar months (time series, spatial and vertical

A Novel Wind Speed Forecasting Model for Wind Farms of Northwest China

NASA Astrophysics Data System (ADS)

Wang, Jian-Zhou; Wang, Yun

2017-01-01

Wind resources are becoming increasingly significant due to their clean and renewable characteristics, and the integration of wind power into existing electricity systems is imminent. To maintain a stable power supply system that takes into account the stochastic nature of wind speed, accurate wind speed forecasting is pivotal. However, no single model can be applied to all cases. Recent studies show that wind speed forecasting errors are approximately 25% to 40% in Chinese wind farms. Presently, hybrid wind speed forecasting models are widely used and have been verified to perform better than conventional single forecasting models, not only in short-term wind speed forecasting but also in long-term forecasting. In this paper, a hybrid forecasting model is developed, the Similar Coefficient Sum (SCS) and Hermite Interpolation are exploited to process the original wind speed data, and the SVM model whose parameters are tuned by an artificial intelligence model is built to make forecast. The results of case studies show that the MAPE value of the hybrid model varies from 22.96% to 28.87 %, and the MAE value varies from 0.47 m/s to 1.30 m/s. Generally, Sign test, Wilcoxon's Signed-Rank test, and Morgan-Granger-Newbold test tell us that the proposed model is different from the compared models.

Multi-parametric variational data assimilation for hydrological forecasting

NASA Astrophysics Data System (ADS)

Alvarado-Montero, R.; Schwanenberg, D.; Krahe, P.; Helmke, P.; Klein, B.

2017-12-01

Ensemble forecasting is increasingly applied in flow forecasting systems to provide users with a better understanding of forecast uncertainty and consequently to take better-informed decisions. A common practice in probabilistic streamflow forecasting is to force deterministic hydrological model with an ensemble of numerical weather predictions. This approach aims at the representation of meteorological uncertainty but neglects uncertainty of the hydrological model as well as its initial conditions. Complementary approaches use probabilistic data assimilation techniques to receive a variety of initial states or represent model uncertainty by model pools instead of single deterministic models. This paper introduces a novel approach that extends a variational data assimilation based on Moving Horizon Estimation to enable the assimilation of observations into multi-parametric model pools. It results in a probabilistic estimate of initial model states that takes into account the parametric model uncertainty in the data assimilation. The assimilation technique is applied to the uppermost area of River Main in Germany. We use different parametric pools, each of them with five parameter sets, to assimilate streamflow data, as well as remotely sensed data from the H-SAF project. We assess the impact of the assimilation in the lead time performance of perfect forecasts (i.e. observed data as forcing variables) as well as deterministic and probabilistic forecasts from ECMWF. The multi-parametric assimilation shows an improvement of up to 23% for CRPS performance and approximately 20% in Brier Skill Scores with respect to the deterministic approach. It also improves the skill of the forecast in terms of rank histogram and produces a narrower ensemble spread.

Multimodel hydrological ensemble forecasts for the Baskatong catchment in Canada using the TIGGE database.

NASA Astrophysics Data System (ADS)

Tito Arandia Martinez, Fabian

2014-05-01

Adequate uncertainty assessment is an important issue in hydrological modelling. An important issue for hydropower producers is to obtain ensemble forecasts which truly grasp the uncertainty linked to upcoming streamflows. If properly assessed, this uncertainty can lead to optimal reservoir management and energy production (ex. [1]). The meteorological inputs to the hydrological model accounts for an important part of the total uncertainty in streamflow forecasting. Since the creation of the THORPEX initiative and the TIGGE database, access to meteorological ensemble forecasts from nine agencies throughout the world have been made available. This allows for hydrological ensemble forecasts based on multiple meteorological ensemble forecasts. Consequently, both the uncertainty linked to the architecture of the meteorological model and the uncertainty linked to the initial condition of the atmosphere can be accounted for. The main objective of this work is to show that a weighted combination of meteorological ensemble forecasts based on different atmospheric models can lead to improved hydrological ensemble forecasts, for horizons from one to ten days. This experiment is performed for the Baskatong watershed, a head subcatchment of the Gatineau watershed in the province of Quebec, in Canada. Baskatong watershed is of great importance for hydro-power production, as it comprises the main reservoir for the Gatineau watershed, on which there are six hydropower plants managed by Hydro-Québec. Since the 70's, they have been using pseudo ensemble forecast based on deterministic meteorological forecasts to which variability derived from past forecasting errors is added. We use a combination of meteorological ensemble forecasts from different models (precipitation and temperature) as the main inputs for hydrological model HSAMI ([2]). The meteorological ensembles from eight of the nine agencies available through TIGGE are weighted according to their individual performance and combined to form a grand ensemble. Results show that the hydrological forecasts derived from the grand ensemble perform better than the pseudo ensemble forecasts actually used operationally at Hydro-Québec. References: [1] M. Verbunt, A. Walser, J. Gurtz et al., "Probabilistic flood forecasting with a limited-area ensemble prediction system: Selected case studies," Journal of Hydrometeorology, vol. 8, no. 4, pp. 897-909, Aug, 2007. [2] N. Evora, Valorisation des prévisions météorologiques d'ensemble, Institu de recherceh d'Hydro-Québec 2005. [3] V. Fortin, Le modèle météo-apport HSAMI: historique, théorie et application, Institut de recherche d'Hydro-Québec, 2000.

Operational seasonal forecasting of crop performance.

PubMed

Stone, Roger C; Meinke, Holger

2005-11-29

Integrated, interdisciplinary crop performance forecasting systems, linked with appropriate decision and discussion support tools, could substantially improve operational decision making in agricultural management. Recent developments in connecting numerical weather prediction models and general circulation models with quantitative crop growth models offer the potential for development of integrated systems that incorporate components of long-term climate change. However, operational seasonal forecasting systems have little or no value unless they are able to change key management decisions. Changed decision making through incorporation of seasonal forecasting ultimately has to demonstrate improved long-term performance of the cropping enterprise. Simulation analyses conducted on specific production scenarios are especially useful in improving decisions, particularly if this is done in conjunction with development of decision-support systems and associated facilitated discussion groups. Improved management of the overall crop production system requires an interdisciplinary approach, where climate scientists, agricultural scientists and extension specialists are intimately linked with crop production managers in the development of targeted seasonal forecast systems. The same principle applies in developing improved operational management systems for commodity trading organizations, milling companies and agricultural marketing organizations. Application of seasonal forecast systems across the whole value chain in agricultural production offers considerable benefits in improving overall operational management of agricultural production.

Operational seasonal forecasting of crop performance

PubMed Central

Stone, Roger C; Meinke, Holger

2005-01-01

Integrated, interdisciplinary crop performance forecasting systems, linked with appropriate decision and discussion support tools, could substantially improve operational decision making in agricultural management. Recent developments in connecting numerical weather prediction models and general circulation models with quantitative crop growth models offer the potential for development of integrated systems that incorporate components of long-term climate change. However, operational seasonal forecasting systems have little or no value unless they are able to change key management decisions. Changed decision making through incorporation of seasonal forecasting ultimately has to demonstrate improved long-term performance of the cropping enterprise. Simulation analyses conducted on specific production scenarios are especially useful in improving decisions, particularly if this is done in conjunction with development of decision-support systems and associated facilitated discussion groups. Improved management of the overall crop production system requires an interdisciplinary approach, where climate scientists, agricultural scientists and extension specialists are intimately linked with crop production managers in the development of targeted seasonal forecast systems. The same principle applies in developing improved operational management systems for commodity trading organizations, milling companies and agricultural marketing organizations. Application of seasonal forecast systems across the whole value chain in agricultural production offers considerable benefits in improving overall operational management of agricultural production. PMID:16433097

Forecasting of Water Consumptions Expenditure Using Holt-Winter’s and ARIMA

NASA Astrophysics Data System (ADS)

Razali, S. N. A. M.; Rusiman, M. S.; Zawawi, N. I.; Arbin, N.

2018-04-01

This study is carried out to forecast water consumption expenditure of Malaysian university specifically at University Tun Hussein Onn Malaysia (UTHM). The proposed Holt-Winter’s and Auto-Regressive Integrated Moving Average (ARIMA) models were applied to forecast the water consumption expenditure in Ringgit Malaysia from year 2006 until year 2014. The two models were compared and performance measurement of the Mean Absolute Percentage Error (MAPE) and Mean Absolute Deviation (MAD) were used. It is found that ARIMA model showed better results regarding the accuracy of forecast with lower values of MAPE and MAD. Analysis showed that ARIMA (2,1,4) model provided a reasonable forecasting tool for university campus water usage.

Small area population forecasting: some experience with British models.

PubMed

Openshaw, S; Van Der Knaap, G A

1983-01-01

This study is concerned with the evaluation of the various models including time-series forecasts, extrapolation, and projection procedures, that have been developed to prepare population forecasts for planning purposes. These models are evaluated using data for the Netherlands. "As part of a research project at the Erasmus University, space-time population data has been assembled in a geographically consistent way for the period 1950-1979. These population time series are of sufficient length for the first 20 years to be used to build models and then evaluate the performance of the model for the next 10 years. Some 154 different forecasting models for 832 municipalities have been evaluated. It would appear that the best forecasts are likely to be provided by either a Holt-Winters model, or a ratio-correction model, or a low order exponential-smoothing model." excerpt

Value of long-term streamflow forecast to reservoir operations for water supply in snow-dominated catchments

DOE Office of Scientific and Technical Information (OSTI.GOV)

Anghileri, Daniela; Voisin, Nathalie; Castelletti, Andrea F.

In this study, we develop a forecast-based adaptive control framework for Oroville reservoir, California, to assess the value of seasonal and inter-annual forecasts for reservoir operation.We use an Ensemble Streamflow Prediction (ESP) approach to generate retrospective, one-year-long streamflow forecasts based on the Variable Infiltration Capacity hydrology model. The optimal sequence of daily release decisions from the reservoir is then determined by Model Predictive Control, a flexible and adaptive optimization scheme.We assess the forecast value by comparing system performance based on the ESP forecasts with that based on climatology and a perfect forecast. In addition, we evaluate system performance based onmore » a synthetic forecast, which is designed to isolate the contribution of seasonal and inter-annual forecast skill to the overall value of the ESP forecasts.Using the same ESP forecasts, we generalize our results by evaluating forecast value as a function of forecast skill, reservoir features, and demand. Our results show that perfect forecasts are valuable when the water demand is high and the reservoir is sufficiently large to allow for annual carry-over. Conversely, ESP forecast value is highest when the reservoir can shift water on a seasonal basis.On average, for the system evaluated here, the overall ESP value is 35% less than the perfect forecast value. The inter-annual component of the ESP forecast contributes 20-60% of the total forecast value. Improvements in the seasonal component of the ESP forecast would increase the overall ESP forecast value between 15 and 20%.« less

Development of speed models for improving travel forecasting and highway performance evaluation : [technical summary].

DOT National Transportation Integrated Search

2013-12-01

Travel forecasting models predict travel demand based on the present transportation system and its use. Transportation modelers must develop, validate, and calibrate models to ensure that predicted travel demand is as close to reality as possible. Mo...

CEREF: A hybrid data-driven model for forecasting annual streamflow from a socio-hydrological system

NASA Astrophysics Data System (ADS)

Zhang, Hongbo; Singh, Vijay P.; Wang, Bin; Yu, Yinghao

2016-09-01

Hydrological forecasting is complicated by flow regime alterations in a coupled socio-hydrologic system, encountering increasingly non-stationary, nonlinear and irregular changes, which make decision support difficult for future water resources management. Currently, many hybrid data-driven models, based on the decomposition-prediction-reconstruction principle, have been developed to improve the ability to make predictions of annual streamflow. However, there exist many problems that require further investigation, the chief among which is the direction of trend components decomposed from annual streamflow series and is always difficult to ascertain. In this paper, a hybrid data-driven model was proposed to capture this issue, which combined empirical mode decomposition (EMD), radial basis function neural networks (RBFNN), and external forces (EF) variable, also called the CEREF model. The hybrid model employed EMD for decomposition and RBFNN for intrinsic mode function (IMF) forecasting, and determined future trend component directions by regression with EF as basin water demand representing the social component in the socio-hydrologic system. The Wuding River basin was considered for the case study, and two standard statistical measures, root mean squared error (RMSE) and mean absolute error (MAE), were used to evaluate the performance of CEREF model and compare with other models: the autoregressive (AR), RBFNN and EMD-RBFNN. Results indicated that the CEREF model had lower RMSE and MAE statistics, 42.8% and 7.6%, respectively, than did other models, and provided a superior alternative for forecasting annual runoff in the Wuding River basin. Moreover, the CEREF model can enlarge the effective intervals of streamflow forecasting compared to the EMD-RBFNN model by introducing the water demand planned by the government department to improve long-term prediction accuracy. In addition, we considered the high-frequency component, a frequent subject of concern in EMD-based forecasting, and results showed that removing high-frequency component is an effective measure to improve forecasting precision and is suggested for use with the CEREF model for better performance. Finally, the study concluded that the CEREF model can be used to forecast non-stationary annual streamflow change as a co-evolution of hydrologic and social systems with better accuracy. Also, the modification about removing high-frequency can further improve the performance of the CEREF model. It should be noted that the CEREF model is beneficial for data-driven hydrologic forecasting in complex socio-hydrologic systems, and as a simple data-driven socio-hydrologic forecasting model, deserves more attention.

The net benefits of human-ignited wildfire forecasting: the case of Tribal land units in the United States

PubMed Central

Prestemon, Jeffrey P.; Butry, David T.; Thomas, Douglas S.

2017-01-01

Research shows that some categories of human-ignited wildfires might be forecastable, due to their temporal clustering, with the possibility that resources could be pre-deployed to help reduce the incidence of such wildfires. We estimated several kinds of incendiary and other human-ignited wildfire forecast models at the weekly time step for tribal land units in the United States, evaluating their forecast skill out of sample. Analyses show that an Autoregressive Conditional Poisson (ACP) model of both incendiary and non-incendiary human-ignited wildfires is more accurate out of sample compared to alternatives, and the simplest of the ACP models performed the best. Additionally, an ensemble of these and simpler, less analytically intensive approaches performed even better. Wildfire hotspot forecast models using all model types were evaluated in a simulation mode to assess the net benefits of forecasts in the context of law enforcement resource reallocations. Our analyses show that such hotspot tools could yield large positive net benefits for the tribes in terms of suppression expenditures averted for incendiary wildfires but that the hotspot tools were less likely to be beneficial for addressing outbreaks of non-incendiary human-ignited wildfires. PMID:28769549

The net benefits of human-ignited wildfire forecasting: the case of Tribal land units in the United States.

PubMed

Prestemon, Jeffrey P; Butry, David T; Thomas, Douglas S

2016-01-01

Research shows that some categories of human-ignited wildfires might be forecastable, due to their temporal clustering, with the possibility that resources could be pre-deployed to help reduce the incidence of such wildfires. We estimated several kinds of incendiary and other human-ignited wildfire forecast models at the weekly time step for tribal land units in the United States, evaluating their forecast skill out of sample. Analyses show that an Autoregressive Conditional Poisson (ACP) model of both incendiary and non-incendiary human-ignited wildfires is more accurate out of sample compared to alternatives, and the simplest of the ACP models performed the best. Additionally, an ensemble of these and simpler, less analytically intensive approaches performed even better. Wildfire hotspot forecast models using all model types were evaluated in a simulation mode to assess the net benefits of forecasts in the context of law enforcement resource reallocations. Our analyses show that such hotspot tools could yield large positive net benefits for the tribes in terms of suppression expenditures averted for incendiary wildfires but that the hotspot tools were less likely to be beneficial for addressing outbreaks of non-incendiary human-ignited wildfires.

Research and application of a novel hybrid decomposition-ensemble learning paradigm with error correction for daily PM10 forecasting

NASA Astrophysics Data System (ADS)

Luo, Hongyuan; Wang, Deyun; Yue, Chenqiang; Liu, Yanling; Guo, Haixiang

2018-03-01

In this paper, a hybrid decomposition-ensemble learning paradigm combining error correction is proposed for improving the forecast accuracy of daily PM10 concentration. The proposed learning paradigm is consisted of the following two sub-models: (1) PM10 concentration forecasting model; (2) error correction model. In the proposed model, fast ensemble empirical mode decomposition (FEEMD) and variational mode decomposition (VMD) are applied to disassemble original PM10 concentration series and error sequence, respectively. The extreme learning machine (ELM) model optimized by cuckoo search (CS) algorithm is utilized to forecast the components generated by FEEMD and VMD. In order to prove the effectiveness and accuracy of the proposed model, two real-world PM10 concentration series respectively collected from Beijing and Harbin located in China are adopted to conduct the empirical study. The results show that the proposed model performs remarkably better than all other considered models without error correction, which indicates the superior performance of the proposed model.

Improvement of forecast skill for severe weather by merging radar-based extrapolation and storm-scale NWP corrected forecast

NASA Astrophysics Data System (ADS)

Wang, Gaili; Wong, Wai-Kin; Hong, Yang; Liu, Liping; Dong, Jili; Xue, Ming

2015-03-01

The primary objective of this study is to improve the performance of deterministic high resolution rainfall forecasts caused by severe storms by merging an extrapolation radar-based scheme with a storm-scale Numerical Weather Prediction (NWP) model. Effectiveness of Multi-scale Tracking and Forecasting Radar Echoes (MTaRE) model was compared with that of a storm-scale NWP model named Advanced Regional Prediction System (ARPS) for forecasting a violent tornado event that developed over parts of western and much of central Oklahoma on May 24, 2011. Then the bias corrections were performed to improve the forecast accuracy of ARPS forecasts. Finally, the corrected ARPS forecast and radar-based extrapolation were optimally merged by using a hyperbolic tangent weight scheme. The comparison of forecast skill between MTaRE and ARPS in high spatial resolution of 0.01° × 0.01° and high temporal resolution of 5 min showed that MTaRE outperformed ARPS in terms of index of agreement and mean absolute error (MAE). MTaRE had a better Critical Success Index (CSI) for less than 20-min lead times and was comparable to ARPS for 20- to 50-min lead times, while ARPS had a better CSI for more than 50-min lead times. Bias correction significantly improved ARPS forecasts in terms of MAE and index of agreement, although the CSI of corrected ARPS forecasts was similar to that of the uncorrected ARPS forecasts. Moreover, optimally merging results using hyperbolic tangent weight scheme further improved the forecast accuracy and became more stable.

A Comparison of the Performance of Advanced Statistical Techniques for the Refinement of Day-ahead and Longer NWP-based Wind Power Forecasts

NASA Astrophysics Data System (ADS)

Zack, J. W.

2015-12-01

Predictions from Numerical Weather Prediction (NWP) models are the foundation for wind power forecasts for day-ahead and longer forecast horizons. The NWP models directly produce three-dimensional wind forecasts on their respective computational grids. These can be interpolated to the location and time of interest. However, these direct predictions typically contain significant systematic errors ("biases"). This is due to a variety of factors including the limited space-time resolution of the NWP models and shortcomings in the model's representation of physical processes. It has become common practice to attempt to improve the raw NWP forecasts by statistically adjusting them through a procedure that is widely known as Model Output Statistics (MOS). The challenge is to identify complex patterns of systematic errors and then use this knowledge to adjust the NWP predictions. The MOS-based improvements are the basis for much of the value added by commercial wind power forecast providers. There are an enormous number of statistical approaches that can be used to generate the MOS adjustments to the raw NWP forecasts. In order to obtain insight into the potential value of some of the newer and more sophisticated statistical techniques often referred to as "machine learning methods" a MOS-method comparison experiment has been performed for wind power generation facilities in 6 wind resource areas of California. The underlying NWP models that provided the raw forecasts were the two primary operational models of the US National Weather Service: the GFS and NAM models. The focus was on 1- and 2-day ahead forecasts of the hourly wind-based generation. The statistical methods evaluated included: (1) screening multiple linear regression, which served as a baseline method, (2) artificial neural networks, (3) a decision-tree approach called random forests, (4) gradient boosted regression based upon an decision-tree algorithm, (5) support vector regression and (6) analog ensemble, which is a case-matching scheme. The presentation will provide (1) an overview of each method and the experimental design, (2) performance comparisons based on standard metrics such as bias, MAE and RMSE, (3) a summary of the performance characteristics of each approach and (4) a preview of further experiments to be conducted.

Short-Term Forecasting of Loads and Wind Power for Latvian Power System: Accuracy and Capacity of the Developed Tools

NASA Astrophysics Data System (ADS)

Radziukynas, V.; Klementavičius, A.

2016-04-01

The paper analyses the performance results of the recently developed short-term forecasting suit for the Latvian power system. The system load and wind power are forecasted using ANN and ARIMA models, respectively, and the forecasting accuracy is evaluated in terms of errors, mean absolute errors and mean absolute percentage errors. The investigation of influence of additional input variables on load forecasting errors is performed. The interplay of hourly loads and wind power forecasting errors is also evaluated for the Latvian power system with historical loads (the year 2011) and planned wind power capacities (the year 2023).

Robustness of disaggregate oil and gas discovery forecasting models

USGS Publications Warehouse

Attanasi, E.D.; Schuenemeyer, J.H.

1989-01-01

The trend in forecasting oil and gas discoveries has been to develop and use models that allow forecasts of the size distribution of future discoveries. From such forecasts, exploration and development costs can more readily be computed. Two classes of these forecasting models are the Arps-Roberts type models and the 'creaming method' models. This paper examines the robustness of the forecasts made by these models when the historical data on which the models are based have been subject to economic upheavals or when historical discovery data are aggregated from areas having widely differing economic structures. Model performance is examined in the context of forecasting discoveries for offshore Texas State and Federal areas. The analysis shows how the model forecasts are limited by information contained in the historical discovery data. Because the Arps-Roberts type models require more regularity in discovery sequence than the creaming models, prior information had to be introduced into the Arps-Roberts models to accommodate the influence of economic changes. The creaming methods captured the overall decline in discovery size but did not easily allow introduction of exogenous information to compensate for incomplete historical data. Moreover, the predictive log normal distribution associated with the creaming model methods appears to understate the importance of the potential contribution of small fields. ?? 1989.

Automation of energy demand forecasting

NASA Astrophysics Data System (ADS)

Siddique, Sanzad

Automation of energy demand forecasting saves time and effort by searching automatically for an appropriate model in a candidate model space without manual intervention. This thesis introduces a search-based approach that improves the performance of the model searching process for econometrics models. Further improvements in the accuracy of the energy demand forecasting are achieved by integrating nonlinear transformations within the models. This thesis introduces machine learning techniques that are capable of modeling such nonlinearity. Algorithms for learning domain knowledge from time series data using the machine learning methods are also presented. The novel search based approach and the machine learning models are tested with synthetic data as well as with natural gas and electricity demand signals. Experimental results show that the model searching technique is capable of finding an appropriate forecasting model. Further experimental results demonstrate an improved forecasting accuracy achieved by using the novel machine learning techniques introduced in this thesis. This thesis presents an analysis of how the machine learning techniques learn domain knowledge. The learned domain knowledge is used to improve the forecast accuracy.

Stationarity test with a direct test for heteroskedasticity in exchange rate forecasting models

NASA Astrophysics Data System (ADS)

Khin, Aye Aye; Chau, Wong Hong; Seong, Lim Chee; Bin, Raymond Ling Leh; Teng, Kevin Low Lock

2017-05-01

Global economic has been decreasing in the recent years, manifested by the greater exchange rates volatility on international commodity market. This study attempts to analyze some prominent exchange rate forecasting models on Malaysian commodity trading: univariate ARIMA, ARCH and GARCH models in conjunction with stationarity test on residual diagnosis direct testing of heteroskedasticity. All forecasting models utilized the monthly data from 1990 to 2015. Given a total of 312 observations, the data used to forecast both short-term and long-term exchange rate. The forecasting power statistics suggested that the forecasting performance of ARIMA (1, 1, 1) model is more efficient than the ARCH (1) and GARCH (1, 1) models. For ex-post forecast, exchange rate was increased from RM 3.50 per USD in January 2015 to RM 4.47 per USD in December 2015 based on the baseline data. For short-term ex-ante forecast, the analysis results indicate a decrease in exchange rate on 2016 June (RM 4.27 per USD) as compared with 2015 December. A more appropriate forecasting method of exchange rate is vital to aid the decision-making process and planning on the sustainable commodities' production in the world economy.

A framework for improving a seasonal hydrological forecasting system using sensitivity analysis

NASA Astrophysics Data System (ADS)

Arnal, Louise; Pappenberger, Florian; Smith, Paul; Cloke, Hannah

2017-04-01

Seasonal streamflow forecasts are of great value for the socio-economic sector, for applications such as navigation, flood and drought mitigation and reservoir management for hydropower generation and water allocation to agriculture and drinking water. However, as we speak, the performance of dynamical seasonal hydrological forecasting systems (systems based on running seasonal meteorological forecasts through a hydrological model to produce seasonal hydrological forecasts) is still limited in space and time. In this context, the ESP (Ensemble Streamflow Prediction) remains an attractive forecasting method for seasonal streamflow forecasting as it relies on forcing a hydrological model (starting from the latest observed or simulated initial hydrological conditions) with historical meteorological observations. This makes it cheaper to run than a standard dynamical seasonal hydrological forecasting system, for which the seasonal meteorological forecasts will first have to be produced, while still producing skilful forecasts. There is thus the need to focus resources and time towards improvements in dynamical seasonal hydrological forecasting systems which will eventually lead to significant improvements in the skill of the streamflow forecasts generated. Sensitivity analyses are a powerful tool that can be used to disentangle the relative contributions of the two main sources of errors in seasonal streamflow forecasts, namely the initial hydrological conditions (IHC; e.g., soil moisture, snow cover, initial streamflow, among others) and the meteorological forcing (MF; i.e., seasonal meteorological forecasts of precipitation and temperature, input to the hydrological model). Sensitivity analyses are however most useful if they inform and change current operational practices. To this end, we propose a method to improve the design of a seasonal hydrological forecasting system. This method is based on sensitivity analyses, informing the forecasters as to which element of the forecasting chain (i.e., IHC or MF) could potentially lead to the highest increase in seasonal hydrological forecasting performance, after each forecast update.

Extending flood forecasting lead time in large basin by coupling bias-corrected WRF QPF with distributed hydrological model

NASA Astrophysics Data System (ADS)

LI, J.; Chen, Y.; Wang, H. Y.

2016-12-01

In large basin flood forecasting, the forecasting lead time is very important. Advances in numerical weather forecasting in the past decades provides new input to extend flood forecasting lead time in large rivers. Challenges for fulfilling this goal currently is that the uncertainty of QPF with these kinds of NWP models are still high, so controlling the uncertainty of QPF is an emerging technique requirement.The Weather Research and Forecasting (WRF) model is one of these NWPs, and how to control the QPF uncertainty of WRF is the research topic of many researchers among the meteorological community. In this study, the QPF products in the Liujiang river basin, a big river with a drainage area of 56,000 km2, was compared with the ground observation precipitation from a rain gauge networks firstly, and the results show that the uncertainty of the WRF QPF is relatively high. So a post-processed algorithm by correlating the QPF with the observed precipitation is proposed to remove the systematical bias in QPF. With this algorithm, the post-processed WRF QPF is close to the ground observed precipitation in area-averaged precipitation. Then the precipitation is coupled with the Liuxihe model, a physically based distributed hydrological model that is widely used in small watershed flash flood forecasting. The Liuxihe Model has the advantage with gridded precipitation from NWP and could optimize model parameters when there are some observed hydrological data even there is only a few, it also has very high model resolution to improve model performance, and runs on high performance supercomputer with parallel algorithm if executed in large rivers. Two flood events in the Liujiang River were collected, one was used to optimize the model parameters and another is used to validate the model. The results show that the river flow simulation has been improved largely, and could be used for real-time flood forecasting trail in extending flood forecasting leading time.

Forecasting Natural Gas Prices Using Wavelets, Time Series, and Artificial Neural Networks

PubMed Central

2015-01-01

Following the unconventional gas revolution, the forecasting of natural gas prices has become increasingly important because the association of these prices with those of crude oil has weakened. With this as motivation, we propose some modified hybrid models in which various combinations of the wavelet approximation, detail components, autoregressive integrated moving average, generalized autoregressive conditional heteroskedasticity, and artificial neural network models are employed to predict natural gas prices. We also emphasize the boundary problem in wavelet decomposition, and compare results that consider the boundary problem case with those that do not. The empirical results show that our suggested approach can handle the boundary problem, such that it facilitates the extraction of the appropriate forecasting results. The performance of the wavelet-hybrid approach was superior in all cases, whereas the application of detail components in the forecasting was only able to yield a small improvement in forecasting performance. Therefore, forecasting with only an approximation component would be acceptable, in consideration of forecasting efficiency. PMID:26539722

Forecasting Natural Gas Prices Using Wavelets, Time Series, and Artificial Neural Networks.

PubMed

Jin, Junghwan; Kim, Jinsoo

2015-01-01

Following the unconventional gas revolution, the forecasting of natural gas prices has become increasingly important because the association of these prices with those of crude oil has weakened. With this as motivation, we propose some modified hybrid models in which various combinations of the wavelet approximation, detail components, autoregressive integrated moving average, generalized autoregressive conditional heteroskedasticity, and artificial neural network models are employed to predict natural gas prices. We also emphasize the boundary problem in wavelet decomposition, and compare results that consider the boundary problem case with those that do not. The empirical results show that our suggested approach can handle the boundary problem, such that it facilitates the extraction of the appropriate forecasting results. The performance of the wavelet-hybrid approach was superior in all cases, whereas the application of detail components in the forecasting was only able to yield a small improvement in forecasting performance. Therefore, forecasting with only an approximation component would be acceptable, in consideration of forecasting efficiency.

Performance of fuzzy approach in Malaysia short-term electricity load forecasting

NASA Astrophysics Data System (ADS)

Mansor, Rosnalini; Zulkifli, Malina; Yusof, Muhammad Mat; Ismail, Mohd Isfahani; Ismail, Suzilah; Yin, Yip Chee

2014-12-01

Many activities such as economic, education and manafucturing would paralyse with limited supply of electricity but surplus contribute to high operating cost. Therefore electricity load forecasting is important in order to avoid shortage or excess. Previous finding showed festive celebration has effect on short-term electricity load forecasting. Being a multi culture country Malaysia has many major festive celebrations such as Eidul Fitri, Chinese New Year and Deepavali but they are moving holidays due to non-fixed dates on the Gregorian calendar. This study emphasis on the performance of fuzzy approach in forecasting electricity load when considering the presence of moving holidays. Autoregressive Distributed Lag model was estimated using simulated data by including model simplification concept (manual or automatic), day types (weekdays or weekend), public holidays and lags of electricity load. The result indicated that day types, public holidays and several lags of electricity load were significant in the model. Overall, model simplification improves fuzzy performance due to less variables and rules.

A Comparison of the Forecast Skills among Three Numerical Models

NASA Astrophysics Data System (ADS)

Lu, D.; Reddy, S. R.; White, L. J.

2003-12-01

Three numerical weather forecast models, MM5, COAMPS and WRF, operating with a joint effort of NOAA HU-NCAS and Jackson State University (JSU) during summer 2003 have been chosen to study their forecast skills against observations. The models forecast over the same region with the same initialization, boundary condition, forecast length and spatial resolution. AVN global dataset have been ingested as initial conditions. Grib resolution of 27 km is chosen to represent the current mesoscale model. The forecasts with the length of 36h are performed to output the result with 12h interval. The key parameters used to evaluate the forecast skill include 12h accumulated precipitation, sea level pressure, wind, surface temperature and dew point. Precipitation is evaluated statistically using conventional skill scores, Threat Score (TS) and Bias Score (BS), for different threshold values based on 12h rainfall observations whereas other statistical methods such as Mean Error (ME), Mean Absolute Error(MAE) and Root Mean Square Error (RMSE) are applied to other forecast parameters.

An Objective Verification of the North American Mesoscale Model for Kennedy Space Center and Cape Canaveral Air Force Station

NASA Technical Reports Server (NTRS)

Bauman, William H., III

2010-01-01

The 45th Weather Squadron (45 WS) Launch Weather Officers (LWO's) use the 12-km resolution North American Mesoscale (NAM) model (MesoNAM) text and graphical product forecasts extensively to support launch weather operations. However, the actual performance of the model at Kennedy Space Center (KSC) and Cape Canaveral Air Force Station (CCAFS) has not been measured objectively. In order to have tangible evidence of model performance, the 45 WS tasked the Applied Meteorology Unit (AMU; Bauman et ai, 2004) to conduct a detailed statistical analysis of model output compared to observed values. The model products are provided to the 45 WS by ACTA, Inc. and include hourly forecasts from 0 to 84 hours based on model initialization times of 00, 06, 12 and 18 UTC. The objective analysis compared the MesoNAM forecast winds, temperature (T) and dew pOint (T d), as well as the changes in these parameters over time, to the observed values from the sensors in the KSC/CCAFS wind tower network shown in Table 1. These objective statistics give the forecasters knowledge of the model's strengths and weaknesses, which will result in improved forecasts for operations.

Towards smart energy systems: application of kernel machine regression for medium term electricity load forecasting.

PubMed

Alamaniotis, Miltiadis; Bargiotas, Dimitrios; Tsoukalas, Lefteri H

2016-01-01

Integration of energy systems with information technologies has facilitated the realization of smart energy systems that utilize information to optimize system operation. To that end, crucial in optimizing energy system operation is the accurate, ahead-of-time forecasting of load demand. In particular, load forecasting allows planning of system expansion, and decision making for enhancing system safety and reliability. In this paper, the application of two types of kernel machines for medium term load forecasting (MTLF) is presented and their performance is recorded based on a set of historical electricity load demand data. The two kernel machine models and more specifically Gaussian process regression (GPR) and relevance vector regression (RVR) are utilized for making predictions over future load demand. Both models, i.e., GPR and RVR, are equipped with a Gaussian kernel and are tested on daily predictions for a 30-day-ahead horizon taken from the New England Area. Furthermore, their performance is compared to the ARMA(2,2) model with respect to mean average percentage error and squared correlation coefficient. Results demonstrate the superiority of RVR over the other forecasting models in performing MTLF.

Adaptive Blending of Model and Observations for Automated Short-Range Forecasting: Examples from the Vancouver 2010 Olympic and Paralympic Winter Games

NASA Astrophysics Data System (ADS)

Bailey, Monika E.; Isaac, George A.; Gultepe, Ismail; Heckman, Ivan; Reid, Janti

2014-01-01

An automated short-range forecasting system, adaptive blending of observations and model (ABOM), was tested in real time during the 2010 Vancouver Olympic and Paralympic Winter Games in British Columbia. Data at 1-min time resolution were available from a newly established, dense network of surface observation stations. Climatological data were not available at these new stations. This, combined with output from new high-resolution numerical models, provided a unique and exciting setting to test nowcasting systems in mountainous terrain during winter weather conditions. The ABOM method blends extrapolations in time of recent local observations with numerical weather predictions (NWP) model predictions to generate short-range point forecasts of surface variables out to 6 h. The relative weights of the model forecast and the observation extrapolation are based on performance over recent history. The average performance of ABOM nowcasts during February and March 2010 was evaluated using standard scores and thresholds important for Olympic events. Significant improvements over the model forecasts alone were obtained for continuous variables such as temperature, relative humidity and wind speed. The small improvements to forecasts of variables such as visibility and ceiling, subject to discontinuous changes, are attributed to the persistence component of ABOM.

Ensemble Solar Forecasting Statistical Quantification and Sensitivity Analysis: Preprint

DOE Office of Scientific and Technical Information (OSTI.GOV)

Cheung, WanYin; Zhang, Jie; Florita, Anthony

2015-12-08

Uncertainties associated with solar forecasts present challenges to maintain grid reliability, especially at high solar penetrations. This study aims to quantify the errors associated with the day-ahead solar forecast parameters and the theoretical solar power output for a 51-kW solar power plant in a utility area in the state of Vermont, U.S. Forecasts were generated by three numerical weather prediction (NWP) models, including the Rapid Refresh, the High Resolution Rapid Refresh, and the North American Model, and a machine-learning ensemble model. A photovoltaic (PV) performance model was adopted to calculate theoretical solar power generation using the forecast parameters (e.g., irradiance,more » cell temperature, and wind speed). Errors of the power outputs were quantified using statistical moments and a suite of metrics, such as the normalized root mean squared error (NRMSE). In addition, the PV model's sensitivity to different forecast parameters was quantified and analyzed. Results showed that the ensemble model yielded forecasts in all parameters with the smallest NRMSE. The NRMSE of solar irradiance forecasts of the ensemble NWP model was reduced by 28.10% compared to the best of the three NWP models. Further, the sensitivity analysis indicated that the errors of the forecasted cell temperature attributed only approximately 0.12% to the NRMSE of the power output as opposed to 7.44% from the forecasted solar irradiance.« less

Impact of multi-resolution analysis of artificial intelligence models inputs on multi-step ahead river flow forecasting

NASA Astrophysics Data System (ADS)

Badrzadeh, Honey; Sarukkalige, Ranjan; Jayawardena, A. W.

2013-12-01

Discrete wavelet transform was applied to decomposed ANN and ANFIS inputs.Novel approach of WNF with subtractive clustering applied for flow forecasting.Forecasting was performed in 1-5 step ahead, using multi-variate inputs.Forecasting accuracy of peak values and longer lead-time significantly improved.

Performance and robustness of probabilistic river forecasts computed with quantile regression based on multiple independent variables in the North Central USA

NASA Astrophysics Data System (ADS)

Hoss, F.; Fischbeck, P. S.

2014-10-01

This study further develops the method of quantile regression (QR) to predict exceedance probabilities of flood stages by post-processing forecasts. Using data from the 82 river gages, for which the National Weather Service's North Central River Forecast Center issues forecasts daily, this is the first QR application to US American river gages. Archived forecasts for lead times up to six days from 2001-2013 were analyzed. Earlier implementations of QR used the forecast itself as the only independent variable (Weerts et al., 2011; López López et al., 2014). This study adds the rise rate of the river stage in the last 24 and 48 h and the forecast error 24 and 48 h ago to the QR model. Including those four variables significantly improved the forecasts, as measured by the Brier Skill Score (BSS). Mainly, the resolution increases, as the original QR implementation already delivered high reliability. Combining the forecast with the other four variables results in much less favorable BSSs. Lastly, the forecast performance does not depend on the size of the training dataset, but on the year, the river gage, lead time and event threshold that are being forecast. We find that each event threshold requires a separate model configuration or at least calibration.

Projecting technology change to improve space technology planning and systems management

NASA Astrophysics Data System (ADS)

Walk, Steven Robert

2011-04-01

Projecting technology performance evolution has been improving over the years. Reliable quantitative forecasting methods have been developed that project the growth, diffusion, and performance of technology in time, including projecting technology substitutions, saturation levels, and performance improvements. These forecasts can be applied at the early stages of space technology planning to better predict available future technology performance, assure the successful selection of technology, and improve technology systems management strategy. Often what is published as a technology forecast is simply scenario planning, usually made by extrapolating current trends into the future, with perhaps some subjective insight added. Typically, the accuracy of such predictions falls rapidly with distance in time. Quantitative technology forecasting (QTF), on the other hand, includes the study of historic data to identify one of or a combination of several recognized universal technology diffusion or substitution patterns. In the same manner that quantitative models of physical phenomena provide excellent predictions of system behavior, so do QTF models provide reliable technological performance trajectories. In practice, a quantitative technology forecast is completed to ascertain with confidence when the projected performance of a technology or system of technologies will occur. Such projections provide reliable time-referenced information when considering cost and performance trade-offs in maintaining, replacing, or migrating a technology, component, or system. This paper introduces various quantitative technology forecasting techniques and illustrates their practical application in space technology and technology systems management.

A novel hybrid ensemble learning paradigm for tourism forecasting

NASA Astrophysics Data System (ADS)

Shabri, Ani

2015-02-01

In this paper, a hybrid forecasting model based on Empirical Mode Decomposition (EMD) and Group Method of Data Handling (GMDH) is proposed to forecast tourism demand. This methodology first decomposes the original visitor arrival series into several Intrinsic Model Function (IMFs) components and one residual component by EMD technique. Then, IMFs components and the residual components is forecasted respectively using GMDH model whose input variables are selected by using Partial Autocorrelation Function (PACF). The final forecasted result for tourism series is produced by aggregating all the forecasted results. For evaluating the performance of the proposed EMD-GMDH methodologies, the monthly data of tourist arrivals from Singapore to Malaysia are used as an illustrative example. Empirical results show that the proposed EMD-GMDH model outperforms the EMD-ARIMA as well as the GMDH and ARIMA (Autoregressive Integrated Moving Average) models without time series decomposition.

Electricity Load Forecasting Using Support Vector Regression with Memetic Algorithms

PubMed Central

Hu, Zhongyi; Xiong, Tao

2013-01-01

Electricity load forecasting is an important issue that is widely explored and examined in power systems operation literature and commercial transactions in electricity markets literature as well. Among the existing forecasting models, support vector regression (SVR) has gained much attention. Considering the performance of SVR highly depends on its parameters; this study proposed a firefly algorithm (FA) based memetic algorithm (FA-MA) to appropriately determine the parameters of SVR forecasting model. In the proposed FA-MA algorithm, the FA algorithm is applied to explore the solution space, and the pattern search is used to conduct individual learning and thus enhance the exploitation of FA. Experimental results confirm that the proposed FA-MA based SVR model can not only yield more accurate forecasting results than the other four evolutionary algorithms based SVR models and three well-known forecasting models but also outperform the hybrid algorithms in the related existing literature. PMID:24459425

Electricity load forecasting using support vector regression with memetic algorithms.

PubMed

Hu, Zhongyi; Bao, Yukun; Xiong, Tao

2013-01-01

Electricity load forecasting is an important issue that is widely explored and examined in power systems operation literature and commercial transactions in electricity markets literature as well. Among the existing forecasting models, support vector regression (SVR) has gained much attention. Considering the performance of SVR highly depends on its parameters; this study proposed a firefly algorithm (FA) based memetic algorithm (FA-MA) to appropriately determine the parameters of SVR forecasting model. In the proposed FA-MA algorithm, the FA algorithm is applied to explore the solution space, and the pattern search is used to conduct individual learning and thus enhance the exploitation of FA. Experimental results confirm that the proposed FA-MA based SVR model can not only yield more accurate forecasting results than the other four evolutionary algorithms based SVR models and three well-known forecasting models but also outperform the hybrid algorithms in the related existing literature.

SST-Forced Seasonal Simulation and Prediction Skill for Versions of the NCEP/MRF Model.

NASA Astrophysics Data System (ADS)

Livezey, Robert E.; Masutani, Michiko; Jil, Ming

1996-03-01

The feasibility of using a two-tier approach to provide guidance to operational long-lead seasonal prediction is explored. The approach includes first a forecast of global sea surface temperatures (SSTs) using a coupled general circulation model, followed by an atmospheric forecast using an atmospheric general circulation model (AGCM). For this exploration, ensembles of decade-long integrations of the AGCM driven by observed SSTs and ensembles of integrations of select cases driven by forecast SSTs have been conducted. The ability of the model in these sets of runs to reproduce observed atmospheric conditions has been evaluated with a multiparameter performance analysis.Results have identified performance and skill levels in the specified SST runs, for winters and springs over the Pacific/North America region, that are sufficient to impact operational seasonal predictions in years with major El Niño-Southern Oscillation (ENSO) episodes. Further, these levels were substantially reproduced in the forecast SST runs for 1-month leads and in many instances for up to one-season leads. In fact, overall the 0- and 1-month-lead forecasts of seasonal temperature over the United States for three falls and winters with major ENSO episodes were substantially better than corresponding official forecasts. Thus, there is considerable reason to develop a dynamical component for the official seasonal forecast process.

Verifying Operational and Developmental Air Force Weather Cloud Analysis and Forecast Products Using Lidar Data from Department of Energy Atmospheric Radiation Measurement (ARM) Sites

NASA Astrophysics Data System (ADS)

Hildebrand, E. P.

2017-12-01

Air Force Weather has developed various cloud analysis and forecast products designed to support global Department of Defense (DoD) missions. A World-Wide Merged Cloud Analysis (WWMCA) and short term Advected Cloud (ADVCLD) forecast is generated hourly using data from 16 geostationary and polar-orbiting satellites. Additionally, WWMCA and Numerical Weather Prediction (NWP) data are used in a statistical long-term (out to five days) cloud forecast model known as the Diagnostic Cloud Forecast (DCF). The WWMCA and ADVCLD are generated on the same polar stereographic 24 km grid for each hemisphere, whereas the DCF is generated on the same grid as its parent NWP model. When verifying the cloud forecast models, the goal is to understand not only the ability to detect cloud, but also the ability to assign it to the correct vertical layer. ADVCLD and DCF forecasts traditionally have been verified using WWMCA data as truth, but this might over-inflate the performance of those models because WWMCA also is a primary input dataset for those models. Because of this, in recent years, a WWMCA Reanalysis product has been developed, but this too is not a fully independent dataset. This year, work has been done to incorporate data from external, independent sources to verify not only the cloud forecast products, but the WWMCA data itself. One such dataset that has been useful for examining the 3-D performance of the cloud analysis and forecast models is Atmospheric Radiation Measurement (ARM) data from various sites around the globe. This presentation will focus on the use of the Department of Energy (DoE) ARM data to verify Air Force Weather cloud analysis and forecast products. Results will be presented to show relative strengths and weaknesses of the analyses and forecasts.

Assessing the skill of seasonal precipitation and streamflow forecasts in sixteen French catchments

NASA Astrophysics Data System (ADS)

Crochemore, Louise; Ramos, Maria-Helena; Pappenberger, Florian

2015-04-01

Meteorological centres make sustained efforts to provide seasonal forecasts that are increasingly skilful. Streamflow forecasting is one of the many applications than can benefit from these efforts. Seasonal flow forecasts generated using seasonal ensemble precipitation forecasts as input to a hydrological model can help to take anticipatory measures for water supply reservoir operation or drought risk management. The objective of the study is to assess the skill of seasonal precipitation and streamflow forecasts in France. First, we evaluated the skill of ECMWF SYS4 seasonal precipitation forecasts for streamflow forecasting in sixteen French catchments. Daily flow forecasts were produced using raw seasonal precipitation forecasts as input to the GR6J hydrological model. Ensemble forecasts are issued every month with 15 or 51 members according to the month of the year and evaluated for up to 90 days ahead. In a second step, we applied eight variants of bias correction approaches to the precipitation forecasts prior to generating the flow forecasts. The approaches were based on the linear scaling and the distribution mapping methods. The skill of the ensemble forecasts was assessed in accuracy (MAE), reliability (PIT Diagram) and overall performance (CRPS). The results show that, in most catchments, raw seasonal precipitation and streamflow forecasts are more skilful in terms of accuracy and overall performance than a reference prediction based on historic observed precipitation and watershed initial conditions at the time of forecast. Reliability is the only attribute that is not significantly improved. The skill of the forecasts is, in general, improved when applying bias correction. Two bias correction methods showed the best performance for the studied catchments: the simple linear scaling of monthly values and the empirical distribution mapping of daily values. L. Crochemore is funded by the Interreg IVB DROP Project (Benefit of governance in DROught adaPtation).

Using Bayesian Model Averaging (BMA) to calibrate probabilistic surface temperature forecasts over Iran

NASA Astrophysics Data System (ADS)

Soltanzadeh, I.; Azadi, M.; Vakili, G. A.

2011-07-01

Using Bayesian Model Averaging (BMA), an attempt was made to obtain calibrated probabilistic numerical forecasts of 2-m temperature over Iran. The ensemble employs three limited area models (WRF, MM5 and HRM), with WRF used with five different configurations. Initial and boundary conditions for MM5 and WRF are obtained from the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) and for HRM the initial and boundary conditions come from analysis of Global Model Europe (GME) of the German Weather Service. The resulting ensemble of seven members was run for a period of 6 months (from December 2008 to May 2009) over Iran. The 48-h raw ensemble outputs were calibrated using BMA technique for 120 days using a 40 days training sample of forecasts and relative verification data. The calibrated probabilistic forecasts were assessed using rank histogram and attribute diagrams. Results showed that application of BMA improved the reliability of the raw ensemble. Using the weighted ensemble mean forecast as a deterministic forecast it was found that the deterministic-style BMA forecasts performed usually better than the best member's deterministic forecast.

Forecasting Daily Volume and Acuity of Patients in the Emergency Department.

PubMed

Calegari, Rafael; Fogliatto, Flavio S; Lucini, Filipe R; Neyeloff, Jeruza; Kuchenbecker, Ricardo S; Schaan, Beatriz D

2016-01-01

This study aimed at analyzing the performance of four forecasting models in predicting the demand for medical care in terms of daily visits in an emergency department (ED) that handles high complexity cases, testing the influence of climatic and calendrical factors on demand behavior. We tested different mathematical models to forecast ED daily visits at Hospital de Clínicas de Porto Alegre (HCPA), which is a tertiary care teaching hospital located in Southern Brazil. Model accuracy was evaluated using mean absolute percentage error (MAPE), considering forecasting horizons of 1, 7, 14, 21, and 30 days. The demand time series was stratified according to patient classification using the Manchester Triage System's (MTS) criteria. Models tested were the simple seasonal exponential smoothing (SS), seasonal multiplicative Holt-Winters (SMHW), seasonal autoregressive integrated moving average (SARIMA), and multivariate autoregressive integrated moving average (MSARIMA). Performance of models varied according to patient classification, such that SS was the best choice when all types of patients were jointly considered, and SARIMA was the most accurate for modeling demands of very urgent (VU) and urgent (U) patients. The MSARIMA models taking into account climatic factors did not improve the performance of the SARIMA models, independent of patient classification.

Forecasting Daily Volume and Acuity of Patients in the Emergency Department

PubMed Central

Fogliatto, Flavio S.; Neyeloff, Jeruza; Kuchenbecker, Ricardo S.; Schaan, Beatriz D.

2016-01-01

This study aimed at analyzing the performance of four forecasting models in predicting the demand for medical care in terms of daily visits in an emergency department (ED) that handles high complexity cases, testing the influence of climatic and calendrical factors on demand behavior. We tested different mathematical models to forecast ED daily visits at Hospital de Clínicas de Porto Alegre (HCPA), which is a tertiary care teaching hospital located in Southern Brazil. Model accuracy was evaluated using mean absolute percentage error (MAPE), considering forecasting horizons of 1, 7, 14, 21, and 30 days. The demand time series was stratified according to patient classification using the Manchester Triage System's (MTS) criteria. Models tested were the simple seasonal exponential smoothing (SS), seasonal multiplicative Holt-Winters (SMHW), seasonal autoregressive integrated moving average (SARIMA), and multivariate autoregressive integrated moving average (MSARIMA). Performance of models varied according to patient classification, such that SS was the best choice when all types of patients were jointly considered, and SARIMA was the most accurate for modeling demands of very urgent (VU) and urgent (U) patients. The MSARIMA models taking into account climatic factors did not improve the performance of the SARIMA models, independent of patient classification. PMID:27725842

Improving wave forecasting by integrating ensemble modelling and machine learning

NASA Astrophysics Data System (ADS)

O'Donncha, F.; Zhang, Y.; James, S. C.

2017-12-01

Modern smart-grid networks use technologies to instantly relay information on supply and demand to support effective decision making. Integration of renewable-energy resources with these systems demands accurate forecasting of energy production (and demand) capacities. For wave-energy converters, this requires wave-condition forecasting to enable estimates of energy production. Current operational wave forecasting systems exhibit substantial errors with wave-height RMSEs of 40 to 60 cm being typical, which limits the reliability of energy-generation predictions thereby impeding integration with the distribution grid. In this study, we integrate physics-based models with statistical learning aggregation techniques that combine forecasts from multiple, independent models into a single "best-estimate" prediction of the true state. The Simulating Waves Nearshore physics-based model is used to compute wind- and currents-augmented waves in the Monterey Bay area. Ensembles are developed based on multiple simulations perturbing input data (wave characteristics supplied at the model boundaries and winds) to the model. A learning-aggregation technique uses past observations and past model forecasts to calculate a weight for each model. The aggregated forecasts are compared to observation data to quantify the performance of the model ensemble and aggregation techniques. The appropriately weighted ensemble model outperforms an individual ensemble member with regard to forecasting wave conditions.

Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS)

NASA Astrophysics Data System (ADS)

OConnor, A.; Kirtman, B. P.; Harrison, S.; Gorman, J.

2016-02-01

Current US Navy forecasting systems cannot easily incorporate extended-range forecasts that can improve mission readiness and effectiveness; ensure safety; and reduce cost, labor, and resource requirements. If Navy operational planners had systems that incorporated these forecasts, they could plan missions using more reliable and longer-term weather and climate predictions. Further, using multi-model forecast ensembles instead of single forecasts would produce higher predictive performance. Extended-range multi-model forecast ensembles, such as those available in the North American Multi-Model Ensemble (NMME), are ideal for system integration because of their high skill predictions; however, even higher skill predictions can be produced if forecast model ensembles are combined correctly. While many methods for weighting models exist, the best method in a given environment requires expert knowledge of the models and combination methods.We present an innovative approach that uses machine learning to combine extended-range predictions from multi-model forecast ensembles and generate a probabilistic forecast for any region of the globe up to 12 months in advance. Our machine-learning approach uses 30 years of hindcast predictions to learn patterns of forecast model successes and failures. Each model is assigned a weight for each environmental condition, 100 km2 region, and day given any expected environmental information. These weights are then applied to the respective predictions for the region and time of interest to effectively stitch together a single, coherent probabilistic forecast. Our experimental results demonstrate the benefits of our approach to produce extended-range probabilistic forecasts for regions and time periods of interest that are superior, in terms of skill, to individual NMME forecast models and commonly weighted models. The probabilistic forecast leverages the strengths of three NMME forecast models to predict environmental conditions for an area spanning from San Diego, CA to Honolulu, HI, seven months in-advance. Key findings include: weighted combinations of models are strictly better than individual models; machine-learned combinations are especially better; and forecasts produced using our approach have the highest rank probability skill score most often.

Evaluating the performance of the Lee-Carter method and its variants in modelling and forecasting Malaysian mortality

NASA Astrophysics Data System (ADS)

Zakiyatussariroh, W. H. Wan; Said, Z. Mohammad; Norazan, M. R.

2014-12-01

This study investigated the performance of the Lee-Carter (LC) method and it variants in modeling and forecasting Malaysia mortality. These include the original LC, the Lee-Miller (LM) variant and the Booth-Maindonald-Smith (BMS) variant. These methods were evaluated using Malaysia's mortality data which was measured based on age specific death rates (ASDR) for 1971 to 2009 for overall population while those for 1980-2009 were used in separate models for male and female population. The performance of the variants has been examined in term of the goodness of fit of the models and forecasting accuracy. Comparison was made based on several criteria namely, mean square error (MSE), root mean square error (RMSE), mean absolute deviation (MAD) and mean absolute percentage error (MAPE). The results indicate that BMS method was outperformed in in-sample fitting for overall population and when the models were fitted separately for male and female population. However, in the case of out-sample forecast accuracy, BMS method only best when the data were fitted to overall population. When the data were fitted separately for male and female, LCnone performed better for male population and LM method is good for female population.

Empirical forecast of quiet time ionospheric Total Electron Content maps over Europe

NASA Astrophysics Data System (ADS)

Badeke, Ronny; Borries, Claudia; Hoque, Mainul M.; Minkwitz, David

2018-06-01

An accurate forecast of the atmospheric Total Electron Content (TEC) is helpful to investigate space weather influences on the ionosphere and technical applications like satellite-receiver radio links. The purpose of this work is to compare four empirical methods for a 24-h forecast of vertical TEC maps over Europe under geomagnetically quiet conditions. TEC map data are obtained from the Space Weather Application Center Ionosphere (SWACI) and the Universitat Politècnica de Catalunya (UPC). The time-series methods Standard Persistence Model (SPM), a 27 day median model (MediMod) and a Fourier Series Expansion are compared to maps for the entire year of 2015. As a representative of the climatological coefficient models the forecast performance of the Global Neustrelitz TEC model (NTCM-GL) is also investigated. Time periods of magnetic storms, which are identified with the Dst index, are excluded from the validation. By calculating the TEC values with the most recent maps, the time-series methods perform slightly better than the coefficient model NTCM-GL. The benefit of NTCM-GL is its independence on observational TEC data. Amongst the time-series methods mentioned, MediMod delivers the best overall performance regarding accuracy and data gap handling. Quiet-time SWACI maps can be forecasted accurately and in real-time by the MediMod time-series approach.

A forecast experiment of earthquake activity in Japan under Collaboratory for the Study of Earthquake Predictability (CSEP)

NASA Astrophysics Data System (ADS)

Hirata, N.; Yokoi, S.; Nanjo, K. Z.; Tsuruoka, H.

2012-04-01

One major focus of the current Japanese earthquake prediction research program (2009-2013), which is now integrated with the research program for prediction of volcanic eruptions, is to move toward creating testable earthquake forecast models. For this purpose we started an experiment of forecasting earthquake activity in Japan under the framework of the Collaboratory for the Study of Earthquake Predictability (CSEP) through an international collaboration. We established the CSEP Testing Centre, an infrastructure to encourage researchers to develop testable models for Japan, and to conduct verifiable prospective tests of their model performance. We started the 1st earthquake forecast testing experiment in Japan within the CSEP framework. We use the earthquake catalogue maintained and provided by the Japan Meteorological Agency (JMA). The experiment consists of 12 categories, with 4 testing classes with different time spans (1 day, 3 months, 1 year, and 3 years) and 3 testing regions called "All Japan," "Mainland," and "Kanto." A total of 105 models were submitted, and are currently under the CSEP official suite of tests for evaluating the performance of forecasts. The experiments were completed for 92 rounds for 1-day, 6 rounds for 3-month, and 3 rounds for 1-year classes. For 1-day testing class all models passed all the CSEP's evaluation tests at more than 90% rounds. The results of the 3-month testing class also gave us new knowledge concerning statistical forecasting models. All models showed a good performance for magnitude forecasting. On the other hand, observation is hardly consistent in space distribution with most models when many earthquakes occurred at a spot. Now we prepare the 3-D forecasting experiment with a depth range of 0 to 100 km in Kanto region. The testing center is improving an evaluation system for 1-day class experiment to finish forecasting and testing results within one day. The special issue of 1st part titled Earthquake Forecast Testing Experiment in Japan was published on the Earth, Planets and Space Vol. 63, No.3, 2011 on March, 2011. The 2nd part of this issue, which is now on line, will be published soon. An outline of the experiment and activities of the Japanese Testing Center are published on our WEB site; http://wwweic.eri.u-tokyo.ac.jp/ZISINyosoku/wiki.en/wiki.cgi

An enhanced PM 2.5 air quality forecast model based on nonlinear regression and back-trajectory concentrations

NASA Astrophysics Data System (ADS)

Cobourn, W. Geoffrey

2010-08-01

An enhanced PM 2.5 air quality forecast model based on nonlinear regression (NLR) and back-trajectory concentrations has been developed for use in the Louisville, Kentucky metropolitan area. The PM 2.5 air quality forecast model is designed for use in the warm season, from May through September, when PM 2.5 air quality is more likely to be critical for human health. The enhanced PM 2.5 model consists of a basic NLR model, developed for use with an automated air quality forecast system, and an additional parameter based on upwind PM 2.5 concentration, called PM24. The PM24 parameter is designed to be determined manually, by synthesizing backward air trajectory and regional air quality information to compute 24-h back-trajectory concentrations. The PM24 parameter may be used by air quality forecasters to adjust the forecast provided by the automated forecast system. In this study of the 2007 and 2008 forecast seasons, the enhanced model performed well using forecasted meteorological data and PM24 as input. The enhanced PM 2.5 model was compared with three alternative models, including the basic NLR model, the basic NLR model with a persistence parameter added, and the NLR model with persistence and PM24. The two models that included PM24 were of comparable accuracy. The two models incorporating back-trajectory concentrations had lower mean absolute errors and higher rates of detecting unhealthy PM2.5 concentrations compared to the other models.

Applying different independent component analysis algorithms and support vector regression for IT chain store sales forecasting.

PubMed

Dai, Wensheng; Wu, Jui-Yu; Lu, Chi-Jie

2014-01-01

Sales forecasting is one of the most important issues in managing information technology (IT) chain store sales since an IT chain store has many branches. Integrating feature extraction method and prediction tool, such as support vector regression (SVR), is a useful method for constructing an effective sales forecasting scheme. Independent component analysis (ICA) is a novel feature extraction technique and has been widely applied to deal with various forecasting problems. But, up to now, only the basic ICA method (i.e., temporal ICA model) was applied to sale forecasting problem. In this paper, we utilize three different ICA methods including spatial ICA (sICA), temporal ICA (tICA), and spatiotemporal ICA (stICA) to extract features from the sales data and compare their performance in sales forecasting of IT chain store. Experimental results from a real sales data show that the sales forecasting scheme by integrating stICA and SVR outperforms the comparison models in terms of forecasting error. The stICA is a promising tool for extracting effective features from branch sales data and the extracted features can improve the prediction performance of SVR for sales forecasting.

Applying Different Independent Component Analysis Algorithms and Support Vector Regression for IT Chain Store Sales Forecasting

PubMed Central

Dai, Wensheng

2014-01-01

Sales forecasting is one of the most important issues in managing information technology (IT) chain store sales since an IT chain store has many branches. Integrating feature extraction method and prediction tool, such as support vector regression (SVR), is a useful method for constructing an effective sales forecasting scheme. Independent component analysis (ICA) is a novel feature extraction technique and has been widely applied to deal with various forecasting problems. But, up to now, only the basic ICA method (i.e., temporal ICA model) was applied to sale forecasting problem. In this paper, we utilize three different ICA methods including spatial ICA (sICA), temporal ICA (tICA), and spatiotemporal ICA (stICA) to extract features from the sales data and compare their performance in sales forecasting of IT chain store. Experimental results from a real sales data show that the sales forecasting scheme by integrating stICA and SVR outperforms the comparison models in terms of forecasting error. The stICA is a promising tool for extracting effective features from branch sales data and the extracted features can improve the prediction performance of SVR for sales forecasting. PMID:25165740

Palm oil price forecasting model: An autoregressive distributed lag (ARDL) approach

NASA Astrophysics Data System (ADS)

Hamid, Mohd Fahmi Abdul; Shabri, Ani

2017-05-01

Palm oil price fluctuated without any clear trend or cyclical pattern in the last few decades. The instability of food commodities price causes it to change rapidly over time. This paper attempts to develop Autoregressive Distributed Lag (ARDL) model in modeling and forecasting the price of palm oil. In order to use ARDL as a forecasting model, this paper modifies the data structure where we only consider lagged explanatory variables to explain the variation in palm oil price. We then compare the performance of this ARDL model with a benchmark model namely ARIMA in term of their comparative forecasting accuracy. This paper also utilize ARDL bound testing approach to co-integration in examining the short run and long run relationship between palm oil price and its determinant; production, stock, and price of soybean as the substitute of palm oil and price of crude oil. The comparative forecasting accuracy suggests that ARDL model has a better forecasting accuracy compared to ARIMA.

Verification of ECMWF System 4 for seasonal hydrological forecasting in a northern climate

NASA Astrophysics Data System (ADS)

Bazile, Rachel; Boucher, Marie-Amélie; Perreault, Luc; Leconte, Robert

2017-11-01

Hydropower production requires optimal dam and reservoir management to prevent flooding damage and avoid operation losses. In a northern climate, where spring freshet constitutes the main inflow volume, seasonal forecasts can help to establish a yearly strategy. Long-term hydrological forecasts often rely on past observations of streamflow or meteorological data. Another alternative is to use ensemble meteorological forecasts produced by climate models. In this paper, those produced by the ECMWF (European Centre for Medium-Range Forecast) System 4 are examined and bias is characterized. Bias correction, through the linear scaling method, improves the performance of the raw ensemble meteorological forecasts in terms of continuous ranked probability score (CRPS). Then, three seasonal ensemble hydrological forecasting systems are compared: (1) the climatology of simulated streamflow, (2) the ensemble hydrological forecasts based on climatology (ESP) and (3) the hydrological forecasts based on bias-corrected ensemble meteorological forecasts from System 4 (corr-DSP). Simulated streamflow computed using observed meteorological data is used as benchmark. Accounting for initial conditions is valuable even for long-term forecasts. ESP and corr-DSP both outperform the climatology of simulated streamflow for lead times from 1 to 5 months depending on the season and watershed. Integrating information about future meteorological conditions also improves monthly volume forecasts. For the 1-month lead time, a gain exists for almost all watersheds during winter, summer and fall. However, volume forecasts performance for spring varies from one watershed to another. For most of them, the performance is close to the performance of ESP. For longer lead times, the CRPS skill score is mostly in favour of ESP, even if for many watersheds, ESP and corr-DSP have comparable skill. Corr-DSP appears quite reliable but, in some cases, under-dispersion or bias is observed. A more complex bias-correction method should be further investigated to remedy this weakness and take more advantage of the ensemble forecasts produced by the climate model. Overall, in this study, bias-corrected ensemble meteorological forecasts appear to be an interesting source of information for hydrological forecasting for lead times up to 1 month. They could also complement ESP for longer lead times.

The Use of Ambient Humidity Conditions to Improve Influenza Forecast

NASA Astrophysics Data System (ADS)

Shaman, J. L.; Kandula, S.; Yang, W.; Karspeck, A. R.

2017-12-01

Laboratory and epidemiological evidence indicate that ambient humidity modulates the survival and transmission of influenza. Here we explore whether the inclusion of humidity forcing in mathematical models describing influenza transmission improves the accuracy of forecasts generated with those models. We generate retrospective forecasts for 95 cities over 10 seasons in the United States and assess both forecast accuracy and error. Overall, we find that humidity forcing improves forecast performance and that forecasts generated using daily climatological humidity forcing generally outperform forecasts that utilize daily observed humidity forcing. These findings hold for predictions of outbreak peak intensity, peak timing, and incidence over 2- and 4-week horizons. The results indicate that use of climatological humidity forcing is warranted for current operational influenza forecast and provide further evidence that humidity modulates rates of influenza transmission.

Extending flood forecasting lead time in a large watershed by coupling WRF QPF with a distributed hydrological model

NASA Astrophysics Data System (ADS)

Li, Ji; Chen, Yangbo; Wang, Huanyu; Qin, Jianming; Li, Jie; Chiao, Sen

2017-03-01

Long lead time flood forecasting is very important for large watershed flood mitigation as it provides more time for flood warning and emergency responses. The latest numerical weather forecast model could provide 1-15-day quantitative precipitation forecasting products in grid format, and by coupling this product with a distributed hydrological model could produce long lead time watershed flood forecasting products. This paper studied the feasibility of coupling the Liuxihe model with the Weather Research and Forecasting quantitative precipitation forecast (WRF QPF) for large watershed flood forecasting in southern China. The QPF of WRF products has three lead times, including 24, 48 and 72 h, with the grid resolution being 20 km  × 20 km. The Liuxihe model is set up with freely downloaded terrain property; the model parameters were previously optimized with rain gauge observed precipitation, and re-optimized with the WRF QPF. Results show that the WRF QPF has bias with the rain gauge precipitation, and a post-processing method is proposed to post-process the WRF QPF products, which improves the flood forecasting capability. With model parameter re-optimization, the model's performance improves also. This suggests that the model parameters be optimized with QPF, not the rain gauge precipitation. With the increasing of lead time, the accuracy of the WRF QPF decreases, as does the flood forecasting capability. Flood forecasting products produced by coupling the Liuxihe model with the WRF QPF provide a good reference for large watershed flood warning due to its long lead time and rational results.

Statistical post-processing of seasonal multi-model forecasts: Why is it so hard to beat the multi-model mean?

NASA Astrophysics Data System (ADS)

Siegert, Stefan

2017-04-01

Initialised climate forecasts on seasonal time scales, run several months or even years ahead, are now an integral part of the battery of products offered by climate services world-wide. The availability of seasonal climate forecasts from various modeling centres gives rise to multi-model ensemble forecasts. Post-processing such seasonal-to-decadal multi-model forecasts is challenging 1) because the cross-correlation structure between multiple models and observations can be complicated, 2) because the amount of training data to fit the post-processing parameters is very limited, and 3) because the forecast skill of numerical models tends to be low on seasonal time scales. In this talk I will review new statistical post-processing frameworks for multi-model ensembles. I will focus particularly on Bayesian hierarchical modelling approaches, which are flexible enough to capture commonly made assumptions about collective and model-specific biases of multi-model ensembles. Despite the advances in statistical methodology, it turns out to be very difficult to out-perform the simplest post-processing method, which just recalibrates the multi-model ensemble mean by linear regression. I will discuss reasons for this, which are closely linked to the specific characteristics of seasonal multi-model forecasts. I explore possible directions for improvements, for example using informative priors on the post-processing parameters, and jointly modelling forecasts and observations.

Mixture EMOS model for calibrating ensemble forecasts of wind speed.

PubMed

Baran, S; Lerch, S

2016-03-01

Ensemble model output statistics (EMOS) is a statistical tool for post-processing forecast ensembles of weather variables obtained from multiple runs of numerical weather prediction models in order to produce calibrated predictive probability density functions. The EMOS predictive probability density function is given by a parametric distribution with parameters depending on the ensemble forecasts. We propose an EMOS model for calibrating wind speed forecasts based on weighted mixtures of truncated normal (TN) and log-normal (LN) distributions where model parameters and component weights are estimated by optimizing the values of proper scoring rules over a rolling training period. The new model is tested on wind speed forecasts of the 50 member European Centre for Medium-range Weather Forecasts ensemble, the 11 member Aire Limitée Adaptation dynamique Développement International-Hungary Ensemble Prediction System ensemble of the Hungarian Meteorological Service, and the eight-member University of Washington mesoscale ensemble, and its predictive performance is compared with that of various benchmark EMOS models based on single parametric families and combinations thereof. The results indicate improved calibration of probabilistic and accuracy of point forecasts in comparison with the raw ensemble and climatological forecasts. The mixture EMOS model significantly outperforms the TN and LN EMOS methods; moreover, it provides better calibrated forecasts than the TN-LN combination model and offers an increased flexibility while avoiding covariate selection problems. © 2016 The Authors Environmetrics Published by JohnWiley & Sons Ltd.

Skill of Ensemble Seasonal Probability Forecasts

NASA Astrophysics Data System (ADS)

Smith, Leonard A.; Binter, Roman; Du, Hailiang; Niehoerster, Falk

2010-05-01

In operational forecasting, the computational complexity of large simulation models is, ideally, justified by enhanced performance over simpler models. We will consider probability forecasts and contrast the skill of ENSEMBLES-based seasonal probability forecasts of interest to the finance sector (specifically temperature forecasts for Nino 3.4 and the Atlantic Main Development Region (MDR)). The ENSEMBLES model simulations will be contrasted against forecasts from statistical models based on the observations (climatological distributions) and empirical dynamics based on the observations but conditioned on the current state (dynamical climatology). For some start dates, individual ENSEMBLES models yield significant skill even at a lead-time of 14 months. The nature of this skill is discussed, and chances of application are noted. Questions surrounding the interpretation of probability forecasts based on these multi-model ensemble simulations are then considered; the distributions considered are formed by kernel dressing the ensemble and blending with the climatology. The sources of apparent (RMS) skill in distributions based on multi-model simulations is discussed, and it is demonstrated that the inclusion of "zero-skill" models in the long range can improve Root-Mean-Square-Error scores, casting some doubt on the common justification for the claim that all models should be included in forming an operational probability forecast. It is argued that the rational response varies with lead time.

Daily River Flow Forecasting with Hybrid Support Vector Machine – Particle Swarm Optimization

NASA Astrophysics Data System (ADS)

Zaini, N.; Malek, M. A.; Yusoff, M.; Mardi, N. H.; Norhisham, S.

2018-04-01

The application of artificial intelligence techniques for river flow forecasting can further improve the management of water resources and flood prevention. This study concerns the development of support vector machine (SVM) based model and its hybridization with particle swarm optimization (PSO) to forecast short term daily river flow at Upper Bertam Catchment located in Cameron Highland, Malaysia. Ten years duration of historical rainfall, antecedent river flow data and various meteorology parameters data from 2003 to 2012 are used in this study. Four SVM based models are proposed which are SVM1, SVM2, SVM-PSO1 and SVM-PSO2 to forecast 1 to 7 day ahead of river flow. SVM1 and SVM-PSO1 are the models with historical rainfall and antecedent river flow as its input, while SVM2 and SVM-PSO2 are the models with historical rainfall, antecedent river flow data and additional meteorological parameters as input. The performances of the proposed model are measured in term of RMSE and R2 . It is found that, SVM2 outperformed SVM1 and SVM-PSO2 outperformed SVM-PSO1 which meant the additional meteorology parameters used as input to the proposed models significantly affect the model performances. Hybrid models SVM-PSO1 and SVM-PSO2 yield higher performances as compared to SVM1 and SVM2. It is found that hybrid models are more effective in forecasting river flow at 1 to 7 day ahead at the study area.

A perturbative approach for enhancing the performance of time series forecasting.

PubMed

de Mattos Neto, Paulo S G; Ferreira, Tiago A E; Lima, Aranildo R; Vasconcelos, Germano C; Cavalcanti, George D C

2017-04-01

This paper proposes a method to perform time series prediction based on perturbation theory. The approach is based on continuously adjusting an initial forecasting model to asymptotically approximate a desired time series model. First, a predictive model generates an initial forecasting for a time series. Second, a residual time series is calculated as the difference between the original time series and the initial forecasting. If that residual series is not white noise, then it can be used to improve the accuracy of the initial model and a new predictive model is adjusted using residual series. The whole process is repeated until convergence or the residual series becomes white noise. The output of the method is then given by summing up the outputs of all trained predictive models in a perturbative sense. To test the method, an experimental investigation was conducted on six real world time series. A comparison was made with six other methods experimented and ten other results found in the literature. Results show that not only the performance of the initial model is significantly improved but also the proposed method outperforms the other results previously published. Copyright © 2017 Elsevier Ltd. All rights reserved.

Hurricane Intensity Forecasts with a Global Mesoscale Model on the NASA Columbia Supercomputer

NASA Technical Reports Server (NTRS)

Shen, Bo-Wen; Tao, Wei-Kuo; Atlas, Robert

2006-01-01

It is known that General Circulation Models (GCMs) have insufficient resolution to accurately simulate hurricane near-eye structure and intensity. The increasing capabilities of high-end computers (e.g., the NASA Columbia Supercomputer) have changed this. In 2004, the finite-volume General Circulation Model at a 1/4 degree resolution, doubling the resolution used by most of operational NWP center at that time, was implemented and run to obtain promising landfall predictions for major hurricanes (e.g., Charley, Frances, Ivan, and Jeanne). In 2005, we have successfully implemented the 1/8 degree version, and demonstrated its performance on intensity forecasts with hurricane Katrina (2005). It is found that the 1/8 degree model is capable of simulating the radius of maximum wind and near-eye wind structure, and thereby promising intensity forecasts. In this study, we will further evaluate the model s performance on intensity forecasts of hurricanes Ivan, Jeanne, Karl in 2004. Suggestions for further model development will be made in the end.

The FAST-T approach for operational, real time, short term hydrological forecasting: Results from the Betania Hydropower Reservoir case study

NASA Astrophysics Data System (ADS)

Domínguez, Efraín; Angarita, Hector; Rosmann, Thomas; Mendez, Zulma; Angulo, Gustavo

2013-04-01

A viable quantitative hydrological forecasting service is a combination of technological elements, personnel and knowledge, working together to establish a stable operational cycle of forecasts emission, dissemination and assimilation; hence, the process for establishing such system usually requires significant resources and time to reach an adequate development and integration in order to produce forecasts with acceptable levels of performance. Here are presented the results of this process for the recently implemented Operational Forecast Service for the Betania's Hydropower Reservoir - or SPHEB, located at the Upper-Magdalena River Basin (Colombia). The current scope of the SPHEB includes forecasting of water levels and discharge for the three main streams affluent to the reservoir, for lead times between +1 to +57 hours, and +1 to +10 days. The core of the SPHEB is the Flexible, Adaptive, Simple and Transient Time forecasting approach, namely FAST-T. This comprises of a set of data structures, mathematical kernel, distributed computing and network infrastructure designed to provide seamless real-time operational forecast and automatic model adjustment in case of failures in data transmission or assimilation. Among FAST-T main features are: an autonomous evaluation and detection of the most relevant information for the later configuration of forecasting models; an adaptively linearized mathematical kernel, the optimal adaptive linear combination or OALC, which provides a computationally simple and efficient algorithm for real-time applications; and finally, a meta-model catalog, containing prioritized forecast models at given stream conditions. The SPHEB is at present feed by the fraction of hydrological monitoring network installed at the basin that has telemetric capabilities via NOAA-GOES satellites (8 stages, approximately 47%) with data availability of about a 90% at one hour intervals. However, there is a dense network of 'conventional' hydro-meteorological stages -read manually once or twice per day - that, despite not ideal in the context of real-time system, improve model performance significantly, and therefore are entered into the system by manual input. At its current configuration, the SPHEB performance objectives are fulfilled for 90% of the forecasts with lead times up to +2 days and +15 hours (using the predictability criteria of the Russian Hydrometeorological Center S/?Δ) and the average accuracy is in the range 70-99% ( r2 criteria). However, longer lead times are at present not satisfactory in terms of forecasts accuracy.

Integrating Wind Profiling Radars and Radiosonde Observations with Model Point Data to Develop a Decision Support Tool to Assess Upper-Level Winds for Space Launch

NASA Technical Reports Server (NTRS)

Bauman, William H., III; Flinn, Clay

2013-01-01

On the day-of-launch, the 45th Weather Squadron (45 WS) Launch Weather Officers (LWOs) monitor the upper-level winds for their launch customers to include NASA's Launch Services Program and NASA's Ground Systems Development and Operations Program. They currently do not have the capability to display and overlay profiles of upper-level observations and numerical weather prediction model forecasts. The LWOs requested the Applied Meteorology Unit (AMU) develop a tool in the form of a graphical user interface (GUI) that will allow them to plot upper-level wind speed and direction observations from the Kennedy Space Center (KSC) 50 MHz tropospheric wind profiling radar, KSC Shuttle Landing Facility 915 MHz boundary layer wind profiling radar and Cape Canaveral Air Force Station (CCAFS) Automated Meteorological Processing System (AMPS) radiosondes, and then overlay forecast wind profiles from the model point data including the North American Mesoscale (NAM) model, Rapid Refresh (RAP) model and Global Forecast System (GFS) model to assess the performance of these models. The AMU developed an Excel-based tool that provides an objective method for the LWOs to compare the model-forecast upper-level winds to the KSC wind profiling radars and CCAFS AMPS observations to assess the model potential to accurately forecast changes in the upperlevel profile through the launch count. The AMU wrote Excel Visual Basic for Applications (VBA) scripts to automatically retrieve model point data for CCAFS (XMR) from the Iowa State University Archive Data Server (http://mtarchive.qeol.iastate.edu) and the 50 MHz, 915 MHz and AMPS observations from the NASA/KSC Spaceport Weather Data Archive web site (http://trmm.ksc.nasa.gov). The AMU then developed code in Excel VBA to automatically ingest and format the observations and model point data in Excel to ready the data for generating Excel charts for the LWO's. The resulting charts allow the LWOs to independently initialize the three models 0-hour forecasts against the observations to determine which is the best performing model and then overlay the model forecasts on time-matched observations during the launch countdown to further assess the model performance and forecasts. This paper will demonstrate integration of observed and predicted atmospheric conditions into a decision support tool and demonstrate how the GUI is implemented in operations.

Utilizing Climate Forecasts for Improving Water and Power Systems Coordination

NASA Astrophysics Data System (ADS)

Arumugam, S.; Queiroz, A.; Patskoski, J.; Mahinthakumar, K.; DeCarolis, J.

2016-12-01

Climate forecasts, typically monthly-to-seasonal precipitation forecasts, are commonly used to develop streamflow forecasts for improving reservoir management. Irrespective of their high skill in forecasting, temperature forecasts in developing power demand forecasts are not often considered along with streamflow forecasts for improving water and power systems coordination. In this study, we consider a prototype system to analyze the utility of climate forecasts, both precipitation and temperature, for improving water and power systems coordination. The prototype system, a unit-commitment model that schedules power generation from various sources, is considered and its performance is compared with an energy system model having an equivalent reservoir representation. Different skill sets of streamflow forecasts and power demand forecasts are forced on both water and power systems representations for understanding the level of model complexity required for utilizing monthly-to-seasonal climate forecasts to improve coordination between these two systems. The analyses also identify various decision-making strategies - forward purchasing of fuel stocks, scheduled maintenance of various power systems and tradeoff on water appropriation between hydropower and other uses - in the context of various water and power systems configurations. Potential application of such analyses for integrating large power systems with multiple river basins is also discussed.

Statistical Modeling and Prediction for Tourism Economy Using Dendritic Neural Network

PubMed Central

Yu, Ying; Wang, Yirui; Tang, Zheng

2017-01-01

With the impact of global internationalization, tourism economy has also been a rapid development. The increasing interest aroused by more advanced forecasting methods leads us to innovate forecasting methods. In this paper, the seasonal trend autoregressive integrated moving averages with dendritic neural network model (SA-D model) is proposed to perform the tourism demand forecasting. First, we use the seasonal trend autoregressive integrated moving averages model (SARIMA model) to exclude the long-term linear trend and then train the residual data by the dendritic neural network model and make a short-term prediction. As the result showed in this paper, the SA-D model can achieve considerably better predictive performances. In order to demonstrate the effectiveness of the SA-D model, we also use the data that other authors used in the other models and compare the results. It also proved that the SA-D model achieved good predictive performances in terms of the normalized mean square error, absolute percentage of error, and correlation coefficient. PMID:28246527

Statistical Modeling and Prediction for Tourism Economy Using Dendritic Neural Network.

PubMed

Yu, Ying; Wang, Yirui; Gao, Shangce; Tang, Zheng

2017-01-01

With the impact of global internationalization, tourism economy has also been a rapid development. The increasing interest aroused by more advanced forecasting methods leads us to innovate forecasting methods. In this paper, the seasonal trend autoregressive integrated moving averages with dendritic neural network model (SA-D model) is proposed to perform the tourism demand forecasting. First, we use the seasonal trend autoregressive integrated moving averages model (SARIMA model) to exclude the long-term linear trend and then train the residual data by the dendritic neural network model and make a short-term prediction. As the result showed in this paper, the SA-D model can achieve considerably better predictive performances. In order to demonstrate the effectiveness of the SA-D model, we also use the data that other authors used in the other models and compare the results. It also proved that the SA-D model achieved good predictive performances in terms of the normalized mean square error, absolute percentage of error, and correlation coefficient.

Three-model ensemble wind prediction in southern Italy

NASA Astrophysics Data System (ADS)

Torcasio, Rosa Claudia; Federico, Stefano; Calidonna, Claudia Roberta; Avolio, Elenio; Drofa, Oxana; Landi, Tony Christian; Malguzzi, Piero; Buzzi, Andrea; Bonasoni, Paolo

2016-03-01

Quality of wind prediction is of great importance since a good wind forecast allows the prediction of available wind power, improving the penetration of renewable energies into the energy market. Here, a 1-year (1 December 2012 to 30 November 2013) three-model ensemble (TME) experiment for wind prediction is considered. The models employed, run operationally at National Research Council - Institute of Atmospheric Sciences and Climate (CNR-ISAC), are RAMS (Regional Atmospheric Modelling System), BOLAM (BOlogna Limited Area Model), and MOLOCH (MOdello LOCale in H coordinates). The area considered for the study is southern Italy and the measurements used for the forecast verification are those of the GTS (Global Telecommunication System). Comparison with observations is made every 3 h up to 48 h of forecast lead time. Results show that the three-model ensemble outperforms the forecast of each individual model. The RMSE improvement compared to the best model is between 22 and 30 %, depending on the season. It is also shown that the three-model ensemble outperforms the IFS (Integrated Forecasting System) of the ECMWF (European Centre for Medium-Range Weather Forecast) for the surface wind forecasts. Notably, the three-model ensemble forecast performs better than each unbiased model, showing the added value of the ensemble technique. Finally, the sensitivity of the three-model ensemble RMSE to the length of the training period is analysed.

Comparison of the performance and reliability of 18 lumped hydrological models driven by ECMWF rainfall ensemble forecasts: a case study on 29 French catchments

NASA Astrophysics Data System (ADS)

Velázquez, Juan Alberto; Anctil, François; Ramos, Maria-Helena; Perrin, Charles

2010-05-01

An ensemble forecasting system seeks to assess and to communicate the uncertainty of hydrological predictions by proposing, at each time step, an ensemble of forecasts from which one can estimate the probability distribution of the predictant (the probabilistic forecast), in contrast with a single estimate of the flow, for which no distribution is obtainable (the deterministic forecast). In the past years, efforts towards the development of probabilistic hydrological prediction systems were made with the adoption of ensembles of numerical weather predictions (NWPs). The additional information provided by the different available Ensemble Prediction Systems (EPS) was evaluated in a hydrological context on various case studies (see the review by Cloke and Pappenberger, 2009). For example, the European ECMWF-EPS was explored in case studies by Roulin et al. (2005), Bartholmes et al. (2005), Jaun et al. (2008), and Renner et al. (2009). The Canadian EC-EPS was also evaluated by Velázquez et al. (2009). Most of these case studies investigate the ensemble predictions of a given hydrological model, set up over a limited number of catchments. Uncertainty from weather predictions is assessed through the use of meteorological ensembles. However, uncertainty from the tested hydrological model and statistical robustness of the forecasting system when coping with different hydro-meteorological conditions are less frequently evaluated. The aim of this study is to evaluate and compare the performance and the reliability of 18 lumped hydrological models applied to a large number of catchments in an operational ensemble forecasting context. Some of these models were evaluated in a previous study (Perrin et al. 2001) for their ability to simulate streamflow. Results demonstrated that very simple models can achieve a level of performance almost as high (sometimes higher) as models with more parameters. In the present study, we focus on the ability of the hydrological models to provide reliable probabilistic forecasts of streamflow, based on ensemble weather predictions. The models were therefore adapted to run in a forecasting mode, i.e., to update initial conditions according to the last observed discharge at the time of the forecast, and to cope with ensemble weather scenarios. All models are lumped, i.e., the hydrological behavior is integrated over the spatial scale of the catchment, and run at daily time steps. The complexity of tested models varies between 3 and 13 parameters. The models are tested on 29 French catchments. Daily streamflow time series extend over 17 months, from March 2005 to July 2006. Catchment areas range between 1470 km2 and 9390 km2, and represent a variety of hydrological and meteorological conditions. The 12 UTC 10-day ECMWF rainfall ensemble (51 members) was used, which led to daily streamflow forecasts for a 9-day lead time. In order to assess the performance and reliability of the hydrological ensemble predictions, we computed the Continuous Ranked probability Score (CRPS) (Matheson and Winkler, 1976), as well as the reliability diagram (e.g. Wilks, 1995) and the rank histogram (Talagrand et al., 1999). Since the ECMWF deterministic forecasts are also available, the performance of the hydrological forecasting systems was also evaluated by comparing the deterministic score (MAE) with the probabilistic score (CRPS). The results obtained for the 18 hydrological models and the 29 studied catchments are discussed in the perspective of improving the operational use of ensemble forecasting in hydrology. References Bartholmes, J. and Todini, E.: Coupling meteorological and hydrological models for flood forecasting, Hydrol. Earth Syst. Sci., 9, 333-346, 2005. Cloke, H. and Pappenberger, F.: Ensemble Flood Forecasting: A Review. Journal of Hydrology 375 (3-4): 613-626, 2009. Jaun, S., Ahrens, B., Walser, A., Ewen, T., and Schär, C.: A probabilistic view on the August 2005 floods in the upper Rhine catchment, Nat. Hazards Earth Syst. Sci., 8, 281-291, 2008. Matheson, J. E. and Winkler, R. L.: Scoring rules for continuous probability distributions, Manage Sci., 22, 1087-1096, 1976. Perrin, C., Michel C. and Andréassian,V. Does a large number of parameters enhance model performance? Comparative assessment of common catchment model structures on 429 catchments, J. Hydrol., 242, 275-301, 2001. Renner, M., Werner, M. G. F., Rademacher, S., and Sprokkereef, E.: Verification of ensemble flow forecast for the River Rhine, J. Hydrol., 376, 463-475, 2009. Roulin, E. and Vannitsem, S.: Skill of medium-range hydrological ensemble predictions, J. Hydrometeorol., 6, 729-744, 2005. Talagrand, O., Vautard, R., and Strauss, B.: Evaluation of the probabilistic prediction systems, in: Proceedings, ECMWF Workshop on Predictability, Shinfield Park, Reading, Berkshire, ECMWF, 1-25, 1999. Velázquez, J.A., Petit, T., Lavoie, A., Boucher M.-A., Turcotte R., Fortin V., and Anctil, F. : An evaluation of the Canadian global meteorological ensemble prediction system for short-term hydrological forecasting, Hydrol. Earth Syst. Sci., 13, 2221-2231, 2009. Wilks, D. S.: Statistical Methods in the Atmospheric Sciences, Academic Press, San Diego, CA, 465 pp., 1995.

Medium-range Performance of the Global NWP Model

NASA Astrophysics Data System (ADS)

Kim, J.; Jang, T.; Kim, J.; Kim, Y.

2017-12-01

The medium-range performance of the global numerical weather prediction (NWP) model in the Korea Meteorological Administration (KMA) is investigated. The performance is based on the prediction of the extratropical circulation. The mean square error is expressed by sum of spatial variance of discrepancy between forecasts and observations and the square of the mean error (ME). Thus, it is important to investigate the ME effect in order to understand the model performance. The ME is expressed by the subtraction of an anomaly from forecast difference against the real climatology. It is found that the global model suffers from a severe systematic ME in medium-range forecasts. The systematic ME is dominant in the entire troposphere in all months. Such ME can explain at most 25% of root mean square error. We also compare the extratropical ME distribution with that from other NWP centers. NWP models exhibit similar spatial ME structure each other. It is found that the spatial ME pattern is highly correlated to that of an anomaly, implying that the ME varies with seasons. For example, the correlation coefficient between ME and anomaly ranges from -0.51 to -0.85 by months. The pattern of the extratropical circulation also has a high correlation to an anomaly. The global model has trouble in faithfully simulating extratropical cyclones and blockings in the medium-range forecast. In particular, the model has a hard to simulate an anomalous event in medium-range forecasts. If we choose an anomalous period for a test-bed experiment, we will suffer from a large error due to an anomaly.

Using Deep Learning Model for Meteorological Satellite Cloud Image Prediction

NASA Astrophysics Data System (ADS)

Su, X.

2017-12-01

A satellite cloud image contains much weather information such as precipitation information. Short-time cloud movement forecast is important for precipitation forecast and is the primary means for typhoon monitoring. The traditional methods are mostly using the cloud feature matching and linear extrapolation to predict the cloud movement, which makes that the nonstationary process such as inversion and deformation during the movement of the cloud is basically not considered. It is still a hard task to predict cloud movement timely and correctly. As deep learning model could perform well in learning spatiotemporal features, to meet this challenge, we could regard cloud image prediction as a spatiotemporal sequence forecasting problem and introduce deep learning model to solve this problem. In this research, we use a variant of Gated-Recurrent-Unit(GRU) that has convolutional structures to deal with spatiotemporal features and build an end-to-end model to solve this forecast problem. In this model, both the input and output are spatiotemporal sequences. Compared to Convolutional LSTM(ConvLSTM) model, this model has lower amount of parameters. We imply this model on GOES satellite data and the model perform well.

Cloud Computing Applications in Support of Earth Science Activities at Marshall Space Flight Center

NASA Astrophysics Data System (ADS)

Molthan, A.; Limaye, A. S.

2011-12-01

Currently, the NASA Nebula Cloud Computing Platform is available to Agency personnel in a pre-release status as the system undergoes a formal operational readiness review. Over the past year, two projects within the Earth Science Office at NASA Marshall Space Flight Center have been investigating the performance and value of Nebula's "Infrastructure as a Service", or "IaaS" concept and applying cloud computing concepts to advance their respective mission goals. The Short-term Prediction Research and Transition (SPoRT) Center focuses on the transition of unique NASA satellite observations and weather forecasting capabilities for use within the operational forecasting community through partnerships with NOAA's National Weather Service (NWS). SPoRT has evaluated the performance of the Weather Research and Forecasting (WRF) model on virtual machines deployed within Nebula and used Nebula instances to simulate local forecasts in support of regional forecast studies of interest to select NWS forecast offices. In addition to weather forecasting applications, rapidly deployable Nebula virtual machines have supported the processing of high resolution NASA satellite imagery to support disaster assessment following the historic severe weather and tornado outbreak of April 27, 2011. Other modeling and satellite analysis activities are underway in support of NASA's SERVIR program, which integrates satellite observations, ground-based data and forecast models to monitor environmental change and improve disaster response in Central America, the Caribbean, Africa, and the Himalayas. Leveraging SPoRT's experience, SERVIR is working to establish a real-time weather forecasting model for Central America. Other modeling efforts include hydrologic forecasts for Kenya, driven by NASA satellite observations and reanalysis data sets provided by the broader meteorological community. Forecast modeling efforts are supplemented by short-term forecasts of convective initiation, determined by geostationary satellite observations processed on virtual machines powered by Nebula. This presentation will provide an overview of these activities from a scientific and cloud computing applications perspective, identifying the strengths and weaknesses for deploying each project within an IaaS environment, and ways to collaborate with the Nebula or other cloud-user communities to collaborate on projects as they go forward.

The use of ambient humidity conditions to improve influenza forecast.

PubMed

Shaman, Jeffrey; Kandula, Sasikiran; Yang, Wan; Karspeck, Alicia

2017-11-01

Laboratory and epidemiological evidence indicate that ambient humidity modulates the survival and transmission of influenza. Here we explore whether the inclusion of humidity forcing in mathematical models describing influenza transmission improves the accuracy of forecasts generated with those models. We generate retrospective forecasts for 95 cities over 10 seasons in the United States and assess both forecast accuracy and error. Overall, we find that humidity forcing improves forecast performance (at 1-4 lead weeks, 3.8% more peak week and 4.4% more peak intensity forecasts are accurate than with no forcing) and that forecasts generated using daily climatological humidity forcing generally outperform forecasts that utilize daily observed humidity forcing (4.4% and 2.6% respectively). These findings hold for predictions of outbreak peak intensity, peak timing, and incidence over 2- and 4-week horizons. The results indicate that use of climatological humidity forcing is warranted for current operational influenza forecast.

The use of ambient humidity conditions to improve influenza forecast

PubMed Central

Kandula, Sasikiran; Karspeck, Alicia

2017-01-01

Laboratory and epidemiological evidence indicate that ambient humidity modulates the survival and transmission of influenza. Here we explore whether the inclusion of humidity forcing in mathematical models describing influenza transmission improves the accuracy of forecasts generated with those models. We generate retrospective forecasts for 95 cities over 10 seasons in the United States and assess both forecast accuracy and error. Overall, we find that humidity forcing improves forecast performance (at 1–4 lead weeks, 3.8% more peak week and 4.4% more peak intensity forecasts are accurate than with no forcing) and that forecasts generated using daily climatological humidity forcing generally outperform forecasts that utilize daily observed humidity forcing (4.4% and 2.6% respectively). These findings hold for predictions of outbreak peak intensity, peak timing, and incidence over 2- and 4-week horizons. The results indicate that use of climatological humidity forcing is warranted for current operational influenza forecast. PMID:29145389

Impact of the initial specification of moisture and vertical motion on precipitation forecasts with a mesoscale model Implications for a satellite mesoscale data base

NASA Technical Reports Server (NTRS)

Mlynczak, Pamela E.; Houghton, David D.; Diak, George R.

1986-01-01

Using a numerical mesoscale model, four simulations were performed to determine the effects of suppressing the initial mesoscale information in the moisture and wind fields on the precipitation forecasts. The simulations included a control forecast 12-h simulation that began at 1200 GMT March 1982 and three experiment simulations with modifications to the moisture and vertical motion fields incorporated at 1800 GMT. The forecasts from 1800 GMT were compared to the second half of the control forecast. It was found that, compared to the control forecast, suppression of the moisture and/or wind initial field(s) produces a drier forecast. However, the characteristics of the precipitation forecasts of the experiments were not different enough to conclude that either mesoscale moisture or mesoscale vertical velocity at the initial time are more important for producing a forecast closer to that of the control.

Improving the effectiveness of real-time flood forecasting through Predictive Uncertainty estimation: the multi-temporal approach

NASA Astrophysics Data System (ADS)

Barbetta, Silvia; Coccia, Gabriele; Moramarco, Tommaso; Todini, Ezio

2015-04-01

The negative effects of severe flood events are usually contrasted through structural measures that, however, do not fully eliminate flood risk. Non-structural measures, such as real-time flood forecasting and warning, are also required. Accurate stage/discharge future predictions with appropriate forecast lead-time are sought by decision-makers for implementing strategies to mitigate the adverse effects of floods. Traditionally, flood forecasting has been approached by using rainfall-runoff and/or flood routing modelling. Indeed, both types of forecasts, cannot be considered perfectly representing future outcomes because of lacking of a complete knowledge of involved processes (Todini, 2004). Nonetheless, although aware that model forecasts are not perfectly representing future outcomes, decision makers are de facto implicitly assuming the forecast of water level/discharge/volume, etc. as "deterministic" and coinciding with what is going to occur. Recently the concept of Predictive Uncertainty (PU) was introduced in hydrology (Krzysztofowicz, 1999), and several uncertainty processors were developed (Todini, 2008). PU is defined as the probability of occurrence of the future realization of a predictand (water level/discharge/volume) conditional on: i) prior observations and knowledge, ii) the available information obtained on the future value, typically provided by one or more forecast models. Unfortunately, PU has been frequently interpreted as a measure of lack of accuracy rather than the appropriate tool allowing to take the most appropriate decisions, given a model or several models' forecasts. With the aim to shed light on the benefits for appropriately using PU, a multi-temporal approach based on the MCP approach (Todini, 2008; Coccia and Todini, 2011) is here applied to stage forecasts at sites along the Upper Tiber River. Specifically, the STAge Forecasting-Rating Curve Model Muskingum-based (STAFOM-RCM) (Barbetta et al., 2014) along with the Rating-Curve Model in Real Time (RCM-RT) (Barbetta and Moramarco, 2014) are used to this end. Both models without considering rainfall information explicitly considers, at each time of forecast, the estimate of lateral contribution along the river reach for which the stage forecast is performed at downstream end. The analysis is performed for several reaches using different lead times according to the channel length. Barbetta, S., Moramarco, T., Brocca, L., Franchini, M. and Melone, F. 2014. Confidence interval of real-time forecast stages provided by the STAFOM-RCM model: the case study of the Tiber River (Italy). Hydrological Processes, 28(3),729-743. Barbetta, S. and Moramarco, T. 2014. Real-time flood forecasting by relating local stage and remote discharge. Hydrological Sciences Journal, 59(9 ), 1656-1674. Coccia, G. and Todini, E. 2011. Recent developments in predictive uncertainty assessment based on the Model Conditional Processor approach. Hydrology and Earth System Sciences, 15, 3253-3274. doi:10.5194/hess-15-3253-2011. Krzysztofowicz, R. 1999. Bayesian theory of probabilistic forecasting via deterministic hydrologic model, Water Resour. Res., 35, 2739-2750. Todini, E. 2004. Role and treatment of uncertainty in real-time flood forecasting. Hydrological Processes 18(14), 2743_2746. Todini, E. 2008. A model conditional processor to assess predictive uncertainty in flood forecasting. Intl. J. River Basin Management, 6(2): 123-137.

Time series model for forecasting the number of new admission inpatients.

PubMed

Zhou, Lingling; Zhao, Ping; Wu, Dongdong; Cheng, Cheng; Huang, Hao

2018-06-15

Hospital crowding is a rising problem, effective predicting and detecting managment can helpful to reduce crowding. Our team has successfully proposed a hybrid model combining both the autoregressive integrated moving average (ARIMA) and the nonlinear autoregressive neural network (NARNN) models in the schistosomiasis and hand, foot, and mouth disease forecasting study. In this paper, our aim is to explore the application of the hybrid ARIMA-NARNN model to track the trends of the new admission inpatients, which provides a methodological basis for reducing crowding. We used the single seasonal ARIMA (SARIMA), NARNN and the hybrid SARIMA-NARNN model to fit and forecast the monthly and daily number of new admission inpatients. The root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) were used to compare the forecasting performance among the three models. The modeling time range of monthly data included was from January 2010 to June 2016, July to October 2016 as the corresponding testing data set. The daily modeling data set was from January 4 to September 4, 2016, while the testing time range included was from September 5 to October 2, 2016. For the monthly data, the modeling RMSE and the testing RMSE, MAE and MAPE of SARIMA-NARNN model were less than those obtained from the single SARIMA or NARNN model, but the MAE and MAPE of modeling performance of SARIMA-NARNN model did not improve. For the daily data, all RMSE, MAE and MAPE of NARNN model were the lowest both in modeling stage and testing stage. Hybrid model does not necessarily outperform its constituents' performances. It is worth attempting to explore the reliable model to forecast the number of new admission inpatients from different data.

Monthly streamflow forecasting using continuous wavelet and multi-gene genetic programming combination

NASA Astrophysics Data System (ADS)

Hadi, Sinan Jasim; Tombul, Mustafa

2018-06-01

Streamflow is an essential component of the hydrologic cycle in the regional and global scale and the main source of fresh water supply. It is highly associated with natural disasters, such as droughts and floods. Therefore, accurate streamflow forecasting is essential. Forecasting streamflow in general and monthly streamflow in particular is a complex process that cannot be handled by data-driven models (DDMs) only and requires pre-processing. Wavelet transformation is a pre-processing technique; however, application of continuous wavelet transformation (CWT) produces many scales that cause deterioration in the performance of any DDM because of the high number of redundant variables. This study proposes multigene genetic programming (MGGP) as a selection tool. After the CWT analysis, it selects important scales to be imposed into the artificial neural network (ANN). A basin located in the southeast of Turkey is selected as case study to prove the forecasting ability of the proposed model. One month ahead downstream flow is used as output, and downstream flow, upstream, rainfall, temperature, and potential evapotranspiration with associated lags are used as inputs. Before modeling, wavelet coherence transformation (WCT) analysis was conducted to analyze the relationship between variables in the time-frequency domain. Several combinations were developed to investigate the effect of the variables on streamflow forecasting. The results indicated a high localized correlation between the streamflow and other variables, especially the upstream. In the models of the standalone layout where the data were entered to ANN and MGGP without CWT, the performance is found poor. In the best-scale layout, where the best scale of the CWT identified as the highest correlated scale is chosen and enters to ANN and MGGP, the performance increased slightly. Using the proposed model, the performance improved dramatically particularly in forecasting the peak values because of the inclusion of several scales in which seasonality and irregularity can be captured. Using hydrological and meteorological variables also improved the ability to forecast the streamflow.

Statistical and dynamical forecast of regional precipitation after mature phase of ENSO

NASA Astrophysics Data System (ADS)

Sohn, S.; Min, Y.; Lee, J.; Tam, C.; Ahn, J.

2010-12-01

While the seasonal predictability of general circulation models (GCMs) has been improved, the current model atmosphere in the mid-latitude does not respond correctly to external forcing such as tropical sea surface temperature (SST), particularly over the East Asia and western North Pacific summer monsoon regions. In addition, the time-scale of prediction scope is considerably limited and the model forecast skill still is very poor beyond two weeks. Although recent studies indicate that coupled model based multi-model ensemble (MME) forecasts show the better performance, the long-lead forecasts exceeding 9 months still show a dramatic decrease of the seasonal predictability. This study aims at diagnosing the dynamical MME forecasts comprised of the state of art 1-tier models as well as comparing them with the statistical model forecasts, focusing on the East Asian summer precipitation predictions after mature phase of ENSO. The lagged impact of El Nino as major climate contributor on the summer monsoon in model environments is also evaluated, in the sense of the conditional probabilities. To evaluate the probability forecast skills, the reliability (attributes) diagram and the relative operating characteristics following the recommendations of the World Meteorological Organization (WMO) Standardized Verification System for Long-Range Forecasts are used in this study. The results should shed light on the prediction skill for dynamical model and also for the statistical model, in forecasting the East Asian summer monsoon rainfall with a long-lead time.

A Bias and Variance Analysis for Multistep-Ahead Time Series Forecasting.

PubMed

Ben Taieb, Souhaib; Atiya, Amir F

2016-01-01

Multistep-ahead forecasts can either be produced recursively by iterating a one-step-ahead time series model or directly by estimating a separate model for each forecast horizon. In addition, there are other strategies; some of them combine aspects of both aforementioned concepts. In this paper, we present a comprehensive investigation into the bias and variance behavior of multistep-ahead forecasting strategies. We provide a detailed review of the different multistep-ahead strategies. Subsequently, we perform a theoretical study that derives the bias and variance for a number of forecasting strategies. Finally, we conduct a Monte Carlo experimental study that compares and evaluates the bias and variance performance of the different strategies. From the theoretical and the simulation studies, we analyze the effect of different factors, such as the forecast horizon and the time series length, on the bias and variance components, and on the different multistep-ahead strategies. Several lessons are learned, and recommendations are given concerning the advantages, disadvantages, and best conditions of use of each strategy.

Model of medicines sales forecasting taking into account factors of influence

NASA Astrophysics Data System (ADS)

Kravets, A. G.; Al-Gunaid, M. A.; Loshmanov, V. I.; Rasulov, S. S.; Lempert, L. B.

2018-05-01

The article describes a method for forecasting sales of medicines in conditions of data sampling, which is insufficient for building a model based on historical data alone. The developed method is applicable mainly to new drugs that are already licensed and released for sale but do not yet have stable sales performance in the market. The purpose of this study is to prove the effectiveness of the suggested method forecasting drug sales, taking into account the selected factors of influence, revealed during the review of existing solutions and analysis of the specificity of the area under study. Three experiments were performed on samples of different volumes, which showed an improvement in the accuracy of forecasting sales in small samples.

Network bandwidth utilization forecast model on high bandwidth networks

DOE Office of Scientific and Technical Information (OSTI.GOV)

Yoo, Wuchert; Sim, Alex

With the increasing number of geographically distributed scientific collaborations and the scale of the data size growth, it has become more challenging for users to achieve the best possible network performance on a shared network. We have developed a forecast model to predict expected bandwidth utilization for high-bandwidth wide area network. The forecast model can improve the efficiency of resource utilization and scheduling data movements on high-bandwidth network to accommodate ever increasing data volume for large-scale scientific data applications. Univariate model is developed with STL and ARIMA on SNMP path utilization data. Compared with traditional approach such as Box-Jenkins methodology,more » our forecast model reduces computation time by 83.2%. It also shows resilience against abrupt network usage change. The accuracy of the forecast model is within the standard deviation of the monitored measurements.« less

Network Bandwidth Utilization Forecast Model on High Bandwidth Network

DOE Office of Scientific and Technical Information (OSTI.GOV)

Yoo, Wucherl; Sim, Alex

With the increasing number of geographically distributed scientific collaborations and the scale of the data size growth, it has become more challenging for users to achieve the best possible network performance on a shared network. We have developed a forecast model to predict expected bandwidth utilization for high-bandwidth wide area network. The forecast model can improve the efficiency of resource utilization and scheduling data movements on high-bandwidth network to accommodate ever increasing data volume for large-scale scientific data applications. Univariate model is developed with STL and ARIMA on SNMP path utilization data. Compared with traditional approach such as Box-Jenkins methodology,more » our forecast model reduces computation time by 83.2percent. It also shows resilience against abrupt network usage change. The accuracy of the forecast model is within the standard deviation of the monitored measurements.« less

Comparison of Conventional and ANN Models for River Flow Forecasting

NASA Astrophysics Data System (ADS)

Jain, A.; Ganti, R.

2011-12-01

Hydrological models are useful in many water resources applications such as flood control, irrigation and drainage, hydro power generation, water supply, erosion and sediment control, etc. Estimates of runoff are needed in many water resources planning, design development, operation and maintenance activities. River flow is generally estimated using time series or rainfall-runoff models. Recently, soft artificial intelligence tools such as Artificial Neural Networks (ANNs) have become popular for research purposes but have not been extensively adopted in operational hydrological forecasts. There is a strong need to develop ANN models based on real catchment data and compare them with the conventional models. In this paper, a comparative study has been carried out for river flow forecasting using the conventional and ANN models. Among the conventional models, multiple linear, and non linear regression, and time series models of auto regressive (AR) type have been developed. Feed forward neural network model structure trained using the back propagation algorithm, a gradient search method, was adopted. The daily river flow data derived from Godavari Basin @ Polavaram, Andhra Pradesh, India have been employed to develop all the models included here. Two inputs, flows at two past time steps, (Q(t-1) and Q(t-2)) were selected using partial auto correlation analysis for forecasting flow at time t, Q(t). A wide range of error statistics have been used to evaluate the performance of all the models developed in this study. It has been found that the regression and AR models performed comparably, and the ANN model performed the best amongst all the models investigated in this study. It is concluded that ANN model should be adopted in real catchments for hydrological modeling and forecasting.

Spatio-temporal pattern clustering for skill assessment of the Korea Operational Oceanographic System

NASA Astrophysics Data System (ADS)

Kim, J.; Park, K.

2016-12-01

In order to evaluate the performance of operational forecast models in the Korea operational oceanographic system (KOOS) which has been developed by Korea Institute of Ocean Science and Technology (KIOST), a skill assessment (SA) tool has developed and provided multiple skill metrics including not only correlation and error skills by comparing predictions and observation but also pattern clustering with numerical models, satellite, and observation. The KOOS has produced 72 hours forecast information on atmospheric and hydrodynamic forecast variables of wind, pressure, current, tide, wave, temperature, and salinity at every 12 hours per day produced by operating numerical models such as WRF, ROMS, MOM5, WW-III, and SWAN and the SA has conducted to evaluate the forecasts. We have been operationally operated several kinds of numerical models such as WRF, ROMS, MOM5, MOHID, WW-III. Quantitative assessment of operational ocean forecast model is very important to provide accurate ocean forecast information not only to general public but also to support ocean-related problems. In this work, we propose a method of pattern clustering using machine learning method and GIS-based spatial analytics to evaluate spatial distribution of numerical models and spatial observation data such as satellite and HF radar. For the clustering, we use 10 or 15 years-long reanalysis data which was computed by the KOOS, ECMWF, and HYCOM to make best matching clusters which are classified physical meaning with time variation and then we compare it with forecast data. Moreover, for evaluating current, we develop extraction method of dominant flow and apply it to hydrodynamic models and HF radar's sea surface current data. By applying pattern clustering method, it allows more accurate and effective assessment of ocean forecast models' performance by comparing not only specific observation positions which are determined by observation stations but also spatio-temporal distribution of whole model areas. We believe that our proposed method will be very useful to examine and evaluate large amount of numerical modeling data as well as satellite data.

Development and Implementation of Dynamic Scripts to Support Local Model Verification at National Weather Service Weather Forecast Offices

NASA Technical Reports Server (NTRS)

Zavordsky, Bradley; Case, Jonathan L.; Gotway, John H.; White, Kristopher; Medlin, Jeffrey; Wood, Lance; Radell, Dave

2014-01-01

Local modeling with a customized configuration is conducted at National Weather Service (NWS) Weather Forecast Offices (WFOs) to produce high-resolution numerical forecasts that can better simulate local weather phenomena and complement larger scale global and regional models. The advent of the Environmental Modeling System (EMS), which provides a pre-compiled version of the Weather Research and Forecasting (WRF) model and wrapper Perl scripts, has enabled forecasters to easily configure and execute the WRF model on local workstations. NWS WFOs often use EMS output to help in forecasting highly localized, mesoscale features such as convective initiation, the timing and inland extent of lake effect snow bands, lake and sea breezes, and topographically-modified winds. However, quantitatively evaluating model performance to determine errors and biases still proves to be one of the challenges in running a local model. Developed at the National Center for Atmospheric Research (NCAR), the Model Evaluation Tools (MET) verification software makes performing these types of quantitative analyses easier, but operational forecasters do not generally have time to familiarize themselves with navigating the sometimes complex configurations associated with the MET tools. To assist forecasters in running a subset of MET programs and capabilities, the Short-term Prediction Research and Transition (SPoRT) Center has developed and transitioned a set of dynamic, easily configurable Perl scripts to collaborating NWS WFOs. The objective of these scripts is to provide SPoRT collaborating partners in the NWS with the ability to evaluate the skill of their local EMS model runs in near real time with little prior knowledge of the MET package. The ultimate goal is to make these verification scripts available to the broader NWS community in a future version of the EMS software. This paper provides an overview of the SPoRT MET scripts, instructions for how the scripts are run, and example use cases.

The Ensemble Space Weather Modeling System (eSWMS): Status, Capabilities and Challenges

NASA Astrophysics Data System (ADS)

Fry, C. D.; Eccles, J. V.; Reich, J. P.

2010-12-01

Marking a milestone in space weather forecasting, the Space Weather Modeling System (SWMS) successfully completed validation testing in advance of operational testing at Air Force Weather Agency’s primary space weather production center. This is the first coupling of stand-alone, physics-based space weather models that are currently in operations at AFWA supporting the warfighter. Significant development effort went into ensuring the component models were portable and scalable while maintaining consistent results across diverse high performance computing platforms. Coupling was accomplished under the Earth System Modeling Framework (ESMF). The coupled space weather models are the Hakamada-Akasofu-Fry version 2 (HAFv2) solar wind model and GAIM1, the ionospheric forecast component of the Global Assimilation of Ionospheric Measurements (GAIM) model. The SWMS was developed by team members from AFWA, Explorations Physics International, Inc. (EXPI) and Space Environment Corporation (SEC). The successful development of the SWMS provides new capabilities beyond enabling extended lead-time, data-driven ionospheric forecasts. These include ingesting diverse data sets at higher resolution, incorporating denser computational grids at finer time steps, and performing probability-based ensemble forecasts. Work of the SWMS development team now focuses on implementing the ensemble-based probability forecast capability by feeding multiple scenarios of 5 days of solar wind forecasts to the GAIM1 model based on the variation of the input fields to the HAFv2 model. The ensemble SWMS (eSWMS) will provide the most-likely space weather scenario with uncertainty estimates for important forecast fields. The eSWMS will allow DoD mission planners to consider the effects of space weather on their systems with more advance warning than is currently possible. The payoff is enhanced, tailored support to the warfighter with improved capabilities, such as point-to-point HF propagation forecasts, single-frequency GPS error corrections, and high cadence, high-resolution Space Situational Awareness (SSA) products. We present the current status of eSWMS, its capabilities, limitations and path of transition to operational use.

Measuring the effectiveness of earthquake forecasting in insurance strategies

NASA Astrophysics Data System (ADS)

Mignan, A.; Muir-Wood, R.

2009-04-01

Given the difficulty of judging whether the skill of a particular methodology of earthquake forecasts is offset by the inevitable false alarms and missed predictions, it is important to find a means to weigh the successes and failures according to a common currency. Rather than judge subjectively the relative costs and benefits of predictions, we develop a simple method to determine if the use of earthquake forecasts can increase the profitability of active financial risk management strategies employed in standard insurance procedures. Three types of risk management transactions are employed: (1) insurance underwriting, (2) reinsurance purchasing and (3) investment in CAT bonds. For each case premiums are collected based on modelled technical risk costs and losses are modelled for the portfolio in force at the time of the earthquake. A set of predetermined actions follow from the announcement of any change in earthquake hazard, so that, for each earthquake forecaster, the financial performance of an active risk management strategy can be compared with the equivalent passive strategy in which no notice is taken of earthquake forecasts. Overall performance can be tracked through time to determine which strategy gives the best long term financial performance. This will be determined by whether the skill in forecasting the location and timing of a significant earthquake (where loss is avoided) is outweighed by false predictions (when no premium is collected). This methodology is to be tested in California, where catastrophe modeling is reasonably mature and where a number of researchers issue earthquake forecasts.

Performance of stochastic approaches for forecasting river water quality.

PubMed

Ahmad, S; Khan, I H; Parida, B P

2001-12-01

This study analysed water quality data collected from the river Ganges in India from 1981 to 1990 for forecasting using stochastic models. Initially the box and whisker plots and Kendall's tau test were used to identify the trends during the study period. For detecting the possible intervention in the data the time series plots and cusum charts were used. The three approaches of stochastic modelling which account for the effect of seasonality in different ways. i.e. multiplicative autoregressive integrated moving average (ARIMA) model. deseasonalised model and Thomas-Fiering model were used to model the observed pattern in water quality. The multiplicative ARIMA model having both nonseasonal and seasonal components were, in general, identified as appropriate models. In the deseasonalised modelling approach, the lower order ARIMA models were found appropriate for the stochastic component. The set of Thomas-Fiering models were formed for each month for all water quality parameters. These models were then used to forecast the future values. The error estimates of forecasts from the three approaches were compared to identify the most suitable approach for the reliable forecast. The deseasonalised modelling approach was recommended for forecasting of water quality parameters of a river.

An Assessment of the Subseasonal Forecast Performance in the Extended Global Ensemble Forecast System (GEFS)

NASA Astrophysics Data System (ADS)

Sinsky, E.; Zhu, Y.; Li, W.; Guan, H.; Melhauser, C.

2017-12-01

Optimal forecast quality is crucial for the preservation of life and property. Improving monthly forecast performance over both the tropics and extra-tropics requires attention to various physical aspects such as the representation of the underlying SST, model physics and the representation of the model physics uncertainty for an ensemble forecast system. This work focuses on the impact of stochastic physics, SST and the convection scheme on forecast performance for the sub-seasonal scale over the tropics and extra-tropics with emphasis on the Madden-Julian Oscillation (MJO). A 2-year period is evaluated using the National Centers for Environmental Prediction (NCEP) Global Ensemble Forecast System (GEFS). Three experiments with different configurations than the operational GEFS were performed to illustrate the impact of the stochastic physics, SST and convection scheme. These experiments are compared against a control experiment (CTL) which consists of the operational GEFS but its integration is extended from 16 to 35 days. The three configurations are: 1) SPs, which uses a Stochastically Perturbed Physics Tendencies (SPPT), Stochastic Perturbed Humidity (SHUM) and Stochastic Kinetic Energy Backscatter (SKEB); 2) SPs+SST_bc, which uses a combination of SPs and a bias-corrected forecast SST from the NCEP Climate Forecast System Version 2 (CFSv2); and 3) SPs+SST_bc+SA_CV, which combines SPs, a bias-corrected forecast SST and a scale aware convection scheme. When comparing to the CTL experiment, SPs shows substantial improvement. The MJO skill has improved by about 4 lead days during the 2-year period. Improvement is also seen over the extra-tropics due to the updated stochastic physics, where there is a 3.1% and a 4.2% improvement during weeks 3 and 4 over the northern hemisphere and southern hemisphere, respectively. Improvement is also seen when the bias-corrected CFSv2 SST is combined with SPs. Additionally, forecast performance enhances when the scale aware convection scheme (SPs+SST_bc+SA_CV) is added, especially over the tropics. Among the three experiments, the SPs+SST_bc+SA_CV is the best configuration in MJO forecast skill.

Forecasting United States heartworm Dirofilaria immitis prevalence in dogs.

PubMed

Bowman, Dwight D; Liu, Yan; McMahan, Christopher S; Nordone, Shila K; Yabsley, Michael J; Lund, Robert B

2016-10-10

This paper forecasts next year's canine heartworm prevalence in the United States from 16 climate, geographic and societal factors. The forecast's construction and an assessment of its performance are described. The forecast is based on a spatial-temporal conditional autoregressive model fitted to over 31 million antigen heartworm tests conducted in the 48 contiguous United States during 2011-2015. The forecast uses county-level data on 16 predictive factors, including temperature, precipitation, median household income, local forest and surface water coverage, and presence/absence of eight mosquito species. Non-static factors are extrapolated into the forthcoming year with various statistical methods. The fitted model and factor extrapolations are used to estimate next year's regional prevalence. The correlation between the observed and model-estimated county-by-county heartworm prevalence for the 5-year period 2011-2015 is 0.727, demonstrating reasonable model accuracy. The correlation between 2015 observed and forecasted county-by-county heartworm prevalence is 0.940, demonstrating significant skill and showing that heartworm prevalence can be forecasted reasonably accurately. The forecast presented herein can a priori alert veterinarians to areas expected to see higher than normal heartworm activity. The proposed methods may prove useful for forecasting other diseases.

Two-step forecast of geomagnetic storm using coronal mass ejection and solar wind condition

PubMed Central

Kim, R-S; Moon, Y-J; Gopalswamy, N; Park, Y-D; Kim, Y-H

2014-01-01

To forecast geomagnetic storms, we had examined initially observed parameters of coronal mass ejections (CMEs) and introduced an empirical storm forecast model in a previous study. Now we suggest a two-step forecast considering not only CME parameters observed in the solar vicinity but also solar wind conditions near Earth to improve the forecast capability. We consider the empirical solar wind criteria derived in this study (Bz ≤ −5 nT or Ey ≥ 3 mV/m for t≥ 2 h for moderate storms with minimum Dst less than −50 nT) and a Dst model developed by Temerin and Li (2002, 2006) (TL model). Using 55 CME-Dst pairs during 1997 to 2003, our solar wind criteria produce slightly better forecasts for 31 storm events (90%) than the forecasts based on the TL model (87%). However, the latter produces better forecasts for 24 nonstorm events (88%), while the former correctly forecasts only 71% of them. We then performed the two-step forecast. The results are as follows: (i) for 15 events that are incorrectly forecasted using CME parameters, 12 cases (80%) can be properly predicted based on solar wind conditions; (ii) if we forecast a storm when both CME and solar wind conditions are satisfied (∩), the critical success index becomes higher than that from the forecast using CME parameters alone, however, only 25 storm events (81%) are correctly forecasted; and (iii) if we forecast a storm when either set of these conditions is satisfied (∪), all geomagnetic storms are correctly forecasted. PMID:26213515

Two-step forecast of geomagnetic storm using coronal mass ejection and solar wind condition.

PubMed

Kim, R-S; Moon, Y-J; Gopalswamy, N; Park, Y-D; Kim, Y-H

2014-04-01

To forecast geomagnetic storms, we had examined initially observed parameters of coronal mass ejections (CMEs) and introduced an empirical storm forecast model in a previous study. Now we suggest a two-step forecast considering not only CME parameters observed in the solar vicinity but also solar wind conditions near Earth to improve the forecast capability. We consider the empirical solar wind criteria derived in this study ( B z  ≤ -5 nT or E y  ≥ 3 mV/m for t ≥ 2 h for moderate storms with minimum Dst less than -50 nT) and a Dst model developed by Temerin and Li (2002, 2006) (TL model). Using 55 CME- Dst pairs during 1997 to 2003, our solar wind criteria produce slightly better forecasts for 31 storm events (90%) than the forecasts based on the TL model (87%). However, the latter produces better forecasts for 24 nonstorm events (88%), while the former correctly forecasts only 71% of them. We then performed the two-step forecast. The results are as follows: (i) for 15 events that are incorrectly forecasted using CME parameters, 12 cases (80%) can be properly predicted based on solar wind conditions; (ii) if we forecast a storm when both CME and solar wind conditions are satisfied (∩), the critical success index becomes higher than that from the forecast using CME parameters alone, however, only 25 storm events (81%) are correctly forecasted; and (iii) if we forecast a storm when either set of these conditions is satisfied (∪), all geomagnetic storms are correctly forecasted.

NASA Products to Enhance Energy Utility Load Forecasting

NASA Technical Reports Server (NTRS)

Lough, G.; Zell, E.; Engel-Cox, J.; Fungard, Y.; Jedlovec, G.; Stackhouse, P.; Homer, R.; Biley, S.

2012-01-01

Existing energy load forecasting tools rely upon historical load and forecasted weather to predict load within energy company service areas. The shortcomings of load forecasts are often the result of weather forecasts that are not at a fine enough spatial or temporal resolution to capture local-scale weather events. This project aims to improve the performance of load forecasting tools through the integration of high-resolution, weather-related NASA Earth Science Data, such as temperature, relative humidity, and wind speed. Three companies are participating in operational testing one natural gas company, and two electric providers. Operational results comparing load forecasts with and without NASA weather forecasts have been generated since March 2010. We have worked with end users at the three companies to refine selection of weather forecast information and optimize load forecast model performance. The project will conclude in 2012 with transitioning documented improvements from the inclusion of NASA forecasts for sustained use by energy utilities nationwide in a variety of load forecasting tools. In addition, Battelle has consulted with energy companies nationwide to document their information needs for long-term planning, in light of climate change and regulatory impacts.

Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networks.

PubMed

Vlachas, Pantelis R; Byeon, Wonmin; Wan, Zhong Y; Sapsis, Themistoklis P; Koumoutsakos, Petros

2018-05-01

We introduce a data-driven forecasting method for high-dimensional chaotic systems using long short-term memory (LSTM) recurrent neural networks. The proposed LSTM neural networks perform inference of high-dimensional dynamical systems in their reduced order space and are shown to be an effective set of nonlinear approximators of their attractor. We demonstrate the forecasting performance of the LSTM and compare it with Gaussian processes (GPs) in time series obtained from the Lorenz 96 system, the Kuramoto-Sivashinsky equation and a prototype climate model. The LSTM networks outperform the GPs in short-term forecasting accuracy in all applications considered. A hybrid architecture, extending the LSTM with a mean stochastic model (MSM-LSTM), is proposed to ensure convergence to the invariant measure. This novel hybrid method is fully data-driven and extends the forecasting capabilities of LSTM networks.

Impact of single-point GPS integrated water vapor estimates on short-range WRF model forecasts over southern India

NASA Astrophysics Data System (ADS)

Kumar, Prashant; Gopalan, Kaushik; Shukla, Bipasha Paul; Shyam, Abhineet

2017-11-01

Specifying physically consistent and accurate initial conditions is one of the major challenges of numerical weather prediction (NWP) models. In this study, ground-based global positioning system (GPS) integrated water vapor (IWV) measurements available from the International Global Navigation Satellite Systems (GNSS) Service (IGS) station in Bangalore, India, are used to assess the impact of GPS data on NWP model forecasts over southern India. Two experiments are performed with and without assimilation of GPS-retrieved IWV observations during the Indian winter monsoon period (November-December, 2012) using a four-dimensional variational (4D-Var) data assimilation method. Assimilation of GPS data improved the model IWV analysis as well as the subsequent forecasts. There is a positive impact of ˜10 % over Bangalore and nearby regions. The Weather Research and Forecasting (WRF) model-predicted 24-h surface temperature forecasts have also improved when compared with observations. Small but significant improvements were found in the rainfall forecasts compared to control experiments.

A Space Weather Forecasting System with Multiple Satellites Based on a Self-Recognizing Network

PubMed Central

Tokumitsu, Masahiro; Ishida, Yoshiteru

2014-01-01

This paper proposes a space weather forecasting system at geostationary orbit for high-energy electron flux (>2 MeV). The forecasting model involves multiple sensors on multiple satellites. The sensors interconnect and evaluate each other to predict future conditions at geostationary orbit. The proposed forecasting model is constructed using a dynamic relational network for sensor diagnosis and event monitoring. The sensors of the proposed model are located at different positions in space. The satellites for solar monitoring equip with monitoring devices for the interplanetary magnetic field and solar wind speed. The satellites orbit near the Earth monitoring high-energy electron flux. We investigate forecasting for typical two examples by comparing the performance of two models with different numbers of sensors. We demonstrate the prediction by the proposed model against coronal mass ejections and a coronal hole. This paper aims to investigate a possibility of space weather forecasting based on the satellite network with in-situ sensing. PMID:24803190

A space weather forecasting system with multiple satellites based on a self-recognizing network.

PubMed

Tokumitsu, Masahiro; Ishida, Yoshiteru

2014-05-05

This paper proposes a space weather forecasting system at geostationary orbit for high-energy electron flux (>2 MeV). The forecasting model involves multiple sensors on multiple satellites. The sensors interconnect and evaluate each other to predict future conditions at geostationary orbit. The proposed forecasting model is constructed using a dynamic relational network for sensor diagnosis and event monitoring. The sensors of the proposed model are located at different positions in space. The satellites for solar monitoring equip with monitoring devices for the interplanetary magnetic field and solar wind speed. The satellites orbit near the Earth monitoring high-energy electron flux. We investigate forecasting for typical two examples by comparing the performance of two models with different numbers of sensors. We demonstrate the prediction by the proposed model against coronal mass ejections and a coronal hole. This paper aims to investigate a possibility of space weather forecasting based on the satellite network with in-situ sensing.

Prospective Evaluation of the Global Earthquake Activity Rate Model (GEAR1) Earthquake Forecast: Preliminary Results

NASA Astrophysics Data System (ADS)

Strader, Anne; Schorlemmer, Danijel; Beutin, Thomas

2017-04-01

The Global Earthquake Activity Rate Model (GEAR1) is a hybrid seismicity model, constructed from a loglinear combination of smoothed seismicity from the Global Centroid Moment Tensor (CMT) earthquake catalog and geodetic strain rates (Global Strain Rate Map, version 2.1). For the 2005-2012 retrospective evaluation period, GEAR1 outperformed both parent strain rate and smoothed seismicity forecasts. Since 1. October 2015, GEAR1 has been prospectively evaluated by the Collaboratory for the Study of Earthquake Predictability (CSEP) testing center. Here, we present initial one-year test results of the GEAR1, GSRM and GSRM2.1, as well as localized evaluation of GEAR1 performance. The models were evaluated on the consistency in number (N-test), spatial (S-test) and magnitude (M-test) distribution of forecasted and observed earthquakes, as well as overall data consistency (CL-, L-tests). Performance at target earthquake locations was compared between models using the classical paired T-test and its non-parametric equivalent, the W-test, to determine if one model could be rejected in favor of another at the 0.05 significance level. For the evaluation period from 1. October 2015 to 1. October 2016, the GEAR1, GSRM and GSRM2.1 forecasts pass all CSEP likelihood tests. Comparative test results show statistically significant improvement of GEAR1 performance over both strain rate-based forecasts, both of which can be rejected in favor of GEAR1. Using point process residual analysis, we investigate the spatial distribution of differences in GEAR1, GSRM and GSRM2 model performance, to identify regions where the GEAR1 model should be adjusted, that could not be inferred from CSEP test results. Furthermore, we investigate whether the optimal combination of smoothed seismicity and strain rates remains stable over space and time.

Interevent times in a new alarm-based earthquake forecasting model

NASA Astrophysics Data System (ADS)

Talbi, Abdelhak; Nanjo, Kazuyoshi; Zhuang, Jiancang; Satake, Kenji; Hamdache, Mohamed

2013-09-01

This study introduces a new earthquake forecasting model that uses the moment ratio (MR) of the first to second order moments of earthquake interevent times as a precursory alarm index to forecast large earthquake events. This MR model is based on the idea that the MR is associated with anomalous long-term changes in background seismicity prior to large earthquake events. In a given region, the MR statistic is defined as the inverse of the index of dispersion or Fano factor, with MR values (or scores) providing a biased estimate of the relative regional frequency of background events, here termed the background fraction. To test the forecasting performance of this proposed MR model, a composite Japan-wide earthquake catalogue for the years between 679 and 2012 was compiled using the Japan Meteorological Agency catalogue for the period between 1923 and 2012, and the Utsu historical seismicity records between 679 and 1922. MR values were estimated by sampling interevent times from events with magnitude M ≥ 6 using an earthquake random sampling (ERS) algorithm developed during previous research. Three retrospective tests of M ≥ 7 target earthquakes were undertaken to evaluate the long-, intermediate- and short-term performance of MR forecasting, using mainly Molchan diagrams and optimal spatial maps obtained by minimizing forecasting error defined by miss and alarm rate addition. This testing indicates that the MR forecasting technique performs well at long-, intermediate- and short-term. The MR maps produced during long-term testing indicate significant alarm levels before 15 of the 18 shallow earthquakes within the testing region during the past two decades, with an alarm region covering about 20 per cent (alarm rate) of the testing region. The number of shallow events missed by forecasting was reduced by about 60 per cent after using the MR method instead of the relative intensity (RI) forecasting method. At short term, our model succeeded in forecasting the occurrence region of the 2011 Mw 9.0 Tohoku earthquake, whereas the RI method did not. Cases where a period of quiescent seismicity occurred before the target event often lead to low MR scores, meaning that the target event was not predicted and indicating that our model could be further improved by taking into account quiescent periods in the alarm strategy.

Next Day Price Forecasting in Deregulated Market by Combination of Artificial Neural Network and ARIMA Time Series Models

NASA Astrophysics Data System (ADS)

Areekul, Phatchakorn; Senjyu, Tomonobu; Urasaki, Naomitsu; Yona, Atsushi

Electricity price forecasting is becoming increasingly relevant to power producers and consumers in the new competitive electric power markets, when planning bidding strategies in order to maximize their benefits and utilities, respectively. This paper proposed a method to predict hourly electricity prices for next-day electricity markets by combination methodology of ARIMA and ANN models. The proposed method is examined on the Australian National Electricity Market (NEM), New South Wales regional in year 2006. Comparison of forecasting performance with the proposed ARIMA, ANN and combination (ARIMA-ANN) models are presented. Empirical results indicate that an ARIMA-ANN model can improve the price forecasting accuracy.

Should we use seasonnal meteorological ensemble forecasts for hydrological forecasting? A case study for nordic watersheds in Canada.

NASA Astrophysics Data System (ADS)

Bazile, Rachel; Boucher, Marie-Amélie; Perreault, Luc; Leconte, Robert; Guay, Catherine

2017-04-01

Hydro-electricity is a major source of energy for many countries throughout the world, including Canada. Long lead-time streamflow forecasts are all the more valuable as they help decision making and dam management. Different techniques exist for long-term hydrological forecasting. Perhaps the most well-known is 'Extended Streamflow Prediction' (ESP), which considers past meteorological scenarios as possible, often equiprobable, future scenarios. In the ESP framework, those past-observed meteorological scenarios (climatology) are used in turn as the inputs of a chosen hydrological model to produce ensemble forecasts (one member corresponding to each year in the available database). Many hydropower companies, including Hydro-Québec (province of Quebec, Canada) use variants of the above described ESP system operationally for long-term operation planning. The ESP system accounts for the hydrological initial conditions and for the natural variability of the meteorological variables. However, it cannot consider the current initial state of the atmosphere. Climate models can help remedy this drawback. In the context of a changing climate, dynamical forecasts issued from climate models seem to be an interesting avenue to improve upon the ESP method and could help hydropower companies to adapt their management practices to an evolving climate. Long-range forecasts from climate models can also be helpful for water management at locations where records of past meteorological conditions are short or nonexistent. In this study, we compare 7-month hydrological forecasts obtained from climate model outputs to an ESP system. The ESP system mimics the one used operationally at Hydro-Québec. The dynamical climate forecasts are produced by the European Center for Medium range Weather Forecasts (ECMWF) System4. Forecasts quality is assessed using numerical scores such as the Continuous Ranked Probability Score (CRPS) and the Ignorance score and also graphical tools such as the reliability diagram. This study covers 10 nordic watersheds. We show that forecast performance according to the CRPS varies with lead-time but also with the period of the year. The raw forecasts from the ECMWF System4 display important biases for both temperature and precipitation, which need to be corrected. The linear scaling method is used for this purpose and is found effective. Bias correction improves forecasts performance, especially during the summer when the precipitations are over-estimated. According to the CRPS, bias corrected forecasts from System4 show performances comparable to those of the ESP system. However, the Ignorance score, which penalizes the lack of calibration (under-dispersive forecasts in this case) more severely than the CRPS, provides a different outlook for the comparison of the two systems. In fact, according to the Ignorance score, the ESP system outperforms forecasts based on System4 in most cases. This illustrates that the joint use of several metrics is crucial to assess the quality of a forecasts system thoroughly. Globally, ESP provide reliable forecasts which can be over-dispersed whereas bias corrected ECMWF System4 forecasts are sharper but at the risk of missing events.

Bayesian flood forecasting methods: A review

NASA Astrophysics Data System (ADS)

Han, Shasha; Coulibaly, Paulin

2017-08-01

Over the past few decades, floods have been seen as one of the most common and largely distributed natural disasters in the world. If floods could be accurately forecasted in advance, then their negative impacts could be greatly minimized. It is widely recognized that quantification and reduction of uncertainty associated with the hydrologic forecast is of great importance for flood estimation and rational decision making. Bayesian forecasting system (BFS) offers an ideal theoretic framework for uncertainty quantification that can be developed for probabilistic flood forecasting via any deterministic hydrologic model. It provides suitable theoretical structure, empirically validated models and reasonable analytic-numerical computation method, and can be developed into various Bayesian forecasting approaches. This paper presents a comprehensive review on Bayesian forecasting approaches applied in flood forecasting from 1999 till now. The review starts with an overview of fundamentals of BFS and recent advances in BFS, followed with BFS application in river stage forecasting and real-time flood forecasting, then move to a critical analysis by evaluating advantages and limitations of Bayesian forecasting methods and other predictive uncertainty assessment approaches in flood forecasting, and finally discusses the future research direction in Bayesian flood forecasting. Results show that the Bayesian flood forecasting approach is an effective and advanced way for flood estimation, it considers all sources of uncertainties and produces a predictive distribution of the river stage, river discharge or runoff, thus gives more accurate and reliable flood forecasts. Some emerging Bayesian forecasting methods (e.g. ensemble Bayesian forecasting system, Bayesian multi-model combination) were shown to overcome limitations of single model or fixed model weight and effectively reduce predictive uncertainty. In recent years, various Bayesian flood forecasting approaches have been developed and widely applied, but there is still room for improvements. Future research in the context of Bayesian flood forecasting should be on assimilation of various sources of newly available information and improvement of predictive performance assessment methods.

A seasonal hydrologic ensemble prediction system for water resource management

NASA Astrophysics Data System (ADS)

Luo, L.; Wood, E. F.

2006-12-01

A seasonal hydrologic ensemble prediction system, developed for the Ohio River basin, has been improved and expanded to several other regions including the Eastern U.S., Africa and East Asia. The prediction system adopts the traditional Extended Streamflow Prediction (ESP) approach, utilizing the VIC (Variable Infiltration Capacity) hydrological model as the central tool for producing ensemble prediction of soil moisture, snow and streamflow with lead times up to 6-month. VIC is forced by observed meteorology to estimate the hydrological initial condition prior to the forecast, but during the forecast period the atmospheric forcing comes from statistically downscaled, seasonal forecast from dynamic climate models. The seasonal hydrologic ensemble prediction system is currently producing realtime seasonal hydrologic forecast for these regions on a monthly basis. Using hindcasts from a 19-year period (1981-1999), during which seasonal hindcasts from NCEP Climate Forecast System (CFS) and European Union DEMETER project are available, we evaluate the performance of the forecast system over our forecast regions. The evaluation shows that the prediction system using the current forecast approach is able to produce reliable and accurate precipitation, soil moisture and streamflow predictions. The overall skill is much higher then the traditional ESP. In particular, forecasts based on multiple climate model forecast are more skillful than single model-based forecast. This emphasizes the significant need for producing seasonal climate forecast with multiple climate models for hydrologic applications. Forecast from this system is expected to provide very valuable information about future hydrologic states and associated risks for end users, including water resource management and financial sectors.

Prospective forecasts of annual dengue hemorrhagic fever incidence in Thailand, 2010–2014

PubMed Central

Sakrejda, Krzysztof; Ray, Evan L.; Keegan, Lindsay T.; Bi, Qifang; Suangtho, Paphanij; Hinjoy, Soawapak; Iamsirithaworn, Sopon; Suthachana, Suthanun; Laosiritaworn, Yongjua; Lessler, Justin; Reich, Nicholas G.

2018-01-01

Dengue hemorrhagic fever (DHF), a severe manifestation of dengue viral infection that can cause severe bleeding, organ impairment, and even death, affects between 15,000 and 105,000 people each year in Thailand. While all Thai provinces experience at least one DHF case most years, the distribution of cases shifts regionally from year to year. Accurately forecasting where DHF outbreaks occur before the dengue season could help public health officials prioritize public health activities. We develop statistical models that use biologically plausible covariates, observed by April each year, to forecast the cumulative DHF incidence for the remainder of the year. We perform cross-validation during the training phase (2000–2009) to select the covariates for these models. A parsimonious model based on preseason incidence outperforms the 10-y median for 65% of province-level annual forecasts, reduces the mean absolute error by 19%, and successfully forecasts outbreaks (area under the receiver operating characteristic curve = 0.84) over the testing period (2010–2014). We find that functions of past incidence contribute most strongly to model performance, whereas the importance of environmental covariates varies regionally. This work illustrates that accurate forecasts of dengue risk are possible in a policy-relevant timeframe. PMID:29463757

Prospective forecasts of annual dengue hemorrhagic fever incidence in Thailand, 2010-2014.

PubMed

Lauer, Stephen A; Sakrejda, Krzysztof; Ray, Evan L; Keegan, Lindsay T; Bi, Qifang; Suangtho, Paphanij; Hinjoy, Soawapak; Iamsirithaworn, Sopon; Suthachana, Suthanun; Laosiritaworn, Yongjua; Cummings, Derek A T; Lessler, Justin; Reich, Nicholas G

2018-03-06

Dengue hemorrhagic fever (DHF), a severe manifestation of dengue viral infection that can cause severe bleeding, organ impairment, and even death, affects between 15,000 and 105,000 people each year in Thailand. While all Thai provinces experience at least one DHF case most years, the distribution of cases shifts regionally from year to year. Accurately forecasting where DHF outbreaks occur before the dengue season could help public health officials prioritize public health activities. We develop statistical models that use biologically plausible covariates, observed by April each year, to forecast the cumulative DHF incidence for the remainder of the year. We perform cross-validation during the training phase (2000-2009) to select the covariates for these models. A parsimonious model based on preseason incidence outperforms the 10-y median for 65% of province-level annual forecasts, reduces the mean absolute error by 19%, and successfully forecasts outbreaks (area under the receiver operating characteristic curve = 0.84) over the testing period (2010-2014). We find that functions of past incidence contribute most strongly to model performance, whereas the importance of environmental covariates varies regionally. This work illustrates that accurate forecasts of dengue risk are possible in a policy-relevant timeframe. Copyright © 2018 the Author(s). Published by PNAS.

Evaluation of Wind Power Forecasts from the Vermont Weather Analytics Center and Identification of Improvements

DOE Office of Scientific and Technical Information (OSTI.GOV)

Optis, Michael; Scott, George N.; Draxl, Caroline

The goal of this analysis was to assess the wind power forecast accuracy of the Vermont Weather Analytics Center (VTWAC) forecast system and to identify potential improvements to the forecasts. Based on the analysis at Georgia Mountain, the following recommendations for improving forecast performance were made: 1. Resolve the significant negative forecast bias in February-March 2017 (50% underprediction on average) 2. Improve the ability of the forecast model to capture the strong diurnal cycle of wind power 3. Add ability for forecast model to assess internal wake loss, particularly at sites where strong diurnal shifts in wind direction are present.more » Data availability and quality limited the robustness of this forecast assessment. A more thorough analysis would be possible given a longer period of record for the data (at least one full year), detailed supervisory control and data acquisition data for each wind plant, and more detailed information on the forecast system input data and methodologies.« less

Towards a better knowledge of flash flood forecasting at the Three Gorges Region: Progress over the past decade and challenges ahead

NASA Astrophysics Data System (ADS)

Li, Zhe; Yang, Dawen; Yang, Hanbo; Wu, Tianjiao; Xu, Jijun; Gao, Bing; Xu, Tao

2015-04-01

The study area, the Three Gorges Region (TGR), plays a critical role in predicting the floods drained into the Three Gorges Reservoir, as reported local floods often exceed 10000m3/s during rainstorm events and trigger fast as well as significant impacts on the Three Gorges Reservoir's regulation. Meanwhile, it is one of typical mountainous areas in China, which is located in the transition zone between two monsoon systems: the East Asian monsoon and the South Asian (Indian) monsoon. This climatic feature, combined with local irregular terrains, has shaped complicated rainfall-runoff regimes in this focal region. However, due to the lack of high-resolution hydrometeorological data and physically-based hydrologic modeling framework, there was little knowledge about rainfall variability and flood pattern in this historically ungauged region, which posed great uncertainties to flash flood forecasting in the past. The present study summarize latest progresses of regional flash floods monitoring and prediction, including installation of a ground-based Hydrometeorological Observation Network (TGR-HMON), application of a regional geomorphology-based hydrological model (TGR-GBHM), development of an integrated forecasting and modeling system (TGR-INFORMS), and evaluation of quantitative precipitation estimations (QPE) and quantitative precipitation forecasting (QPF) products in TGR flash flood forecasting. With these continuing efforts to improve the forecasting performance of flash floods in TGR, we have addressed several critical issues: (1) Current observation network is still insufficient to capture localized rainstorms, and weather radar provides valuable information to forecast flash floods induced by localized rainstorms, although current radar QPE products can be improved substantially in future; (2) Long-term evaluation shows that the geomorphology-based distributed hydrologic model (GBHM) is able to simulate flash flooding processes reasonably, while model performance will decline at hourly scale with larger uncertainties. However, model comparison suggests that this physically-based distributed model (GBHM), compared with a traditional lumped model (Xin'anjiang model), shows more robust performance and larger transferability for prediction in those ungauged basins in TGR; (3) Operational test of our integrated forecasting system (TRG-INFORMS) shows that it works reasonably to simulate the flood routing in Three Gorges reservoir, indicating the accuracy of simulation of total floods generated at region scale; (4) Current operational QPF is too coarse to provide valuable information even for flood forecasting of whole TGR, thus, downscaling and high-resolution QPF are necessary to unravel the potentials of weather forecasting. Finally, according to these results, we also discuss about some possible solutions with high priority for future advanced forecasting scheme of local flash floods in TGR.

Reconstructing Common Era Climate Fields Using Data Assimilation, Proxies, and a Linear Climate Model

NASA Astrophysics Data System (ADS)

Perkins, W. A.; Hakim, G. J.

2016-12-01

In this work, we examine the skill of a new approach to performing climate field reconstructions (CFRs) using a form of online paleoclimate data assimilation (PDA). Many previous studies have foregone climate model forecasts during assimilation due to the computational expense of running coupled global climate models (CGCMs), and the relatively low skill of these forecasts on longer timescales. Here we greatly diminish the computational costs by employing an empirical forecast model (known as a linear inverse model; LIM), which has been shown to have comparable skill to CGCMs. CFRs of annually averaged 2m air temperature anomalies are compared between the Last Millennium Reanalysis framework (no forecasting or "offline"), a persistence forecast, and four LIM forecasting experiments over the instrumental period (1850 - 2000). We test LIM calibrations for observational (Berkeley Earth), reanalysis (20th Century Reanalysis), and CMIP5 climate model (CCSM4 and MPI) data. Generally, we find that the usage of LIM forecasts for online PDA increases reconstruction agreement with the instrumental record for both spatial and global mean temperature (GMT). The detrended GMT skill metrics show the most dramatic increases in skill with coefficient of efficiency (CE) improvements over the no-forecasting benchmark averaging 57%. LIM experiments display a common pattern of spatial field increases in CE skill over northern hemisphere land areas and in the high-latitude North Atlantic - Barents Sea corridor (Figure 1). However, the non-GCM-calibrated LIMs introduce other deficiencies into the spatial skill of these reconstructions, likely due to aspects of the LIM calibration process. Overall, the CMIP5 LIMs have the best performance when considering both spatial fields and GMT. A comparison with the persistence forecast experiment suggests that improvements are associated with the usage of the LIM forecasts, and not simple persistence of temperature anomalies over time. These results show that the use of LIM forecasting can help add further dynamical constraint to CFRs. As we move forward, this will be an important factor in fully utilizing dynamically consistent information from the proxy record while reconstructing the past millennium.

Forecasting monthly inflow discharge of the Iffezheim reservoir using data-driven models

NASA Astrophysics Data System (ADS)

Zhang, Qing; Aljoumani, Basem; Hillebrand, Gudrun; Hoffmann, Thomas; Hinkelmann, Reinhard

2017-04-01

River stream flow is an essential element in hydrology study fields, especially for reservoir management, since it defines input into reservoirs. Forecasting this stream flow plays an important role in short or long-term planning and management in the reservoir, e.g. optimized reservoir and hydroelectric operation or agricultural irrigation. Highly accurate flow forecasting can significantly reduce economic losses and is always pursued by reservoir operators. Therefore, hydrologic time series forecasting has received tremendous attention of researchers. Many models have been proposed to improve the hydrological forecasting. Due to the fact that most natural phenomena occurring in environmental systems appear to behave in random or probabilistic ways, different cases may need a different methods to forecast the inflow and even a unique treatment to improve the forecast accuracy. The purpose of this study is to determine an appropriate model for forecasting monthly inflow to the Iffezheim reservoir in Germany, which is the last of the barrages in the Upper Rhine. Monthly time series of discharges, measured from 1946 to 2001 at the Plittersdorf station, which is located 6 km downstream of the Iffezheim reservoir, were applied. The accuracies of the used stochastic models - Fiering model and Auto-Regressive Integrated Moving Average models (ARIMA) are compared with Artificial Intelligence (AI) models - single Artificial Neural Network (ANN) and Wavelet ANN models (WANN). The Fiering model is a linear stochastic model and used for generating synthetic monthly data. The basic idea in modeling time series using ARIMA is to identify a simple model with as few model parameters as possible in order to provide a good statistical fit to the data. To identify and fit the ARIMA models, four phase approaches were used: identification, parameter estimation, diagnostic checking, and forecasting. An automatic selection criterion, such as the Akaike information criterion, is utilized to enhance this flexible approach to set up the model. As distinct from both stochastic models, the ANN and its related conjunction methods Wavelet-ANN (WANN) models are effective to handle non-linear systems and have been developed with antecedent flows as inputs to forecast up to 12-months lead-time for the Iffezheim reservoir. In the ANN and WANN models, the Feed Forward Back Propagation method (FFBP) is applied. The sigmoid activity and linear functions were used with several different neurons for the hidden layers and for the output layer, respectively. To compare the accuracy of the different models and identify the most suitable model for reliable forecasting, four quantitative standard statistical performance evaluation measures, the root mean square error (RMSE), the mean bias error (MAE) and the determination correlation coefficient (DC), are employed. The results reveal that the ARIMA (2, 1, 2) performs better than Fiering, ANN and WANN models. Further, the WANN model is found to be slightly better than the ANN model for forecasting monthly inflow of the Iffezheim reservoir. As a result, by using the ARIMA model, the predicted and observed values agree reasonably well.

Use of observational and model-derived fields and regime model output statistics in mesoscale forecasting

NASA Technical Reports Server (NTRS)

Forbes, G. S.; Pielke, R. A.

1985-01-01

Various empirical and statistical weather-forecasting studies which utilize stratification by weather regime are described. Objective classification was used to determine weather regime in some studies. In other cases the weather pattern was determined on the basis of a parameter representing the physical and dynamical processes relevant to the anticipated mesoscale phenomena, such as low level moisture convergence and convective precipitation, or the Froude number and the occurrence of cold-air damming. For mesoscale phenomena already in existence, new forecasting techniques were developed. The use of cloud models in operational forecasting is discussed. Models to calculate the spatial scales of forcings and resultant response for mesoscale systems are presented. The use of these models to represent the climatologically most prevalent systems, and to perform case-by-case simulations is reviewed. Operational implementation of mesoscale data into weather forecasts, using both actual simulation output and method-output statistics is discussed.

State space model approach for forecasting the use of electrical energy (a case study on: PT. PLN (Persero) district of Kroya)

NASA Astrophysics Data System (ADS)

Kurniati, Devi; Hoyyi, Abdul; Widiharih, Tatik

2018-05-01

Time series data is a series of data taken or measured based on observations at the same time interval. Time series data analysis is used to perform data analysis considering the effect of time. The purpose of time series analysis is to know the characteristics and patterns of a data and predict a data value in some future period based on data in the past. One of the forecasting methods used for time series data is the state space model. This study discusses the modeling and forecasting of electric energy consumption using the state space model for univariate data. The modeling stage is began with optimal Autoregressive (AR) order selection, determination of state vector through canonical correlation analysis, estimation of parameter, and forecasting. The result of this research shows that modeling of electric energy consumption using state space model of order 4 with Mean Absolute Percentage Error (MAPE) value 3.655%, so the model is very good forecasting category.

Adaptation of Mesoscale Weather Models to Local Forecasting

NASA Technical Reports Server (NTRS)

Manobianco, John T.; Taylor, Gregory E.; Case, Jonathan L.; Dianic, Allan V.; Wheeler, Mark W.; Zack, John W.; Nutter, Paul A.

2003-01-01

Methodologies have been developed for (1) configuring mesoscale numerical weather-prediction models for execution on high-performance computer workstations to make short-range weather forecasts for the vicinity of the Kennedy Space Center (KSC) and the Cape Canaveral Air Force Station (CCAFS) and (2) evaluating the performances of the models as configured. These methodologies have been implemented as part of a continuing effort to improve weather forecasting in support of operations of the U.S. space program. The models, methodologies, and results of the evaluations also have potential value for commercial users who could benefit from tailoring their operations and/or marketing strategies based on accurate predictions of local weather. More specifically, the purpose of developing the methodologies for configuring the models to run on computers at KSC and CCAFS is to provide accurate forecasts of winds, temperature, and such specific thunderstorm-related phenomena as lightning and precipitation. The purpose of developing the evaluation methodologies is to maximize the utility of the models by providing users with assessments of the capabilities and limitations of the models. The models used in this effort thus far include the Mesoscale Atmospheric Simulation System (MASS), the Regional Atmospheric Modeling System (RAMS), and the National Centers for Environmental Prediction Eta Model ( Eta for short). The configuration of the MASS and RAMS is designed to run the models at very high spatial resolution and incorporate local data to resolve fine-scale weather features. Model preprocessors were modified to incorporate surface, ship, buoy, and rawinsonde data as well as data from local wind towers, wind profilers, and conventional or Doppler radars. The overall evaluation of the MASS, Eta, and RAMS was designed to assess the utility of these mesoscale models for satisfying the weather-forecasting needs of the U.S. space program. The evaluation methodology includes objective and subjective verification methodologies. Objective (e.g., statistical) verification of point forecasts is a stringent measure of model performance, but when used alone, it is not usually sufficient for quantifying the value of the overall contribution of the model to the weather-forecasting process. This is especially true for mesoscale models with enhanced spatial and temporal resolution that may be capable of predicting meteorologically consistent, though not necessarily accurate, fine-scale weather phenomena. Therefore, subjective (phenomenological) evaluation, focusing on selected case studies and specific weather features, such as sea breezes and precipitation, has been performed to help quantify the added value that cannot be inferred solely from objective evaluation.

Status of Air Quality in Central California and Needs for Further Study

NASA Astrophysics Data System (ADS)

Tanrikulu, S.; Beaver, S.; Soong, S.; Tran, C.; Jia, Y.; Matsuoka, J.; McNider, R. T.; Biazar, A. P.; Palazoglu, A.; Lee, P.; Wang, J.; Kang, D.; Aneja, V. P.

2012-12-01

Ozone and PM2.5 levels frequently exceed NAAQS in central California (CC). Additional emission reductions are needed to attain and maintain the standards there. Agencies are developing cost-effective emission control strategies along with complementary incentive programs to reduce emissions when exceedances are forecasted. These approaches require accurate modeling and forecasting capabilities. A variety of models have been rigorously applied (MM5, WRF, CMAQ, CAMx) over CC. Despite the vast amount of land-based measurements from special field programs and significant effort, models have historically exhibited marginal performance. Satellite data may improve model performance by: establishing IC/BC over outlying areas of the modeling domain having unknown conditions; enabling FDDA over the Pacific Ocean to characterize important marine inflows and pollutant outflows; and filling in the gaps of the land-based monitoring network. BAAQMD, in collaboration with the NASA AQAST, plans to conduct four studies that include satellite-based data in CC air quality analysis and modeling: The first project enhances and refines weather patterns, especially aloft, impacting summer ozone formation. Surface analyses were unable to characterize the strong attenuating effect of the complex terrain to steer marine winds impinging on the continent. The dense summer clouds and fog over the Pacific Ocean form spatial patterns that can be related to the downstream air flows through polluted areas. The goal of this project is to explore, characterize, and quantify these relationships using cloud cover data. Specifically, cloud agreement statistics will be developed using satellite data and model clouds. Model skin temperature predictions will be compared to both MODIS and GOES skin temperatures. The second project evaluates and improves the initial and simulated fields of meteorological models that provide inputs to air quality models. The study will attempt to determine whether a cloud dynamical adjustment developed by UAHuntsville can improve model performance for maritime stratus and whether a moisture adjustment scheme in the Pleim-Xiu boundary layer scheme can use satellite data in place of coarse surface air temperature measurements. The goal is to improve meteorological model performance that leads to improved air quality model performance. The third project evaluates and improves forecasting skills of the National Air Quality Forecasting Model in CC by using land-based routine measurements as well as satellite data. Local forecasts are mostly based on surface meteorological and air quality measurements and weather charts provided by NWS. The goal is to improve the average accuracy in forecasting exceedances, which is around 60%. The fourth project uses satellite data for monitoring trends in fine particulate matter (PM2.5) in the San Francisco Bay Area. It evaluates the effectiveness of a rule adopted in 2008 that restricts household wood burning on days forecasted to have high PM2.5 levels. The goal is to complement current analyses based on surface data covering the largest sub-regions and population centers. The overall goal is to use satellite data to overcome limitations of land-based measurements. The outcomes will be further conceptual understanding of pollutant formation, improved regulatory model performance, and better optimized forecasting programs.

Forecasting SPEI and SPI Drought Indices Using the Integrated Artificial Neural Networks

PubMed Central

Maca, Petr; Pech, Pavel

2016-01-01

The presented paper compares forecast of drought indices based on two different models of artificial neural networks. The first model is based on feedforward multilayer perceptron, sANN, and the second one is the integrated neural network model, hANN. The analyzed drought indices are the standardized precipitation index (SPI) and the standardized precipitation evaporation index (SPEI) and were derived for the period of 1948–2002 on two US catchments. The meteorological and hydrological data were obtained from MOPEX experiment. The training of both neural network models was made by the adaptive version of differential evolution, JADE. The comparison of models was based on six model performance measures. The results of drought indices forecast, explained by the values of four model performance indices, show that the integrated neural network model was superior to the feedforward multilayer perceptron with one hidden layer of neurons. PMID:26880875

Forecasting SPEI and SPI Drought Indices Using the Integrated Artificial Neural Networks.

PubMed

Maca, Petr; Pech, Pavel

2016-01-01

The presented paper compares forecast of drought indices based on two different models of artificial neural networks. The first model is based on feedforward multilayer perceptron, sANN, and the second one is the integrated neural network model, hANN. The analyzed drought indices are the standardized precipitation index (SPI) and the standardized precipitation evaporation index (SPEI) and were derived for the period of 1948-2002 on two US catchments. The meteorological and hydrological data were obtained from MOPEX experiment. The training of both neural network models was made by the adaptive version of differential evolution, JADE. The comparison of models was based on six model performance measures. The results of drought indices forecast, explained by the values of four model performance indices, show that the integrated neural network model was superior to the feedforward multilayer perceptron with one hidden layer of neurons.

Forecasting the ocean optical environment in support of Navy mine warfare operations

NASA Astrophysics Data System (ADS)

Ladner, S. D.; Arnone, R.; Jolliff, J.; Casey, B.; Matulewski, K.

2012-06-01

A 3D ocean optical forecast system called TODS (Tactical Ocean Data System) has been developed to determine the performance of underwater LIDAR detection/identification systems. TODS fuses optical measurements from gliders, surface satellite optical properties, and 3D ocean forecast circulation models to extend the 2-dimensional surface satellite optics into a 3-dimensional optical volume including subsurface optical layers of beam attenuation coefficient (c) and diver visibility. Optical 3D nowcast and forecasts are combined with electro-optical identification (EOID) models to determine the underwater LIDAR imaging performance field used to identify subsurface mine threats in rapidly changing coastal regions. TODS was validated during a recent mine warfare exercise with Helicopter Mine Countermeasures Squadron (HM-14). Results include the uncertainties in the optical forecast and lidar performance and sensor tow height predictions that are based on visual detection and identification metrics using actual mine target images from the EOID system. TODS is a new capability of coupling the 3D optical environment and EOID system performance and is proving important for the MIW community as both a tactical decision aid and for use in operational planning, improving timeliness and efficiency in clearance operations.

The Past, Present and Future of the Meteorological Phenomena Identification Near the Ground (mPING) Project

NASA Astrophysics Data System (ADS)

Elmore, K. L.

2016-12-01

The Metorological Phenomemna Identification NeartheGround (mPING) project is an example of a crowd-sourced, citizen science effort to gather data of sufficeint quality and quantity needed by new post processing methods that use machine learning. Transportation and infrastructure are particularly sensitive to precipitation type in winter weather. We extract attributes from operational numerical forecast models and use them in a random forest to generate forecast winter precipitation types. We find that random forests applied to forecast soundings are effective at generating skillful forecasts of surface ptype with consideralbly more skill than the current algorithms, especuially for ice pellets and freezing rain. We also find that three very different forecast models yuield similar overall results, showing that random forests are able to extract essentially equivalent information from different forecast models. We also show that the random forest for each model, and each profile type is unique to the particular forecast model and that the random forests developed using a particular model suffer significant degradation when given attributes derived from a different model. This implies that no single algorithm can perform well across all forecast models. Clearly, random forests extract information unavailable to "physically based" methods because the physical information in the models does not appear as we expect. One intersting result is that results from the classic "warm nose" sounding profile are, by far, the most sensitive to the particular forecast model, but this profile is also the one for which random forests are most skillful. Finally, a method for calibrarting probabilties for each different ptype using multinomial logistic regression is shown.

The Super Tuesday Outbreak: Forecast Sensitivities to Single-Moment Microphysics Schemes

NASA Technical Reports Server (NTRS)

Molthan, Andrew L.; Case, Jonathan L.; Dembek, Scott R.; Jedlovec, Gary J.; Lapenta, William M.

2008-01-01

Forecast precipitation and radar characteristics are used by operational centers to guide the issuance of advisory products. As operational numerical weather prediction is performed at increasingly finer spatial resolution, convective precipitation traditionally represented by sub-grid scale parameterization schemes is now being determined explicitly through single- or multi-moment bulk water microphysics routines. Gains in forecasting skill are expected through improved simulation of clouds and their microphysical processes. High resolution model grids and advanced parameterizations are now available through steady increases in computer resources. As with any parameterization, their reliability must be measured through performance metrics, with errors noted and targeted for improvement. Furthermore, the use of these schemes within an operational framework requires an understanding of limitations and an estimate of biases so that forecasters and model development teams can be aware of potential errors. The National Severe Storms Laboratory (NSSL) Spring Experiments have produced daily, high resolution forecasts used to evaluate forecast skill among an ensemble with varied physical parameterizations and data assimilation techniques. In this research, high resolution forecasts of the 5-6 February 2008 Super Tuesday Outbreak are replicated using the NSSL configuration in order to evaluate two components of simulated convection on a large domain: sensitivities of quantitative precipitation forecasts to assumptions within a single-moment bulk water microphysics scheme, and to determine if these schemes accurately depict the reflectivity characteristics of well-simulated, organized, cold frontal convection. As radar returns are sensitive to the amount of hydrometeor mass and the distribution of mass among variably sized targets, radar comparisons may guide potential improvements to a single-moment scheme. In addition, object-based verification metrics are evaluated for their utility in gauging model performance and QPF variability.

Performance of time-series methods in forecasting the demand for red blood cell transfusion.

PubMed

Pereira, Arturo

2004-05-01

Planning the future blood collection efforts must be based on adequate forecasts of transfusion demand. In this study, univariate time-series methods were investigated for their performance in forecasting the monthly demand for RBCs at one tertiary-care, university hospital. Three time-series methods were investigated: autoregressive integrated moving average (ARIMA), the Holt-Winters family of exponential smoothing models, and one neural-network-based method. The time series consisted of the monthly demand for RBCs from January 1988 to December 2002 and was divided into two segments: the older one was used to fit or train the models, and the younger to test for the accuracy of predictions. Performance was compared across forecasting methods by calculating goodness-of-fit statistics, the percentage of months in which forecast-based supply would have met the RBC demand (coverage rate), and the outdate rate. The RBC transfusion series was best fitted by a seasonal ARIMA(0,1,1)(0,1,1)(12) model. Over 1-year time horizons, forecasts generated by ARIMA or exponential smoothing laid within the +/- 10 percent interval of the real RBC demand in 79 percent of months (62% in the case of neural networks). The coverage rate for the three methods was 89, 91, and 86 percent, respectively. Over 2-year time horizons, exponential smoothing largely outperformed the other methods. Predictions by exponential smoothing laid within the +/- 10 percent interval of real values in 75 percent of the 24 forecasted months, and the coverage rate was 87 percent. Over 1-year time horizons, predictions of RBC demand generated by ARIMA or exponential smoothing are accurate enough to be of help in the planning of blood collection efforts. For longer time horizons, exponential smoothing outperforms the other forecasting methods.

Intercomparison of air quality data using principal component analysis, and forecasting of PM₁₀ and PM₂.₅ concentrations using artificial neural networks, in Thessaloniki and Helsinki.

PubMed

Voukantsis, Dimitris; Karatzas, Kostas; Kukkonen, Jaakko; Räsänen, Teemu; Karppinen, Ari; Kolehmainen, Mikko

2011-03-01

In this paper we propose a methodology consisting of specific computational intelligence methods, i.e. principal component analysis and artificial neural networks, in order to inter-compare air quality and meteorological data, and to forecast the concentration levels for environmental parameters of interest (air pollutants). We demonstrate these methods to data monitored in the urban areas of Thessaloniki and Helsinki in Greece and Finland, respectively. For this purpose, we applied the principal component analysis method in order to inter-compare the patterns of air pollution in the two selected cities. Then, we proceeded with the development of air quality forecasting models for both studied areas. On this basis, we formulated and employed a novel hybrid scheme in the selection process of input variables for the forecasting models, involving a combination of linear regression and artificial neural networks (multi-layer perceptron) models. The latter ones were used for the forecasting of the daily mean concentrations of PM₁₀ and PM₂.₅ for the next day. Results demonstrated an index of agreement between measured and modelled daily averaged PM₁₀ concentrations, between 0.80 and 0.85, while the kappa index for the forecasting of the daily averaged PM₁₀ concentrations reached 60% for both cities. Compared with previous corresponding studies, these statistical parameters indicate an improved performance of air quality parameters forecasting. It was also found that the performance of the models for the forecasting of the daily mean concentrations of PM₁₀ was not substantially different for both cities, despite the major differences of the two urban environments under consideration. Copyright © 2011 Elsevier B.V. All rights reserved.

A comparison of performance of several artificial intelligence methods for forecasting monthly discharge time series

NASA Astrophysics Data System (ADS)

Wang, Wen-Chuan; Chau, Kwok-Wing; Cheng, Chun-Tian; Qiu, Lin

2009-08-01

SummaryDeveloping a hydrological forecasting model based on past records is crucial to effective hydropower reservoir management and scheduling. Traditionally, time series analysis and modeling is used for building mathematical models to generate hydrologic records in hydrology and water resources. Artificial intelligence (AI), as a branch of computer science, is capable of analyzing long-series and large-scale hydrological data. In recent years, it is one of front issues to apply AI technology to the hydrological forecasting modeling. In this paper, autoregressive moving-average (ARMA) models, artificial neural networks (ANNs) approaches, adaptive neural-based fuzzy inference system (ANFIS) techniques, genetic programming (GP) models and support vector machine (SVM) method are examined using the long-term observations of monthly river flow discharges. The four quantitative standard statistical performance evaluation measures, the coefficient of correlation ( R), Nash-Sutcliffe efficiency coefficient ( E), root mean squared error (RMSE), mean absolute percentage error (MAPE), are employed to evaluate the performances of various models developed. Two case study river sites are also provided to illustrate their respective performances. The results indicate that the best performance can be obtained by ANFIS, GP and SVM, in terms of different evaluation criteria during the training and validation phases.

Artificial neural networks applied to forecasting time series.

PubMed

Montaño Moreno, Juan J; Palmer Pol, Alfonso; Muñoz Gracia, Pilar

2011-04-01

This study offers a description and comparison of the main models of Artificial Neural Networks (ANN) which have proved to be useful in time series forecasting, and also a standard procedure for the practical application of ANN in this type of task. The Multilayer Perceptron (MLP), Radial Base Function (RBF), Generalized Regression Neural Network (GRNN), and Recurrent Neural Network (RNN) models are analyzed. With this aim in mind, we use a time series made up of 244 time points. A comparative study establishes that the error made by the four neural network models analyzed is less than 10%. In accordance with the interpretation criteria of this performance, it can be concluded that the neural network models show a close fit regarding their forecasting capacity. The model with the best performance is the RBF, followed by the RNN and MLP. The GRNN model is the one with the worst performance. Finally, we analyze the advantages and limitations of ANN, the possible solutions to these limitations, and provide an orientation towards future research.

[Predicting Incidence of Hepatitis E in Chinausing Fuzzy Time Series Based on Fuzzy C-Means Clustering Analysis].

PubMed

Luo, Yi; Zhang, Tao; Li, Xiao-song

2016-05-01

To explore the application of fuzzy time series model based on fuzzy c-means clustering in forecasting monthly incidence of Hepatitis E in mainland China. Apredictive model (fuzzy time series method based on fuzzy c-means clustering) was developed using Hepatitis E incidence data in mainland China between January 2004 and July 2014. The incidence datafrom August 2014 to November 2014 were used to test the fitness of the predictive model. The forecasting results were compared with those resulted from traditional fuzzy time series models. The fuzzy time series model based on fuzzy c-means clustering had 0.001 1 mean squared error (MSE) of fitting and 6.977 5 x 10⁻⁴ MSE of forecasting, compared with 0.0017 and 0.0014 from the traditional forecasting model. The results indicate that the fuzzy time series model based on fuzzy c-means clustering has a better performance in forecasting incidence of Hepatitis E.

Adjusting Wavelet-based Multiresolution Analysis Boundary Conditions for Robust Long-term Streamflow Forecasting Model

NASA Astrophysics Data System (ADS)

Maslova, I.; Ticlavilca, A. M.; McKee, M.

2012-12-01

There has been an increased interest in wavelet-based streamflow forecasting models in recent years. Often overlooked in this approach are the circularity assumptions of the wavelet transform. We propose a novel technique for minimizing the wavelet decomposition boundary condition effect to produce long-term, up to 12 months ahead, forecasts of streamflow. A simulation study is performed to evaluate the effects of different wavelet boundary rules using synthetic and real streamflow data. A hybrid wavelet-multivariate relevance vector machine model is developed for forecasting the streamflow in real-time for Yellowstone River, Uinta Basin, Utah, USA. The inputs of the model utilize only the past monthly streamflow records. They are decomposed into components formulated in terms of wavelet multiresolution analysis. It is shown that the model model accuracy can be increased by using the wavelet boundary rule introduced in this study. This long-term streamflow modeling and forecasting methodology would enable better decision-making and managing water availability risk.

Nonlinear autoregressive neural networks with external inputs for forecasting of typhoon inundation level.

PubMed

Ouyang, Huei-Tau

2017-08-01

Accurate inundation level forecasting during typhoon invasion is crucial for organizing response actions such as the evacuation of people from areas that could potentially flood. This paper explores the ability of nonlinear autoregressive neural networks with exogenous inputs (NARX) to predict inundation levels induced by typhoons. Two types of NARX architecture were employed: series-parallel (NARX-S) and parallel (NARX-P). Based on cross-correlation analysis of rainfall and water-level data from historical typhoon records, 10 NARX models (five of each architecture type) were constructed. The forecasting ability of each model was assessed by considering coefficient of efficiency (CE), relative time shift error (RTS), and peak water-level error (PE). The results revealed that high CE performance could be achieved by employing more model input variables. Comparisons of the two types of model demonstrated that the NARX-S models outperformed the NARX-P models in terms of CE and RTS, whereas both performed exceptionally in terms of PE and without significant difference. The NARX-S and NARX-P models with the highest overall performance were identified and their predictions were compared with those of traditional ARX-based models. The NARX-S model outperformed the ARX-based models in all three indexes, whereas the NARX-P model exhibited comparable CE performance and superior RTS and PE performance.

Gridded Calibration of Ensemble Wind Vector Forecasts Using Ensemble Model Output Statistics

NASA Astrophysics Data System (ADS)

Lazarus, S. M.; Holman, B. P.; Splitt, M. E.

2017-12-01

A computationally efficient method is developed that performs gridded post processing of ensemble wind vector forecasts. An expansive set of idealized WRF model simulations are generated to provide physically consistent high resolution winds over a coastal domain characterized by an intricate land / water mask. Ensemble model output statistics (EMOS) is used to calibrate the ensemble wind vector forecasts at observation locations. The local EMOS predictive parameters (mean and variance) are then spread throughout the grid utilizing flow-dependent statistical relationships extracted from the downscaled WRF winds. Using data withdrawal and 28 east central Florida stations, the method is applied to one year of 24 h wind forecasts from the Global Ensemble Forecast System (GEFS). Compared to the raw GEFS, the approach improves both the deterministic and probabilistic forecast skill. Analysis of multivariate rank histograms indicate the post processed forecasts are calibrated. Two downscaling case studies are presented, a quiescent easterly flow event and a frontal passage. Strengths and weaknesses of the approach are presented and discussed.

Novel approach for streamflow forecasting using a hybrid ANFIS-FFA model

NASA Astrophysics Data System (ADS)

Yaseen, Zaher Mundher; Ebtehaj, Isa; Bonakdari, Hossein; Deo, Ravinesh C.; Danandeh Mehr, Ali; Mohtar, Wan Hanna Melini Wan; Diop, Lamine; El-shafie, Ahmed; Singh, Vijay P.

2017-11-01

The present study proposes a new hybrid evolutionary Adaptive Neuro-Fuzzy Inference Systems (ANFIS) approach for monthly streamflow forecasting. The proposed method is a novel combination of the ANFIS model with the firefly algorithm as an optimizer tool to construct a hybrid ANFIS-FFA model. The results of the ANFIS-FFA model is compared with the classical ANFIS model, which utilizes the fuzzy c-means (FCM) clustering method in the Fuzzy Inference Systems (FIS) generation. The historical monthly streamflow data for Pahang River, which is a major river system in Malaysia that characterized by highly stochastic hydrological patterns, is used in the study. Sixteen different input combinations with one to five time-lagged input variables are incorporated into the ANFIS-FFA and ANFIS models to consider the antecedent seasonal variations in historical streamflow data. The mean absolute error (MAE), root mean square error (RMSE) and correlation coefficient (r) are used to evaluate the forecasting performance of ANFIS-FFA model. In conjunction with these metrics, the refined Willmott's Index (Drefined), Nash-Sutcliffe coefficient (ENS) and Legates and McCabes Index (ELM) are also utilized as the normalized goodness-of-fit metrics. Comparison of the results reveals that the FFA is able to improve the forecasting accuracy of the hybrid ANFIS-FFA model (r = 1; RMSE = 0.984; MAE = 0.364; ENS = 1; ELM = 0.988; Drefined = 0.994) applied for the monthly streamflow forecasting in comparison with the traditional ANFIS model (r = 0.998; RMSE = 3.276; MAE = 1.553; ENS = 0.995; ELM = 0.950; Drefined = 0.975). The results also show that the ANFIS-FFA is not only superior to the ANFIS model but also exhibits a parsimonious modelling framework for streamflow forecasting by incorporating a smaller number of input variables required to yield the comparatively better performance. It is construed that the FFA optimizer can thus surpass the accuracy of the traditional ANFIS model in general, and is able to remove the false (inaccurately) forecasted data in the ANFIS model for extremely low flows. The present results have wider implications not only for streamflow forecasting purposes, but also for other hydro-meteorological forecasting variables requiring only the historical data input data, and attaining a greater level of predictive accuracy with the incorporation of the FFA algorithm as an optimization tool in an ANFIS model.

Precipitation forecasts for rainfall runoff predictions. A case study in poorly gauged Ribb and Gumara catchments, upper Blue Nile, Ethiopia

NASA Astrophysics Data System (ADS)

Seyoum, Mesgana; van Andel, Schalk Jan; Xuan, Yunqing; Amare, Kibreab

Flow forecasting in poorly gauged, flood-prone Ribb and Gumara sub-catchments of the Blue Nile was studied with the aim of testing the performance of Quantitative Precipitation Forecasts (QPFs). Four types of QPFs namely MM5 forecasts with a spatial resolution of 2 km; the Maximum, Mean and Minimum members (MaxEPS, MeanEPS and MinEPS where EPS stands for Ensemble Prediction System) of the fixed, low resolution (2.5 by 2.5 degrees) National Oceanic and Atmospheric Administration Global Forecast System (NOAA GFS) ensemble forecasts were used. Both the MM5 and the EPS were not calibrated (bias correction, downscaling (for EPS), etc.). In addition, zero forecasts assuming no rainfall in the coming days, and monthly average forecasts assuming average monthly rainfall in the coming days, were used. These rainfall forecasts were then used to drive the Hydrologic Engineering Center’s-Hydrologic Modeling System, HEC-HMS, hydrologic model for flow predictions. The results show that flow predictions using MaxEPS and MM5 precipitation forecasts over-predicted the peak flow for most of the seven events analyzed, whereas under-predicted peak flow was found using zero- and monthly average rainfall. The comparison of observed and predicted flow hydrographs shows that MM5, MaxEPS and MeanEPS precipitation forecasts were able to capture the rainfall signal that caused peak flows. Flow predictions based on MaxEPS and MeanEPS gave results that were quantitatively close to the observed flow for most events, whereas flow predictions based on MM5 resulted in large overestimations for some events. In follow-up research for this particular case study, calibration of the MM5 model will be performed. The overall analysis shows that freely available atmospheric forecasting products can provide additional information on upcoming rainfall and peak flow events in areas where only base-line forecasts such as no-rainfall or climatology are available.

Short-term Forecasting of the Prevalence of Trachoma: Expert Opinion, Statistical Regression, versus Transmission Models

PubMed Central

Liu, Fengchen; Porco, Travis C.; Amza, Abdou; Kadri, Boubacar; Nassirou, Baido; West, Sheila K.; Bailey, Robin L.; Keenan, Jeremy D.; Solomon, Anthony W.; Emerson, Paul M.; Gambhir, Manoj; Lietman, Thomas M.

2015-01-01

Background Trachoma programs rely on guidelines made in large part using expert opinion of what will happen with and without intervention. Large community-randomized trials offer an opportunity to actually compare forecasting methods in a masked fashion. Methods The Program for the Rapid Elimination of Trachoma trials estimated longitudinal prevalence of ocular chlamydial infection from 24 communities treated annually with mass azithromycin. Given antibiotic coverage and biannual assessments from baseline through 30 months, forecasts of the prevalence of infection in each of the 24 communities at 36 months were made by three methods: the sum of 15 experts’ opinion, statistical regression of the square-root-transformed prevalence, and a stochastic hidden Markov model of infection transmission (Susceptible-Infectious-Susceptible, or SIS model). All forecasters were masked to the 36-month results and to the other forecasts. Forecasts of the 24 communities were scored by the likelihood of the observed results and compared using Wilcoxon’s signed-rank statistic. Findings Regression and SIS hidden Markov models had significantly better likelihood than community expert opinion (p = 0.004 and p = 0.01, respectively). All forecasts scored better when perturbed to decrease Fisher’s information. Each individual expert’s forecast was poorer than the sum of experts. Interpretation Regression and SIS models performed significantly better than expert opinion, although all forecasts were overly confident. Further model refinements may score better, although would need to be tested and compared in new masked studies. Construction of guidelines that rely on forecasting future prevalence could consider use of mathematical and statistical models. PMID:26302380

Short-term Forecasting of the Prevalence of Trachoma: Expert Opinion, Statistical Regression, versus Transmission Models.

PubMed

Liu, Fengchen; Porco, Travis C; Amza, Abdou; Kadri, Boubacar; Nassirou, Baido; West, Sheila K; Bailey, Robin L; Keenan, Jeremy D; Solomon, Anthony W; Emerson, Paul M; Gambhir, Manoj; Lietman, Thomas M

2015-08-01

Trachoma programs rely on guidelines made in large part using expert opinion of what will happen with and without intervention. Large community-randomized trials offer an opportunity to actually compare forecasting methods in a masked fashion. The Program for the Rapid Elimination of Trachoma trials estimated longitudinal prevalence of ocular chlamydial infection from 24 communities treated annually with mass azithromycin. Given antibiotic coverage and biannual assessments from baseline through 30 months, forecasts of the prevalence of infection in each of the 24 communities at 36 months were made by three methods: the sum of 15 experts' opinion, statistical regression of the square-root-transformed prevalence, and a stochastic hidden Markov model of infection transmission (Susceptible-Infectious-Susceptible, or SIS model). All forecasters were masked to the 36-month results and to the other forecasts. Forecasts of the 24 communities were scored by the likelihood of the observed results and compared using Wilcoxon's signed-rank statistic. Regression and SIS hidden Markov models had significantly better likelihood than community expert opinion (p = 0.004 and p = 0.01, respectively). All forecasts scored better when perturbed to decrease Fisher's information. Each individual expert's forecast was poorer than the sum of experts. Regression and SIS models performed significantly better than expert opinion, although all forecasts were overly confident. Further model refinements may score better, although would need to be tested and compared in new masked studies. Construction of guidelines that rely on forecasting future prevalence could consider use of mathematical and statistical models. Clinicaltrials.gov NCT00792922.

Hybrid methodology for tuberculosis incidence time-series forecasting based on ARIMA and a NAR neural network.

PubMed

Wang, K W; Deng, C; Li, J P; Zhang, Y Y; Li, X Y; Wu, M C

2017-04-01

Tuberculosis (TB) affects people globally and is being reconsidered as a serious public health problem in China. Reliable forecasting is useful for the prevention and control of TB. This study proposes a hybrid model combining autoregressive integrated moving average (ARIMA) with a nonlinear autoregressive (NAR) neural network for forecasting the incidence of TB from January 2007 to March 2016. Prediction performance was compared between the hybrid model and the ARIMA model. The best-fit hybrid model was combined with an ARIMA (3,1,0) × (0,1,1)12 and NAR neural network with four delays and 12 neurons in the hidden layer. The ARIMA-NAR hybrid model, which exhibited lower mean square error, mean absolute error, and mean absolute percentage error of 0·2209, 0·1373, and 0·0406, respectively, in the modelling performance, could produce more accurate forecasting of TB incidence compared to the ARIMA model. This study shows that developing and applying the ARIMA-NAR hybrid model is an effective method to fit the linear and nonlinear patterns of time-series data, and this model could be helpful in the prevention and control of TB.

Calibration and combination of dynamical seasonal forecasts to enhance the value of predicted probabilities for managing risk

NASA Astrophysics Data System (ADS)

Dutton, John A.; James, Richard P.; Ross, Jeremy D.

2013-06-01

Seasonal probability forecasts produced with numerical dynamics on supercomputers offer great potential value in managing risk and opportunity created by seasonal variability. The skill and reliability of contemporary forecast systems can be increased by calibration methods that use the historical performance of the forecast system to improve the ongoing real-time forecasts. Two calibration methods are applied to seasonal surface temperature forecasts of the US National Weather Service, the European Centre for Medium Range Weather Forecasts, and to a World Climate Service multi-model ensemble created by combining those two forecasts with Bayesian methods. As expected, the multi-model is somewhat more skillful and more reliable than the original models taken alone. The potential value of the multimodel in decision making is illustrated with the profits achieved in simulated trading of a weather derivative. In addition to examining the seasonal models, the article demonstrates that calibrated probability forecasts of weekly average temperatures for leads of 2-4 weeks are also skillful and reliable. The conversion of ensemble forecasts into probability distributions of impact variables is illustrated with degree days derived from the temperature forecasts. Some issues related to loss of stationarity owing to long-term warming are considered. The main conclusion of the article is that properly calibrated probabilistic forecasts possess sufficient skill and reliability to contribute to effective decisions in government and business activities that are sensitive to intraseasonal and seasonal climate variability.

Gas demand forecasting by a new artificial intelligent algorithm

NASA Astrophysics Data System (ADS)

Khatibi. B, Vahid; Khatibi, Elham

2012-01-01

Energy demand forecasting is a key issue for consumers and generators in all energy markets in the world. This paper presents a new forecasting algorithm for daily gas demand prediction. This algorithm combines a wavelet transform and forecasting models such as multi-layer perceptron (MLP), linear regression or GARCH. The proposed method is applied to real data from the UK gas markets to evaluate their performance. The results show that the forecasting accuracy is improved significantly by using the proposed method.

Statistical and Hydrological evaluation of precipitation forecasts from IMD MME and ECMWF numerical weather forecasts for Indian River basins

NASA Astrophysics Data System (ADS)

Mohite, A. R.; Beria, H.; Behera, A. K.; Chatterjee, C.; Singh, R.

2016-12-01

Flood forecasting using hydrological models is an important and cost-effective non-structural flood management measure. For forecasting at short lead times, empirical models using real-time precipitation estimates have proven to be reliable. However, their skill depreciates with increasing lead time. Coupling a hydrologic model with real-time rainfall forecasts issued from numerical weather prediction (NWP) systems could increase the lead time substantially. In this study, we compared 1-5 days precipitation forecasts from India Meteorological Department (IMD) Multi-Model Ensemble (MME) with European Center for Medium Weather forecast (ECMWF) NWP forecasts for over 86 major river basins in India. We then evaluated the hydrologic utility of these forecasts over Basantpur catchment (approx. 59,000 km2) of the Mahanadi River basin. Coupled MIKE 11 RR (NAM) and MIKE 11 hydrodynamic (HD) models were used for the development of flood forecast system (FFS). RR model was calibrated using IMD station rainfall data. Cross-sections extracted from SRTM 30 were used as input to the MIKE 11 HD model. IMD started issuing operational MME forecasts from the year 2008, and hence, both the statistical and hydrologic evaluation were carried out from 2008-2014. The performance of FFS was evaluated using both the NWP datasets separately for the year 2011, which was a large flood year in Mahanadi River basin. We will present figures and metrics for statistical (threshold based statistics, skill in terms of correlation and bias) and hydrologic (Nash Sutcliffe efficiency, mean and peak error statistics) evaluation. The statistical evaluation will be at pan-India scale for all the major river basins and the hydrologic evaluation will be for the Basantpur catchment of the Mahanadi River basin.

Evaluation of the fast orthogonal search method for forecasting chloride levels in the Deltona groundwater supply (Florida, USA)

NASA Astrophysics Data System (ADS)

El-Jaat, Majda; Hulley, Michael; Tétreault, Michel

2018-02-01

Despite the broad impact and importance of saltwater intrusion in coastal aquifers, little research has been directed towards forecasting saltwater intrusion in areas where the source of saltwater is uncertain. Saline contamination in inland groundwater supplies is a concern for numerous communities in the southern US including the city of Deltona, Florida. Furthermore, conventional numerical tools for forecasting saltwater contamination are heavily dependent on reliable characterization of the physical characteristics of underlying aquifers, information that is often absent or challenging to obtain. To overcome these limitations, a reliable alternative data-driven model for forecasting salinity in a groundwater supply was developed for Deltona using the fast orthogonal search (FOS) method. FOS was applied on monthly water-demand data and corresponding chloride concentrations at water supply wells. Groundwater salinity measurements from Deltona water supply wells were applied to evaluate the forecasting capability and accuracy of the FOS model. Accurate and reliable groundwater salinity forecasting is necessary to support effective and sustainable coastal-water resource planning and management. The available (27) water supply wells for Deltona were randomly split into three test groups for the purposes of FOS model development and performance assessment. Based on four performance indices (RMSE, RSR, NSEC, and R), the FOS model proved to be a reliable and robust forecaster of groundwater salinity. FOS is relatively inexpensive to apply, is not based on rigorous physical characterization of the water supply aquifer, and yields reliable estimates of groundwater salinity in active water supply wells.

Earthquake forecasting test for Kanto district to reduce vulnerability of urban mega earthquake disasters

NASA Astrophysics Data System (ADS)

Yokoi, S.; Tsuruoka, H.; Nanjo, K.; Hirata, N.

2012-12-01

Collaboratory for the Study of Earthquake Predictability (CSEP) is a global project on earthquake predictability research. The final goal of this project is to search for the intrinsic predictability of the earthquake rupture process through forecast testing experiments. The Earthquake Research Institute, the University of Tokyo joined CSEP and started the Japanese testing center called as CSEP-Japan. This testing center provides an open access to researchers contributing earthquake forecast models applied to Japan. Now more than 100 earthquake forecast models were submitted on the prospective experiment. The models are separated into 4 testing classes (1 day, 3 months, 1 year and 3 years) and 3 testing regions covering an area of Japan including sea area, Japanese mainland and Kanto district. We evaluate the performance of the models in the official suite of tests defined by CSEP. The total number of experiments was implemented for approximately 300 rounds. These results provide new knowledge concerning statistical forecasting models. We started a study for constructing a 3-dimensional earthquake forecasting model for Kanto district in Japan based on CSEP experiments under the Special Project for Reducing Vulnerability for Urban Mega Earthquake Disasters. Because seismicity of the area ranges from shallower part to a depth of 80 km due to subducting Philippine Sea plate and Pacific plate, we need to study effect of depth distribution. We will develop models for forecasting based on the results of 2-D modeling. We defined the 3D - forecasting area in the Kanto region with test classes of 1 day, 3 months, 1 year and 3 years, and magnitudes from 4.0 to 9.0 as in CSEP-Japan. In the first step of the study, we will install RI10K model (Nanjo, 2011) and the HISTETAS models (Ogata, 2011) to know if those models have good performance as in the 3 months 2-D CSEP-Japan experiments in the Kanto region before the 2011 Tohoku event (Yokoi et al., in preparation). We use CSEP-Japan experiments as a starting model of non-divided column in a depth. In the presentation, we will discuss the performance of the models comparing results of the Kanto district with those obtained in all over Japan by CSEP-Japan and also add to discuss the results of the 3-month experiments after the 2011 Tohoku earthquake to understand the learning ability of the models associated with recent seismicity of the area.

Forecasting surface-layer atmospheric parameters at the Large Binocular Telescope site

NASA Astrophysics Data System (ADS)

Turchi, Alessio; Masciadri, Elena; Fini, Luca

2017-04-01

In this paper, we quantify the performance of an automated weather forecast system implemented on the Large Binocular Telescope (LBT) site at Mt Graham (Arizona) in forecasting the main atmospheric parameters close to the ground. The system employs a mesoscale non-hydrostatic numerical model (Meso-Nh). To validate the model, we compare the forecasts of wind speed, wind direction, temperature and relative humidity close to the ground with the respective values measured by instrumentation installed on the telescope dome. The study is performed over a large sample of nights uniformly distributed over 2 yr. The quantitative analysis is done using classical statistical operators [bias, root-mean-square error (RMSE) and σ] and contingency tables, which allows us to extract complementary key information, such as the percentage of correct detections (PC) and the probability of obtaining a correct detection within a defined interval of values (POD). The results of our study indicate that the model performance in forecasting the atmospheric parameters we have just cited are very good, in some cases excellent: RMSE for temperature is below 1°C, for relative humidity it is 14 per cent and for the wind speed it is around 2.5 m s-1. The relative error of the RMSE for wind direction varies from 9 to 17 per cent depending on the wind speed conditions. This work is performed in the context of the ALTA (Advanced LBT Turbulence and Atmosphere) Center project, whose final goal is to provide forecasts of all the atmospheric parameters and the optical turbulence to support LBT observations, adaptive optics facilities and interferometric facilities.

Soil Moisture Drought Monitoring and Forecasting Using Satellite and Climate Model Data over Southwestern China

DOE Office of Scientific and Technical Information (OSTI.GOV)

Zhang, Xuejun; Tang, Qiuhong; Liu, Xingcai

Real-time monitoring and predicting drought development with several months in advance is of critical importance for drought risk adaptation and mitigation. In this paper, we present a drought monitoring and seasonal forecasting framework based on the Variable Infiltration Capacity (VIC) hydrologic model over Southwest China (SW). The satellite precipitation data are used to force VIC model for near real-time estimate of land surface hydrologic conditions. As initialized with satellite-aided monitoring, the climate model-based forecast (CFSv2_VIC) and ensemble streamflow prediction (ESP)-based forecast (ESP_VIC) are both performed and evaluated through their ability in reproducing the evolution of the 2009/2010 severe drought overmore » SW. The results show that the satellite-aided monitoring is able to provide reasonable estimate of forecast initial conditions (ICs) in a real-time manner. Both of CFSv2_VIC and ESP_VIC exhibit comparable performance against the observation-based estimates for the first month, whereas the predictive skill largely drops beyond 1-month. Compared to ESP_VIC, CFSv2_VIC shows better performance as indicated by the smaller ensemble range. This study highlights the value of this operational framework in generating near real-time ICs and giving a reliable prediction with 1-month ahead, which has great implications for drought risk assessment, preparation and relief.« less

Increased performance in the short-term water demand forecasting through the use of a parallel adaptive weighting strategy

NASA Astrophysics Data System (ADS)

Sardinha-Lourenço, A.; Andrade-Campos, A.; Antunes, A.; Oliveira, M. S.

2018-03-01

Recent research on water demand short-term forecasting has shown that models using univariate time series based on historical data are useful and can be combined with other prediction methods to reduce errors. The behavior of water demands in drinking water distribution networks focuses on their repetitive nature and, under meteorological conditions and similar consumers, allows the development of a heuristic forecast model that, in turn, combined with other autoregressive models, can provide reliable forecasts. In this study, a parallel adaptive weighting strategy of water consumption forecast for the next 24-48 h, using univariate time series of potable water consumption, is proposed. Two Portuguese potable water distribution networks are used as case studies where the only input data are the consumption of water and the national calendar. For the development of the strategy, the Autoregressive Integrated Moving Average (ARIMA) method and a short-term forecast heuristic algorithm are used. Simulations with the model showed that, when using a parallel adaptive weighting strategy, the prediction error can be reduced by 15.96% and the average error by 9.20%. This reduction is important in the control and management of water supply systems. The proposed methodology can be extended to other forecast methods, especially when it comes to the availability of multiple forecast models.

Maximizing Statistical Power When Verifying Probabilistic Forecasts of Hydrometeorological Events

NASA Astrophysics Data System (ADS)

DeChant, C. M.; Moradkhani, H.

2014-12-01

Hydrometeorological events (i.e. floods, droughts, precipitation) are increasingly being forecasted probabilistically, owing to the uncertainties in the underlying causes of the phenomenon. In these forecasts, the probability of the event, over some lead time, is estimated based on some model simulations or predictive indicators. By issuing probabilistic forecasts, agencies may communicate the uncertainty in the event occurring. Assuming that the assigned probability of the event is correct, which is referred to as a reliable forecast, the end user may perform some risk management based on the potential damages resulting from the event. Alternatively, an unreliable forecast may give false impressions of the actual risk, leading to improper decision making when protecting resources from extreme events. Due to this requisite for reliable forecasts to perform effective risk management, this study takes a renewed look at reliability assessment in event forecasts. Illustrative experiments will be presented, showing deficiencies in the commonly available approaches (Brier Score, Reliability Diagram). Overall, it is shown that the conventional reliability assessment techniques do not maximize the ability to distinguish between a reliable and unreliable forecast. In this regard, a theoretical formulation of the probabilistic event forecast verification framework will be presented. From this analysis, hypothesis testing with the Poisson-Binomial distribution is the most exact model available for the verification framework, and therefore maximizes one's ability to distinguish between a reliable and unreliable forecast. Application of this verification system was also examined within a real forecasting case study, highlighting the additional statistical power provided with the use of the Poisson-Binomial distribution.

A national-scale seasonal hydrological forecast system: development and evaluation over Britain

NASA Astrophysics Data System (ADS)

Bell, Victoria A.; Davies, Helen N.; Kay, Alison L.; Brookshaw, Anca; Scaife, Adam A.

2017-09-01

Skilful winter seasonal predictions for the North Atlantic circulation and northern Europe have now been demonstrated and the potential for seasonal hydrological forecasting in the UK is now being explored. One of the techniques being used combines seasonal rainfall forecasts provided by operational weather forecast systems with hydrological modelling tools to provide estimates of seasonal mean river flows up to a few months ahead. The work presented here shows how spatial information contained in a distributed hydrological model typically requiring high-resolution (daily or better) rainfall data can be used to provide an initial condition for a much simpler forecast model tailored to use low-resolution monthly rainfall forecasts. Rainfall forecasts (hindcasts) from the GloSea5 model (1996 to 2009) are used to provide the first assessment of skill in these national-scale flow forecasts. The skill in the combined modelling system is assessed for different seasons and regions of Britain, and compared to what might be achieved using other approaches such as use of an ensemble of historical rainfall in a hydrological model, or a simple flow persistence forecast. The analysis indicates that only limited forecast skill is achievable for Spring and Summer seasonal hydrological forecasts; however, Autumn and Winter flows can be reasonably well forecast using (ensemble mean) rainfall forecasts based on either GloSea5 forecasts or historical rainfall (the preferred type of forecast depends on the region). Flow forecasts using ensemble mean GloSea5 rainfall perform most consistently well across Britain, and provide the most skilful forecasts overall at the 3-month lead time. Much of the skill (64 %) in the 1-month ahead seasonal flow forecasts can be attributed to the hydrological initial condition (particularly in regions with a significant groundwater contribution to flows), whereas for the 3-month ahead lead time, GloSea5 forecasts account for ˜ 70 % of the forecast skill (mostly in areas of high rainfall to the north and west) and only 30 % of the skill arises from hydrological memory (typically groundwater-dominated areas). Given the high spatial heterogeneity in typical patterns of UK rainfall and evaporation, future development of skilful spatially distributed seasonal forecasts could lead to substantial improvements in seasonal flow forecast capability, potentially benefitting practitioners interested in predicting hydrological extremes, not only in the UK but also across Europe.

Multi-step-ahead crude oil price forecasting using a hybrid grey wave model

NASA Astrophysics Data System (ADS)

Chen, Yanhui; Zhang, Chuan; He, Kaijian; Zheng, Aibing

2018-07-01

Crude oil is crucial to the operation and economic well-being of the modern society. Huge changes of crude oil price always cause panics to the global economy. There are many factors influencing crude oil price. Crude oil price prediction is still a difficult research problem widely discussed among researchers. Based on the researches on Heterogeneous Market Hypothesis and the relationship between crude oil price and macroeconomic factors, exchange market, stock market, this paper proposes a hybrid grey wave forecasting model, which combines Random Walk (RW)/ARMA to forecast multi-step-ahead crude oil price. More specifically, we use grey wave forecasting model to model the periodical characteristics of crude oil price and ARMA/RW to simulate the daily random movements. The innovation also comes from using the information of the time series graph to forecast crude oil price, since grey wave forecasting is a graphical prediction method. The empirical results demonstrate that based on the daily data of crude oil price, the hybrid grey wave forecasting model performs well in 15- to 20-step-ahead prediction and it always dominates ARMA and Random Walk in correct direction prediction.

Improving forecasting accuracy of medium and long-term runoff using artificial neural network based on EEMD decomposition.

PubMed

Wang, Wen-chuan; Chau, Kwok-wing; Qiu, Lin; Chen, Yang-bo

2015-05-01

Hydrological time series forecasting is one of the most important applications in modern hydrology, especially for the effective reservoir management. In this research, an artificial neural network (ANN) model coupled with the ensemble empirical mode decomposition (EEMD) is presented for forecasting medium and long-term runoff time series. First, the original runoff time series is decomposed into a finite and often small number of intrinsic mode functions (IMFs) and a residual series using EEMD technique for attaining deeper insight into the data characteristics. Then all IMF components and residue are predicted, respectively, through appropriate ANN models. Finally, the forecasted results of the modeled IMFs and residual series are summed to formulate an ensemble forecast for the original annual runoff series. Two annual reservoir runoff time series from Biuliuhe and Mopanshan in China, are investigated using the developed model based on four performance evaluation measures (RMSE, MAPE, R and NSEC). The results obtained in this work indicate that EEMD can effectively enhance forecasting accuracy and the proposed EEMD-ANN model can attain significant improvement over ANN approach in medium and long-term runoff time series forecasting. Copyright © 2015 Elsevier Inc. All rights reserved.

Prospective Tests of Southern California Earthquake Forecasts

NASA Astrophysics Data System (ADS)

Jackson, D. D.; Schorlemmer, D.; Gerstenberger, M.; Kagan, Y. Y.; Helmstetter, A.; Wiemer, S.; Field, N.

2004-12-01

We are testing earthquake forecast models prospectively using likelihood ratios. Several investigators have developed such models as part of the Southern California Earthquake Center's project called Regional Earthquake Likelihood Models (RELM). Various models are based on fault geometry and slip rates, seismicity, geodetic strain, and stress interactions. Here we describe the testing procedure and present preliminary results. Forecasts are expressed as the yearly rate of earthquakes within pre-specified bins of longitude, latitude, magnitude, and focal mechanism parameters. We test models against each other in pairs, which requires that both forecasts in a pair be defined over the same set of bins. For this reason we specify a standard "menu" of bins and ground rules to guide forecasters in using common descriptions. One menu category includes five-year forecasts of magnitude 5.0 and larger. Contributors will be requested to submit forecasts in the form of a vector of yearly earthquake rates on a 0.1 degree grid at the beginning of the test. Focal mechanism forecasts, when available, are also archived and used in the tests. Interim progress will be evaluated yearly, but final conclusions would be made on the basis of cumulative five-year performance. The second category includes forecasts of earthquakes above magnitude 4.0 on a 0.1 degree grid, evaluated and renewed daily. Final evaluation would be based on cumulative performance over five years. Other types of forecasts with different magnitude, space, and time sampling are welcome and will be tested against other models with shared characteristics. Tests are based on the log likelihood scores derived from the probability that future earthquakes would occur where they do if a given forecast were true [Kagan and Jackson, J. Geophys. Res.,100, 3,943-3,959, 1995]. For each pair of forecasts, we compute alpha, the probability that the first would be wrongly rejected in favor of the second, and beta, the probability that the second would be wrongly rejected in favor of the first. Computing alpha and beta requires knowing the theoretical distribution of likelihood scores under each hypothesis, which we estimate by simulations. In this scheme, each forecast is given equal status; there is no "null hypothesis" which would be accepted by default. Forecasts and test results will be archived and posted on the RELM web site. Major problems under discussion include how to treat aftershocks, which clearly violate the variable-rate Poissonian hypotheses that we employ, and how to deal with the temporal variations in catalog completeness that follow large earthquakes.

A Bayesian spatio-temporal model for forecasting Anaplasma species seroprevalence in domestic dogs within the contiguous United States.

PubMed

Liu, Yan; Watson, Stella C; Gettings, Jenna R; Lund, Robert B; Nordone, Shila K; Yabsley, Michael J; McMahan, Christopher S

2017-01-01

This paper forecasts the 2016 canine Anaplasma spp. seroprevalence in the United States from eight climate, geographic and societal factors. The forecast's construction and an assessment of its performance are described. The forecast is based on a spatial-temporal conditional autoregressive model fitted to over 11 million Anaplasma spp. seroprevalence test results for dogs conducted in the 48 contiguous United States during 2011-2015. The forecast uses county-level data on eight predictive factors, including annual temperature, precipitation, relative humidity, county elevation, forestation coverage, surface water coverage, population density and median household income. Non-static factors are extrapolated into the forthcoming year with various statistical methods. The fitted model and factor extrapolations are used to estimate next year's regional prevalence. The correlation between the observed and model-estimated county-by-county Anaplasma spp. seroprevalence for the five-year period 2011-2015 is 0.902, demonstrating reasonable model accuracy. The weighted correlation (accounting for different sample sizes) between 2015 observed and forecasted county-by-county Anaplasma spp. seroprevalence is 0.987, exhibiting that the proposed approach can be used to accurately forecast Anaplasma spp. seroprevalence. The forecast presented herein can a priori alert veterinarians to areas expected to see Anaplasma spp. seroprevalence beyond the accepted endemic range. The proposed methods may prove useful for forecasting other diseases.

Real-time forecasting at weekly timescales of the SST and SLA of the Ligurian Sea with a satellite-based ocean forecasting (SOFT) system

NASA Astrophysics Data System (ADS)

ÁLvarez, A.; Orfila, A.; Tintoré, J.

2004-03-01

Satellites are the only systems able to provide continuous information on the spatiotemporal variability of vast areas of the ocean. Relatively long-term time series of satellite data are nowadays available. These spatiotemporal time series of satellite observations can be employed to build empirical models, called satellite-based ocean forecasting (SOFT) systems, to forecast certain aspects of future ocean states. SOFT systems can predict satellite-observed fields at different timescales. The forecast skill of SOFT systems forecasting the sea surface temperature (SST) at monthly timescales has been extensively explored in previous works. In this work we study the performance of two SOFT systems forecasting, respectively, the SST and sea level anomaly (SLA) at weekly timescales, that is, providing forecasts of the weekly averaged SST and SLA fields with 1 week in advance. The SOFT systems were implemented in the Ligurian Sea (Western Mediterranean Sea). Predictions from the SOFT systems are compared with observations and with the predictions obtained from persistence models. Results indicate that the SOFT system forecasting the SST field is always superior in terms of predictability to persistence. Minimum prediction errors in the SST are obtained during winter and spring seasons. On the other hand, the biggest differences between the performance of SOFT and persistence models are found during summer and autumn. These changes in the predictability are explained on the basis of the particular variability of the SST field in the Ligurian Sea. Concerning the SLA field, no improvements with respect to persistence have been found for the SOFT system forecasting the SLA field.

Forecasting sea fog on the coast of southern China

NASA Astrophysics Data System (ADS)

Huang, H.; Huang, B.; Liu, C.; Tu, J.; Wen, G.; Mao, W.

2016-12-01

Forecast sea fog is still full of challenges. We have performed the numerical forecasting of sea fog on the coast of southern China by using the operational meso-scale regional model GRAPES (Global/Regional assimilation and prediction system). The GRAPES model horizontal resolution was 3km and with 66 vertical levels. A total of 72 hours forecasting of sea fog was conducted with hourly outputs over the sea fog event. The results show that the model system can predict reasonable characteristics of typical sea fog events on the coast of southern China. The scope of sea fog coincides with the observations of meteorological stations, the observations of the Marine Meteorological Science Experiment Base (MMSEB) at Bohe, Maoming and satellite products of sea fog. The goal of this study is to establish an operational numerical forecasting model system of sea fog on the coast of southern China.

Optimized Structure of the Traffic Flow Forecasting Model With a Deep Learning Approach.

PubMed

Yang, Hao-Fan; Dillon, Tharam S; Chen, Yi-Ping Phoebe

2017-10-01

Forecasting accuracy is an important issue for successful intelligent traffic management, especially in the domain of traffic efficiency and congestion reduction. The dawning of the big data era brings opportunities to greatly improve prediction accuracy. In this paper, we propose a novel model, stacked autoencoder Levenberg-Marquardt model, which is a type of deep architecture of neural network approach aiming to improve forecasting accuracy. The proposed model is designed using the Taguchi method to develop an optimized structure and to learn traffic flow features through layer-by-layer feature granulation with a greedy layerwise unsupervised learning algorithm. It is applied to real-world data collected from the M6 freeway in the U.K. and is compared with three existing traffic predictors. To the best of our knowledge, this is the first time that an optimized structure of the traffic flow forecasting model with a deep learning approach is presented. The evaluation results demonstrate that the proposed model with an optimized structure has superior performance in traffic flow forecasting.

Seasonal forecasting of dolphinfish distribution in eastern Australia to aid recreational fishers and managers

NASA Astrophysics Data System (ADS)

Brodie, Stephanie; Hobday, Alistair J.; Smith, James A.; Spillman, Claire M.; Hartog, Jason R.; Everett, Jason D.; Taylor, Matthew D.; Gray, Charles A.; Suthers, Iain M.

2017-06-01

Seasonal forecasting of environmental conditions and marine species distribution has been used as a decision support tool in commercial and aquaculture fisheries. These tools may also be applicable to species targeted by the recreational fisheries sector, a sector that is increasing its use of marine resources, and making important economic and social contributions to coastal communities around the world. Here, a seasonal forecast of the habitat and density of dolphinfish (Coryphaena hippurus), based on sea surface temperatures, was developed for the east coast of New South Wales (NSW), Australia. Two prototype forecast products were created; geographic spatial forecasts of dolphinfish habitat and a latitudinal summary identifying the location of fish density peaks. The less detailed latitudinal summary was created to limit the resolution of habitat information to prevent potential resource over-exploitation by fishers in the absence of total catch controls. The forecast dolphinfish habitat model was accurate at the start of the annual dolphinfish migration in NSW (December) but other months (January - May) showed poor performance due to spatial and temporal variability in the catch data used in model validation. Habitat forecasts for December were useful up to five months ahead, with performance decreasing as forecast were made further into the future. The continued development and sound application of seasonal forecasts will help fishery industries cope with future uncertainty and promote dynamic and sustainable marine resource management.

Training the next generation of scientists in Weather Forecasting: new approaches with real models

NASA Astrophysics Data System (ADS)

Carver, Glenn; Váňa, Filip; Siemen, Stephan; Kertesz, Sandor; Keeley, Sarah

2014-05-01

The European Centre for Medium Range Weather Forecasts operationally produce medium range forecasts using what is internationally acknowledged as the world leading global weather forecast model. Future development of this scientifically advanced model relies on a continued availability of experts in the field of meteorological science and with high-level software skills. ECMWF therefore has a vested interest in young scientists and University graduates developing the necessary skills in numerical weather prediction including both scientific and technical aspects. The OpenIFS project at ECMWF maintains a portable version of the ECMWF forecast model (known as IFS) for use in education and research at Universities, National Meteorological Services and other research and education organisations. OpenIFS models can be run on desktop or high performance computers to produce weather forecasts in a similar way to the operational forecasts at ECMWF. ECMWF also provide the Metview desktop application, a modern, graphical, and easy to use tool for analysing and visualising forecasts that is routinely used by scientists and forecasters at ECMWF and other institutions. The combination of Metview with the OpenIFS models has the potential to deliver classroom-friendly tools allowing students to apply their theoretical knowledge to real-world examples using a world-leading weather forecasting model. In this paper we will describe how the OpenIFS model has been used for teaching. We describe the use of Linux based 'virtual machines' pre-packaged on USB sticks that support a technically easy and safe way of providing 'classroom-on-a-stick' learning environments for advanced training in numerical weather prediction. We welcome discussions with interested parties.

Long-range forecast of all India summer monsoon rainfall using adaptive neuro-fuzzy inference system: skill comparison with CFSv2 model simulation and real-time forecast for the year 2015

NASA Astrophysics Data System (ADS)

Chaudhuri, S.; Das, D.; Goswami, S.; Das, S. K.

2016-11-01

All India summer monsoon rainfall (AISMR) characteristics play a vital role for the policy planning and national economy of the country. In view of the significant impact of monsoon system on regional as well as global climate systems, accurate prediction of summer monsoon rainfall has become a challenge. The objective of this study is to develop an adaptive neuro-fuzzy inference system (ANFIS) for long range forecast of AISMR. The NCEP/NCAR reanalysis data of temperature, zonal and meridional wind at different pressure levels have been taken to construct the input matrix of ANFIS. The membership of the input parameters for AISMR as high, medium or low is estimated with trapezoidal membership function. The fuzzified standardized input parameters and the de-fuzzified target output are trained with artificial neural network models. The forecast of AISMR with ANFIS is compared with non-hybrid multi-layer perceptron model (MLP), radial basis functions network (RBFN) and multiple linear regression (MLR) models. The forecast error analyses of the models reveal that ANFIS provides the best forecast of AISMR with minimum prediction error of 0.076, whereas the errors with MLP, RBFN and MLR models are 0.22, 0.18 and 0.73 respectively. During validation with observations, ANFIS shows its potency over the said comparative models. Performance of the ANFIS model is verified through different statistical skill scores, which also confirms the aptitude of ANFIS in forecasting AISMR. The forecast skill of ANFIS is also observed to be better than Climate Forecast System version 2. The real-time forecast with ANFIS shows possibility of deficit (65-75 cm) AISMR in the year 2015.

Seasonal precipitation forecasting for the Melbourne region using a Self-Organizing Maps approach

NASA Astrophysics Data System (ADS)

Pidoto, Ross; Wallner, Markus; Haberlandt, Uwe

2017-04-01

The Melbourne region experiences highly variable inter-annual rainfall. For close to a decade during the 2000s, below average rainfall seriously affected the environment, water supplies and agriculture. A seasonal rainfall forecasting model for the Melbourne region based on the novel approach of a Self-Organizing Map has been developed and tested for its prediction performance. Predictor variables at varying lead times were first assessed for inclusion within the model by calculating their importance via Random Forests. Predictor variables tested include the climate indices SOI, DMI and N3.4, in addition to gridded global sea surface temperature data. Five forecasting models were developed: an annual model and four seasonal models, each individually optimized for performance through Pearson's correlation r and the Nash-Sutcliffe Efficiency. The annual model showed a prediction performance of r = 0.54 and NSE = 0.14. The best seasonal model was for spring, with r = 0.61 and NSE = 0.31. Autumn was the worst performing seasonal model. The sea surface temperature data contributed fewer predictor variables compared to climate indices. Most predictor variables were supplied at a minimum lead, however some predictors were found at lead times of up to a year.

Forecasting the mortality rates using Lee-Carter model and Heligman-Pollard model

NASA Astrophysics Data System (ADS)

Ibrahim, R. I.; Ngataman, N.; Abrisam, W. N. A. Wan Mohd

2017-09-01

Improvement in life expectancies has driven further declines in mortality. The sustained reduction in mortality rates and its systematic underestimation has been attracting the significant interest of researchers in recent years because of its potential impact on population size and structure, social security systems, and (from an actuarial perspective) the life insurance and pensions industry worldwide. Among all forecasting methods, the Lee-Carter model has been widely accepted by the actuarial community and Heligman-Pollard model has been widely used by researchers in modelling and forecasting future mortality. Therefore, this paper only focuses on Lee-Carter model and Heligman-Pollard model. The main objective of this paper is to investigate how accurately these two models will perform using Malaysian data. Since these models involves nonlinear equations that are explicitly difficult to solve, the Matrix Laboratory Version 8.0 (MATLAB 8.0) software will be used to estimate the parameters of the models. Autoregressive Integrated Moving Average (ARIMA) procedure is applied to acquire the forecasted parameters for both models as the forecasted mortality rates are obtained by using all the values of forecasted parameters. To investigate the accuracy of the estimation, the forecasted results will be compared against actual data of mortality rates. The results indicate that both models provide better results for male population. However, for the elderly female population, Heligman-Pollard model seems to underestimate to the mortality rates while Lee-Carter model seems to overestimate to the mortality rates.

The development and evaluation of a hydrological seasonal forecast system prototype for predicting spring flood volumes in Swedish rivers

NASA Astrophysics Data System (ADS)

Foster, Kean; Bertacchi Uvo, Cintia; Olsson, Jonas

2018-05-01

Hydropower makes up nearly half of Sweden's electrical energy production. However, the distribution of the water resources is not aligned with demand, as most of the inflows to the reservoirs occur during the spring flood period. This means that carefully planned reservoir management is required to help redistribute water resources to ensure optimal production and accurate forecasts of the spring flood volume (SFV) is essential for this. The current operational SFV forecasts use a historical ensemble approach where the HBV model is forced with historical observations of precipitation and temperature. In this work we develop and test a multi-model prototype, building on previous work, and evaluate its ability to forecast the SFV in 84 sub-basins in northern Sweden. The hypothesis explored in this work is that a multi-model seasonal forecast system incorporating different modelling approaches is generally more skilful at forecasting the SFV in snow dominated regions than a forecast system that utilises only one approach. The testing is done using cross-validated hindcasts for the period 1981-2015 and the results are evaluated against both climatology and the current system to determine skill. Both the multi-model methods considered showed skill over the reference forecasts. The version that combined the historical modelling chain, dynamical modelling chain, and statistical modelling chain performed better than the other and was chosen for the prototype. The prototype was able to outperform the current operational system 57 % of the time on average and reduce the error in the SFV by ˜ 6 % across all sub-basins and forecast dates.

Value-at-risk estimation with wavelet-based extreme value theory: Evidence from emerging markets

NASA Astrophysics Data System (ADS)

Cifter, Atilla

2011-06-01

This paper introduces wavelet-based extreme value theory (EVT) for univariate value-at-risk estimation. Wavelets and EVT are combined for volatility forecasting to estimate a hybrid model. In the first stage, wavelets are used as a threshold in generalized Pareto distribution, and in the second stage, EVT is applied with a wavelet-based threshold. This new model is applied to two major emerging stock markets: the Istanbul Stock Exchange (ISE) and the Budapest Stock Exchange (BUX). The relative performance of wavelet-based EVT is benchmarked against the Riskmetrics-EWMA, ARMA-GARCH, generalized Pareto distribution, and conditional generalized Pareto distribution models. The empirical results show that the wavelet-based extreme value theory increases predictive performance of financial forecasting according to number of violations and tail-loss tests. The superior forecasting performance of the wavelet-based EVT model is also consistent with Basel II requirements, and this new model can be used by financial institutions as well.

The potential of radar-based ensemble forecasts for flash-flood early warning in the southern Swiss Alps

NASA Astrophysics Data System (ADS)

Liechti, K.; Panziera, L.; Germann, U.; Zappa, M.

2013-10-01

This study explores the limits of radar-based forecasting for hydrological runoff prediction. Two novel radar-based ensemble forecasting chains for flash-flood early warning are investigated in three catchments in the southern Swiss Alps and set in relation to deterministic discharge forecasts for the same catchments. The first radar-based ensemble forecasting chain is driven by NORA (Nowcasting of Orographic Rainfall by means of Analogues), an analogue-based heuristic nowcasting system to predict orographic rainfall for the following eight hours. The second ensemble forecasting system evaluated is REAL-C2, where the numerical weather prediction COSMO-2 is initialised with 25 different initial conditions derived from a four-day nowcast with the radar ensemble REAL. Additionally, three deterministic forecasting chains were analysed. The performance of these five flash-flood forecasting systems was analysed for 1389 h between June 2007 and December 2010 for which NORA forecasts were issued, due to the presence of orographic forcing. A clear preference was found for the ensemble approach. Discharge forecasts perform better when forced by NORA and REAL-C2 rather then by deterministic weather radar data. Moreover, it was observed that using an ensemble of initial conditions at the forecast initialisation, as in REAL-C2, significantly improved the forecast skill. These forecasts also perform better then forecasts forced by ensemble rainfall forecasts (NORA) initialised form a single initial condition of the hydrological model. Thus the best results were obtained with the REAL-C2 forecasting chain. However, for regions where REAL cannot be produced, NORA might be an option for forecasting events triggered by orographic precipitation.

Comparison of Filtering Methods for the Modeling and Retrospective Forecasting of Influenza Epidemics

PubMed Central

Yang, Wan; Karspeck, Alicia; Shaman, Jeffrey

2014-01-01

A variety of filtering methods enable the recursive estimation of system state variables and inference of model parameters. These methods have found application in a range of disciplines and settings, including engineering design and forecasting, and, over the last two decades, have been applied to infectious disease epidemiology. For any system of interest, the ideal filter depends on the nonlinearity and complexity of the model to which it is applied, the quality and abundance of observations being entrained, and the ultimate application (e.g. forecast, parameter estimation, etc.). Here, we compare the performance of six state-of-the-art filter methods when used to model and forecast influenza activity. Three particle filters—a basic particle filter (PF) with resampling and regularization, maximum likelihood estimation via iterated filtering (MIF), and particle Markov chain Monte Carlo (pMCMC)—and three ensemble filters—the ensemble Kalman filter (EnKF), the ensemble adjustment Kalman filter (EAKF), and the rank histogram filter (RHF)—were used in conjunction with a humidity-forced susceptible-infectious-recovered-susceptible (SIRS) model and weekly estimates of influenza incidence. The modeling frameworks, first validated with synthetic influenza epidemic data, were then applied to fit and retrospectively forecast the historical incidence time series of seven influenza epidemics during 2003–2012, for 115 cities in the United States. Results suggest that when using the SIRS model the ensemble filters and the basic PF are more capable of faithfully recreating historical influenza incidence time series, while the MIF and pMCMC do not perform as well for multimodal outbreaks. For forecast of the week with the highest influenza activity, the accuracies of the six model-filter frameworks are comparable; the three particle filters perform slightly better predicting peaks 1–5 weeks in the future; the ensemble filters are more accurate predicting peaks in the past. PMID:24762780

Stratospheric wind errors, initial states and forecast skill in the GLAS general circulation model

NASA Technical Reports Server (NTRS)

Tenenbaum, J.

1983-01-01

Relations between stratospheric wind errors, initial states and 500 mb skill are investigated using the GLAS general circulation model initialized with FGGE data. Erroneous stratospheric winds are seen in all current general circulation models, appearing also as weak shear above the subtropical jet and as cold polar stratospheres. In this study it is shown that the more anticyclonic large-scale flows are correlated with large forecast stratospheric winds. In addition, it is found that for North America the resulting errors are correlated with initial state jet stream accelerations while for East Asia the forecast winds are correlated with initial state jet strength. Using 500 mb skill scores over Europe at day 5 to measure forecast performance, it is found that both poor forecast skill and excessive stratospheric winds are correlated with more anticyclonic large-scale flows over North America. It is hypothesized that the resulting erroneous kinetic energy contributes to the poor forecast skill, and that the problem is caused by a failure in the modeling of the stratospheric energy cycle in current general circulation models independent of vertical resolution.

Short-term forecasting of meteorological time series using Nonparametric Functional Data Analysis (NPFDA)

NASA Astrophysics Data System (ADS)

Curceac, S.; Ternynck, C.; Ouarda, T.

2015-12-01

Over the past decades, a substantial amount of research has been conducted to model and forecast climatic variables. In this study, Nonparametric Functional Data Analysis (NPFDA) methods are applied to forecast air temperature and wind speed time series in Abu Dhabi, UAE. The dataset consists of hourly measurements recorded for a period of 29 years, 1982-2010. The novelty of the Functional Data Analysis approach is in expressing the data as curves. In the present work, the focus is on daily forecasting and the functional observations (curves) express the daily measurements of the above mentioned variables. We apply a non-linear regression model with a functional non-parametric kernel estimator. The computation of the estimator is performed using an asymmetrical quadratic kernel function for local weighting based on the bandwidth obtained by a cross validation procedure. The proximities between functional objects are calculated by families of semi-metrics based on derivatives and Functional Principal Component Analysis (FPCA). Additionally, functional conditional mode and functional conditional median estimators are applied and the advantages of combining their results are analysed. A different approach employs a SARIMA model selected according to the minimum Akaike (AIC) and Bayessian (BIC) Information Criteria and based on the residuals of the model. The performance of the models is assessed by calculating error indices such as the root mean square error (RMSE), relative RMSE, BIAS and relative BIAS. The results indicate that the NPFDA models provide more accurate forecasts than the SARIMA models. Key words: Nonparametric functional data analysis, SARIMA, time series forecast, air temperature, wind speed

Time series analysis of temporal trends in the pertussis incidence in Mainland China from 2005 to 2016.

PubMed

Zeng, Qianglin; Li, Dandan; Huang, Gui; Xia, Jin; Wang, Xiaoming; Zhang, Yamei; Tang, Wanping; Zhou, Hui

2016-08-31

Short-term forecast of pertussis incidence is helpful for advanced warning and planning resource needs for future epidemics. By utilizing the Auto-Regressive Integrated Moving Average (ARIMA) model and Exponential Smoothing (ETS) model as alterative models with R software, this paper analyzed data from Chinese Center for Disease Control and Prevention (China CDC) between January 2005 and June 2016. The ARIMA (0,1,0)(1,1,1)12 model (AICc = 1342.2 BIC = 1350.3) was selected as the best performing ARIMA model and the ETS (M,N,M) model (AICc = 1678.6, BIC = 1715.4) was selected as the best performing ETS model, and the ETS (M,N,M) model with the minimum RMSE was finally selected for in-sample-simulation and out-of-sample forecasting. Descriptive statistics showed that the reported number of pertussis cases by China CDC increased by 66.20% from 2005 (4058 cases) to 2015 (6744 cases). According to Hodrick-Prescott filter, there was an apparent cyclicity and seasonality in the pertussis reports. In out of sample forecasting, the model forecasted a relatively high incidence cases in 2016, which predicates an increasing risk of ongoing pertussis resurgence in the near future. In this regard, the ETS model would be a useful tool in simulating and forecasting the incidence of pertussis, and helping decision makers to take efficient decisions based on the advanced warning of disease incidence.

Ridge Polynomial Neural Network with Error Feedback for Time Series Forecasting

PubMed Central

Ghazali, Rozaida; Herawan, Tutut

2016-01-01

Time series forecasting has gained much attention due to its many practical applications. Higher-order neural network with recurrent feedback is a powerful technique that has been used successfully for time series forecasting. It maintains fast learning and the ability to learn the dynamics of the time series over time. Network output feedback is the most common recurrent feedback for many recurrent neural network models. However, not much attention has been paid to the use of network error feedback instead of network output feedback. In this study, we propose a novel model, called Ridge Polynomial Neural Network with Error Feedback (RPNN-EF) that incorporates higher order terms, recurrence and error feedback. To evaluate the performance of RPNN-EF, we used four univariate time series with different forecasting horizons, namely star brightness, monthly smoothed sunspot numbers, daily Euro/Dollar exchange rate, and Mackey-Glass time-delay differential equation. We compared the forecasting performance of RPNN-EF with the ordinary Ridge Polynomial Neural Network (RPNN) and the Dynamic Ridge Polynomial Neural Network (DRPNN). Simulation results showed an average 23.34% improvement in Root Mean Square Error (RMSE) with respect to RPNN and an average 10.74% improvement with respect to DRPNN. That means that using network errors during training helps enhance the overall forecasting performance for the network. PMID:27959927

Ridge Polynomial Neural Network with Error Feedback for Time Series Forecasting.

PubMed

Waheeb, Waddah; Ghazali, Rozaida; Herawan, Tutut

2016-01-01

Time series forecasting has gained much attention due to its many practical applications. Higher-order neural network with recurrent feedback is a powerful technique that has been used successfully for time series forecasting. It maintains fast learning and the ability to learn the dynamics of the time series over time. Network output feedback is the most common recurrent feedback for many recurrent neural network models. However, not much attention has been paid to the use of network error feedback instead of network output feedback. In this study, we propose a novel model, called Ridge Polynomial Neural Network with Error Feedback (RPNN-EF) that incorporates higher order terms, recurrence and error feedback. To evaluate the performance of RPNN-EF, we used four univariate time series with different forecasting horizons, namely star brightness, monthly smoothed sunspot numbers, daily Euro/Dollar exchange rate, and Mackey-Glass time-delay differential equation. We compared the forecasting performance of RPNN-EF with the ordinary Ridge Polynomial Neural Network (RPNN) and the Dynamic Ridge Polynomial Neural Network (DRPNN). Simulation results showed an average 23.34% improvement in Root Mean Square Error (RMSE) with respect to RPNN and an average 10.74% improvement with respect to DRPNN. That means that using network errors during training helps enhance the overall forecasting performance for the network.

A comparison of the stochastic and machine learning approaches in hydrologic time series forecasting

NASA Astrophysics Data System (ADS)

Kim, T.; Joo, K.; Seo, J.; Heo, J. H.

2016-12-01

Hydrologic time series forecasting is an essential task in water resources management and it becomes more difficult due to the complexity of runoff process. Traditional stochastic models such as ARIMA family has been used as a standard approach in time series modeling and forecasting of hydrological variables. Due to the nonlinearity in hydrologic time series data, machine learning approaches has been studied with the advantage of discovering relevant features in a nonlinear relation among variables. This study aims to compare the predictability between the traditional stochastic model and the machine learning approach. Seasonal ARIMA model was used as the traditional time series model, and Random Forest model which consists of decision tree and ensemble method using multiple predictor approach was applied as the machine learning approach. In the application, monthly inflow data from 1986 to 2015 of Chungju dam in South Korea were used for modeling and forecasting. In order to evaluate the performances of the used models, one step ahead and multi-step ahead forecasting was applied. Root mean squared error and mean absolute error of two models were compared.

Metrics for the Evaluation the Utility of Air Quality Forecasting

NASA Astrophysics Data System (ADS)

Sumo, T. M.; Stockwell, W. R.

2013-12-01

Global warming is expected to lead to higher levels of air pollution and therefore the forecasting of both long-term and daily air quality is an important component for the assessment of the costs of climate change and its impact on human health. Some of the risks associated with poor air quality days (where the Air Pollution Index is greater than 100), include hospital visits and mortality. Accurate air quality forecasting has the potential to allow sensitive groups to take appropriate precautions. This research builds metrics for evaluating the utility of air quality forecasting in terms of its potential impacts. Our analysis of air quality models focuses on the Washington, DC/Baltimore, MD region over the summertime ozone seasons between 2010 and 2012. The metrics that are relevant to our analysis include: (1) The number of times that a high ozone or particulate matter (PM) episode is correctly forecasted, (2) the number of times that high ozone or PM episode is forecasted when it does not occur and (3) the number of times when the air quality forecast predicts a cleaner air episode when the air was observed to have high ozone or PM. Our evaluation of the performance of air quality forecasts include those forecasts of ozone and particulate matter and data available from the U.S. Environmental Protection Agency (EPA)'s AIRNOW. We also examined observational ozone and particulate matter data available from Clean Air Partners. Overall the forecast models perform well for our region and time interval.

Evaluation of ensemble precipitation forecasts generated through post-processing in a Canadian catchment

NASA Astrophysics Data System (ADS)

Jha, Sanjeev K.; Shrestha, Durga L.; Stadnyk, Tricia A.; Coulibaly, Paulin

2018-03-01

Flooding in Canada is often caused by heavy rainfall during the snowmelt period. Hydrologic forecast centers rely on precipitation forecasts obtained from numerical weather prediction (NWP) models to enforce hydrological models for streamflow forecasting. The uncertainties in raw quantitative precipitation forecasts (QPFs) are enhanced by physiography and orography effects over a diverse landscape, particularly in the western catchments of Canada. A Bayesian post-processing approach called rainfall post-processing (RPP), developed in Australia (Robertson et al., 2013; Shrestha et al., 2015), has been applied to assess its forecast performance in a Canadian catchment. Raw QPFs obtained from two sources, Global Ensemble Forecasting System (GEFS) Reforecast 2 project, from the National Centers for Environmental Prediction, and Global Deterministic Forecast System (GDPS), from Environment and Climate Change Canada, are used in this study. The study period from January 2013 to December 2015 covered a major flood event in Calgary, Alberta, Canada. Post-processed results show that the RPP is able to remove the bias and reduce the errors of both GEFS and GDPS forecasts. Ensembles generated from the RPP reliably quantify the forecast uncertainty.

Addressing forecast uncertainty impact on CSP annual performance

NASA Astrophysics Data System (ADS)

Ferretti, Fabio; Hogendijk, Christopher; Aga, Vipluv; Ehrsam, Andreas

2017-06-01

This work analyzes the impact of weather forecast uncertainty on the annual performance of a Concentrated Solar Power (CSP) plant. Forecast time series has been produced by a commercial forecast provider using the technique of hindcasting for the full year 2011 in hourly resolution for Ouarzazate, Morocco. Impact of forecast uncertainty has been measured on three case studies, representing typical tariff schemes observed in recent CSP projects plus a spot market price scenario. The analysis has been carried out using an annual performance model and a standard dispatch optimization algorithm based on dynamic programming. The dispatch optimizer has been demonstrated to be a key requisite to maximize the annual revenues depending on the price scenario, harvesting the maximum potential out of the CSP plant. Forecasting uncertainty affects the revenue enhancement outcome of a dispatch optimizer depending on the error level and the price function. Results show that forecasting accuracy of direct solar irradiance (DNI) is important to make best use of an optimized dispatch but also that a higher number of calculation updates can partially compensate this uncertainty. Improvement in revenues can be significant depending on the price profile and the optimal operation strategy. Pathways to achieve better performance are presented by having more updates both by repeatedly generating new optimized trajectories but also more often updating weather forecasts. This study shows the importance of working on DNI weather forecasting for revenue enhancement as well as selecting weather services that can provide multiple updates a day and probabilistic forecast information.

Evaluating simplified methods for liquefaction assessment for loss estimation

NASA Astrophysics Data System (ADS)

Kongar, Indranil; Rossetto, Tiziana; Giovinazzi, Sonia

2017-06-01

Currently, some catastrophe models used by the insurance industry account for liquefaction by applying a simple factor to shaking-induced losses. The factor is based only on local liquefaction susceptibility and this highlights the need for a more sophisticated approach to incorporating the effects of liquefaction in loss models. This study compares 11 unique models, each based on one of three principal simplified liquefaction assessment methods: liquefaction potential index (LPI) calculated from shear-wave velocity, the HAZUS software method and a method created specifically to make use of USGS remote sensing data. Data from the September 2010 Darfield and February 2011 Christchurch earthquakes in New Zealand are used to compare observed liquefaction occurrences to forecasts from these models using binary classification performance measures. The analysis shows that the best-performing model is the LPI calculated using known shear-wave velocity profiles, which correctly forecasts 78 % of sites where liquefaction occurred and 80 % of sites where liquefaction did not occur, when the threshold is set at 7. However, these data may not always be available to insurers. The next best model is also based on LPI but uses shear-wave velocity profiles simulated from the combination of USGS VS30 data and empirical functions that relate VS30 to average shear-wave velocities at shallower depths. This model correctly forecasts 58 % of sites where liquefaction occurred and 84 % of sites where liquefaction did not occur, when the threshold is set at 4. These scores increase to 78 and 86 %, respectively, when forecasts are based on liquefaction probabilities that are empirically related to the same values of LPI. This model is potentially more useful for insurance since the input data are publicly available. HAZUS models, which are commonly used in studies where no local model is available, perform poorly and incorrectly forecast 87 % of sites where liquefaction occurred, even at optimal thresholds. This paper also considers two models (HAZUS and EPOLLS) for estimation of the scale of liquefaction in terms of permanent ground deformation but finds that both models perform poorly, with correlations between observations and forecasts lower than 0.4 in all cases. Therefore these models potentially provide negligible additional value to loss estimation analysis outside of the regions for which they have been developed.

Forecasting urban PM10 and PM2.5 pollution episodes in very stable nocturnal conditions and complex terrain using WRF-Chem CO tracer model

NASA Astrophysics Data System (ADS)

Saide, Pablo E.; Carmichael, Gregory R.; Spak, Scott N.; Gallardo, Laura; Osses, Axel E.; Mena-Carrasco, Marcelo A.; Pagowski, Mariusz

2011-05-01

This study presents a system to predict high pollution events that develop in connection with enhanced subsidence due to coastal lows, particularly in winter over Santiago de Chile. An accurate forecast of these episodes is of interest since the local government is entitled by law to take actions in advance to prevent public exposure to PM10 concentrations in excess of 150 μg m -3 (24 h running averages). The forecasting system is based on accurately simulating carbon monoxide (CO) as a PM10/PM2.5 surrogate, since during episodes and within the city there is a high correlation (over 0.95) among these pollutants. Thus, by accurately forecasting CO, which behaves closely to a tracer on this scale, a PM estimate can be made without involving aerosol-chemistry modeling. Nevertheless, the very stable nocturnal conditions over steep topography associated with maxima in concentrations are hard to represent in models. Here we propose a forecast system based on the WRF-Chem model with optimum settings, determined through extensive testing, that best describe both meteorological and air quality available measurements. Some of the important configurations choices involve the boundary layer (PBL) scheme, model grid resolution (both vertical and horizontal), meteorological initial and boundary conditions and spatial and temporal distribution of the emissions. A forecast for the 2008 winter is performed showing that this forecasting system is able to perform similarly to the authority decision for PM10 and better than persistence when forecasting PM10 and PM2.5 high pollution episodes. Problems regarding false alarm predictions could be related to different uncertainties in the model such as day to day emission variability, inability of the model to completely resolve the complex topography and inaccuracy in meteorological initial and boundary conditions. Finally, according to our simulations, emissions from previous days dominate episode concentrations, which highlights the need for 48 h forecasts that can be achieved by the system presented here. This is in fact the largest advantage of the proposed system.

Ensemble flare forecasting: using numerical weather prediction techniques to improve space weather operations

NASA Astrophysics Data System (ADS)

Murray, S.; Guerra, J. A.

2017-12-01

One essential component of operational space weather forecasting is the prediction of solar flares. Early flare forecasting work focused on statistical methods based on historical flaring rates, but more complex machine learning methods have been developed in recent years. A multitude of flare forecasting methods are now available, however it is still unclear which of these methods performs best, and none are substantially better than climatological forecasts. Current operational space weather centres cannot rely on automated methods, and generally use statistical forecasts with a little human intervention. Space weather researchers are increasingly looking towards methods used in terrestrial weather to improve current forecasting techniques. Ensemble forecasting has been used in numerical weather prediction for many years as a way to combine different predictions in order to obtain a more accurate result. It has proved useful in areas such as magnetospheric modelling and coronal mass ejection arrival analysis, however has not yet been implemented in operational flare forecasting. Here we construct ensemble forecasts for major solar flares by linearly combining the full-disk probabilistic forecasts from a group of operational forecasting methods (ASSA, ASAP, MAG4, MOSWOC, NOAA, and Solar Monitor). Forecasts from each method are weighted by a factor that accounts for the method's ability to predict previous events, and several performance metrics (both probabilistic and categorical) are considered. The results provide space weather forecasters with a set of parameters (combination weights, thresholds) that allow them to select the most appropriate values for constructing the 'best' ensemble forecast probability value, according to the performance metric of their choice. In this way different forecasts can be made to fit different end-user needs.

Short-range solar radiation forecasts over Sweden

NASA Astrophysics Data System (ADS)

Landelius, Tomas; Lindskog, Magnus; Körnich, Heiner; Andersson, Sandra

2018-04-01

In this article the performance for short-range solar radiation forecasts by the global deterministic and ensemble models from the European Centre for Medium-Range Weather Forecasts (ECMWF) is compared with an ensemble of the regional mesoscale model HARMONIE-AROME used by the national meteorological services in Sweden, Norway and Finland. Note however that only the control members and the ensemble means are included in the comparison. The models resolution differs considerably with 18 km for the ECMWF ensemble, 9 km for the ECMWF deterministic model, and 2.5 km for the HARMONIE-AROME ensemble. The models share the same radiation code. It turns out that they all underestimate systematically the Direct Normal Irradiance (DNI) for clear-sky conditions. Except for this shortcoming, the HARMONIE-AROME ensemble model shows the best agreement with the distribution of observed Global Horizontal Irradiance (GHI) and DNI values. During mid-day the HARMONIE-AROME ensemble mean performs best. The control member of the HARMONIE-AROME ensemble also scores better than the global deterministic ECMWF model. This is an interesting result since mesoscale models have so far not shown good results when compared to the ECMWF models. Three days with clear, mixed and cloudy skies are used to illustrate the possible added value of a probabilistic forecast. It is shown that in these cases the mesoscale ensemble could provide decision support to a grid operator in terms of forecasts of both the amount of solar power and its probabilities.

Challenges for operational forecasting and early warning of rainfall induced landslides

NASA Astrophysics Data System (ADS)

Guzzetti, Fausto

2017-04-01

In many areas of the world, landslides occur every year, claiming lives and producing severe economic and environmental damage. Many of the landslides with human or economic consequences are the result of intense or prolonged rainfall. For this reason, in many areas the timely forecast of rainfall-induced landslides is of both scientific interest and social relevance. In the recent years, there has been a mounting interest and an increasing demand for operational landslide forecasting, and for associated landslide early warning systems. Despite the relevance of the problem, and the increasing interest and demand, only a few systems have been designed, and are currently operated. Inspection of the - limited - literature on operational landslide forecasting, and on the associated early warning systems, reveals that common criteria and standards for the design, the implementation, the operation, and the evaluation of the performances of the systems, are lacking. This limits the possibility to compare and to evaluate the systems critically, to identify their inherent strengths and weaknesses, and to improve the performance of the systems. Lack of common criteria and of established standards can also limit the credibility of the systems, and consequently their usefulness and potential practical impact. Landslides are very diversified phenomena, and the information and the modelling tools used to attempt landslide forecasting vary largely, depending on the type and size of the landslides, the extent of the geographical area considered, the timeframe of the forecasts, and the scope of the predictions. Consequently, systems for landslide forecasting and early warning can be designed and implemented at several different geographical scales, from the local (site or slope specific) to the regional, or even national scale. The talk focuses on regional to national scale landslide forecasting systems, and specifically on operational systems based on empirical rainfall threshold models. Building on the experience gained in designing, implementing, and operating national and regional landslide forecasting systems in Italy, and on a preliminary review of the existing literature on regional landslide early warning systems, the talk discusses concepts, limitations and challenges inherent to the design of reliable forecasting and early warning systems for rainfall-triggered landslides, the evaluation of the performances of the systems, and on problems related to the use of the forecasts and the issuing of landslide warnings. Several of the typical elements of an operational landslide forecasting system are considered, including: (i) the rainfall and landslide information used to establish the threshold models, (ii) the methods and tools used to define the empirical rainfall thresholds, and their associated uncertainty, (iii) the quality (e.g., the temporal and spatial resolution) of the rainfall information used for operational forecasting, including rain gauge and radar measurements, satellite estimates, and quantitative weather forecasts, (iv) the ancillary information used to prepare the forecasts, including e.g., the terrain subdivisions and the landslide susceptibility zonations, (v) the criteria used to transform the forecasts into landslide warnings and the methods used to communicate the warnings, and (vi) the criteria and strategies adopted to evaluate the performances of the systems, and to define minimum or optimal performance levels.

Evaluation of medium-range ensemble flood forecasting based on calibration strategies and ensemble methods in Lanjiang Basin, Southeast China

NASA Astrophysics Data System (ADS)

Liu, Li; Gao, Chao; Xuan, Weidong; Xu, Yue-Ping

2017-11-01

Ensemble flood forecasts by hydrological models using numerical weather prediction products as forcing data are becoming more commonly used in operational flood forecasting applications. In this study, a hydrological ensemble flood forecasting system comprised of an automatically calibrated Variable Infiltration Capacity model and quantitative precipitation forecasts from TIGGE dataset is constructed for Lanjiang Basin, Southeast China. The impacts of calibration strategies and ensemble methods on the performance of the system are then evaluated. The hydrological model is optimized by the parallel programmed ε-NSGA II multi-objective algorithm. According to the solutions by ε-NSGA II, two differently parameterized models are determined to simulate daily flows and peak flows at each of the three hydrological stations. Then a simple yet effective modular approach is proposed to combine these daily and peak flows at the same station into one composite series. Five ensemble methods and various evaluation metrics are adopted. The results show that ε-NSGA II can provide an objective determination on parameter estimation, and the parallel program permits a more efficient simulation. It is also demonstrated that the forecasts from ECMWF have more favorable skill scores than other Ensemble Prediction Systems. The multimodel ensembles have advantages over all the single model ensembles and the multimodel methods weighted on members and skill scores outperform other methods. Furthermore, the overall performance at three stations can be satisfactory up to ten days, however the hydrological errors can degrade the skill score by approximately 2 days, and the influence persists until a lead time of 10 days with a weakening trend. With respect to peak flows selected by the Peaks Over Threshold approach, the ensemble means from single models or multimodels are generally underestimated, indicating that the ensemble mean can bring overall improvement in forecasting of flows. For peak values taking flood forecasts from each individual member into account is more appropriate.

Water and Power Systems Co-optimization under a High Performance Computing Framework

NASA Astrophysics Data System (ADS)

Xuan, Y.; Arumugam, S.; DeCarolis, J.; Mahinthakumar, K.

2016-12-01

Water and energy systems optimizations are traditionally being treated as two separate processes, despite their intrinsic interconnections (e.g., water is used for hydropower generation, and thermoelectric cooling requires a large amount of water withdrawal). Given the challenges of urbanization, technology uncertainty and resource constraints, and the imminent threat of climate change, a cyberinfrastructure is needed to facilitate and expedite research into the complex management of these two systems. To address these issues, we developed a High Performance Computing (HPC) framework for stochastic co-optimization of water and energy resources to inform water allocation and electricity demand. The project aims to improve conjunctive management of water and power systems under climate change by incorporating improved ensemble forecast models of streamflow and power demand. First, by downscaling and spatio-temporally disaggregating multimodel climate forecasts from General Circulation Models (GCMs), temperature and precipitation forecasts are obtained and input into multi-reservoir and power systems models. Extended from Optimus (Optimization Methods for Universal Simulators), the framework drives the multi-reservoir model and power system model, Temoa (Tools for Energy Model Optimization and Analysis), and uses Particle Swarm Optimization (PSO) algorithm to solve high dimensional stochastic problems. The utility of climate forecasts on the cost of water and power systems operations is assessed and quantified based on different forecast scenarios (i.e., no-forecast, multimodel forecast and perfect forecast). Analysis of risk management actions and renewable energy deployments will be investigated for the Catawba River basin, an area with adequate hydroclimate predicting skill and a critical basin with 11 reservoirs that supplies water and generates power for both North and South Carolina. Further research using this scalable decision supporting framework will provide understanding and elucidate the intricate and interdependent relationship between water and energy systems and enhance the security of these two critical public infrastructures.

Small-time Scale Network Traffic Prediction Based on Complex-valued Neural Network

NASA Astrophysics Data System (ADS)

Yang, Bin

2017-07-01

Accurate models play an important role in capturing the significant characteristics of the network traffic, analyzing the network dynamic, and improving the forecasting accuracy for system dynamics. In this study, complex-valued neural network (CVNN) model is proposed to further improve the accuracy of small-time scale network traffic forecasting. Artificial bee colony (ABC) algorithm is proposed to optimize the complex-valued and real-valued parameters of CVNN model. Small-scale traffic measurements data namely the TCP traffic data is used to test the performance of CVNN model. Experimental results reveal that CVNN model forecasts the small-time scale network traffic measurement data very accurately

CSEP-Japan: The Japanese node of the collaboratory for the study of earthquake predictability

NASA Astrophysics Data System (ADS)

Yokoi, S.; Tsuruoka, H.; Nanjo, K.; Hirata, N.

2011-12-01

Collaboratory for the Study of Earthquake Predictability (CSEP) is a global project of earthquake predictability research. The final goal of this project is to have a look for the intrinsic predictability of the earthquake rupture process through forecast testing experiments. The Earthquake Research Institute, the University of Tokyo joined the CSEP and started the Japanese testing center called as CSEP-Japan. This testing center constitutes an open access to researchers contributing earthquake forecast models for applied to Japan. A total of 91 earthquake forecast models were submitted on the prospective experiment starting from 1 November 2009. The models are separated into 4 testing classes (1 day, 3 months, 1 year and 3 years) and 3 testing regions covering an area of Japan including sea area, Japanese mainland and Kanto district. We evaluate the performance of the models in the official suite of tests defined by the CSEP. The experiments of 1-day, 3-month, 1-year and 3-year forecasting classes were implemented for 92 rounds, 4 rounds, 1round and 0 round (now in progress), respectively. The results of the 3-month class gave us new knowledge concerning statistical forecasting models. All models showed a good performance for magnitude forecasting. On the other hand, observation is hardly consistent in space-distribution with most models in some cases where many earthquakes occurred at the same spot. Throughout the experiment, it has been clarified that some properties of the CSEP's evaluation tests such as the L-test show strong correlation with the N-test. We are now processing to own (cyber-) infrastructure to support the forecast experiment as follows. (1) Japanese seismicity has changed since the 2011 Tohoku earthquake. The 3rd call for forecasting models was announced in order to promote model improvement for forecasting earthquakes after this earthquake. So, we provide Japanese seismicity catalog maintained by JMA for modelers to study how seismicity changes in Japan. (2) Now we prepare the 3-D forecasting experiment with a depth range of 0 to 100 km in Kanto region. (3) The testing center improved an evaluation system for 1-day class experiment because this testing class required fast calculation ability to finish forecasting and testing results within one day. This development will make a real-time forecasting system come true. (4) The special issue of 1st part titled Earthquake Forecast Testing Experiment in Japan was published on the Earth, Planets and Space Vol. 63, No.3, 2011 on March, 2011. This issue includes papers of algorithm of statistical models participating our experiment and outline of the experiment in Japan. The 2nd part of this issue, which is now on line, will be published soon. In this presentation, we will overview CSEP-Japan and results of the experiments, and discuss direction of our activity. An outline of the experiment and activities of the Japanese Testing Center are published on our WEB site;

Verification of Advances in a Coupled Snow-runoff Modeling Framework for Operational Streamflow Forecasts

NASA Astrophysics Data System (ADS)

Barik, M. G.; Hogue, T. S.; Franz, K. J.; He, M.

2011-12-01

The National Oceanic and Atmospheric Administration's (NOAA's) River Forecast Centers (RFCs) issue hydrologic forecasts related to flood events, reservoir operations for water supply, streamflow regulation, and recreation on the nation's streams and rivers. The RFCs use the National Weather Service River Forecast System (NWSRFS) for streamflow forecasting which relies on a coupled snow model (i.e. SNOW17) and rainfall-runoff model (i.e. SAC-SMA) in snow-dominated regions of the US. Errors arise in various steps of the forecasting system from input data, model structure, model parameters, and initial states. The goal of the current study is to undertake verification of potential improvements in the SNOW17-SAC-SMA modeling framework developed for operational streamflow forecasts. We undertake verification for a range of parameters sets (i.e. RFC, DREAM (Differential Evolution Adaptive Metropolis)) as well as a data assimilation (DA) framework developed for the coupled models. Verification is also undertaken for various initial conditions to observe the influence of variability in initial conditions on the forecast. The study basin is the North Fork America River Basin (NFARB) located on the western side of the Sierra Nevada Mountains in northern California. Hindcasts are verified using both deterministic (i.e. Nash Sutcliffe efficiency, root mean square error, and joint distribution) and probabilistic (i.e. reliability diagram, discrimination diagram, containing ratio, and Quantile plots) statistics. Our presentation includes comparison of the performance of different optimized parameters and the DA framework as well as assessment of the impact associated with the initial conditions used for streamflow forecasts for the NFARB.

Forecasting European Droughts using the North American Multi-Model Ensemble (NMME)

NASA Astrophysics Data System (ADS)

Thober, Stephan; Kumar, Rohini; Samaniego, Luis; Sheffield, Justin; Schäfer, David; Mai, Juliane

2015-04-01

Soil moisture droughts have the potential to diminish crop yields causing economic damage or even threatening the livelihood of societies. State-of-the-art drought forecasting systems incorporate seasonal meteorological forecasts to estimate future drought conditions. Meteorological forecasting skill (in particular that of precipitation), however, is limited to a few weeks because of the chaotic behaviour of the atmosphere. One of the most important challenges in drought forecasting is to understand how the uncertainty in the atmospheric forcings (e.g., precipitation and temperature) is further propagated into hydrologic variables such as soil moisture. The North American Multi-Model Ensemble (NMME) provides the latest collection of a multi-institutional seasonal forecasting ensemble for precipitation and temperature. In this study, we analyse the skill of NMME forecasts for predicting European drought events. The monthly NMME forecasts are downscaled to daily values to force the mesoscale hydrological model (mHM). The mHM soil moisture forecasts obtained with the forcings of the dynamical models are then compared against those obtained with the Ensemble Streamflow Prediction (ESP) approach. ESP recombines historical meteorological forcings to create a new ensemble forecast. Both forecasts are compared against reference soil moisture conditions obtained using observation based meteorological forcings. The study is conducted for the period from 1982 to 2009 and covers a large part of the Pan-European domain (10°W to 40°E and 35°N to 55°N). Results indicate that NMME forecasts are better at predicting the reference soil moisture variability as compared to ESP. For example, NMME explains 50% of the variability in contrast to only 31% by ESP at a six-month lead time. The Equitable Threat Skill Score (ETS), which combines the hit and false alarm rates, is analysed for drought events using a 0.2 threshold of a soil moisture percentile index. On average, the NMME based ensemble forecasts have consistently higher skill than the ESP based ones (ETS of 13% as compared to 5% at a six-month lead time). Additionally, the ETS ensemble spread of NMME forecasts is considerably narrower than that of ESP; the lower boundary of the NMME ensemble spread coincides most of the time with the ensemble median of ESP. Among the NMME models, NCEP-CFSv2 outperforms the other models in terms of ETS most of the time. Removing the three worst performing models does not deteriorate the ensemble performance (neither in skill nor in spread), but would substantially reduce the computational resources required in an operational forecasting system. For major European drought events (e.g., 1990, 1992, 2003, and 2007), NMME forecasts tend to underestimate area under drought and drought magnitude during times of drought development. During drought recovery, this underestimation is weaker for area under drought or even reversed into an overestimation for drought magnitude. This indicates that the NMME models are too wet during drought development and too dry during drought recovery. In summary, soil moisture drought forecasts by NMME are more skillful than those of an ESP based approach. However, they still show systematic biases in reproducing the observed drought dynamics during drought development and recovery.

Long-term ensemble forecast of snowmelt inflow into the Cheboksary Reservoir under two different weather scenarios

NASA Astrophysics Data System (ADS)

Gelfan, Alexander; Moreydo, Vsevolod; Motovilov, Yury; Solomatine, Dimitri P.

2018-04-01

A long-term forecasting ensemble methodology, applied to water inflows into the Cheboksary Reservoir (Russia), is presented. The methodology is based on a version of the semi-distributed hydrological model ECOMAG (ECOlogical Model for Applied Geophysics) that allows for the calculation of an ensemble of inflow hydrographs using two different sets of weather ensembles for the lead time period: observed weather data, constructed on the basis of the Ensemble Streamflow Prediction methodology (ESP-based forecast), and synthetic weather data, simulated by a multi-site weather generator (WG-based forecast). We have studied the following: (1) whether there is any advantage of the developed ensemble forecasts in comparison with the currently issued operational forecasts of water inflow into the Cheboksary Reservoir, and (2) whether there is any noticeable improvement in probabilistic forecasts when using the WG-simulated ensemble compared to the ESP-based ensemble. We have found that for a 35-year period beginning from the reservoir filling in 1982, both continuous and binary model-based ensemble forecasts (issued in the deterministic form) outperform the operational forecasts of the April-June inflow volume actually used and, additionally, provide acceptable forecasts of additional water regime characteristics besides the inflow volume. We have also demonstrated that the model performance measures (in the verification period) obtained from the WG-based probabilistic forecasts, which are based on a large number of possible weather scenarios, appeared to be more statistically reliable than the corresponding measures calculated from the ESP-based forecasts based on the observed weather scenarios.

Uncertainty in flood forecasting: A distributed modeling approach in a sparse data catchment

NASA Astrophysics Data System (ADS)

Mendoza, Pablo A.; McPhee, James; Vargas, Ximena

2012-09-01

Data scarcity has traditionally precluded the application of advanced hydrologic techniques in developing countries. In this paper, we evaluate the performance of a flood forecasting scheme in a sparsely monitored catchment based on distributed hydrologic modeling, discharge assimilation, and numerical weather predictions with explicit validation uncertainty analysis. For the hydrologic component of our framework, we apply TopNet to the Cautin River basin, located in southern Chile, using a fully distributed a priori parameterization based on both literature-suggested values and data gathered during field campaigns. Results obtained from this step indicate that the incremental effort spent in measuring directly a set of model parameters was insufficient to represent adequately the most relevant hydrologic processes related to spatiotemporal runoff patterns. Subsequent uncertainty validation performed over a six month ensemble simulation shows that streamflow uncertainty is better represented during flood events, due to both the increase of state perturbation introduced by rainfall and the flood-oriented calibration strategy adopted here. Results from different assimilation configurations suggest that the upper part of the basin is the major source of uncertainty in hydrologic process representation and hint at the usefulness of interpreting assimilation results in terms of model input and parameterization inadequacy. Furthermore, in this case study the violation of Markovian state properties by the Ensemble Kalman filter did affect the numerical results, showing that an explicit treatment of the time delay between the generation of surface runoff and the arrival at the basin outlet is required in the assimilation scheme. Peak flow forecasting results demonstrate that there is a major problem with the Weather Research and Forecasting model outputs, which systematically overestimate precipitation over the catchment. A final analysis performed for a large flooding event that occurred in July 2006 shows that, in the absence of bias introduced by an incorrect model calibration, the updating of both model states and meteorological forecasts contributes to a better representation of streamflow uncertainty and to better hydrologic forecasts.

The Automation of Nowcast Model Assessment Processes

DTIC Science & Technology

2016-09-01

that will automate real-time WRE-N model simulations, collect and quality control check weather observations for assimilation and verification, and...domains centered near White Sands Missile Range, New Mexico, where the Meteorological Sensor Array (MSA) will be located. The MSA will provide...observations and performing quality -control checks for the pre-forecast data assimilation period. 2. Run the WRE-N model to generate model forecast data

Development of seasonal flow outlook model for Ganges-Brahmaputra Basins in Bangladesh

NASA Astrophysics Data System (ADS)

Hossain, Sazzad; Haque Khan, Raihanul; Gautum, Dilip Kumar; Karmaker, Ripon; Hossain, Amirul

2016-10-01

Bangladesh is crisscrossed by the branches and tributaries of three main river systems, the Ganges, Bramaputra and Meghna (GBM). The temporal variation of water availability of those rivers has an impact on the different water usages such as irrigation, urban water supply, hydropower generation, navigation etc. Thus, seasonal flow outlook can play important role in various aspects of water management. The Flood Forecasting and Warning Center (FFWC) in Bangladesh provides short term and medium term flood forecast, and there is a wide demand from end-users about seasonal flow outlook for agricultural purposes. The objective of this study is to develop a seasonal flow outlook model in Bangladesh based on rainfall forecast. It uses European Centre for Medium-Range Weather Forecasts (ECMWF) seasonal precipitation, temperature forecast to simulate HYDROMAD hydrological model. Present study is limited for Ganges and Brahmaputra River Basins. ARIMA correction is applied to correct the model error. The performance of the model is evaluated using coefficient of determination (R2) and Nash-Sutcliffe Efficiency (NSE). The model result shows good performance with R2 value of 0.78 and NSE of 0.61 for the Brahmaputra River Basin, and R2 value of 0.72 and NSE of 0.59 for the Ganges River Basin for the period of May to July 2015. The result of the study indicates strong potential to make seasonal outlook to be operationalized.

The diagnosis and forecast system of hydrometeorological characteristics for the White, Barents, Kara and Pechora Seas

NASA Astrophysics Data System (ADS)

Fomin, Vladimir; Diansky, Nikolay; Gusev, Anatoly; Kabatchenko, Ilia; Panasenkova, Irina

2017-04-01

The diagnosis and forecast system for simulating hydrometeorological characteristics of the Russian Western Arctic seas is presented. It performs atmospheric forcing computation with the regional non-hydrostatic atmosphere model Weather Research and Forecasting model (WRF) with spatial resolution 15 km, as well as computation of circulation, sea level, temperature, salinity and sea ice with the marine circulation model INMOM (Institute of Numerical Mathematics Ocean Model) with spatial resolution 2.7 km, and the computation of wind wave parameters using the Russian wind-wave model (RWWM) with spatial resolution 5 km. Verification of the meteorological characteristics is done for air temperature, air pressure, wind velocity, water temperature, currents, sea level anomaly, wave characteristics such as wave height and wave period. The results of the hydrometeorological characteristic verification are presented for both retrospective and forecast computations. The retrospective simulation of the hydrometeorological characteristics for the White, Barents, Kara and Pechora Seas was performed with the diagnosis and forecast system for the period 1986-2015. The important features of the Kara Sea circulation are presented. Water exchange between Pechora and Kara Seas is described. The importance is shown of using non-hydrostatic atmospheric circulation model for the atmospheric forcing computation in coastal areas. According to the computation results, extreme values of hydrometeorological characteristics were obtained for the Russian Western Arctic seas.

Nonmechanistic forecasts of seasonal influenza with iterative one-week-ahead distributions.

PubMed

Brooks, Logan C; Farrow, David C; Hyun, Sangwon; Tibshirani, Ryan J; Rosenfeld, Roni

2018-06-15

Accurate and reliable forecasts of seasonal epidemics of infectious disease can assist in the design of countermeasures and increase public awareness and preparedness. This article describes two main contributions we made recently toward this goal: a novel approach to probabilistic modeling of surveillance time series based on "delta densities", and an optimization scheme for combining output from multiple forecasting methods into an adaptively weighted ensemble. Delta densities describe the probability distribution of the change between one observation and the next, conditioned on available data; chaining together nonparametric estimates of these distributions yields a model for an entire trajectory. Corresponding distributional forecasts cover more observed events than alternatives that treat the whole season as a unit, and improve upon multiple evaluation metrics when extracting key targets of interest to public health officials. Adaptively weighted ensembles integrate the results of multiple forecasting methods, such as delta density, using weights that can change from situation to situation. We treat selection of optimal weightings across forecasting methods as a separate estimation task, and describe an estimation procedure based on optimizing cross-validation performance. We consider some details of the data generation process, including data revisions and holiday effects, both in the construction of these forecasting methods and when performing retrospective evaluation. The delta density method and an adaptively weighted ensemble of other forecasting methods each improve significantly on the next best ensemble component when applied separately, and achieve even better cross-validated performance when used in conjunction. We submitted real-time forecasts based on these contributions as part of CDC's 2015/2016 FluSight Collaborative Comparison. Among the fourteen submissions that season, this system was ranked by CDC as the most accurate.

Using a Hybrid Model to Forecast the Prevalence of Schistosomiasis in Humans.

PubMed

Zhou, Lingling; Xia, Jing; Yu, Lijing; Wang, Ying; Shi, Yun; Cai, Shunxiang; Nie, Shaofa

2016-03-23

We previously proposed a hybrid model combining both the autoregressive integrated moving average (ARIMA) and the nonlinear autoregressive neural network (NARNN) models in forecasting schistosomiasis. Our purpose in the current study was to forecast the annual prevalence of human schistosomiasis in Yangxin County, using our ARIMA-NARNN model, thereby further certifying the reliability of our hybrid model. We used the ARIMA, NARNN and ARIMA-NARNN models to fit and forecast the annual prevalence of schistosomiasis. The modeling time range included was the annual prevalence from 1956 to 2008 while the testing time range included was from 2009 to 2012. The mean square error (MSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) were used to measure the model performance. We reconstructed the hybrid model to forecast the annual prevalence from 2013 to 2016. The modeling and testing errors generated by the ARIMA-NARNN model were lower than those obtained from either the single ARIMA or NARNN models. The predicted annual prevalence from 2013 to 2016 demonstrated an initial decreasing trend, followed by an increase. The ARIMA-NARNN model can be well applied to analyze surveillance data for early warning systems for the control and elimination of schistosomiasis.

AROME-Arctic: New operational NWP model for the Arctic region

NASA Astrophysics Data System (ADS)

Süld, Jakob; Dale, Knut S.; Myrland, Espen; Batrak, Yurii; Homleid, Mariken; Valkonen, Teresa; Seierstad, Ivar A.; Randriamampianina, Roger

2016-04-01

In the frame of the EU-funded project ACCESS (Arctic Climate Change, Economy and Society), MET Norway aimed 1) to describe the present monitoring and forecasting capabilities in the Arctic; and 2) to identify the key factors limiting the forecasting capabilities and to give recommendations on key areas to improve the forecasting capabilities in the Arctic. We have observed that the NWP forecast quality is lower in the Arctic than in the regions further south. Earlier research indicated that one of the factors behind this is the composition of the observing system in the Arctic, in particular the scarceness of conventional observations. To further assess possible strategies for alleviating the situation and propose scenarios for a future Arctic observing system, we have performed a set of experiments to gain a more detailed insight in the contribution of the components of the present observing system in a regional state-of-the-art non-hydrostatic NWP model using the AROME physics (Seity et al, 2011) at 2.5 km horizontal resolution - AROME-Arctic. Our observing system experiment studies showed that conventional observations (Synop, Buoys) can play an important role in correcting the surface state of the model, but prove that the present upper-air conventional (Radiosondes, Aircraft) observations in the area are too scarce to have a significant effect on forecasts. We demonstrate that satellite sounding data play an important role in improving forecast quality. This is the case with satellite temperature sounding data (AMSU-A, IASI), as well as with the satellite moisture sounding data (AMSU-B/MHS, IASI). With these sets of observations, the AROME-Arctic clearly performs better in forecasting extreme events, like for example polar lows. For more details see presentation by Randriamampianina et al. in this session. The encouraging performance of AROME-Arctic lead us to implement it with more observations and improved settings into daily runs with the objective to substitute our actual operational Arctic mesoscale HIRLAM (High Resolution Limited Area Model) NWP model. This presentation will discuss in detail the operational implementation of the AROME-Arctic model together with post-processing methods. Aimed services in the Arctic region covered by the model, such as online weather forecasting (yr.no) and tracking of polar lows (barentswatch.no), is also included.

Probabilistic postprocessing models for flow forecasts for a system of catchments and several lead times

NASA Astrophysics Data System (ADS)

Engeland, Kolbjorn; Steinsland, Ingelin

2014-05-01

This study introduces a methodology for the construction of probabilistic inflow forecasts for multiple catchments and lead times, and investigates criterions for evaluation of multi-variate forecasts. A post-processing approach is used, and a Gaussian model is applied for transformed variables. The post processing model has two main components, the mean model and the dependency model. The mean model is used to estimate the marginal distributions for forecasted inflow for each catchment and lead time, whereas the dependency models was used to estimate the full multivariate distribution of forecasts, i.e. co-variances between catchments and lead times. In operational situations, it is a straightforward task to use the models to sample inflow ensembles which inherit the dependencies between catchments and lead times. The methodology was tested and demonstrated in the river systems linked to the Ulla-Førre hydropower complex in southern Norway, where simultaneous probabilistic forecasts for five catchments and ten lead times were constructed. The methodology exhibits sufficient flexibility to utilize deterministic flow forecasts from a numerical hydrological model as well as statistical forecasts such as persistent forecasts and sliding window climatology forecasts. It also deals with variation in the relative weights of these forecasts with both catchment and lead time. When evaluating predictive performance in original space using cross validation, the case study found that it is important to include the persistent forecast for the initial lead times and the hydrological forecast for medium-term lead times. Sliding window climatology forecasts become more important for the latest lead times. Furthermore, operationally important features in this case study such as heteroscedasticity, lead time varying between lead time dependency and lead time varying between catchment dependency are captured. Two criterions were used for evaluating the added value of the dependency model. The first one was the Energy score (ES) that is a multi-dimensional generalization of continuous rank probability score (CRPS). ES was calculated for all lead-times and catchments together, for each catchment across all lead times and for each lead time across all catchments. The second criterion was to use CRPS for forecasted inflows accumulated over several lead times and catchments. The results showed that ES was not very sensitive to correct covariance structure, whereas CRPS for accumulated flows where more suitable for evaluating the dependency model. This indicates that it is more appropriate to evaluate relevant univariate variables that depends on the dependency structure then to evaluate the multivariate forecast directly.

Impact of Reservoir Operation to the Inflow Flood - a Case Study of Xinfengjiang Reservoir

NASA Astrophysics Data System (ADS)

Chen, L.

2017-12-01

Building of reservoir shall impact the runoff production and routing characteristics, and changes the flood formation. This impact, called as reservoir flood effect, could be divided into three parts, including routing effect, volume effect and peak flow effect, and must be evaluated in a whole by using hydrological model. After analyzing the reservoir flood formation, the Liuxihe Model for reservoir flood forecasting is proposed. The Xinfengjiang Reservoir is studied as a case. Results show that the routing effect makes peak flow appear 4 to 6 hours in advance, volume effect is bigger for large flood than small one, and when rainfall focus on the reservoir area, this effect also increases peak flow largely, peak flow effect makes peak flow increase 6.63% to 8.95%. Reservoir flood effect is obvious, which have significant impact to reservoir flood. If this effect is not considered in the flood forecasting model, the flood could not be forecasted accurately, particularly the peak flow. Liuxihe Model proposed for Xinfengjiang Reservoir flood forecasting has a good performance, and could be used for real-time flood forecasting of Xinfengjiang Reservoir.Key words: Reservoir flood effect, reservoir flood forecasting, physically based distributed hydrological model, Liuxihe Model, parameter optimization

Multicomponent ensemble models to forecast induced seismicity

NASA Astrophysics Data System (ADS)

Király-Proag, E.; Gischig, V.; Zechar, J. D.; Wiemer, S.

2018-01-01

In recent years, human-induced seismicity has become a more and more relevant topic due to its economic and social implications. Several models and approaches have been developed to explain underlying physical processes or forecast induced seismicity. They range from simple statistical models to coupled numerical models incorporating complex physics. We advocate the need for forecast testing as currently the best method for ascertaining if models are capable to reasonably accounting for key physical governing processes—or not. Moreover, operational forecast models are of great interest to help on-site decision-making in projects entailing induced earthquakes. We previously introduced a standardized framework following the guidelines of the Collaboratory for the Study of Earthquake Predictability, the Induced Seismicity Test Bench, to test, validate, and rank induced seismicity models. In this study, we describe how to construct multicomponent ensemble models based on Bayesian weightings that deliver more accurate forecasts than individual models in the case of Basel 2006 and Soultz-sous-Forêts 2004 enhanced geothermal stimulation projects. For this, we examine five calibrated variants of two significantly different model groups: (1) Shapiro and Smoothed Seismicity based on the seismogenic index, simple modified Omori-law-type seismicity decay, and temporally weighted smoothed seismicity; (2) Hydraulics and Seismicity based on numerically modelled pore pressure evolution that triggers seismicity using the Mohr-Coulomb failure criterion. We also demonstrate how the individual and ensemble models would perform as part of an operational Adaptive Traffic Light System. Investigating seismicity forecasts based on a range of potential injection scenarios, we use forecast periods of different durations to compute the occurrence probabilities of seismic events M ≥ 3. We show that in the case of the Basel 2006 geothermal stimulation the models forecast hazardous levels of seismicity days before the occurrence of felt events.

Evaluation of the synoptic and mesoscale predictive capabilities of a mesoscale atmospheric simulation system

NASA Technical Reports Server (NTRS)

Koch, S. E.; Skillman, W. C.; Kocin, P. J.; Wetzel, P. J.; Brill, K.; Keyser, D. A.; Mccumber, M. C.

1983-01-01

The overall performance characteristics of a limited area, hydrostatic, fine (52 km) mesh, primitive equation, numerical weather prediction model are determined in anticipation of satellite data assimilations with the model. The synoptic and mesoscale predictive capabilities of version 2.0 of this model, the Mesoscale Atmospheric Simulation System (MASS 2.0), were evaluated. The two part study is based on a sample of approximately thirty 12h and 24h forecasts of atmospheric flow patterns during spring and early summer. The synoptic scale evaluation results benchmark the performance of MASS 2.0 against that of an operational, synoptic scale weather prediction model, the Limited area Fine Mesh (LFM). The large sample allows for the calculation of statistically significant measures of forecast accuracy and the determination of systematic model errors. The synoptic scale benchmark is required before unsmoothed mesoscale forecast fields can be seriously considered.

Forecasting stochastic neural network based on financial empirical mode decomposition.

PubMed

Wang, Jie; Wang, Jun

2017-06-01

In an attempt to improve the forecasting accuracy of stock price fluctuations, a new one-step-ahead model is developed in this paper which combines empirical mode decomposition (EMD) with stochastic time strength neural network (STNN). The EMD is a processing technique introduced to extract all the oscillatory modes embedded in a series, and the STNN model is established for considering the weight of occurrence time of the historical data. The linear regression performs the predictive availability of the proposed model, and the effectiveness of EMD-STNN is revealed clearly through comparing the predicted results with the traditional models. Moreover, a new evaluated method (q-order multiscale complexity invariant distance) is applied to measure the predicted results of real stock index series, and the empirical results show that the proposed model indeed displays a good performance in forecasting stock market fluctuations. Copyright © 2017 Elsevier Ltd. All rights reserved.

The value of the North American Multi Model Ensemble phase 2 for sub-seasonal hydrological forecasting

NASA Astrophysics Data System (ADS)

Wanders, Niko; Wood, Eric

2016-04-01

Sub-seasonal to seasonal weather and hydrological forecasts have the potential to provide vital information for a variety of water-related decision makers. For example, seasonal forecasts of drought risk can enable farmers to make adaptive choices on crop varieties, labour usage, and technology investments. Seasonal and sub-seasonal predictions can increase preparedness to hydrological extremes that regularly occur in all regions of the world with large impacts on society. We investigated the skill of six seasonal forecast models from the NMME-2 ensemble coupled to two global hydrological models (VIC and PCRGLOBWB) for the period 1982-2012. The 31 years of NNME-2 hindcast data is used in combination with an ensemble mean and ESP forecast, to forecast important hydrological variables (e.g. soil moisture, groundwater storage, snow, reservoir levels and river discharge). By using two global hydrological models we are able to quantify both the uncertainty in the meteorological input and the uncertainty created by the different hydrological models. We show that the NMME-2 forecast outperforms the ESP forecasts in terms of anomaly correlation and brier skill score for all forecasted hydrological variables, with a low uncertainty in the performance amongst the hydrological models. However, the continuous ranked probability score (CRPS) of the NMME-2 ensemble is inferior to the ESP due to a large spread between the individual ensemble members. We use a cost analysis to show that the damage caused by floods and droughts in large scale rivers can globally be reduced by 48% (for leads from 1-2 months) to 20% (for leads between 6-9 months) when precautions are taken based on the NMME-2 ensemble instead of an ESP forecast. In collaboration with our local partner in West Africa (AGHRYMET), we looked at the performance of the sub-seasonal forecasts for crop planting dates and high flow season in West Africa. We show that the uncertainty in the optimal planting date is reduced from 30 days to 12 days (2.5 month lead) and an increased predictability of the high flow season from 45 days to 20 days (3-4 months lead). Additionally, we show that snow accumulation and melt onset in the Northern hemisphere can be forecasted with an uncertainty of 10 days (2.5 months lead). Both the overall skill, and the skill found in these last two examples, indicates that the new NMME-2 forecast dataset is valuable for sub-seasonal forecast applications. The high temporal resolution (daily), long leads (one year leads) and large hindcast archive enable new sub-seasonal forecasting applications to be explored. We show that the NMME-2 has a large potential for sub-seasonal hydrological forecasting and other potential hydrological applications (e.g. reservoir management), which could benefit from these new forecasts.

A comparative verification of high resolution precipitation forecasts using model output statistics

NASA Astrophysics Data System (ADS)

van der Plas, Emiel; Schmeits, Maurice; Hooijman, Nicolien; Kok, Kees

2017-04-01

Verification of localized events such as precipitation has become even more challenging with the advent of high-resolution meso-scale numerical weather prediction (NWP). The realism of a forecast suggests that it should compare well against precipitation radar imagery with similar resolution, both spatially and temporally. Spatial verification methods solve some of the representativity issues that point verification gives rise to. In this study a verification strategy based on model output statistics is applied that aims to address both double penalty and resolution effects that are inherent to comparisons of NWP models with different resolutions. Using predictors based on spatial precipitation patterns around a set of stations, an extended logistic regression (ELR) equation is deduced, leading to a probability forecast distribution of precipitation for each NWP model, analysis and lead time. The ELR equations are derived for predictands based on areal calibrated radar precipitation and SYNOP observations. The aim is to extract maximum information from a series of precipitation forecasts, like a trained forecaster would. The method is applied to the non-hydrostatic model Harmonie (2.5 km resolution), Hirlam (11 km resolution) and the ECMWF model (16 km resolution), overall yielding similar Brier skill scores for the 3 post-processed models, but larger differences for individual lead times. Besides, the Fractions Skill Score is computed using the 3 deterministic forecasts, showing somewhat better skill for the Harmonie model. In other words, despite the realism of Harmonie precipitation forecasts, they only perform similarly or somewhat better than precipitation forecasts from the 2 lower resolution models, at least in the Netherlands.

Comparative assessment of several post-processing methods for correcting evapotranspiration forecasts derived from TIGGE datasets.

NASA Astrophysics Data System (ADS)

Tian, D.; Medina, H.

2017-12-01

Post-processing of medium range reference evapotranspiration (ETo) forecasts based on numerical weather prediction (NWP) models has the potential of improving the quality and utility of these forecasts. This work compares the performance of several post-processing methods for correcting ETo forecasts over the continental U.S. generated from The Observing System Research and Predictability Experiment (THORPEX) Interactive Grand Global Ensemble (TIGGE) database using data from Europe (EC), the United Kingdom (MO), and the United States (NCEP). The pondered post-processing techniques are: simple bias correction, the use of multimodels, the Ensemble Model Output Statistics (EMOS, Gneitting et al., 2005) and the Bayesian Model Averaging (BMA, Raftery et al., 2005). ETo estimates based on quality-controlled U.S. Regional Climate Reference Network measurements, and computed with the FAO 56 Penman Monteith equation, are adopted as baseline. EMOS and BMA are generally the most efficient post-processing techniques of the ETo forecasts. Nevertheless, the simple bias correction of the best model is commonly much more rewarding than using multimodel raw forecasts. Our results demonstrate the potential of different forecasting and post-processing frameworks in operational evapotranspiration and irrigation advisory systems at national scale.

Enhanced seasonal forecast skill following stratospheric sudden warmings

NASA Astrophysics Data System (ADS)

Sigmond, M.; Scinocca, J. F.; Kharin, V. V.; Shepherd, T. G.

2013-02-01

Advances in seasonal forecasting have brought widespread socio-economic benefits. However, seasonal forecast skill in the extratropics is relatively modest, prompting the seasonal forecasting community to search for additional sources of predictability. For over a decade it has been suggested that knowledge of the state of the stratosphere can act as a source of enhanced seasonal predictability; long-lived circulation anomalies in the lower stratosphere that follow stratospheric sudden warmings are associated with circulation anomalies in the troposphere that can last up to two months. Here, we show by performing retrospective ensemble model forecasts that such enhanced predictability can be realized in a dynamical seasonal forecast system with a good representation of the stratosphere. When initialized at the onset date of stratospheric sudden warmings, the model forecasts faithfully reproduce the observed mean tropospheric conditions in the months following the stratospheric sudden warmings. Compared with an equivalent set of forecasts that are not initialized during stratospheric sudden warmings, we document enhanced forecast skill for atmospheric circulation patterns, surface temperatures over northern Russia and eastern Canada and North Atlantic precipitation. We suggest that seasonal forecast systems initialized during stratospheric sudden warmings are likely to yield significantly greater forecast skill in some regions.

Operational foreshock forecasting: Fifteen years after

NASA Astrophysics Data System (ADS)

Ogata, Y.

2010-12-01

We are concerned with operational forecasting of the probability that events are foreshocks of a forthcoming earthquake that is significantly larger (mainshock). Specifically, we define foreshocks as the preshocks substantially smaller than the mainshock by a magnitude gap of 0.5 or larger. The probability gain of foreshock forecast is extremely high compare to long-term forecast by renewal processes or various alarm-based intermediate-term forecasts because of a large event’s low occurrence rate in a short period and a narrow target region. Thus, it is desired to establish operational foreshock probability forecasting as seismologists have done for aftershocks. When a series of earthquakes occurs in a region, we attempt to discriminate foreshocks from a swarm or mainshock-aftershock sequence. Namely, after real time identification of an earthquake cluster using methods such as the single-link algorithm, the probability is calculated by applying statistical features that discriminate foreshocks from other types of clusters, by considering the events' stronger proximity in time and space and tendency towards chronologically increasing magnitudes. These features were modeled for probability forecasting and the coefficients of the model were estimated in Ogata et al. (1996) for the JMA hypocenter data (M≧4, 1926-1993). Currently, fifteen years has passed since the publication of the above-stated work so that we are able to present the performance and validation of the forecasts (1994-2009) by using the same model. Taking isolated events into consideration, the probability of the first events in a potential cluster being a foreshock vary in a range between 0+% and 10+% depending on their locations. This conditional forecasting performs significantly better than the unconditional (average) foreshock probability of 3.7% throughout Japan region. Furthermore, when we have the additional events in a cluster, the forecast probabilities range more widely from nearly 0% to about 40% depending on the discrimination features among the events in the cluster. This conditional forecasting further performs significantly better than the unconditional foreshock probability of 7.3%, which is the average probability of the plural events in the earthquake clusters. Indeed, the frequency ratios of the actual foreshocks are consistent with the forecasted probabilities. Reference: Ogata, Y., Utsu, T. and Katsura, K. (1996). Statistical discrimination of foreshocks from other earthquake clusters, Geophys. J. Int. 127, 17-30.

Determination of sample size for higher volatile data using new framework of Box-Jenkins model with GARCH: A case study on gold price

NASA Astrophysics Data System (ADS)

Roslindar Yaziz, Siti; Zakaria, Roslinazairimah; Hura Ahmad, Maizah

2017-09-01

The model of Box-Jenkins - GARCH has been shown to be a promising tool for forecasting higher volatile time series. In this study, the framework of determining the optimal sample size using Box-Jenkins model with GARCH is proposed for practical application in analysing and forecasting higher volatile data. The proposed framework is employed to daily world gold price series from year 1971 to 2013. The data is divided into 12 different sample sizes (from 30 to 10200). Each sample is tested using different combination of the hybrid Box-Jenkins - GARCH model. Our study shows that the optimal sample size to forecast gold price using the framework of the hybrid model is 1250 data of 5-year sample. Hence, the empirical results of model selection criteria and 1-step-ahead forecasting evaluations suggest that the latest 12.25% (5-year data) of 10200 data is sufficient enough to be employed in the model of Box-Jenkins - GARCH with similar forecasting performance as by using 41-year data.

The Role of Model and Initial Condition Error in Numerical Weather Forecasting Investigated with an Observing System Simulation Experiment

NASA Technical Reports Server (NTRS)

Prive, Nikki C.; Errico, Ronald M.

2013-01-01

A series of experiments that explore the roles of model and initial condition error in numerical weather prediction are performed using an observing system simulation experiment (OSSE) framework developed at the National Aeronautics and Space Administration Global Modeling and Assimilation Office (NASA/GMAO). The use of an OSSE allows the analysis and forecast errors to be explicitly calculated, and different hypothetical observing networks can be tested with ease. In these experiments, both a full global OSSE framework and an 'identical twin' OSSE setup are utilized to compare the behavior of the data assimilation system and evolution of forecast skill with and without model error. The initial condition error is manipulated by varying the distribution and quality of the observing network and the magnitude of observation errors. The results show that model error has a strong impact on both the quality of the analysis field and the evolution of forecast skill, including both systematic and unsystematic model error components. With a realistic observing network, the analysis state retains a significant quantity of error due to systematic model error. If errors of the analysis state are minimized, model error acts to rapidly degrade forecast skill during the first 24-48 hours of forward integration. In the presence of model error, the impact of observation errors on forecast skill is small, but in the absence of model error, observation errors cause a substantial degradation of the skill of medium range forecasts.

Forecast errors in dust vertical distributions over Rome (Italy): Multiple particle size representation and cloud contributions

NASA Astrophysics Data System (ADS)

Kishcha, P.; Alpert, P.; Shtivelman, A.; Krichak, S. O.; Joseph, J. H.; Kallos, G.; Katsafados, P.; Spyrou, C.; Gobbi, G. P.; Barnaba, F.; Nickovic, S.; PéRez, C.; Baldasano, J. M.

2007-08-01

In this study, forecast errors in dust vertical distributions were analyzed. This was carried out by using quantitative comparisons between dust vertical profiles retrieved from lidar measurements over Rome, Italy, performed from 2001 to 2003, and those predicted by models. Three models were used: the four-particle-size Dust Regional Atmospheric Model (DREAM), the older one-particle-size version of the SKIRON model from the University of Athens (UOA), and the pre-2006 one-particle-size Tel Aviv University (TAU) model. SKIRON and DREAM are initialized on a daily basis using the dust concentration from the previous forecast cycle, while the TAU model initialization is based on the Total Ozone Mapping Spectrometer aerosol index (TOMS AI). The quantitative comparison shows that (1) the use of four-particle-size bins in the dust modeling instead of only one-particle-size bins improves dust forecasts; (2) cloud presence could contribute to noticeable dust forecast errors in SKIRON and DREAM; and (3) as far as the TAU model is concerned, its forecast errors were mainly caused by technical problems with TOMS measurements from the Earth Probe satellite. As a result, dust forecast errors in the TAU model could be significant even under cloudless conditions. The DREAM versus lidar quantitative comparisons at different altitudes show that the model predictions are more accurate in the middle part of dust layers than in the top and bottom parts of dust layers.

Regional Model Nesting Within GFS Daily Forecasts Over West Africa

NASA Technical Reports Server (NTRS)

Druyan, Leonard M.; Fulakeza, Matthew; Lonergan, Patrick; Worrell, Ruben

2010-01-01

The study uses the RM3, the regional climate model at the Center for Climate Systems Research of Columbia University and the NASA/Goddard Institute for Space Studies (CCSR/GISS). The paper evaluates 30 48-hour RM3 weather forecasts over West Africa during September 2006 made on a 0.5 grid nested within 1 Global Forecast System (GFS) global forecasts. September 2006 was the Special Observing Period #3 of the African Monsoon Multidisciplinary Analysis (AMMA). Archived GFS initial conditions and lateral boundary conditions for the simulations from the US National Weather Service, National Oceanographic and Atmospheric Administration were interpolated four times daily. Results for precipitation forecasts are validated against Tropical Rainfall Measurement Mission (TRMM) satellite estimates and data from the Famine Early Warning System (FEWS), which includes rain gauge measurements, and forecasts of circulation are compared to reanalysis 2. Performance statistics for the precipitation forecasts include bias, root-mean-square errors and spatial correlation coefficients. The nested regional model forecasts are compared to GFS forecasts to gauge whether nesting provides additional realistic information. They are also compared to RM3 simulations driven by reanalysis 2, representing high potential skill forecasts, to gauge the sensitivity of results to lateral boundary conditions. Nested RM3/GFS forecasts generate excessive moisture advection toward West Africa, which in turn causes prodigious amounts of model precipitation. This problem is corrected by empirical adjustments in the preparation of lateral boundary conditions and initial conditions. The resulting modified simulations improve on the GFS precipitation forecasts, achieving time-space correlations with TRMM of 0.77 on the first day and 0.63 on the second day. One realtime RM3/GFS precipitation forecast made at and posted by the African Centre of Meteorological Application for Development (ACMAD) in Niamey, Niger is shown.

Evaluation and intercomparison of air quality forecasts over Korea during the KORUS-AQ campaign

NASA Astrophysics Data System (ADS)

Lee, Seungun; Park, Rokjin J.; Kim, Soontae; Song, Chul H.; Kim, Cheol-Hee; Woo, Jung-Hun

2017-04-01

We evaluate and intercompare ozone and aerosol simulations over Korea during the KORUS-AQ campaign, which was conducted in May-June 2016. Four global and regional air quality models participated in the campaign and provided daily air quality forecasts over Korea to guide aircraft flight paths for detecting air pollution events over Korean peninsula and its nearby oceans. We first evaluate the model performance by comparing simulated and observed hourly surface ozone and PM2.5 concentrations at ground sites in Korea and find that the models successfully capture intermittent air pollution events and reproduce the daily variation of ozone and PM2.5 concentrations. However, significant underestimates of peak ozone concentrations in the afternoon are also found in most models. Among chemical constituents of PM2.5, the models typically overestimate observed nitrate aerosol concentrations and underestimate organic aerosol concentrations, although the observed mass concentrations of PM2.5 are seemingly reproduced by the models. In particular, all models used the same anthropogenic emission inventory (KU-CREATE) for daily air quality forecast, but they show a considerable discrepancy for ozone and aerosols. Compared to individual model results, the ensemble mean of all models shows the best performance with correlation coefficients of 0.73 for ozone and 0.57 for PM2.5. We here investigate contributing factors to the discrepancy, which will serve as a guidance to improve the performance of the air quality forecast.

Winter precipitation forecast in the European and Mediterranean regions using cluster analysis

NASA Astrophysics Data System (ADS)

Molnos, S.

2017-12-01

The European and Mediterranean climates are sensitive to large-scale circulation of the atmosphere andocean making it difficult to forecast precipitation or temperature on seasonal time-scales. In addition, theMediterranean region has been identified as a hotspot for climate change and already today a drying in theMediterranean region is observed.Thus, it is critically important to predict seasonal droughts as early as possible such that water managersand stakeholders can mitigate impacts.We developed a novel cluster-based forecast method to empirically predict winter's precipitationanomalies in European and Mediterranean regions using precursors in autumn. This approach does notonly utilizes the amplitude but also the pattern of the precursors in generating the forecast.Using a toy model we show that it achieves a better forecast skill than more traditional regression models. Furthermore, we compare our algorithm with dynamic forecast models demonstrating that our prediction method performs better in terms of time and pattern correlation in the Mediterranean and European regions.

A Bayesian spatio-temporal model for forecasting Anaplasma species seroprevalence in domestic dogs within the contiguous United States

PubMed Central

Liu, Yan; Watson, Stella C.; Gettings, Jenna R.; Lund, Robert B.; Nordone, Shila K.; McMahan, Christopher S.

2017-01-01

This paper forecasts the 2016 canine Anaplasma spp. seroprevalence in the United States from eight climate, geographic and societal factors. The forecast’s construction and an assessment of its performance are described. The forecast is based on a spatial-temporal conditional autoregressive model fitted to over 11 million Anaplasma spp. seroprevalence test results for dogs conducted in the 48 contiguous United States during 2011–2015. The forecast uses county-level data on eight predictive factors, including annual temperature, precipitation, relative humidity, county elevation, forestation coverage, surface water coverage, population density and median household income. Non-static factors are extrapolated into the forthcoming year with various statistical methods. The fitted model and factor extrapolations are used to estimate next year’s regional prevalence. The correlation between the observed and model-estimated county-by-county Anaplasma spp. seroprevalence for the five-year period 2011–2015 is 0.902, demonstrating reasonable model accuracy. The weighted correlation (accounting for different sample sizes) between 2015 observed and forecasted county-by-county Anaplasma spp. seroprevalence is 0.987, exhibiting that the proposed approach can be used to accurately forecast Anaplasma spp. seroprevalence. The forecast presented herein can a priori alert veterinarians to areas expected to see Anaplasma spp. seroprevalence beyond the accepted endemic range. The proposed methods may prove useful for forecasting other diseases. PMID:28738085

Prediction of fog/visibility over India using NWP Model

NASA Astrophysics Data System (ADS)

Singh, Aditi; George, John P.; Iyengar, Gopal Raman

2018-03-01

Frequent occurrence of fog in different parts of northern India is common during the winter months of December and January. Low visibility conditions due to fog disrupt normal public life. Visibility conditions heavily affect both surface and air transport. A number of flights are either diverted or cancelled every year during the winter season due to low visibility conditions, experienced at different airports of north India. Thus, fog and visibility forecasts over plains of north India become very important during winter months. This study aims to understand the ability of a NWP model (NCMRWF, Unified Model, NCUM) with a diagnostic visibility scheme to forecast visibility over plains of north India. The present study verifies visibility forecasts obtained from NCUM against the INSAT-3D fog images and visibility observations from the METAR reports of different stations in the plains of north India. The study shows that the visibility forecast obtained from NCUM can provide reasonably good indication of the spatial extent of fog in advance of one day. The fog intensity is also predicted fairly well. The study also verifies the simple diagnostic model for fog which is driven by NWP model forecast of surface relative humidity and wind speed. The performance of NWP model forecast of visibility is found comparable to that from simple fog model driven by NWP forecast of relative humidity and wind speed.

Forecasting hourly PM(10) concentration in Cyprus through artificial neural networks and multiple regression models: implications to local environmental management.

PubMed

Paschalidou, Anastasia K; Karakitsios, Spyridon; Kleanthous, Savvas; Kassomenos, Pavlos A

2011-02-01

In the present work, two types of artificial neural network (NN) models using the multilayer perceptron (MLP) and the radial basis function (RBF) techniques, as well as a model based on principal component regression analysis (PCRA), are employed to forecast hourly PM(10) concentrations in four urban areas (Larnaca, Limassol, Nicosia and Paphos) in Cyprus. The model development is based on a variety of meteorological and pollutant parameters corresponding to the 2-year period between July 2006 and June 2008, and the model evaluation is achieved through the use of a series of well-established evaluation instruments and methodologies. The evaluation reveals that the MLP NN models display the best forecasting performance with R (2) values ranging between 0.65 and 0.76, whereas the RBF NNs and the PCRA models reveal a rather weak performance with R (2) values between 0.37-0.43 and 0.33-0.38, respectively. The derived MLP models are also used to forecast Saharan dust episodes with remarkable success (probability of detection ranging between 0.68 and 0.71). On the whole, the analysis shows that the models introduced here could provide local authorities with reliable and precise predictions and alarms about air quality if used on an operational basis.

Automated time series forecasting for biosurveillance.

PubMed

Burkom, Howard S; Murphy, Sean Patrick; Shmueli, Galit

2007-09-30

For robust detection performance, traditional control chart monitoring for biosurveillance is based on input data free of trends, day-of-week effects, and other systematic behaviour. Time series forecasting methods may be used to remove this behaviour by subtracting forecasts from observations to form residuals for algorithmic input. We describe three forecast methods and compare their predictive accuracy on each of 16 authentic syndromic data streams. The methods are (1) a non-adaptive regression model using a long historical baseline, (2) an adaptive regression model with a shorter, sliding baseline, and (3) the Holt-Winters method for generalized exponential smoothing. Criteria for comparing the forecasts were the root-mean-square error, the median absolute per cent error (MedAPE), and the median absolute deviation. The median-based criteria showed best overall performance for the Holt-Winters method. The MedAPE measures over the 16 test series averaged 16.5, 11.6, and 9.7 for the non-adaptive regression, adaptive regression, and Holt-Winters methods, respectively. The non-adaptive regression forecasts were degraded by changes in the data behaviour in the fixed baseline period used to compute model coefficients. The mean-based criterion was less conclusive because of the effects of poor forecasts on a small number of calendar holidays. The Holt-Winters method was also most effective at removing serial autocorrelation, with most 1-day-lag autocorrelation coefficients below 0.15. The forecast methods were compared without tuning them to the behaviour of individual series. We achieved improved predictions with such tuning of the Holt-Winters method, but practical use of such improvements for routine surveillance will require reliable data classification methods.

A Feature Fusion Based Forecasting Model for Financial Time Series

PubMed Central

Guo, Zhiqiang; Wang, Huaiqing; Liu, Quan; Yang, Jie

2014-01-01

Predicting the stock market has become an increasingly interesting research area for both researchers and investors, and many prediction models have been proposed. In these models, feature selection techniques are used to pre-process the raw data and remove noise. In this paper, a prediction model is constructed to forecast stock market behavior with the aid of independent component analysis, canonical correlation analysis, and a support vector machine. First, two types of features are extracted from the historical closing prices and 39 technical variables obtained by independent component analysis. Second, a canonical correlation analysis method is utilized to combine the two types of features and extract intrinsic features to improve the performance of the prediction model. Finally, a support vector machine is applied to forecast the next day's closing price. The proposed model is applied to the Shanghai stock market index and the Dow Jones index, and experimental results show that the proposed model performs better in the area of prediction than other two similar models. PMID:24971455

Forecasting the realized volatility of the Chinese stock market: Do the G7 stock markets help?

NASA Astrophysics Data System (ADS)

Peng, Huan; Chen, Ruoxun; Mei, Dexiang; Diao, Xiaohua

2018-07-01

In this paper, we use a comprehensive look to investigate whether the G7 stock markets can contain predictive information to help in forecasting the Chinese stock market volatility. Our out-of-sample empirical results indicate the kitchen sink (HAR-RV-SK) model is able to attain better performance than the benchmark model (HAR-RV) and other models, implying that the G7 stock markets can help in predicting the one-day volatility of the Chinese stock market. Moreover, the kitchen sink strategy can beat the strategy of the simple combination forecasts. Finally, the G7 stock markets can indeed contain useful information, which can increase the accuracy forecasts of the Chinese stock market.

DOE Office of Scientific and Technical Information (OSTI.GOV)

The ECE application forecasts annual costs of preventive and corrective maintenance for budgeting purposes. Features within the application enable the user to change the specifications of the model to customize your forecast to best fit their needs and support “what if” analysis. Based on the user's selections, the ECE model forecasts annual maintenance costs. Preventive maintenance costs include the cost of labor to perform preventive maintenance activities at the specific frequency and labor rate. Corrective maintenance costs include the cost of labor and the cost of replacement parts. The application presents forecasted maintenance costs for the next five years inmore » two tables: costs by year and costs by site.« less

Forecasting volatility of SSEC in Chinese stock market using multifractal analysis

NASA Astrophysics Data System (ADS)

Wei, Yu; Wang, Peng

2008-03-01

In this paper, taking about 7 years’ high-frequency data of the Shanghai Stock Exchange Composite Index (SSEC) as an example, we propose a daily volatility measure based on the multifractal spectrum of the high-frequency price variability within a trading day. An ARFIMA model is used to depict the dynamics of this multifractal volatility (MFV) measures. The one-day ahead volatility forecasting performances of the MFV model and some other existing volatility models, such as the realized volatility model, stochastic volatility model and GARCH, are evaluated by the superior prediction ability (SPA) test. The empirical results show that under several loss functions, the MFV model obtains the best forecasting accuracy.

Real-Time Bias-Adjusted O3 and PM2.5 Air Quality Index Forecasts and their Performance Evaluations over the Continental United States

EPA Science Inventory

The National Air Quality Forecast Capacity (NAQFC) system, which links NOAA's North American Mesoscale (NAM) meteorological model with EPA's Community Multiscale Air Quality (CMAQ) model, provided operational ozone (O₃) and experimental fine particular matter (PM_2...

Interactive Vegetation Phenology, Soil Moisture, and Monthly Temperature Forecasts

NASA Technical Reports Server (NTRS)

Koster, R. D.; Walker, G. K.

2015-01-01

The time scales that characterize the variations of vegetation phenology are generally much longer than those that characterize atmospheric processes. The explicit modeling of phenological processes in an atmospheric forecast system thus has the potential to provide skill to subseasonal or seasonal forecasts. We examine this possibility here using a forecast system fitted with a dynamic vegetation phenology model. We perform three experiments, each consisting of 128 independent warm-season monthly forecasts: 1) an experiment in which both soil moisture states and carbon states (e.g., those determining leaf area index) are initialized realistically, 2) an experiment in which the carbon states are prescribed to climatology throughout the forecasts, and 3) an experiment in which both the carbon and soil moisture states are prescribed to climatology throughout the forecasts. Evaluating the monthly forecasts of air temperature in each ensemble against observations, as well as quantifying the inherent predictability of temperature within each ensemble, shows that dynamic phenology can indeed contribute positively to subseasonal forecasts, though only to a small extent, with an impact dwarfed by that of soil moisture.

Solar radio proxies for improved satellite orbit prediction

NASA Astrophysics Data System (ADS)

Yaya, Philippe; Hecker, Louis; Dudok de Wit, Thierry; Fèvre, Clémence Le; Bruinsma, Sean

2017-12-01

Specification and forecasting of solar drivers to thermosphere density models is critical for satellite orbit prediction and debris avoidance. Satellite operators routinely forecast orbits up to 30 days into the future. This requires forecasts of the drivers to these orbit prediction models such as the solar Extreme-UV (EUV) flux and geomagnetic activity. Most density models use the 10.7 cm radio flux (F10.7 index) as a proxy for solar EUV. However, daily measurements at other centimetric wavelengths have also been performed by the Nobeyama Radio Observatory (Japan) since the 1950's, thereby offering prospects for improving orbit modeling. Here we present a pre-operational service at the Collecte Localisation Satellites company that collects these different observations in one single homogeneous dataset and provides a 30 days forecast on a daily basis. Interpolation and preprocessing algorithms were developed to fill in missing data and remove anomalous values. We compared various empirical time series prediction techniques and selected a multi-wavelength non-recursive analogue neural network. The prediction of the 30 cm flux, and to a lesser extent that of the 10.7 cm flux, performs better than NOAA's present prediction of the 10.7 cm flux, especially during periods of high solar activity. In addition, we find that the DTM-2013 density model (Drag Temperature Model) performs better with (past and predicted) values of the 30 cm radio flux than with the 10.7 flux.

A Comparison of Forecast Error Generators for Modeling Wind and Load Uncertainty

DOE Office of Scientific and Technical Information (OSTI.GOV)

Lu, Ning; Diao, Ruisheng; Hafen, Ryan P.

2013-07-25

This paper presents four algorithms to generate random forecast error time series. The performance of four algorithms is compared. The error time series are used to create real-time (RT), hour-ahead (HA), and day-ahead (DA) wind and load forecast time series that statistically match historically observed forecasting data sets used in power grid operation to study the net load balancing need in variable generation integration studies. The four algorithms are truncated-normal distribution models, state-space based Markov models, seasonal autoregressive moving average (ARMA) models, and a stochastic-optimization based approach. The comparison is made using historical DA load forecast and actual load valuesmore » to generate new sets of DA forecasts with similar stoical forecast error characteristics (i.e., mean, standard deviation, autocorrelation, and cross-correlation). The results show that all methods generate satisfactory results. One method may preserve one or two required statistical characteristics better the other methods, but may not preserve other statistical characteristics as well compared with the other methods. Because the wind and load forecast error generators are used in wind integration studies to produce wind and load forecasts time series for stochastic planning processes, it is sometimes critical to use multiple methods to generate the error time series to obtain a statistically robust result. Therefore, this paper discusses and compares the capabilities of each algorithm to preserve the characteristics of the historical forecast data sets.« less

Experiments with Seasonal Forecasts of ocean conditions for the Northern region of the California Current upwelling system

PubMed Central

Siedlecki, Samantha A.; Kaplan, Isaac C.; Hermann, Albert J.; Nguyen, Thanh Tam; Bond, Nicholas A.; Newton, Jan A.; Williams, Gregory D.; Peterson, William T.; Alin, Simone R.; Feely, Richard A.

2016-01-01

Resource managers at the state, federal, and tribal levels make decisions on a weekly to quarterly basis, and fishers operate on a similar timeframe. To determine the potential of a support tool for these efforts, a seasonal forecast system is experimented with here. JISAO’s Seasonal Coastal Ocean Prediction of the Ecosystem (J-SCOPE) features dynamical downscaling of regional ocean conditions in Washington and Oregon waters using a combination of a high-resolution regional model with biogeochemistry and forecasts from NOAA’s Climate Forecast System (CFS). Model performance and predictability were examined for sea surface temperature (SST), bottom temperature, bottom oxygen, pH, and aragonite saturation state through model hindcasts, reforecast, and forecast comparisons with observations. Results indicate J-SCOPE forecasts have measurable skill on seasonal timescales. Experiments suggest that seasonal forecasting of ocean conditions important for fisheries is possible with the right combination of components. Those components include regional predictability on seasonal timescales of the physical environment from a large-scale model, a high-resolution regional model with biogeochemistry that simulates seasonal conditions in hindcasts, a relationship with local stakeholders, and a real-time observational network. Multiple efforts and approaches in different regions would advance knowledge to provide additional tools to fishers and other stakeholders. PMID:27273473

Experiments with Seasonal Forecasts of ocean conditions for the Northern region of the California Current upwelling system

NASA Astrophysics Data System (ADS)

Siedlecki, Samantha A.; Kaplan, Isaac C.; Hermann, Albert J.; Nguyen, Thanh Tam; Bond, Nicholas A.; Newton, Jan A.; Williams, Gregory D.; Peterson, William T.; Alin, Simone R.; Feely, Richard A.

2016-06-01

Resource managers at the state, federal, and tribal levels make decisions on a weekly to quarterly basis, and fishers operate on a similar timeframe. To determine the potential of a support tool for these efforts, a seasonal forecast system is experimented with here. JISAO’s Seasonal Coastal Ocean Prediction of the Ecosystem (J-SCOPE) features dynamical downscaling of regional ocean conditions in Washington and Oregon waters using a combination of a high-resolution regional model with biogeochemistry and forecasts from NOAA’s Climate Forecast System (CFS). Model performance and predictability were examined for sea surface temperature (SST), bottom temperature, bottom oxygen, pH, and aragonite saturation state through model hindcasts, reforecast, and forecast comparisons with observations. Results indicate J-SCOPE forecasts have measurable skill on seasonal timescales. Experiments suggest that seasonal forecasting of ocean conditions important for fisheries is possible with the right combination of components. Those components include regional predictability on seasonal timescales of the physical environment from a large-scale model, a high-resolution regional model with biogeochemistry that simulates seasonal conditions in hindcasts, a relationship with local stakeholders, and a real-time observational network. Multiple efforts and approaches in different regions would advance knowledge to provide additional tools to fishers and other stakeholders.

Experiments with Seasonal Forecasts of ocean conditions for the Northern region of the California Current upwelling system.

PubMed

Siedlecki, Samantha A; Kaplan, Isaac C; Hermann, Albert J; Nguyen, Thanh Tam; Bond, Nicholas A; Newton, Jan A; Williams, Gregory D; Peterson, William T; Alin, Simone R; Feely, Richard A

2016-06-07

Resource managers at the state, federal, and tribal levels make decisions on a weekly to quarterly basis, and fishers operate on a similar timeframe. To determine the potential of a support tool for these efforts, a seasonal forecast system is experimented with here. JISAO's Seasonal Coastal Ocean Prediction of the Ecosystem (J-SCOPE) features dynamical downscaling of regional ocean conditions in Washington and Oregon waters using a combination of a high-resolution regional model with biogeochemistry and forecasts from NOAA's Climate Forecast System (CFS). Model performance and predictability were examined for sea surface temperature (SST), bottom temperature, bottom oxygen, pH, and aragonite saturation state through model hindcasts, reforecast, and forecast comparisons with observations. Results indicate J-SCOPE forecasts have measurable skill on seasonal timescales. Experiments suggest that seasonal forecasting of ocean conditions important for fisheries is possible with the right combination of components. Those components include regional predictability on seasonal timescales of the physical environment from a large-scale model, a high-resolution regional model with biogeochemistry that simulates seasonal conditions in hindcasts, a relationship with local stakeholders, and a real-time observational network. Multiple efforts and approaches in different regions would advance knowledge to provide additional tools to fishers and other stakeholders.

Evaluation and economic value of winter weather forecasts

NASA Astrophysics Data System (ADS)

Snyder, Derrick W.

State and local highway agencies spend millions of dollars each year to deploy winter operation teams to plow snow and de-ice roadways. Accurate and timely weather forecast information is critical for effective decision making. Students from Purdue University partnered with the Indiana Department of Transportation to create an experimental winter weather forecast service for the 2012-2013 winter season in Indiana to assist in achieving these goals. One forecast product, an hourly timeline of winter weather hazards produced daily, was evaluated for quality and economic value. Verification of the forecasts was performed with data from the Rapid Refresh numerical weather model. Two objective verification criteria were developed to evaluate the performance of the timeline forecasts. Using both criteria, the timeline forecasts had issues with reliability and discrimination, systematically over-forecasting the amount of winter weather that was observed while also missing significant winter weather events. Despite these quality issues, the forecasts still showed significant, but varied, economic value compared to climatology. Economic value of the forecasts was estimated to be 29.5 million or 4.1 million, depending on the verification criteria used. Limitations of this valuation system are discussed and a framework is developed for more thorough studies in the future.

Short-term streamflow forecasting with global climate change implications A comparative study between genetic programming and neural network models

NASA Astrophysics Data System (ADS)

Makkeasorn, A.; Chang, N. B.; Zhou, X.

2008-05-01

SummarySustainable water resources management is a critically important priority across the globe. While water scarcity limits the uses of water in many ways, floods may also result in property damages and the loss of life. To more efficiently use the limited amount of water under the changing world or to resourcefully provide adequate time for flood warning, the issues have led us to seek advanced techniques for improving streamflow forecasting on a short-term basis. This study emphasizes the inclusion of sea surface temperature (SST) in addition to the spatio-temporal rainfall distribution via the Next Generation Radar (NEXRAD), meteorological data via local weather stations, and historical stream data via USGS gage stations to collectively forecast discharges in a semi-arid watershed in south Texas. Two types of artificial intelligence models, including genetic programming (GP) and neural network (NN) models, were employed comparatively. Four numerical evaluators were used to evaluate the validity of a suite of forecasting models. Research findings indicate that GP-derived streamflow forecasting models were generally favored in the assessment in which both SST and meteorological data significantly improve the accuracy of forecasting. Among several scenarios, NEXRAD rainfall data were proven its most effectiveness for a 3-day forecast, and SST Gulf-to-Atlantic index shows larger impacts than the SST Gulf-to-Pacific index on the streamflow forecasts. The most forward looking GP-derived models can even perform a 30-day streamflow forecast ahead of time with an r-square of 0.84 and RMS error 5.4 in our study.

Testing hypotheses of earthquake occurrence

NASA Astrophysics Data System (ADS)

Kagan, Y. Y.; Jackson, D. D.; Schorlemmer, D.; Gerstenberger, M.

2003-12-01

We present a relatively straightforward likelihood method for testing those earthquake hypotheses that can be stated as vectors of earthquake rate density in defined bins of area, magnitude, and time. We illustrate the method as it will be applied to the Regional Earthquake Likelihood Models (RELM) project of the Southern California Earthquake Center (SCEC). Several earthquake forecast models are being developed as part of this project, and additional contributed forecasts are welcome. Various models are based on fault geometry and slip rates, seismicity, geodetic strain, and stress interactions. We would test models in pairs, requiring that both forecasts in a pair be defined over the same set of bins. Thus we offer a standard "menu" of bins and ground rules to encourage standardization. One menu category includes five-year forecasts of magnitude 5.0 and larger. Forecasts would be in the form of a vector of yearly earthquake rates on a 0.05 degree grid at the beginning of the test. Focal mechanism forecasts, when available, would be also be archived and used in the tests. The five-year forecast category may be appropriate for testing hypotheses of stress shadows from large earthquakes. Interim progress will be evaluated yearly, but final conclusions would be made on the basis of cumulative five-year performance. The second category includes forecasts of earthquakes above magnitude 4.0 on a 0.05 degree grid, evaluated and renewed daily. Final evaluation would be based on cumulative performance over five years. Other types of forecasts with different magnitude, space, and time sampling are welcome and will be tested against other models with shared characteristics. All earthquakes would be counted, and no attempt made to separate foreshocks, main shocks, and aftershocks. Earthquakes would be considered as point sources located at the hypocenter. For each pair of forecasts, we plan to compute alpha, the probability that the first would be wrongly rejected in favor of the second, and beta, the probability that the second would be wrongly rejected in favor of the first. Computing alpha and beta requires knowing the theoretical distribution of likelihood scores under each hypothesis, which we will estimate by simulations. Each forecast is given equal status; there is no "null hypothesis" which would be accepted by default. Forecasts and test results would be archived and posted on the RELM web site. The same methods can be applied to any region with adequate monitoring and sufficient earthquakes. If fewer than ten events are forecasted, the likelihood tests may not give definitive results. The tests do force certain requirements on the forecast models. Because the tests are based on absolute rates, stress models must be explicit about how stress increments affect past seismicity rates. Aftershocks of triggered events must be accounted for. Furthermore, the tests are sensitive to magnitude, so forecast models must specify the magnitude distribution of triggered events. Models should account for probable errors in magnitude and location by appropriate smoothing of the probabilities, as the tests will be "cold hearted:" near misses won't count.

Applied Meteorology Unit (AMU) Quarterly Report - Fourth Quarter FY-10

NASA Technical Reports Server (NTRS)

Bauman, William; Crawford, Winifred; Barrett, Joe; Watson, Leela; Wheeler, Mark

2010-01-01

Three AMU tasks were completed in this Quarter, each resulting in a forecast tool now being used in operations and a final report documenting how the work was done. AMU personnel completed the following tasks (1) Phase II of the Peak Wind Tool for General Forecasting task by delivering an improved wind forecasting tool to operations and providing training on its use; (2) a graphical user interface (GUI) she updated with new scripts to complete the ADAS Update and Maintainability task, and delivered the scripts to the Spaceflight Meteorology Group on Johnson Space Center, Texas and National Weather Service in Melbourne, Fla.; and (3) the Verify MesoNAM Performance task after we created and delivered a GUI that forecasters will use to determine the performance of the operational MesoNAM weather model forecast.

A hybrid least squares support vector machines and GMDH approach for river flow forecasting

NASA Astrophysics Data System (ADS)

Samsudin, R.; Saad, P.; Shabri, A.

2010-06-01

This paper proposes a novel hybrid forecasting model, which combines the group method of data handling (GMDH) and the least squares support vector machine (LSSVM), known as GLSSVM. The GMDH is used to determine the useful input variables for LSSVM model and the LSSVM model which works as time series forecasting. In this study the application of GLSSVM for monthly river flow forecasting of Selangor and Bernam River are investigated. The results of the proposed GLSSVM approach are compared with the conventional artificial neural network (ANN) models, Autoregressive Integrated Moving Average (ARIMA) model, GMDH and LSSVM models using the long term observations of monthly river flow discharge. The standard statistical, the root mean square error (RMSE) and coefficient of correlation (R) are employed to evaluate the performance of various models developed. Experiment result indicates that the hybrid model was powerful tools to model discharge time series and can be applied successfully in complex hydrological modeling.

Developing a dengue forecast model using machine learning: A case study in China.

PubMed

Guo, Pi; Liu, Tao; Zhang, Qin; Wang, Li; Xiao, Jianpeng; Zhang, Qingying; Luo, Ganfeng; Li, Zhihao; He, Jianfeng; Zhang, Yonghui; Ma, Wenjun

2017-10-01

In China, dengue remains an important public health issue with expanded areas and increased incidence recently. Accurate and timely forecasts of dengue incidence in China are still lacking. We aimed to use the state-of-the-art machine learning algorithms to develop an accurate predictive model of dengue. Weekly dengue cases, Baidu search queries and climate factors (mean temperature, relative humidity and rainfall) during 2011-2014 in Guangdong were gathered. A dengue search index was constructed for developing the predictive models in combination with climate factors. The observed year and week were also included in the models to control for the long-term trend and seasonality. Several machine learning algorithms, including the support vector regression (SVR) algorithm, step-down linear regression model, gradient boosted regression tree algorithm (GBM), negative binomial regression model (NBM), least absolute shrinkage and selection operator (LASSO) linear regression model and generalized additive model (GAM), were used as candidate models to predict dengue incidence. Performance and goodness of fit of the models were assessed using the root-mean-square error (RMSE) and R-squared measures. The residuals of the models were examined using the autocorrelation and partial autocorrelation function analyses to check the validity of the models. The models were further validated using dengue surveillance data from five other provinces. The epidemics during the last 12 weeks and the peak of the 2014 large outbreak were accurately forecasted by the SVR model selected by a cross-validation technique. Moreover, the SVR model had the consistently smallest prediction error rates for tracking the dynamics of dengue and forecasting the outbreaks in other areas in China. The proposed SVR model achieved a superior performance in comparison with other forecasting techniques assessed in this study. The findings can help the government and community respond early to dengue epidemics.

A novel single-parameter approach for forecasting algal blooms.

PubMed

Xiao, Xi; He, Junyu; Huang, Haomin; Miller, Todd R; Christakos, George; Reichwaldt, Elke S; Ghadouani, Anas; Lin, Shengpan; Xu, Xinhua; Shi, Jiyan

2017-01-01

Harmful algal blooms frequently occur globally, and forecasting could constitute an essential proactive strategy for bloom control. To decrease the cost of aquatic environmental monitoring and increase the accuracy of bloom forecasting, a novel single-parameter approach combining wavelet analysis with artificial neural networks (WNN) was developed and verified based on daily online monitoring datasets of algal density in the Siling Reservoir, China and Lake Winnebago, U.S.A. Firstly, a detailed modeling process was illustrated using the forecasting of cyanobacterial cell density in the Chinese reservoir as an example. Three WNN models occupying various prediction time intervals were optimized through model training using an early stopped training approach. All models performed well in fitting historical data and predicting the dynamics of cyanobacterial cell density, with the best model predicting cyanobacteria density one-day ahead (r = 0.986 and mean absolute error = 0.103 × 10 4  cells mL -1 ). Secondly, the potential of this novel approach was further confirmed by the precise predictions of algal biomass dynamics measured as chl a in both study sites, demonstrating its high performance in forecasting algal blooms, including cyanobacteria as well as other blooming species. Thirdly, the WNN model was compared to current algal forecasting methods (i.e. artificial neural networks, autoregressive integrated moving average model), and was found to be more accurate. In addition, the application of this novel single-parameter approach is cost effective as it requires only a buoy-mounted fluorescent probe, which is merely a fraction (∼15%) of the cost of a typical auto-monitoring system. As such, the newly developed approach presents a promising and cost-effective tool for the future prediction and management of harmful algal blooms. Copyright © 2016 Elsevier Ltd. All rights reserved.

Forecasting Global Horizontal Irradiance Using the LETKF and a Combination of Advected Satellite Images and Sparse Ground Sensors

NASA Astrophysics Data System (ADS)

Harty, T. M.; Lorenzo, A.; Holmgren, W.; Morzfeld, M.

2017-12-01

The irradiance incident on a solar panel is the main factor in determining the power output of that panel. For this reason, accurate global horizontal irradiance (GHI) estimates and forecasts are critical when determining the optimal location for a solar power plant, forecasting utility scale solar power production, or forecasting distributed, behind the meter rooftop solar power production. Satellite images provide a basis for producing the GHI estimates needed to undertake these objectives. The focus of this work is to combine satellite derived GHI estimates with ground sensor measurements and an advection model. The idea is to use accurate but sparsely distributed ground sensors to improve satellite derived GHI estimates which can cover large areas (the size of a city or a region of the United States). We use a Bayesian framework to perform the data assimilation, which enables us to produce irradiance forecasts and associated uncertainties which incorporate both satellite and ground sensor data. Within this framework, we utilize satellite images taken from the GOES-15 geostationary satellite (available every 15-30 minutes) as well as ground data taken from irradiance sensors and rooftop solar arrays (available every 5 minutes). The advection model, driven by wind forecasts from a numerical weather model, simulates cloud motion between measurements. We use the Local Ensemble Transform Kalman Filter (LETKF) to perform the data assimilation. We present preliminary results towards making such a system useful in an operational context. We explain how localization and inflation in the LETKF, perturbations of wind-fields, and random perturbations of the advection model, affect the accuracy of our estimates and forecasts. We present experiments showing the accuracy of our forecasted GHI over forecast-horizons of 15 mins to 1 hr. The limitations of our approach and future improvements are also discussed.

Characterizing Time Series Data Diversity for Wind Forecasting: Preprint

DOE Office of Scientific and Technical Information (OSTI.GOV)

Hodge, Brian S; Chartan, Erol Kevin; Feng, Cong

Wind forecasting plays an important role in integrating variable and uncertain wind power into the power grid. Various forecasting models have been developed to improve the forecasting accuracy. However, it is challenging to accurately compare the true forecasting performances from different methods and forecasters due to the lack of diversity in forecasting test datasets. This paper proposes a time series characteristic analysis approach to visualize and quantify wind time series diversity. The developed method first calculates six time series characteristic indices from various perspectives. Then the principal component analysis is performed to reduce the data dimension while preserving the importantmore » information. The diversity of the time series dataset is visualized by the geometric distribution of the newly constructed principal component space. The volume of the 3-dimensional (3D) convex polytope (or the length of 1D number axis, or the area of the 2D convex polygon) is used to quantify the time series data diversity. The method is tested with five datasets with various degrees of diversity.« less

Integrating Wind Profiling Radars and Radiosonde Observations with Model Point Data to Develop a Decision Support Tool to Assess Upper-Level Winds for Space Launch

NASA Technical Reports Server (NTRS)

Bauman, William H., III; Flinn, Clay

2013-01-01

On the day of launch, the 45th Weather Squadron (45 WS) Launch Weather Officers (LWOs) monitor the upper-level winds for their launch customers. During launch operations, the payload/launch team sometimes asks the LWOs if they expect the upper-level winds to change during the countdown. The LWOs used numerical weather prediction model point forecasts to provide the information, but did not have the capability to quickly retrieve or adequately display the upper-level observations and compare them directly in the same display to the model point forecasts to help them determine which model performed the best. The LWOs requested the Applied Meteorology Unit (AMU) develop a graphical user interface (GUI) that will plot upper-level wind speed and direction observations from the Cape Canaveral Air Force Station (CCAFS) Automated Meteorological Profiling System (AMPS) rawinsondes with point forecast wind profiles from the National Centers for Environmental Prediction (NCEP) North American Mesoscale (NAM), Rapid Refresh (RAP) and Global Forecast System (GFS) models to assess the performance of these models. The AMU suggested adding observations from the NASA 50 MHz wind profiler and one of the US Air Force 915 MHz wind profilers, both located near the Kennedy Space Center (KSC) Shuttle Landing Facility, to supplement the AMPS observations with more frequent upper-level profiles. Figure 1 shows a map of KSC/CCAFS with the locations of the observation sites and the model point forecasts.

WRF model forecasts and their use for hydroclimate monitoring over southern South America

NASA Astrophysics Data System (ADS)

Muller, Omar; Lovino, Miguel; Berbery, E. Hugo

2017-04-01

Weather forecasting and monitoring systems based on regional models are becoming increasingly relevant for decision support in agriculture and water management. This work evaluates the predictive and monitoring capabilities of a system based on WRF model simulations at 15 km grid spacing over a domain that encompasses La Plata Basin (LPB) in southern South America, where agriculture and water resources are essential. The model's skill up to a lead-time of 7 days is evaluated with daily precipitation and 2m temperature in-situ observations. Results show high prediction performance with 7 days lead-time throughout the domain and particularly over LPB, where about 70% of rain and no-rain days are correctly predicted. The scores tend to be better over humid climates than over arid-to-semiarid climates. Compared to the arid-semiarid climate, the humid climate has a higher probability of detection and less false alarms. The ranges of the skill scores are similar to those found over the United States, suggesting that proper choice of parameterizations lead to no loss of performance of the model. Daily mean, minimum and maximum forecast temperatures are highly correlated with observations up to 7 day lead time. The best performance is for daily mean temperature, followed by minimum temperature and a slightly weaker performance for maximum temperature over arid regions. The usefulness of WRF products for hydroclimate monitoring was tested for an unprecedented drought in southern Brazil and for a slightly above normal precipitation season in northeastern Argentina. In both cases the model products reproduce the observed precipitation conditions with consistent impacts on soil moisture, evapotranspiration and runoff. This evaluation validates the model's usefulness to fore-cast weather up to one week and to monitor climate conditions in real time. The scores suggest that the forecast lead-time can be extended into week two, while bias correction methods can reduce part of the systematic errors.

CALIPSO Satellite Lidar Identification Of Elevated Dust Over Australia Compared With Air Quality Model PM60 Forecasts

NASA Technical Reports Server (NTRS)

Young, Stuart A.; Vaughan, Mark; Omar, Ali; Liu, Zhaoyan; Lee, Sunhee; Hu, Youngxiang; Cope, Martin

2008-01-01

Global measurements of the vertical distribution of clouds and aerosols have been recorded by the lidar on board the CALIPSO (Cloud Aerosol Lidar Infrared Pathfinder Satellite Observations) satellite since June 2006. Such extensive, height-resolved measurements provide a rare and valuable opportunity for developing, testing and validating various atmospheric models, including global climate, numerical weather prediction, chemical transport and air quality models. Here we report on the initial results of an investigation into the performance of the Australian Air Quality Forecast System (AAQFS) model in forecasting the distribution of elevated dust over the Australian region. The model forecasts of PM60 dust distribution are compared with the CALIPSO lidar Vertical Feature Mask (VFM) data product. The VFM classifies contiguous atmospheric regions of enhanced backscatter as either cloud or aerosols. Aerosols are further classified into six subtypes. By comparing forecast PM60 concentration profiles to the spatial distribution of dust reported in the CALIPSO VFM, we can assess the model s ability to predict the occurrence and the vertical and horizontal extents of dust events within the study area.

Development of the GEM-MACH-FireWork System: An Air Quality Model with On-line Wildfire Emissions within the Canadian Operational Air Quality Forecast System

NASA Astrophysics Data System (ADS)

Pavlovic, Radenko; Chen, Jack; Beaulieu, Paul-Andre; Anselmp, David; Gravel, Sylvie; Moran, Mike; Menard, Sylvain; Davignon, Didier

2014-05-01

A wildfire emissions processing system has been developed to incorporate near-real-time emissions from wildfires and large prescribed burns into Environment Canada's real-time GEM-MACH air quality (AQ) forecast system. Since the GEM-MACH forecast domain covers Canada and most of the U.S.A., including Alaska, fire location information is needed for both of these large countries. During AQ model runs, emissions from individual fire sources are injected into elevated model layers based on plume-rise calculations and then transport and chemistry calculations are performed. This "on the fly" approach to the insertion of the fire emissions provides flexibility and efficiency since on-line meteorology is used and computational overhead in emissions pre-processing is reduced. GEM-MACH-FireWork, an experimental wildfire version of GEM-MACH, was run in real-time mode for the summers of 2012 and 2013 in parallel with the normal operational version. 48-hour forecasts were generated every 12 hours (at 00 and 12 UTC). Noticeable improvements in the AQ forecasts for PM2.5 were seen in numerous regions where fire activity was high. Case studies evaluating model performance for specific regions and computed objective scores will be included in this presentation. Using the lessons learned from the last two summers, Environment Canada will continue to work towards the goal of incorporating near-real-time intermittent wildfire emissions into the operational air quality forecast system.

Real-Time Kennedy Space Center and Cape Canaveral Air Force Station High-Resolution Model Implementation and Verification

NASA Technical Reports Server (NTRS)

Shafer, Jaclyn A.; Watson, Leela R.

2015-01-01

Customer: NASA's Launch Services Program (LSP), Ground Systems Development and Operations (GSDO), and Space Launch System (SLS) programs. NASA's LSP, GSDO, SLS and other programs at Kennedy Space Center (KSC) and Cape Canaveral Air Force Station (CCAFS) use the daily and weekly weather forecasts issued by the 45th Weather Squadron (45 WS) as decision tools for their day-to-day and launch operations on the Eastern Range (ER). For example, to determine if they need to limit activities such as vehicle transport to the launch pad, protect people, structures or exposed launch vehicles given a threat of severe weather, or reschedule other critical operations. The 45 WS uses numerical weather prediction models as a guide for these weather forecasts, particularly the Air Force Weather Agency (AFWA) 1.67 kilometer Weather Research and Forecasting (WRF) model. Considering the 45 WS forecasters' and Launch Weather Officers' (LWO) extensive use of the AFWA model, the 45 WS proposed a task at the September 2013 Applied Meteorology Unit (AMU) Tasking Meeting requesting the AMU verify this model. Due to the lack of archived model data available from AFWA, verification is not yet possible. Instead, the AMU proposed to implement and verify the performance of an ER version of the AMU high-resolution WRF Environmental Modeling System (EMS) model (Watson 2013) in real-time. The tasking group agreed to this proposal; therefore the AMU implemented the WRF-EMS model on the second of two NASA AMU modeling clusters. The model was set up with a triple-nested grid configuration over KSC/CCAFS based on previous AMU work (Watson 2013). The outer domain (D01) has 12-kilometer grid spacing, the middle domain (D02) has 4-kilometer grid spacing, and the inner domain (D03) has 1.33-kilometer grid spacing. The model runs a 12-hour forecast every hour, D01 and D02 domain outputs are available once an hour and D03 is every 15 minutes during the forecast period. The AMU assessed the WRF-EMS 1.33-kilometer domain model performance for the 2014 warm season (May-September). Verification statistics were computed using the Model Evaluation Tools, which compared the model forecasts to observations. The mean error values were close to 0 and the root mean square error values were less than 1.8 for mean sea-level pressure (millibars), temperature (degrees Kelvin), dewpoint temperature (degrees Kelvin), and wind speed (per millisecond), all very small differences between the forecast and observations considering the normal magnitudes of the parameters. The precipitation forecast verification results showed consistent under-forecasting of the precipitation object size. This could be an artifact of calculating the statistics for each hour rather than for the entire 12-hour period. The AMU will continue to generate verification statistics for the 1.33-kilometer WRF-EMS domain as data become available in future cool and warm seasons. More data will produce more robust statistics and reveal a more accurate assessment of model performance. Once the formal task was complete, the AMU conducted additional work to better understand the wind direction results. The results were stratified diurnally and by wind speed to determine what effects the stratifications would have on the model wind direction verification statistics. The results are summarized in the addendum at the end of this report. In addition to verifying the model's performance, the AMU also made the output available in the Advanced Weather Interactive Processing System II (AWIPS II). This allows the 45 WS and AMU staff to customize the model output display on the AMU and Range Weather Operations AWIPS II client computers and conduct real-time subjective analyses. In the future, the AMU will implement an updated version of the WRF-EMS model that incorporates local data assimilation. This model will also run in real-time and be made available in AWIPS II.

The accuracy comparison between ARFIMA and singular spectrum analysis for forecasting the sales volume of motorcycle in Indonesia

NASA Astrophysics Data System (ADS)

Sitohang, Yosep Oktavianus; Darmawan, Gumgum

2017-08-01

This research attempts to compare between two forecasting models in time series analysis for predicting the sales volume of motorcycle in Indonesia. The first forecasting model used in this paper is Autoregressive Fractionally Integrated Moving Average (ARFIMA). ARFIMA can handle non-stationary data and has a better performance than ARIMA in forecasting accuracy on long memory data. This is because the fractional difference parameter can explain correlation structure in data that has short memory, long memory, and even both structures simultaneously. The second forecasting model is Singular spectrum analysis (SSA). The advantage of the technique is that it is able to decompose time series data into the classic components i.e. trend, cyclical, seasonal and noise components. This makes the forecasting accuracy of this technique significantly better. Furthermore, SSA is a model-free technique, so it is likely to have a very wide range in its application. Selection of the best model is based on the value of the lowest MAPE. Based on the calculation, it is obtained the best model for ARFIMA is ARFIMA (3, d = 0, 63, 0) with MAPE value of 22.95 percent. For SSA with a window length of 53 and 4 group of reconstructed data, resulting MAPE value of 13.57 percent. Based on these results it is concluded that SSA produces better forecasting accuracy.

Three models intercomparison for Quantitative Precipitation Forecast over Calabria

NASA Astrophysics Data System (ADS)

Federico, S.; Avolio, E.; Bellecci, C.; Colacino, M.; Lavagnini, A.; Accadia, C.; Mariani, S.; Casaioli, M.

2004-11-01

In the framework of the National Project “Sviluppo di distretti industriali per le Osservazioni della Terra” (Development of Industrial Districts for Earth Observations) funded by MIUR (Ministero dell'Università e della Ricerca Scientifica --Italian Ministry of the University and Scientific Research) two operational mesoscale models were set-up for Calabria, the southernmost tip of the Italian peninsula. Models are RAMS (Regional Atmospheric Modeling System) and MM5 (Mesoscale Modeling 5) that are run every day at Crati scrl to produce weather forecast over Calabria (http://www.crati.it). This paper reports model intercomparison for Quantitative Precipitation Forecast evaluated for a 20 month period from 1th October 2000 to 31th May 2002. In addition to RAMS and MM5 outputs, QBOLAM rainfall fields are available for the period selected and included in the comparison. This model runs operationally at “Agenzia per la Protezione dell'Ambiente e per i Servizi Tecnici”. Forecasts are verified comparing models outputs with raingauge data recorded by the regional meteorological network, which has 75 raingauges. Large-scale forcing is the same for all models considered and differences are due to physical/numerical parameterizations and horizontal resolutions. QPFs show differences between models. Largest differences are for BIA compared to the other considered scores. Performances decrease with increasing forecast time for RAMS and MM5, whilst QBOLAM scores better for second day forecast.

Empirical seasonal forecasts of the NAO

NASA Astrophysics Data System (ADS)

Sanchezgomez, E.; Ortizbevia, M.

2003-04-01

We present here seasonal forecasts of the North Atlantic Oscillation (NAO) issued from ocean predictors with an empirical procedure. The Singular Values Decomposition (SVD) of the cross-correlation matrix between predictor and predictand fields at the lag used for the forecast lead is at the core of the empirical model. The main predictor field are sea surface temperature anomalies, although sea ice cover anomalies are also used. Forecasts are issued in probabilistic form. The model is an improvement over a previous version (1), where Sea Level Pressure Anomalies were first forecast, and the NAO Index built from this forecast field. Both correlation skill between forecast and observed field, and number of forecasts that hit the correct NAO sign, are used to assess the forecast performance , usually above those values found in the case of forecasts issued assuming persistence. For certain seasons and/or leads, values of the skill are above the .7 usefulness treshold. References (1) SanchezGomez, E. and Ortiz Bevia M., 2002, Estimacion de la evolucion pluviometrica de la Espana Seca atendiendo a diversos pronosticos empiricos de la NAO, in 'El Agua y el Clima', Publicaciones de la AEC, Serie A, N 3, pp 63-73, Palma de Mallorca, Spain

Forecasting the Emergency Department Patients Flow.

PubMed

Afilal, Mohamed; Yalaoui, Farouk; Dugardin, Frédéric; Amodeo, Lionel; Laplanche, David; Blua, Philippe

2016-07-01

Emergency department (ED) have become the patient's main point of entrance in modern hospitals causing it frequent overcrowding, thus hospital managers are increasingly paying attention to the ED in order to provide better quality service for patients. One of the key elements for a good management strategy is demand forecasting. In this case, forecasting patients flow, which will help decision makers to optimize human (doctors, nurses…) and material(beds, boxs…) resources allocation. The main interest of this research is forecasting daily attendance at an emergency department. The study was conducted on the Emergency Department of Troyes city hospital center, France, in which we propose a new practical ED patients classification that consolidate the CCMU and GEMSA categories into one category and innovative time-series based models to forecast long and short term daily attendance. The models we developed for this case study shows very good performances (up to 91,24 % for the annual Total flow forecast) and robustness to epidemic periods.

Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS)

NASA Astrophysics Data System (ADS)

O'Connor, Alison; Kirtman, Benjamin; Harrison, Scott; Gorman, Joe

2016-05-01

The US Navy faces several limitations when planning operations in regard to forecasting environmental conditions. Currently, mission analysis and planning tools rely heavily on short-term (less than a week) forecasts or long-term statistical climate products. However, newly available data in the form of weather forecast ensembles provides dynamical and statistical extended-range predictions that can produce more accurate predictions if ensemble members can be combined correctly. Charles River Analytics is designing the Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS), which performs data fusion over extended-range multi-model ensembles, such as the North American Multi-Model Ensemble (NMME), to produce a unified forecast for several weeks to several seasons in the future. We evaluated thirty years of forecasts using machine learning to select predictions for an all-encompassing and superior forecast that can be used to inform the Navy's decision planning process.

Very short-term reactive forecasting of the solar ultraviolet index using an extreme learning machine integrated with the solar zenith angle.

PubMed

Deo, Ravinesh C; Downs, Nathan; Parisi, Alfio V; Adamowski, Jan F; Quilty, John M

2017-05-01

Exposure to erythemally-effective solar ultraviolet radiation (UVR) that contributes to malignant keratinocyte cancers and associated health-risk is best mitigated through innovative decision-support systems, with global solar UV index (UVI) forecast necessary to inform real-time sun-protection behaviour recommendations. It follows that the UVI forecasting models are useful tools for such decision-making. In this study, a model for computationally-efficient data-driven forecasting of diffuse and global very short-term reactive (VSTR) (10-min lead-time) UVI, enhanced by drawing on the solar zenith angle (θ s ) data, was developed using an extreme learning machine (ELM) algorithm. An ELM algorithm typically serves to address complex and ill-defined forecasting problems. UV spectroradiometer situated in Toowoomba, Australia measured daily cycles (0500-1700h) of UVI over the austral summer period. After trialling activations functions based on sine, hard limit, logarithmic and tangent sigmoid and triangular and radial basis networks for best results, an optimal ELM architecture utilising logarithmic sigmoid equation in hidden layer, with lagged combinations of θ s as the predictor data was developed. ELM's performance was evaluated using statistical metrics: correlation coefficient (r), Willmott's Index (WI), Nash-Sutcliffe efficiency coefficient (E NS ), root mean square error (RMSE), and mean absolute error (MAE) between observed and forecasted UVI. Using these metrics, the ELM model's performance was compared to that of existing methods: multivariate adaptive regression spline (MARS), M5 Model Tree, and a semi-empirical (Pro6UV) clear sky model. Based on RMSE and MAE values, the ELM model (0.255, 0.346, respectively) outperformed the MARS (0.310, 0.438) and M5 Model Tree (0.346, 0.466) models. Concurring with these metrics, the Willmott's Index for the ELM, MARS and M5 Model Tree models were 0.966, 0.942 and 0.934, respectively. About 57% of the ELM model's absolute errors were small in magnitude (±0.25), whereas the MARS and M5 Model Tree models generated 53% and 48% of such errors, respectively, indicating the latter models' errors to be distributed in larger magnitude error range. In terms of peak global UVI forecasting, with half the level of error, the ELM model outperformed MARS and M5 Model Tree. A comparison of the magnitude of hourly-cumulated errors of 10-min lead time forecasts for diffuse and global UVI highlighted ELM model's greater accuracy compared to MARS, M5 Model Tree or Pro6UV models. This confirmed the versatility of an ELM model drawing on θ s data for VSTR forecasting of UVI at near real-time horizon. When applied to the goal of enhancing expert systems, ELM-based accurate forecasts capable of reacting quickly to measured conditions can enhance real-time exposure advice for the public, mitigating the potential for solar UV-exposure-related disease. Crown Copyright © 2017. Published by Elsevier Inc. All rights reserved.

Statistical Correction of Air Temperature Forecasts for City and Road Weather Applications

NASA Astrophysics Data System (ADS)

Mahura, Alexander; Petersen, Claus; Sass, Bent; Gilet, Nicolas

2014-05-01

The method for statistical correction of air /road surface temperatures forecasts was developed based on analysis of long-term time-series of meteorological observations and forecasts (from HIgh Resolution Limited Area Model & Road Conditions Model; 3 km horizontal resolution). It has been tested for May-Aug 2012 & Oct 2012 - Mar 2013, respectively. The developed method is based mostly on forecasted meteorological parameters with a minimal inclusion of observations (covering only a pre-history period). Although the st iteration correction is based taking into account relevant temperature observations, but the further adjustment of air and road temperature forecasts is based purely on forecasted meteorological parameters. The method is model independent, e.g. it can be applied for temperature correction with other types of models having different horizontal resolutions. It is relatively fast due to application of the singular value decomposition method for matrix solution to find coefficients. Moreover, there is always a possibility for additional improvement due to extra tuning of the temperature forecasts for some locations (stations), and in particular, where for example, the MAEs are generally higher compared with others (see Gilet et al., 2014). For the city weather applications, new operationalized procedure for statistical correction of the air temperature forecasts has been elaborated and implemented for the HIRLAM-SKA model runs at 00, 06, 12, and 18 UTCs covering forecast lengths up to 48 hours. The procedure includes segments for extraction of observations and forecast data, assigning these to forecast lengths, statistical correction of temperature, one-&multi-days statistical evaluation of model performance, decision-making on using corrections by stations, interpolation, visualisation and storage/backup. Pre-operational air temperature correction runs were performed for the mainland Denmark since mid-April 2013 and shown good results. Tests also showed that the CPU time required for the operational procedure is relatively short (less than 15 minutes including a large time spent for interpolation). These also showed that in order to start correction of forecasts there is no need to have a long-term pre-historical data (containing forecasts and observations) and, at least, a couple of weeks will be sufficient when a new observational station is included and added to the forecast point. Note for the road weather application, the operationalization of the statistical correction of the road surface temperature forecasts (for the RWM system daily hourly runs covering forecast length up to 5 hours ahead) for the Danish road network (for about 400 road stations) was also implemented, and it is running in a test mode since Sep 2013. The method can also be applied for correction of the dew point temperature and wind speed (as a part of observations/ forecasts at synoptical stations), where these both meteorological parameters are parts of the proposed system of equations. The evaluation of the method performance for improvement of the wind speed forecasts is planned as well, with considering possibilities for the wind direction improvements (which is more complex due to multi-modal types of such data distribution). The method worked for the entire domain of mainland Denmark (tested for 60 synoptical and 395 road stations), and hence, it can be also applied for any geographical point within this domain, as through interpolation to about 100 cities' locations (for Danish national byvejr forecasts). Moreover, we can assume that the same method can be used in other geographical areas. The evaluation for other domains (with a focus on Greenland and Nordic countries) is planned. In addition, a similar approach might be also tested for statistical correction of concentrations of chemical species, but such approach will require additional elaboration and evaluation.

Fuzzy Temporal Logic Based Railway Passenger Flow Forecast Model

PubMed Central

Dou, Fei; Jia, Limin; Wang, Li; Xu, Jie; Huang, Yakun

2014-01-01

Passenger flow forecast is of essential importance to the organization of railway transportation and is one of the most important basics for the decision-making on transportation pattern and train operation planning. Passenger flow of high-speed railway features the quasi-periodic variations in a short time and complex nonlinear fluctuation because of existence of many influencing factors. In this study, a fuzzy temporal logic based passenger flow forecast model (FTLPFFM) is presented based on fuzzy logic relationship recognition techniques that predicts the short-term passenger flow for high-speed railway, and the forecast accuracy is also significantly improved. An applied case that uses the real-world data illustrates the precision and accuracy of FTLPFFM. For this applied case, the proposed model performs better than the k-nearest neighbor (KNN) and autoregressive integrated moving average (ARIMA) models. PMID:25431586

Assessing the Value of Post-processed State-of-the-art Long-term Weather Forecast Ensembles within An Integrated Agronomic Modelling Framework

NASA Astrophysics Data System (ADS)

LI, Y.; Castelletti, A.; Giuliani, M.

2014-12-01

Over recent years, long-term climate forecast from global circulation models (GCMs) has been demonstrated to show increasing skills over the climatology, thanks to the advances in the modelling of coupled ocean-atmosphere dynamics. Improved information from long-term forecast is supposed to be a valuable support to farmers in optimizing farming operations (e.g. crop choice, cropping time) and for more effectively coping with the adverse impacts of climate variability. Yet, evaluating how valuable this information can be is not straightforward and farmers' response must be taken into consideration. Indeed, while long-range forecast are traditionally evaluated in terms of accuracy by comparison of hindcast and observed values, in the context of agricultural systems, potentially useful forecast information should alter the stakeholders' expectation, modify their decisions and ultimately have an impact on their annual benefit. Therefore, it is more desirable to assess the value of those long-term forecasts via decision-making models so as to extract direct indication of probable decision outcomes from farmers, i.e. from an end-to-end perspective. In this work, we evaluate the operational value of thirteen state-of-the-art long-range forecast ensembles against climatology forecast and subjective prediction (i.e. past year climate and historical average) within an integrated agronomic modeling framework embedding an implicit model of farmers' behavior. Collected ensemble datasets are bias-corrected and downscaled using a stochastic weather generator, in order to address the mismatch of the spatio-temporal scale between forecast data from GCMs and distributed crop simulation model. The agronomic model is first simulated using the forecast information (ex-ante), followed by a second run with actual climate (ex-post). Multi-year simulations are performed to account for climate variability and the value of the different climate forecast is evaluated against the perfect foresight scenario based on the expected crop productivity as well as the land-use decisions. Our results show that not all the products generate beneficial effects to farmers and that the forecast errors might be amplified by the farmers decisions.

Research and Application of an Air Quality Early Warning System Based on a Modified Least Squares Support Vector Machine and a Cloud Model.

PubMed

Wang, Jianzhou; Niu, Tong; Wang, Rui

2017-03-02

The worsening atmospheric pollution increases the necessity of air quality early warning systems (EWSs). Despite the fact that a massive amount of investigation about EWS in theory and practicality has been conducted by numerous researchers, studies concerning the quantification of uncertain information and comprehensive evaluation are still lacking, which impedes further development in the area. In this paper, firstly a comprehensive warning system is proposed, which consists of two vital indispensable modules, namely effective forecasting and scientific evaluation, respectively. For the forecasting module, a novel hybrid model combining the theory of data preprocessing and numerical optimization is first developed to implement effective forecasting for air pollutant concentration. Especially, in order to further enhance the accuracy and robustness of the warning system, interval forecasting is implemented to quantify the uncertainties generated by forecasts, which can provide significant risk signals by using point forecasting for decision-makers. For the evaluation module, a cloud model, based on probability and fuzzy set theory, is developed to perform comprehensive evaluations of air quality, which can realize the transformation between qualitative concept and quantitative data. To verify the effectiveness and efficiency of the warning system, extensive simulations based on air pollutants data from Dalian in China were effectively implemented, which illustrate that the warning system is not only remarkably high-performance, but also widely applicable.

Research and Application of an Air Quality Early Warning System Based on a Modified Least Squares Support Vector Machine and a Cloud Model

PubMed Central

Wang, Jianzhou; Niu, Tong; Wang, Rui

2017-01-01

The worsening atmospheric pollution increases the necessity of air quality early warning systems (EWSs). Despite the fact that a massive amount of investigation about EWS in theory and practicality has been conducted by numerous researchers, studies concerning the quantification of uncertain information and comprehensive evaluation are still lacking, which impedes further development in the area. In this paper, firstly a comprehensive warning system is proposed, which consists of two vital indispensable modules, namely effective forecasting and scientific evaluation, respectively. For the forecasting module, a novel hybrid model combining the theory of data preprocessing and numerical optimization is first developed to implement effective forecasting for air pollutant concentration. Especially, in order to further enhance the accuracy and robustness of the warning system, interval forecasting is implemented to quantify the uncertainties generated by forecasts, which can provide significant risk signals by using point forecasting for decision-makers. For the evaluation module, a cloud model, based on probability and fuzzy set theory, is developed to perform comprehensive evaluations of air quality, which can realize the transformation between qualitative concept and quantitative data. To verify the effectiveness and efficiency of the warning system, extensive simulations based on air pollutants data from Dalian in China were effectively implemented, which illustrate that the warning system is not only remarkably high-performance, but also widely applicable. PMID:28257122

Forecasting volcanic eruptions and other material failure phenomena: An evaluation of the failure forecast method

NASA Astrophysics Data System (ADS)

Bell, Andrew F.; Naylor, Mark; Heap, Michael J.; Main, Ian G.

2011-08-01

Power-law accelerations in the mean rate of strain, earthquakes and other precursors have been widely reported prior to material failure phenomena, including volcanic eruptions, landslides and laboratory deformation experiments, as predicted by several theoretical models. The Failure Forecast Method (FFM), which linearizes the power-law trend, has been routinely used to forecast the failure time in retrospective analyses; however, its performance has never been formally evaluated. Here we use synthetic and real data, recorded in laboratory brittle creep experiments and at volcanoes, to show that the assumptions of the FFM are inconsistent with the error structure of the data, leading to biased and imprecise forecasts. We show that a Generalized Linear Model method provides higher-quality forecasts that converge more accurately to the eventual failure time, accounting for the appropriate error distributions. This approach should be employed in place of the FFM to provide reliable quantitative forecasts and estimate their associated uncertainties.

AQA - Air Quality model for Austria - Evaluation and Developments

NASA Astrophysics Data System (ADS)

Hirtl, M.; Krüger, B. C.; Baumann-Stanzer, K.; Skomorowski, P.

2009-04-01

The regional weather forecast model ALADIN of the Central Institute for Meteorology and Geodynamics (ZAMG) is used in combination with the chemical transport model CAMx (www.camx.com) to conduct forecasts of gaseous and particulate air pollution over Europe. The forecasts which are done in cooperation with the University of Natural Resources and Applied Life Sciences in Vienna (BOKU) are supported by the regional governments since 2005 with the main interest on the prediction of tropospheric ozone. The daily ozone forecasts are evaluated for the summer 2008 with the observations of about 150 air quality stations in Austria. In 2008 the emission-model SMOKE was integrated into the modelling system to calculate the biogenic emissions. The anthropogenic emissions are based on the newest EMEP data set as well as on regional inventories for the core domain. The performance of SMOKE is shown for a summer period in 2007. In the frame of the COST-action 728 „Enhancing mesoscale meteorological modelling capabilities for air pollution and dispersion applications", multi-model ensembles are used to conduct an international model evaluation. The model calculations of meteorological- and concentration fields are compared to measurements on the ensemble platform at the Joint Research Centre (JRC) in Ispra. The results for 2 episodes in 2006 show the performance of the different models as well as of the model ensemble.

High-order fuzzy time-series based on multi-period adaptation model for forecasting stock markets

NASA Astrophysics Data System (ADS)

Chen, Tai-Liang; Cheng, Ching-Hsue; Teoh, Hia-Jong

2008-02-01

Stock investors usually make their short-term investment decisions according to recent stock information such as the late market news, technical analysis reports, and price fluctuations. To reflect these short-term factors which impact stock price, this paper proposes a comprehensive fuzzy time-series, which factors linear relationships between recent periods of stock prices and fuzzy logical relationships (nonlinear relationships) mined from time-series into forecasting processes. In empirical analysis, the TAIEX (Taiwan Stock Exchange Capitalization Weighted Stock Index) and HSI (Heng Seng Index) are employed as experimental datasets, and four recent fuzzy time-series models, Chen’s (1996), Yu’s (2005), Cheng’s (2006) and Chen’s (2007), are used as comparison models. Besides, to compare with conventional statistic method, the method of least squares is utilized to estimate the auto-regressive models of the testing periods within the databases. From analysis results, the performance comparisons indicate that the multi-period adaptation model, proposed in this paper, can effectively improve the forecasting performance of conventional fuzzy time-series models which only factor fuzzy logical relationships in forecasting processes. From the empirical study, the traditional statistic method and the proposed model both reveal that stock price patterns in the Taiwan stock and Hong Kong stock markets are short-term.

Forecasting Performance of Grey Prediction for Education Expenditure and School Enrollment

ERIC Educational Resources Information Center

Tang, Hui-Wen Vivian; Yin, Mu-Shang

2012-01-01

GM(1,1) and GM(1,1) rolling models derived from grey system theory were estimated using time-series data from projection studies by National Center for Education Statistics (NCES). An out-of-sample forecasting competition between the two grey prediction models and exponential smoothing used by NCES was conducted for education expenditure and…

The Impact of Microphysics on Intensity and Structure of Hurricanes

NASA Technical Reports Server (NTRS)

Tao, Wei-Kuo; Shi, Jainn; Lang, Steve; Peters-Lidard, Christa

2006-01-01

During the past decade, both research and operational numerical weather prediction models, e.g. Weather Research and Forecast (WRF) model, have started using more complex microphysical schemes originally developed for high-resolution cloud resolving models (CRMs) with a 1-2 km or less horizontal resolutions. WFW is a next-generation mesoscale forecast model and assimilation system that has incorporated modern software framework, advanced dynamics, numeric and data assimilation techniques, a multiple moveable nesting capability, and improved physical packages. WFW model can be used for a wide range of applications, from idealized research to operational forecasting, with an emphasis on horizontal grid sizes in the range of 1-10 km. The current WRF includes several different microphysics options such as Lin et al. (1983), WSM 6-class and Thompson microphysics schemes. We have recently implemented three sophisticated cloud microphysics schemes into WRF. The cloud microphysics schemes have been extensively tested and applied for different mesoscale systems in different geographical locations. The performances of these schemes have been compared to those from other WRF microphysics options. We are performing sensitivity tests in using WW to examine the impact of six different cloud microphysical schemes on hurricane track, intensity and rainfall forecast. We are also performing the inline tracer calculation to comprehend the physical processes @e., boundary layer and each quadrant in the boundary layer) related to the development and structure of hurricanes.

The Economic Value of Air Quality Forecasting

NASA Astrophysics Data System (ADS)

Anderson-Sumo, Tasha

Both long-term and daily air quality forecasts provide an essential component to human health and impact costs. According the American Lung Association, the estimated current annual cost of air pollution related illness in the United States, adjusted for inflation (3% per year), is approximately $152 billion. Many of the risks such as hospital visits and morality are associated with poor air quality days (where the Air Quality Index is greater than 100). Groups such as sensitive groups become more susceptible to the resulting conditions and more accurate forecasts would help to take more appropriate precautions. This research focuses on evaluating the utility of air quality forecasting in terms of its potential impacts by building on air quality forecasting and economical metrics. Our analysis includes data collected during the summertime ozone seasons between 2010 and 2012 from air quality models for the Washington, DC/Baltimore, MD region. The metrics that are relevant to our analysis include: (1) The number of times that a high ozone or particulate matter (PM) episode is correctly forecasted, (2) the number of times that high ozone or PM episode is forecasted when it does not occur and (3) the number of times when the air quality forecast predicts a cleaner air episode when the air was observed to have high ozone or PM. Our collection of data included available air quality model forecasts of ozone and particulate matter data from the U.S. Environmental Protection Agency (EPA)'s AIRNOW as well as observational data of ozone and particulate matter from Clean Air Partners. We evaluated the performance of the air quality forecasts with that of the observational data and found that the forecast models perform well for the Baltimore/Washington region and the time interval observed. We estimate the potential amount for the Baltimore/Washington region accrues to a savings of up to 5,905 lives and 5.9 billion dollars per year. This total assumes perfect compliance with bad air quality warning and forecast air quality forecasts. There is a difficulty presented with evaluating the economic utility of the forecasts. All may not comply and even with a low compliance rate of 5% and 72% as the average probability of detection of poor air quality days by the air quality models, we estimate that the forecasting program saves 412 lives or 412 million dollars per year for the region. The totals we found are great or greater than other typical yearly meteorological hazard programs such as tornado or hurricane forecasting and it is clear that the economic value of air quality forecasting in the Baltimore/Washington region is vital.

Using a Hybrid Model to Forecast the Prevalence of Schistosomiasis in Humans

PubMed Central

Zhou, Lingling; Xia, Jing; Yu, Lijing; Wang, Ying; Shi, Yun; Cai, Shunxiang; Nie, Shaofa

2016-01-01

Background: We previously proposed a hybrid model combining both the autoregressive integrated moving average (ARIMA) and the nonlinear autoregressive neural network (NARNN) models in forecasting schistosomiasis. Our purpose in the current study was to forecast the annual prevalence of human schistosomiasis in Yangxin County, using our ARIMA-NARNN model, thereby further certifying the reliability of our hybrid model. Methods: We used the ARIMA, NARNN and ARIMA-NARNN models to fit and forecast the annual prevalence of schistosomiasis. The modeling time range included was the annual prevalence from 1956 to 2008 while the testing time range included was from 2009 to 2012. The mean square error (MSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) were used to measure the model performance. We reconstructed the hybrid model to forecast the annual prevalence from 2013 to 2016. Results: The modeling and testing errors generated by the ARIMA-NARNN model were lower than those obtained from either the single ARIMA or NARNN models. The predicted annual prevalence from 2013 to 2016 demonstrated an initial decreasing trend, followed by an increase. Conclusions: The ARIMA-NARNN model can be well applied to analyze surveillance data for early warning systems for the control and elimination of schistosomiasis. PMID:27023573

Validation of Model Forecasts of the Ambient Solar Wind

NASA Technical Reports Server (NTRS)

Macneice, P. J.; Hesse, M.; Kuznetsova, M. M.; Rastaetter, L.; Taktakishvili, A.

2009-01-01

Independent and automated validation is a vital step in the progression of models from the research community into operational forecasting use. In this paper we describe a program in development at the CCMC to provide just such a comprehensive validation for models of the ambient solar wind in the inner heliosphere. We have built upon previous efforts published in the community, sharpened their definitions, and completed a baseline study. We also provide first results from this program of the comparative performance of the MHD models available at the CCMC against that of the Wang-Sheeley-Arge (WSA) model. An important goal of this effort is to provide a consistent validation to all available models. Clearly exposing the relative strengths and weaknesses of the different models will enable forecasters to craft more reliable ensemble forecasting strategies. Models of the ambient solar wind are developing rapidly as a result of improvements in data supply, numerical techniques, and computing resources. It is anticipated that in the next five to ten years, the MHD based models will supplant semi-empirical potential based models such as the WSA model, as the best available forecast models. We anticipate that this validation effort will track this evolution and so assist policy makers in gauging the value of past and future investment in modeling support.

Modeling and forecasting of KLCI weekly return using WT-ANN integrated model

NASA Astrophysics Data System (ADS)

Liew, Wei-Thong; Liong, Choong-Yeun; Hussain, Saiful Izzuan; Isa, Zaidi

2013-04-01

The forecasting of weekly return is one of the most challenging tasks in investment since the time series are volatile and non-stationary. In this study, an integrated model of wavelet transform and artificial neural network, WT-ANN is studied for modeling and forecasting of KLCI weekly return. First, the WT is applied to decompose the weekly return time series in order to eliminate noise. Then, a mathematical model of the time series is constructed using the ANN. The performance of the suggested model will be evaluated by root mean squared error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE). The result shows that the WT-ANN model can be considered as a feasible and powerful model for time series modeling and prediction.

Error Estimation of An Ensemble Statistical Seasonal Precipitation Prediction Model

NASA Technical Reports Server (NTRS)

Shen, Samuel S. P.; Lau, William K. M.; Kim, Kyu-Myong; Li, Gui-Long

2001-01-01

This NASA Technical Memorandum describes an optimal ensemble canonical correlation forecasting model for seasonal precipitation. Each individual forecast is based on the canonical correlation analysis (CCA) in the spectral spaces whose bases are empirical orthogonal functions (EOF). The optimal weights in the ensemble forecasting crucially depend on the mean square error of each individual forecast. An estimate of the mean square error of a CCA prediction is made also using the spectral method. The error is decomposed onto EOFs of the predictand and decreases linearly according to the correlation between the predictor and predictand. Since new CCA scheme is derived for continuous fields of predictor and predictand, an area-factor is automatically included. Thus our model is an improvement of the spectral CCA scheme of Barnett and Preisendorfer. The improvements include (1) the use of area-factor, (2) the estimation of prediction error, and (3) the optimal ensemble of multiple forecasts. The new CCA model is applied to the seasonal forecasting of the United States (US) precipitation field. The predictor is the sea surface temperature (SST). The US Climate Prediction Center's reconstructed SST is used as the predictor's historical data. The US National Center for Environmental Prediction's optimally interpolated precipitation (1951-2000) is used as the predictand's historical data. Our forecast experiments show that the new ensemble canonical correlation scheme renders a reasonable forecasting skill. For example, when using September-October-November SST to predict the next season December-January-February precipitation, the spatial pattern correlation between the observed and predicted are positive in 46 years among the 50 years of experiments. The positive correlations are close to or greater than 0.4 in 29 years, which indicates excellent performance of the forecasting model. The forecasting skill can be further enhanced when several predictors are used.

Extending Data Worth Analyses to Select Multiple Observations Targeting Multiple Forecasts.

PubMed

Vilhelmsen, Troels N; Ferré, Ty P A

2018-05-01

Hydrological models are often set up to provide specific forecasts of interest. Owing to the inherent uncertainty in data used to derive model structure and used to constrain parameter variations, the model forecasts will be uncertain. Additional data collection is often performed to minimize this forecast uncertainty. Given our common financial restrictions, it is critical that we identify data with maximal information content with respect to forecast of interest. In practice, this often devolves to qualitative decisions based on expert opinion. However, there is no assurance that this will lead to optimal design, especially for complex hydrogeological problems. Specifically, these complexities include considerations of multiple forecasts, shared information among potential observations, information content of existing data, and the assumptions and simplifications underlying model construction. In the present study, we extend previous data worth analyses to include: simultaneous selection of multiple new measurements and consideration of multiple forecasts of interest. We show how the suggested approach can be used to optimize data collection. This can be used in a manner that suggests specific measurement sets or that produces probability maps indicating areas likely to be informative for specific forecasts. Moreover, we provide examples documenting that sequential measurement election approaches often lead to suboptimal designs and that estimates of data covariance should be included when selecting future measurement sets. © 2017, National Ground Water Association.

Application and evaluation of forecasting methods for municipal solid waste generation in an Eastern-European city.

PubMed

Rimaityte, Ingrida; Ruzgas, Tomas; Denafas, Gintaras; Racys, Viktoras; Martuzevicius, Dainius

2012-01-01

Forecasting of generation of municipal solid waste (MSW) in developing countries is often a challenging task due to the lack of data and selection of suitable forecasting method. This article aimed to select and evaluate several methods for MSW forecasting in a medium-scaled Eastern European city (Kaunas, Lithuania) with rapidly developing economics, with respect to affluence-related and seasonal impacts. The MSW generation was forecast with respect to the economic activity of the city (regression modelling) and using time series analysis. The modelling based on social-economic indicators (regression implemented in LCA-IWM model) showed particular sensitivity (deviation from actual data in the range from 2.2 to 20.6%) to external factors, such as the synergetic effects of affluence parameters or changes in MSW collection system. For the time series analysis, the combination of autoregressive integrated moving average (ARIMA) and seasonal exponential smoothing (SES) techniques were found to be the most accurate (mean absolute percentage error equalled to 6.5). Time series analysis method was very valuable for forecasting the weekly variation of waste generation data (r (2) > 0.87), but the forecast yearly increase should be verified against the data obtained by regression modelling. The methods and findings of this study may assist the experts, decision-makers and scientists performing forecasts of MSW generation, especially in developing countries.

Modeling and Forecasting Mortality With Economic Growth: A Multipopulation Approach.

PubMed

Boonen, Tim J; Li, Hong

2017-10-01

Research on mortality modeling of multiple populations focuses mainly on extrapolating past mortality trends and summarizing these trends by one or more common latent factors. This article proposes a multipopulation stochastic mortality model that uses the explanatory power of economic growth. In particular, we extend the Li and Lee model (Li and Lee 2005) by including economic growth, represented by the real gross domestic product (GDP) per capita, to capture the common mortality trend for a group of populations with similar socioeconomic conditions. We find that our proposed model provides a better in-sample fit and an out-of-sample forecast performance. Moreover, it generates lower (higher) forecasted period life expectancy for countries with high (low) GDP per capita than the Li and Lee model.

The Betting Odds Rating System: Using soccer forecasts to forecast soccer.

PubMed

Wunderlich, Fabian; Memmert, Daniel

2018-01-01

Betting odds are frequently found to outperform mathematical models in sports related forecasting tasks, however the factors contributing to betting odds are not fully traceable and in contrast to rating-based forecasts no straightforward measure of team-specific quality is deducible from the betting odds. The present study investigates the approach of combining the methods of mathematical models and the information included in betting odds. A soccer forecasting model based on the well-known ELO rating system and taking advantage of betting odds as a source of information is presented. Data from almost 15.000 soccer matches (seasons 2007/2008 until 2016/2017) are used, including both domestic matches (English Premier League, German Bundesliga, Spanish Primera Division and Italian Serie A) and international matches (UEFA Champions League, UEFA Europe League). The novel betting odds based ELO model is shown to outperform classic ELO models, thus demonstrating that betting odds prior to a match contain more relevant information than the result of the match itself. It is shown how the novel model can help to gain valuable insights into the quality of soccer teams and its development over time, thus having a practical benefit in performance analysis. Moreover, it is argued that network based approaches might help in further improving rating and forecasting methods.

The Betting Odds Rating System: Using soccer forecasts to forecast soccer

PubMed Central

Memmert, Daniel

2018-01-01

Betting odds are frequently found to outperform mathematical models in sports related forecasting tasks, however the factors contributing to betting odds are not fully traceable and in contrast to rating-based forecasts no straightforward measure of team-specific quality is deducible from the betting odds. The present study investigates the approach of combining the methods of mathematical models and the information included in betting odds. A soccer forecasting model based on the well-known ELO rating system and taking advantage of betting odds as a source of information is presented. Data from almost 15.000 soccer matches (seasons 2007/2008 until 2016/2017) are used, including both domestic matches (English Premier League, German Bundesliga, Spanish Primera Division and Italian Serie A) and international matches (UEFA Champions League, UEFA Europe League). The novel betting odds based ELO model is shown to outperform classic ELO models, thus demonstrating that betting odds prior to a match contain more relevant information than the result of the match itself. It is shown how the novel model can help to gain valuable insights into the quality of soccer teams and its development over time, thus having a practical benefit in performance analysis. Moreover, it is argued that network based approaches might help in further improving rating and forecasting methods. PMID:29870554

Assessment of marine weather forecasts over the Indian sector of Southern Ocean

NASA Astrophysics Data System (ADS)

Gera, Anitha; Mahapatra, D. K.; Sharma, Kuldeep; Prakash, Satya; Mitra, A. K.; Iyengar, G. R.; Rajagopal, E. N.; Anilkumar, N.

2017-09-01

The Southern Ocean (SO) is one of the important regions where significant processes and feedbacks of the Earth's climate take place. Expeditions to the SO provide useful data for improving global weather/climate simulations and understanding many processes. Some of the uncertainties in these weather/climate models arise during the first few days of simulation/forecast and do not grow much further. NCMRWF issued real-time five day weather forecasts of mean sea level pressure, surface winds, winds at 500 hPa & 850 hPa and rainfall, daily to NCAOR to provide guidance for their expedition to Indian sector of SO during the austral summer of 2014-2015. Evaluation of the skill of these forecasts indicates possible error growth in the atmospheric model at shorter time scales. The error growth is assessed using the model analysis/reanalysis, satellite data and observations made during the expedition. The observed variability of sub-seasonal rainfall associated with mid-latitude systems is seen to exhibit eastward propagations and are well reproduced in the model forecasts. All cyclonic disturbances including the sub-polar lows and tropical cyclones that occurred during this period were well captured in the model forecasts. Overall, this model performs reasonably well over the Indian sector of the SO in medium range time scale.

Modeling and forecasting US presidential election using learning algorithms

NASA Astrophysics Data System (ADS)

Zolghadr, Mohammad; Niaki, Seyed Armin Akhavan; Niaki, S. T. A.

2017-09-01

The primary objective of this research is to obtain an accurate forecasting model for the US presidential election. To identify a reliable model, artificial neural networks (ANN) and support vector regression (SVR) models are compared based on some specified performance measures. Moreover, six independent variables such as GDP, unemployment rate, the president's approval rate, and others are considered in a stepwise regression to identify significant variables. The president's approval rate is identified as the most significant variable, based on which eight other variables are identified and considered in the model development. Preprocessing methods are applied to prepare the data for the learning algorithms. The proposed procedure significantly increases the accuracy of the model by 50%. The learning algorithms (ANN and SVR) proved to be superior to linear regression based on each method's calculated performance measures. The SVR model is identified as the most accurate model among the other models as this model successfully predicted the outcome of the election in the last three elections (2004, 2008, and 2012). The proposed approach significantly increases the accuracy of the forecast.

Crash forecasting in the Korean stock market based on the log-periodic structure and pattern recognition

NASA Astrophysics Data System (ADS)

Ko, Bonggyun; Song, Jae Wook; Chang, Woojin

2018-02-01

The aim of this research is to propose an alarm index to forecast the crash of the Korean financial market in extension to the idea of Johansen-Ledoit-Sornette model, which uses the log-periodic functions and pattern recognition algorithm. We discover that the crashes of the Korean financial market can be classified into domestic and global crises where each category requires different window length of fitted datasets. Therefore, we add the window length as a new parameter to enhance the performance of alarm index. Distinguishing the domestic and global crises separately, our alarm index demonstrates more robust forecasting than previous model by showing the error diagram and the results of trading performance.

Initial conditions and ENSO prediction using a coupled ocean-atmosphere model

NASA Astrophysics Data System (ADS)

Larow, T. E.; Krishnamurti, T. N.

1998-01-01

A coupled ocean-atmosphere initialization scheme using Newtonian relaxation has been developed for the Florida State University coupled ocean-atmosphere global general circulation model. The initialization scheme is used to initialize the coupled model for seasonal forecasting the boreal summers of 1987 and 1988. The atmosphere model is a modified version of the Florida State University global spectral model, resolution T-42. The ocean general circulation model consists of a slightly modified version of the Hamburg's climate group model described in Latif (1987) and Latif et al. (1993). The coupling is synchronous with information exchanged every two model hours. Using ECMWF atmospheric daily analysis and observed monthly mean SSTs, two, 1-year, time-dependent, Newtonian relaxation were performed using the coupled model prior to conducting the seasonal forecasts. The coupled initializations were conducted from 1 June 1986 to 1 June 1987 and from 1 June 1987 to 1 June 1988. Newtonian relaxation was applied to the prognostic atmospheric vorticity, divergence, temperature and dew point depression equations. In the ocean model the relaxation was applied to the surface temperature. Two, 10-member ensemble integrations were conducted to examine the impact of the coupled initialization on the seasonal forecasts. The initial conditions used for the ensembles are the ocean's final state after the initialization and the atmospheric initial conditions are ECMWF analysis. Examination of the SST root mean square error and anomaly correlations between observed and forecasted SSTs in the Niño-3 and Niño-4 regions for the 2 seasonal forecasts, show closer agreement between the initialized forecast than two, 10-member non-initialized ensemble forecasts. The main conclusion here is that a single forecast with the coupled initialization outperforms, in SST anomaly prediction, against each of the control forecasts (members of the ensemble) which do not include such an initialization, indicating possible importance for the inclusion of the atmosphere during the coupled initialization.

Essays on oil price volatility and irreversible investment

NASA Astrophysics Data System (ADS)

Pastor, Daniel J.

In chapter 1, we provide an extensive and systematic evaluation of the relative forecasting performance of several models for the volatility of daily spot crude oil prices. Empirical research over the past decades has uncovered significant gains in forecasting performance of Markov Switching GARCH models over GARCH models for the volatility of financial assets and crude oil futures. We find that, for spot oil price returns, non-switching models perform better in the short run, whereas switching models tend to do better at longer horizons. In chapter 2, I investigate the impact of volatility on firms' irreversible investment decisions using real options theory. Cost incurred in oil drilling is considered sunk cost, thus irreversible. I collect detailed data on onshore, development oil well drilling on the North Slope of Alaska from 2003 to 2014. Volatility is modeled by constructing GARCH, EGARCH, and GJR-GARCH forecasts based on monthly real oil prices, and realized volatility from 5-minute intraday returns of oil futures prices. Using a duration model, I show that oil price volatility generally has a negative relationship with the hazard rate of drilling an oil well both when aggregating all the fields, and in individual fields.

Performance of the Prognocean Plus system during the El Niño 2015/2016: predictions of sea level anomalies as tools for forecasting El Niño

NASA Astrophysics Data System (ADS)

Świerczyńska-Chlaściak, Małgorzata; Niedzielski, Tomasz; Miziński, Bartłomiej

2017-04-01

The aim of this paper is to present the performance of the Prognocean Plus system, which produces long-term predictions of sea level anomalies, during the El Niño 2015/2016. The main objective of work is to identify such ocean areas in which long-term forecasts of sea level anomalies during El Niño 2015/2016 reveal a considerable accuracy. At present, the system produces prognoses using four data-based models and their combinations: polynomial-harmonic model, autoregressive model, threshold autoregressive model and multivariate autoregressive model. The system offers weekly forecasts, with lead time up to 12 weeks. Several statistics that describe the efficiency of the available prediction models in four seasons used for estimating Oceanic Niño index (ONI) are calculated. The accuracies/skills of the predicting models were computed in the specific locations in the equatorial Pacific, namely the geometrically-determined central points of all Niño regions. For the said locations, we focused on the forecasts which targeted at the local maximum of sea level, driven by the El Niño 2015/2016. As a result, a series of the "spaghetti" graphs (for each point, season and model) as well as plots presenting the prognostic performance of every model - for all lead times, seasons and locations - were created. It is found that the Prognocean Plus system has a potential to become a new solution which may enhance the diagnostic discussions on the El Niño development. The forecasts produced by the threshold autoregressive model, for lead times of 5-6 weeks and 9 weeks, within the Niño1+2 region for the November-to-January (NDJ) season anticipated the culmination of the El Niño 2015/2016. The longest forecasts (8-12 weeks) were found to be the most accurate in the phase of transition from El Niño to normal conditions (the multivariate autoregressive model, central point of Niño1+2 region, the December-to-February season). The study was conducted to verify the ability and usefulness of sea level anomaly forecasts in predicting phenomena that are controlled by the ocean-atmosphere processes, such as El Niño Southern Oscillation or North Atlantic Oscillation. The results may support further investigations into long-term forecasting of the quantitative indices of these oscillations, solely based on prognoses of sea level change. In particular, comparing the accuracies of prognoses of the North Atlantic Oscillation index remains one of the tasks of the research project no. 2016/21/N/ST10/03231, financed by the National Science Center of Poland.

The Impact of STTP on the GEFS Forecast of Week-2 and Beyond in the Presence of Stochastic Physics

NASA Astrophysics Data System (ADS)

Hou, D.

2015-12-01

The Stochastic Total Tendency Perturbation (STTP) scheme was designed to represent the model related uncertainties not considered in the numerical model itself and the physics based stochastic schemes. It has been applied in NCEP's Global Ensemble Forecast System (GEFS) since 2010, showing significant positive impacts on the forecast with improved spread-error ratio and probabilistic forecast skills. The scheme is robust and it went well with the resolution increases and model improvements in 2012 and 2015 with minimum changes. Recently, a set of stochastic physics schemes are coded in the Global Forecast System model and tested in the GEFS package. With these schemes turned on and STTP off, the forecast performance is comparable or even superior to the operational GEFS, in which STTP is the only contributor to the model related uncertainties. This is true especially in week one. However, over the second week and beyond, both the experimental and the operational GEFS has insufficient spread, especially over the warmer seasons. This is a major challenge when the GEFS is extended to sub-seasonal (week 4-6) time scales. The impact of STTP on the GEFS forecast in the presence of stochastic physics is investigated by turning both the stochastic physics schemes and STTP on and carefully tuning their amplitudes. Analysis will be focused on the forecast of extended range, especially week 2. Its impacts on week 3-4 will also be addressed.

A smoothed stochastic earthquake rate model considering seismicity and fault moment release for Europe

NASA Astrophysics Data System (ADS)

Hiemer, S.; Woessner, J.; Basili, R.; Danciu, L.; Giardini, D.; Wiemer, S.

2014-08-01

We present a time-independent gridded earthquake rate forecast for the European region including Turkey. The spatial component of our model is based on kernel density estimation techniques, which we applied to both past earthquake locations and fault moment release on mapped crustal faults and subduction zone interfaces with assigned slip rates. Our forecast relies on the assumption that the locations of past seismicity is a good guide to future seismicity, and that future large-magnitude events occur more likely in the vicinity of known faults. We show that the optimal weighted sum of the corresponding two spatial densities depends on the magnitude range considered. The kernel bandwidths and density weighting function are optimized using retrospective likelihood-based forecast experiments. We computed earthquake activity rates (a- and b-value) of the truncated Gutenberg-Richter distribution separately for crustal and subduction seismicity based on a maximum likelihood approach that considers the spatial and temporal completeness history of the catalogue. The final annual rate of our forecast is purely driven by the maximum likelihood fit of activity rates to the catalogue data, whereas its spatial component incorporates contributions from both earthquake and fault moment-rate densities. Our model constitutes one branch of the earthquake source model logic tree of the 2013 European seismic hazard model released by the EU-FP7 project `Seismic HAzard haRmonization in Europe' (SHARE) and contributes to the assessment of epistemic uncertainties in earthquake activity rates. We performed retrospective and pseudo-prospective likelihood consistency tests to underline the reliability of our model and SHARE's area source model (ASM) using the testing algorithms applied in the collaboratory for the study of earthquake predictability (CSEP). We comparatively tested our model's forecasting skill against the ASM and find a statistically significant better performance for testing periods of 10-20 yr. The testing results suggest that our model is a viable candidate model to serve for long-term forecasting on timescales of years to decades for the European region.

Exploring the calibration of a wind forecast ensemble for energy applications

NASA Astrophysics Data System (ADS)

Heppelmann, Tobias; Ben Bouallegue, Zied; Theis, Susanne

2015-04-01

In the German research project EWeLiNE, Deutscher Wetterdienst (DWD) and Fraunhofer Institute for Wind Energy and Energy System Technology (IWES) are collaborating with three German Transmission System Operators (TSO) in order to provide the TSOs with improved probabilistic power forecasts. Probabilistic power forecasts are derived from probabilistic weather forecasts, themselves derived from ensemble prediction systems (EPS). Since the considered raw ensemble wind forecasts suffer from underdispersiveness and bias, calibration methods are developed for the correction of the model bias and the ensemble spread bias. The overall aim is to improve the ensemble forecasts such that the uncertainty of the possible weather deployment is depicted by the ensemble spread from the first forecast hours. Additionally, the ensemble members after calibration should remain physically consistent scenarios. We focus on probabilistic hourly wind forecasts with horizon of 21 h delivered by the convection permitting high-resolution ensemble system COSMO-DE-EPS which has become operational in 2012 at DWD. The ensemble consists of 20 ensemble members driven by four different global models. The model area includes whole Germany and parts of Central Europe with a horizontal resolution of 2.8 km and a vertical resolution of 50 model levels. For verification we use wind mast measurements around 100 m height that corresponds to the hub height of wind energy plants that belong to wind farms within the model area. Calibration of the ensemble forecasts can be performed by different statistical methods applied to the raw ensemble output. Here, we explore local bivariate Ensemble Model Output Statistics at individual sites and quantile regression with different predictors. Applying different methods, we already show an improvement of ensemble wind forecasts from COSMO-DE-EPS for energy applications. In addition, an ensemble copula coupling approach transfers the time-dependencies of the raw ensemble to the calibrated ensemble. The calibrated wind forecasts are evaluated first with univariate probabilistic scores and additionally with diagnostics of wind ramps in order to assess the time-consistency of the calibrated ensemble members.

Perspectives on model forecasts of the 2014-2015 Ebola epidemic in West Africa: lessons and the way forward.

PubMed

Chowell, Gerardo; Viboud, Cécile; Simonsen, Lone; Merler, Stefano; Vespignani, Alessandro

2017-03-01

The unprecedented impact and modeling efforts associated with the 2014-2015 Ebola epidemic in West Africa provides a unique opportunity to document the performances and caveats of forecasting approaches used in near-real time for generating evidence and to guide policy. A number of international academic groups have developed and parameterized mathematical models of disease spread to forecast the trajectory of the outbreak. These modeling efforts often relied on limited epidemiological data to derive key transmission and severity parameters, which are needed to calibrate mechanistic models. Here, we provide a perspective on some of the challenges and lessons drawn from these efforts, focusing on (1) data availability and accuracy of early forecasts; (2) the ability of different models to capture the profile of early growth dynamics in local outbreaks and the importance of reactive behavior changes and case clustering; (3) challenges in forecasting the long-term epidemic impact very early in the outbreak; and (4) ways to move forward. We conclude that rapid availability of aggregated population-level data and detailed information on a subset of transmission chains is crucial to characterize transmission patterns, while ensemble-forecasting approaches could limit the uncertainty of any individual model. We believe that coordinated forecasting efforts, combined with rapid dissemination of disease predictions and underlying epidemiological data in shared online platforms, will be critical in optimizing the response to current and future infectious disease emergencies.

Chemical weather forecasting for the Yangtze River Delta

NASA Astrophysics Data System (ADS)

Xie, Y.; Xu, J.; Zhou, G.; Chang, L.; Chen, B.

2016-12-01

Shanghai is one of the largest megacities in the world. With rapid economic growth of the city and its surrounding areas in recent years, air pollution has posed adverse effects on public health and ecosystem. In winter heavy pollution episodes are often associated with PM exceedances under stagnant conditions or transport events, whereas in summer the region frequently experiences elevated O3 levels. Chemical weather prediction systems with the WRF-Chem and CMAQ models are being developed to support air quality and haze forecasting for Shanghai and the Yangtze River Delta region. We will present main components of the modeling system, forecasting products, as well as evaluation results. Evaluation of the WRF-Chem forecasts show the model has generally good ability to capture the temporal variations of O3 and PM2.5. Substantial regional differences exist, with the best performance in Shanghai. Meanwhile, the forecasts tend to degrade during highly polluted episodes and transitional time periods, which highlights the need to improve model representation of key process (e.g. meteorological fields and formation of secondary pollutants). Recent work includes using the ECMWF global model forecasts as chemical boundary conditions for our regional model. We investigate the impact of chemical downscaling, and also compare the results from different models participated in the PANDA (PArtnership with chiNa on space Data) project. Results from ongoing efforts (e.g. chemical weather forecasting driven by SMS regional high resolution NWP) will also be presented.

A domain analysis approach to clear-air turbulence forecasting using high-density in-situ measurements

NASA Astrophysics Data System (ADS)

Abernethy, Jennifer A.

Pilots' ability to avoid clear-air turbulence (CAT) during flight affects the safety of the millions of people who fly commercial airlines and other aircraft, and turbulence costs millions in injuries and aircraft maintenance every year. Forecasting CAT is not straightforward, however; microscale features like the turbulence eddies that affect aircraft (100m) are below the current resolution of operational numerical weather prediction (NWP) models, and the only evidence of CAT episodes, until recently, has been sparse, subjective reports from pilots known as PIREPs. To forecast CAT, researchers use a simple weighted sum of top-performing turbulence indicators derived from NWP model outputs---termed diagnostics---based on their agreement with current PIREPs. However, a new, quantitative source of observation data---high-density measurements made by sensor equipment and software on aircraft, called in-situ measurements---is now available. The main goal of this thesis is to develop new data analysis and processing techniques to apply to the model and new observation data, in order to improve CAT forecasting accuracy. This thesis shows that using in-situ data improves forecasting accuracy and that automated machine learning algorithms such as support vector machines (SVM), logistic regression, and random forests, can match current performance while eliminating almost all hand-tuning. Feature subset selection is paired with the new algorithms to choose diagnostics that predict well as a group rather than individually. Specializing forecasts and choice of diagnostics by geographic region further improves accuracy because of the geographic variation in turbulence sources. This work uses random forests to find climatologically-relevant regions based on these variations and implements a forecasting system testbed which brings these techniques together to rapidly prototype new, regionalized versions of operational CAT forecasting systems.

Validation of Seasonal Forecast of Indian Summer Monsoon Rainfall

NASA Astrophysics Data System (ADS)

Das, Sukanta Kumar; Deb, Sanjib Kumar; Kishtawal, C. M.; Pal, Pradip Kumar

2015-06-01

The experimental seasonal forecast of Indian summer monsoon (ISM) rainfall during June through September using Community Atmosphere Model (CAM) version 3 has been carried out at the Space Applications Centre Ahmedabad since 2009. The forecasts, based on a number of ensemble members (ten minimum) of CAM, are generated in several phases and updated on regular basis. On completion of 5 years of experimental seasonal forecasts in operational mode, it is required that the overall validation or correctness of the forecast system is quantified and that the scope is assessed for further improvements of the forecast over time, if any. The ensemble model climatology generated by a set of 20 identical CAM simulations is considered as the model control simulation. The performance of the forecast has been evaluated by assuming the control simulation as the model reference. The forecast improvement factor shows positive improvements, with higher values for the recent forecasted years as compared to the control experiment over the Indian landmass. The Taylor diagram representation of the Pearson correlation coefficient (PCC), standard deviation and centered root mean square difference has been used to demonstrate the best PCC, in the order of 0.74-0.79, recorded for the seasonal forecast made during 2013. Further, the bias score of different phases of experiment revealed the fact that the ISM rainfall forecast is affected by overestimation in predicting the low rain-rate (less than 7 mm/day), but by underestimation in the medium and high rain-rate (higher than 11 mm/day). Overall, the analysis shows significant improvement of the ISM forecast over the last 5 years, viz. 2009-2013, due to several important modifications that have been implemented in the forecast system. The validation exercise has also pointed out a number of shortcomings in the forecast system; these will be addressed in the upcoming years of experiments to improve the quality of the ISM prediction.

Time Series Forecasting of Daily Reference Evapotranspiration by Neural Network Ensemble Learning for Irrigation System

NASA Astrophysics Data System (ADS)

Manikumari, N.; Murugappan, A.; Vinodhini, G.

2017-07-01

Time series forecasting has gained remarkable interest of researchers in the last few decades. Neural networks based time series forecasting have been employed in various application areas. Reference Evapotranspiration (ETO) is one of the most important components of the hydrologic cycle and its precise assessment is vital in water balance and crop yield estimation, water resources system design and management. This work aimed at achieving accurate time series forecast of ETO using a combination of neural network approaches. This work was carried out using data collected in the command area of VEERANAM Tank during the period 2004 - 2014 in India. In this work, the Neural Network (NN) models were combined by ensemble learning in order to improve the accuracy for forecasting Daily ETO (for the year 2015). Bagged Neural Network (Bagged-NN) and Boosted Neural Network (Boosted-NN) ensemble learning were employed. It has been proved that Bagged-NN and Boosted-NN ensemble models are better than individual NN models in terms of accuracy. Among the ensemble models, Boosted-NN reduces the forecasting errors compared to Bagged-NN and individual NNs. Regression co-efficient, Mean Absolute Deviation, Mean Absolute Percentage error and Root Mean Square Error also ascertain that Boosted-NN lead to improved ETO forecasting performance.

Extreme precipitation forecasting in the Chilean Andean region with complex topography using the Weather Research and Forecasting (WRF) model

NASA Astrophysics Data System (ADS)

Gironás, J.; Yáñez Morroni, G.; Caneo, M.; Delgado, R.

2017-12-01

The Weather Research and Forecasting (WRF) model is broadly used for weather forecasting, hindcasting and researching due to its good performance. However, the atmospheric conditions for simulating are not always optimal when it includes complex topographies: affecting WRF mathematical stability and convergence, therefore, its performance. As Chile is a country strongly characterized by a complex topography and high gradients of elevation, WRF is ineffective resolving Chilean mountainous terrain and foothills. The need to own an effective weather forecasting tool relies on that Chile's main cities are located in these regions. Furthermore, the most intense rainfall events take place here, commonly caused by the presence of cutoff lows. This work analyzes a microphysics scheme ensemble to enhance initial forecasts made by the Chilean Weather Agency (DMC). These forecasts were made over the Santiago piedmont, in Quebrada de Ramón watershed, located upstream an urban area highly populated. In this region a non-existing planning increases the potential damage of a flash flood. An initial testing was made over different vertical levels resolution (39 and 50 levels), and subsequently testing with land use and surface models, and finally with the initial and boundary condition data (GFS/FNL). Our task made emphasis in analyzing microphysics and lead time (3 to 5 days before the storm peak) in the computational simulations over three extreme rainfall events between 2015 and 2017. WRF shortcoming are also related to the complex configuration of the synoptic events, even when the steep topography difficult the rainfall event peak amount, and to a lesser degree, the exact rainfall event beginning prediction. No evident trend was found in the lead time, but as expected, better results in rainfall and zero isotherm height are obtained with smaller anticipation. We found that WRF do predict properly the N-hours with the biggest amount of rainfall (5 hours corresponding to Quebrada de Ramón's time of concentration) and the temperatures during the event. This is a fundamental input to a hydrological model that could forecast flash floods. Finally, WSM-6Class microphysics was chosen as the one with best performance, but a geostatistical approach to countervail WRF forecasts' shortcomings over Andean piedmont is required.

Input selection and performance optimization of ANN-based streamflow forecasts in the drought-prone Murray Darling Basin region using IIS and MODWT algorithm

NASA Astrophysics Data System (ADS)

Prasad, Ramendra; Deo, Ravinesh C.; Li, Yan; Maraseni, Tek

2017-11-01

Forecasting streamflow is vital for strategically planning, utilizing and redistributing water resources. In this paper, a wavelet-hybrid artificial neural network (ANN) model integrated with iterative input selection (IIS) algorithm (IIS-W-ANN) is evaluated for its statistical preciseness in forecasting monthly streamflow, and it is then benchmarked against M5 Tree model. To develop hybrid IIS-W-ANN model, a global predictor matrix is constructed for three local hydrological sites (Richmond, Gwydir, and Darling River) in Australia's agricultural (Murray-Darling) Basin. Model inputs comprised of statistically significant lagged combination of streamflow water level, are supplemented by meteorological data (i.e., precipitation, maximum and minimum temperature, mean solar radiation, vapor pressure and evaporation) as the potential model inputs. To establish robust forecasting models, iterative input selection (IIS) algorithm is applied to screen the best data from the predictor matrix and is integrated with the non-decimated maximum overlap discrete wavelet transform (MODWT) applied on the IIS-selected variables. This resolved the frequencies contained in predictor data while constructing a wavelet-hybrid (i.e., IIS-W-ANN and IIS-W-M5 Tree) model. Forecasting ability of IIS-W-ANN is evaluated via correlation coefficient (r), Willmott's Index (WI), Nash-Sutcliffe Efficiency (ENS), root-mean-square-error (RMSE), and mean absolute error (MAE), including the percentage RMSE and MAE. While ANN models are seen to outperform M5 Tree executed for all hydrological sites, the IIS variable selector was efficient in determining the appropriate predictors, as stipulated by the better performance of the IIS coupled (ANN and M5 Tree) models relative to the models without IIS. When IIS-coupled models are integrated with MODWT, the wavelet-hybrid IIS-W-ANN and IIS-W-M5 Tree are seen to attain significantly accurate performance relative to their standalone counterparts. Importantly, IIS-W-ANN model accuracy outweighs IIS-ANN, as evidenced by a larger r and WI (by 7.5% and 3.8%, respectively) and a lower RMSE (by 21.3%). In comparison to the IIS-W-M5 Tree model, IIS-W-ANN model yielded larger values of WI = 0.936-0.979 and ENS = 0.770-0.920. Correspondingly, the errors (RMSE and MAE) ranged from 0.162-0.487 m and 0.139-0.390 m, respectively, with relative errors, RRMSE = (15.65-21.00) % and MAPE = (14.79-20.78) %. Distinct geographic signature is evident where the most and least accurately forecasted streamflow data is attained for the Gwydir and Darling River, respectively. Conclusively, this study advocates the efficacy of iterative input selection, allowing the proper screening of model predictors, and subsequently, its integration with MODWT resulting in enhanced performance of the models applied in streamflow forecasting.

A new short-term forecasting model for the total electron content storm time disturbances

NASA Astrophysics Data System (ADS)

Tsagouri, Ioanna; Koutroumbas, Konstantinos; Elias, Panagiotis

2018-06-01

This paper aims to introduce a new model for the short-term forecast of the vertical Total Electron Content (vTEC). The basic idea of the proposed model lies on the concept of the Solar Wind driven autoregressive model for Ionospheric short-term Forecast (SWIF). In its original version, the model is operationally implemented in the DIAS system (http://dias.space.noa.gr) and provides alerts and warnings for upcoming ionospheric disturbances, as well as single site and regional forecasts of the foF2 critical frequency over Europe up to 24 h in advance. The forecasts are driven by the real time assessment of the solar wind conditions at ACE location. The comparative analysis of the variations in foF2 and vTEC during eleven geomagnetic storm events that occurred in the present solar cycle 24 reveals similarities but also differences in the storm-time response of the two characteristics with respect to the local time and the latitude of the observation point. Since the aforementioned dependences drive the storm-time forecasts of the SWIF model, the results obtained here support the upgrade of the SWIF's modeling technique in forecasting the storm-time vTEC variation from its onset to full development and recovery. According to the proposed approach, the vTEC storm-time response can be forecasted from 1 to 12-13 h before its onset, depending on the local time of the observation point at storm onset at L1. Preliminary results on the assessment of the performance of the proposed model and further considerations on its potential implementation in operational mode are also discussed.

A Four-Stage Hybrid Model for Hydrological Time Series Forecasting

PubMed Central

Di, Chongli; Yang, Xiaohua; Wang, Xiaochao

2014-01-01

Hydrological time series forecasting remains a difficult task due to its complicated nonlinear, non-stationary and multi-scale characteristics. To solve this difficulty and improve the prediction accuracy, a novel four-stage hybrid model is proposed for hydrological time series forecasting based on the principle of ‘denoising, decomposition and ensemble’. The proposed model has four stages, i.e., denoising, decomposition, components prediction and ensemble. In the denoising stage, the empirical mode decomposition (EMD) method is utilized to reduce the noises in the hydrological time series. Then, an improved method of EMD, the ensemble empirical mode decomposition (EEMD), is applied to decompose the denoised series into a number of intrinsic mode function (IMF) components and one residual component. Next, the radial basis function neural network (RBFNN) is adopted to predict the trend of all of the components obtained in the decomposition stage. In the final ensemble prediction stage, the forecasting results of all of the IMF and residual components obtained in the third stage are combined to generate the final prediction results, using a linear neural network (LNN) model. For illustration and verification, six hydrological cases with different characteristics are used to test the effectiveness of the proposed model. The proposed hybrid model performs better than conventional single models, the hybrid models without denoising or decomposition and the hybrid models based on other methods, such as the wavelet analysis (WA)-based hybrid models. In addition, the denoising and decomposition strategies decrease the complexity of the series and reduce the difficulties of the forecasting. With its effective denoising and accurate decomposition ability, high prediction precision and wide applicability, the new model is very promising for complex time series forecasting. This new forecast model is an extension of nonlinear prediction models. PMID:25111782

A four-stage hybrid model for hydrological time series forecasting.

PubMed

Di, Chongli; Yang, Xiaohua; Wang, Xiaochao

2014-01-01

Hydrological time series forecasting remains a difficult task due to its complicated nonlinear, non-stationary and multi-scale characteristics. To solve this difficulty and improve the prediction accuracy, a novel four-stage hybrid model is proposed for hydrological time series forecasting based on the principle of 'denoising, decomposition and ensemble'. The proposed model has four stages, i.e., denoising, decomposition, components prediction and ensemble. In the denoising stage, the empirical mode decomposition (EMD) method is utilized to reduce the noises in the hydrological time series. Then, an improved method of EMD, the ensemble empirical mode decomposition (EEMD), is applied to decompose the denoised series into a number of intrinsic mode function (IMF) components and one residual component. Next, the radial basis function neural network (RBFNN) is adopted to predict the trend of all of the components obtained in the decomposition stage. In the final ensemble prediction stage, the forecasting results of all of the IMF and residual components obtained in the third stage are combined to generate the final prediction results, using a linear neural network (LNN) model. For illustration and verification, six hydrological cases with different characteristics are used to test the effectiveness of the proposed model. The proposed hybrid model performs better than conventional single models, the hybrid models without denoising or decomposition and the hybrid models based on other methods, such as the wavelet analysis (WA)-based hybrid models. In addition, the denoising and decomposition strategies decrease the complexity of the series and reduce the difficulties of the forecasting. With its effective denoising and accurate decomposition ability, high prediction precision and wide applicability, the new model is very promising for complex time series forecasting. This new forecast model is an extension of nonlinear prediction models.

Hospital daily outpatient visits forecasting using a combinatorial model based on ARIMA and SES models.

PubMed

Luo, Li; Luo, Le; Zhang, Xinli; He, Xiaoli

2017-07-10

Accurate forecasting of hospital outpatient visits is beneficial for the reasonable planning and allocation of healthcare resource to meet the medical demands. In terms of the multiple attributes of daily outpatient visits, such as randomness, cyclicity and trend, time series methods, ARIMA, can be a good choice for outpatient visits forecasting. On the other hand, the hospital outpatient visits are also affected by the doctors' scheduling and the effects are not pure random. Thinking about the impure specialty, this paper presents a new forecasting model that takes cyclicity and the day of the week effect into consideration. We formulate a seasonal ARIMA (SARIMA) model on a daily time series and then a single exponential smoothing (SES) model on the day of the week time series, and finally establish a combinatorial model by modifying them. The models are applied to 1 year of daily visits data of urban outpatients in two internal medicine departments of a large hospital in Chengdu, for forecasting the daily outpatient visits about 1 week ahead. The proposed model is applied to forecast the cross-sectional data for 7 consecutive days of daily outpatient visits over an 8-weeks period based on 43 weeks of observation data during 1 year. The results show that the two single traditional models and the combinatorial model are simplicity of implementation and low computational intensiveness, whilst being appropriate for short-term forecast horizons. Furthermore, the combinatorial model can capture the comprehensive features of the time series data better. Combinatorial model can achieve better prediction performance than the single model, with lower residuals variance and small mean of residual errors which needs to be optimized deeply on the next research step.

Multi-RCM ensemble downscaling of global seasonal forecasts (MRED)

NASA Astrophysics Data System (ADS)

Arritt, R.

2009-04-01

Regional climate models (RCMs) have long been used to downscale global climate simulations. In contrast the ability of RCMs to downscale seasonal climate forecasts has received little attention. The Multi-RCM Ensemble Downscaling (MRED) project was recently initiated to address the question, Does dynamical downscaling using RCMs provide additional useful information for seasonal forecasts made by global models? MRED is using a suite of RCMs to downscale seasonal forecasts produced by the National Centers for Environmental Prediction (NCEP) Climate Forecast System (CFS) seasonal forecast system and the NASA GEOS5 system. The initial focus is on wintertime forecasts in order to evaluate topographic forcing, snowmelt, and the usefulness of higher resolution for near-surface fields influenced by high resolution orography. Each RCM covers the conterminous U.S. at approximately 32 km resolution, comparable to the scale of the North American Regional Reanalysis (NARR) which will be used to evaluate the models. The forecast ensemble for each RCM is comprised of 15 members over a period of 22+ years (from 1982 to 2003+) for the forecast period 1 December - 30 April. Each RCM will create a 15-member lagged ensemble by starting on different dates in the preceding November. This results in a 120-member ensemble for each projection (8 RCMs by 15 members per RCM). The RCMs will be continually updated at their lateral boundaries using 6-hourly output from CFS or GEOS5. Hydrometeorological output will be produced in a standard netCDF-based format for a common analysis grid, which simplifies both model intercomparison and the generation of ensembles. MRED will compare individual RCM and global forecasts as well as ensemble mean precipitation and temperature forecasts, which are currently being used to drive macroscale land surface models (LSMs). Metrics of ensemble spread will also be evaluated. Extensive process-oriented analysis will be performed to link improvements in downscaled forecast skill to regional forcings and physical mechanisms. Our overarching goal is to determine what additional skill can be provided by a community ensemble of high resolution regional models, which we believe will define a strategy for more skillful and useful regional seasonal climate forecasts.

A hybrid spatiotemporal drought forecasting model for operational use

NASA Astrophysics Data System (ADS)

Vasiliades, L.; Loukas, A.

2010-09-01

Drought forecasting plays an important role in the planning and management of natural resources and water resource systems in a river basin. Early and timelines forecasting of a drought event can help to take proactive measures and set out drought mitigation strategies to alleviate the impacts of drought. Spatiotemporal data mining is the extraction of unknown and implicit knowledge, structures, spatiotemporal relationships, or patterns not explicitly stored in spatiotemporal databases. As one of data mining techniques, forecasting is widely used to predict the unknown future based upon the patterns hidden in the current and past data. This study develops a hybrid spatiotemporal scheme for integrated spatial and temporal forecasting. Temporal forecasting is achieved using feed-forward neural networks and the temporal forecasts are extended to the spatial dimension using a spatial recurrent neural network model. The methodology is demonstrated for an operational meteorological drought index the Standardized Precipitation Index (SPI) calculated at multiple timescales. 48 precipitation stations and 18 independent precipitation stations, located at Pinios river basin in Thessaly region, Greece, were used for the development and spatiotemporal validation of the hybrid spatiotemporal scheme. Several quantitative temporal and spatial statistical indices were considered for the performance evaluation of the models. Furthermore, qualitative statistical criteria based on contingency tables between observed and forecasted drought episodes were calculated. The results show that the lead time of forecasting for operational use depends on the SPI timescale. The hybrid spatiotemporal drought forecasting model could be operationally used for forecasting up to three months ahead for SPI short timescales (e.g. 3-6 months) up to six months ahead for large SPI timescales (e.g. 24 months). The above findings could be useful in developing a drought preparedness plan in the region.

Assimilation of water temperature and discharge data for ensemble water temperature forecasting

NASA Astrophysics Data System (ADS)

Ouellet-Proulx, Sébastien; Chimi Chiadjeu, Olivier; Boucher, Marie-Amélie; St-Hilaire, André

2017-11-01

Recent work demonstrated the value of water temperature forecasts to improve water resources allocation and highlighted the importance of quantifying their uncertainty adequately. In this study, we perform a multisite cascading ensemble assimilation of discharge and water temperature on the Nechako River (Canada) using particle filters. Hydrological and thermal initial conditions were provided to a rainfall-runoff model, coupled to a thermal module, using ensemble meteorological forecasts as inputs to produce 5 day ensemble thermal forecasts. Results show good performances of the particle filters with improvements of the accuracy of initial conditions by more than 65% compared to simulations without data assimilation for both the hydrological and the thermal component. All thermal forecasts returned continuous ranked probability scores under 0.8 °C when using a set of 40 initial conditions and meteorological forecasts comprising 20 members. A greater contribution of the initial conditions to the total uncertainty of the system for 1-dayforecasts is observed (mean ensemble spread = 1.1 °C) compared to meteorological forcings (mean ensemble spread = 0.6 °C). The inclusion of meteorological uncertainty is critical to maintain reliable forecasts and proper ensemble spread for lead times of 2 days and more. This work demonstrates the ability of the particle filters to properly update the initial conditions of a coupled hydrological and thermal model and offers insights regarding the contribution of two major sources of uncertainty to the overall uncertainty in thermal forecasts.

The Impact of Lightning on Hurricane Rapid Intensification Forecasts Using the HWRF Model

NASA Astrophysics Data System (ADS)

Rosado, K.; Tallapragada, V.; Jenkins, G. S.

2016-12-01

In 2010, the National Oceanic and Atmospheric Administration (NOAA) created the Hurricane Forecast Improvement Project (HFIP) with the main goal of improving the tropical cyclone intensity and track forecasts by 50% in ten years. One of the focus areas is the improvement of the tropical cyclone rapid intensification (RI) forecasts. In order to contribute to this task, the role of lightning during the life cycle of a tropical cyclone using the NCEP operational HWRF hurricane model has been investigated. We ask two key research questions: (1) What is the functional relationship between atmospheric moisture content, lightning, and intensity in the HWRF model? and (2) How well does the HWRF model forecast the spatial distributions of lightning before, during, and after tropical cyclone intensification, especially for RI events? In order to address those questions, a lightning parameterization scheme called the Lightning Potential Index (LPI) was implemented into the HWRF model. The selected study cases to test the LPI implementation on the 2015 HWRF (operational version) are: Earl and Joaquin (North Atlantic), Haiyan (Western North Pacific), and Patricia (Eastern North Pacific). Five-day forecasts was executed on each case study with emphasis on rapid intensification periods. An extensive analysis between observed "best track" intensity, model intensity forecast, and potential for lightning forecast was performed. Preliminary results show that: (1) strong correlation between lightning and intensity changes does exists; and (2) the potential for lightning increases to its maximum peak a few hours prior to the peak intensity of the tropical cyclone. LPI peak values could potentially serve as indicator for future rapid intensification periods. Results from this investigation are giving us a better understanding of the mechanism behind lightning as a proxy for tropical cyclone steady state intensification and tropical cyclone rapid intensification processes. Improvement of lightning forecast has the potential to improve HWRF hurricane model intensity forecasts.

Global Positioning System (GPS) Precipitable Water in Forecasting Lightning at Spaceport Canaveral

NASA Technical Reports Server (NTRS)

Kehrer, Kristen C.; Graf, Brian; Roeder, William

2006-01-01

This paper evaluates the use of precipitable water (PW) from Global Positioning System (GPS) in lightning prediction. Additional independent verification of an earlier model is performed. This earlier model used binary logistic regression with the following four predictor variables optimally selected from a candidate list of 23 candidate predictors: the current precipitable water value for a given time of the day, the change in GPS-PW over the past 9 hours, the KIndex, and the electric field mill value. This earlier model was not optimized for any specific forecast interval, but showed promise for 6 hour and 1.5 hour forecasts. Two new models were developed and verified. These new models were optimized for two operationally significant forecast intervals. The first model was optimized for the 0.5 hour lightning advisories issued by the 45th Weather Squadron. An additional 1.5 hours was allowed for sensor dwell, communication, calculation, analysis, and advisory decision by the forecaster. Therefore the 0.5 hour advisory model became a 2 hour forecast model for lightning within the 45th Weather Squadron advisory areas. The second model was optimized for major ground processing operations supported by the 45th Weather Squadron, which can require lightning forecasts with a lead-time of up to 7.5 hours. Using the same 1.5 lag as in the other new model, this became a 9 hour forecast model for lightning within 37 km (20 NM)) of the 45th Weather Squadron advisory areas. The two new models were built using binary logistic regression from a list of 26 candidate predictor variables: the current GPS-PW value, the change of GPS-PW over 0.5 hour increments from 0.5 to 12 hours, and the K-index. The new 2 hour model found the following for predictors to be statistically significant, listed in decreasing order of contribution to the forecast: the 0.5 hour change in GPS-PW, the 7.5 hour change in GPS-PW, the current GPS-PW value, and the KIndex. The new 9 hour forecast model found the following five independent variables to be statistically significant, listed in decreasing order of contribution to the forecast: the current GPSPW value, the 8.5 hour change in GPS-PW, the 3.5 hour change in GPS-PW, the 12 hour change in GPS-PW, and the K-Index. In both models, the GPS-PW parameters had better correlation to the lightning forecast than the K-Index, a widely used thunderstorm index. Possible future improvements to this study are discussed.

A New Statistical Model for Eruption Forecasting at Open Conduit Volcanoes: an Application to Mt Etna and Kilauea Volcanoes

NASA Astrophysics Data System (ADS)

Passarelli, Luigi; Sanso, Bruno; Laura, Sandri; Marzocchi, Warner

2010-05-01

One of the main goals in volcanology is to forecast volcanic eruptions. A trenchant forecast should be made before the onset of a volcanic eruption, using the data available at that time, with the aim of mitigating the volcanic risk associated to the volcanic event. In other words, models implemented with forecast purposes have to take into account the possibility to provide "forward" forecasts and should avoid the idea of a merely "retrospective" fitting of the data available. In this perspective, the main idea of the present model is to forecast the next volcanic eruption after the end of the last one, using only the data available at that time. We focus our attention on volcanoes with open conduit regime and high eruption frequency. We assume a generalization of the classical time predictable model to describe the eruptive behavior of open conduit volcanoes and we use a Bayesian hierarchical model to make probabilistic forecast. We apply the model to Kilauea volcano eruptive data and Mt. Etna volcano flank eruption data. The aims of this model are: 1) to test whether or not the Kilauea and Mt Etna volcanoes follow a time predictable behavior; 2) to discuss the volcanological implications of the time predictable model parameters inferred; 3) to compare the forecast capabilities of this model with other models present in literature. The results obtained using the MCMC sampling algorithm show that both volcanoes follow a time predictable behavior. The numerical values of the time predictable model parameters inferred suggest that the amount of the erupted volume could change the dynamics of the magma chamber refilling process during the repose period. The probability gain of this model compared with other models already present in literature is appreciably greater than zero. This means that our model performs better forecast than previous models and it could be used in a probabilistic volcanic hazard assessment scheme. In this perspective, the probability of eruptions given by our model for Mt Etna volcano flank eruption are published on a internet website and are updated after any change in the eruptive activity.

Improving flood forecasting capability of physically based distributed hydrological model by parameter optimization

NASA Astrophysics Data System (ADS)

Chen, Y.; Li, J.; Xu, H.

2015-10-01

Physically based distributed hydrological models discrete the terrain of the whole catchment into a number of grid cells at fine resolution, and assimilate different terrain data and precipitation to different cells, and are regarded to have the potential to improve the catchment hydrological processes simulation and prediction capability. In the early stage, physically based distributed hydrological models are assumed to derive model parameters from the terrain properties directly, so there is no need to calibrate model parameters, but unfortunately, the uncertanties associated with this model parameter deriving is very high, which impacted their application in flood forecasting, so parameter optimization may also be necessary. There are two main purposes for this study, the first is to propose a parameter optimization method for physically based distributed hydrological models in catchment flood forecasting by using PSO algorithm and to test its competence and to improve its performances, the second is to explore the possibility of improving physically based distributed hydrological models capability in cathcment flood forecasting by parameter optimization. In this paper, based on the scalar concept, a general framework for parameter optimization of the PBDHMs for catchment flood forecasting is first proposed that could be used for all PBDHMs. Then, with Liuxihe model as the study model, which is a physically based distributed hydrological model proposed for catchment flood forecasting, the improverd Particle Swarm Optimization (PSO) algorithm is developed for the parameter optimization of Liuxihe model in catchment flood forecasting, the improvements include to adopt the linear decreasing inertia weight strategy to change the inertia weight, and the arccosine function strategy to adjust the acceleration coefficients. This method has been tested in two catchments in southern China with different sizes, and the results show that the improved PSO algorithm could be used for Liuxihe model parameter optimization effectively, and could improve the model capability largely in catchment flood forecasting, thus proven that parameter optimization is necessary to improve the flood forecasting capability of physically based distributed hydrological model. It also has been found that the appropriate particle number and the maximum evolution number of PSO algorithm used for Liuxihe model catchment flood forcasting is 20 and 30, respectively.

Evaluating the extreme precipitation events using a mesoscale atmopshere model

NASA Astrophysics Data System (ADS)

Yucel, I.; Onen, A.

2012-04-01

Evidence is showing that global warming or climate change has a direct influence on changes in precipitation and the hydrological cycle. Extreme weather events such as heavy rainfall and flooding are projected to become much more frequent as climate warms. Mesoscale atmospheric models coupled with land surface models provide efficient forecasts for meteorological events in high lead time and therefore they should be used for flood forecasting and warning issues as they provide more continuous monitoring of precipitation over large areas. This study examines the performance of the Weather Research and Forecasting (WRF) model in producing the temporal and spatial characteristics of the number of extreme precipitation events observed in West Black Sea Region of Turkey. Extreme precipitation events usually resulted in flood conditions as an associated hydrologic response of the basin. The performance of the WRF system is further investigated by using the three dimensional variational (3D-VAR) data assimilation scheme within WRF. WRF performance with and without data assimilation at high spatial resolution (4 km) is evaluated by making comparison with gauge precipitation and satellite-estimated rainfall data from Multi Precipitation Estimates (MPE). WRF-derived precipitation showed capabilities in capturing the timing of the precipitation extremes and in some extent spatial distribution and magnitude of the heavy rainfall events. These precipitation characteristics are enhanced with the use of 3D-VAR scheme in WRF system. Data assimilation improved area-averaged precipitation forecasts by 9 percent and at some points there exists quantitative match in precipitation events, which are critical for hydrologic forecast application.

Coupling fast all-season soil strength land surface model with weather research and forecasting model to assess low-level icing in complex terrain

NASA Astrophysics Data System (ADS)

Sines, Taleena R.

Icing poses as a severe hazard to aircraft safety with financial resources and even human lives hanging in the balance when the decision to ground a flight must be made. When analyzing the effects of ice on aviation, a chief cause for danger is the disruption of smooth airflow, which increases the drag force on the aircraft therefore decreasing its ability to create lift. The Weather Research and Forecast (WRF) model Advanced Research WRF (WRF-ARW) is a collaboratively created, flexible model designed to run on distributed computing systems for a variety of applications including forecasting research, parameterization research, and real-time numerical weather prediction. Land-surface models, one of the physics options available in the WRF-ARW, output surface heat and moisture flux given radiation, precipitation, and surface properties such as soil type. The Fast All-Season Soil STrength (FASST) land-surface model was developed by the U.S. Army ERDC-CRREL in Hanover, New Hampshire. Designed to use both meteorological and terrain data, the model calculates heat and moisture within the surface layer as well as the exchange of these parameters between the soil, surface elements (such as snow and vegetation), and atmosphere. Focusing on the Presidential Mountain Range of New Hampshire under the NASA Experimental Program to Stimulate Competitive Research (EPSCoR) Icing Assessments in Cold and Alpine Environments project, one of the main goals is to create a customized, high resolution model to predict and assess ice accretion in complex terrain. The purpose of this research is to couple the FASST land-surface model with the WRF to improve icing forecasts in complex terrain. Coupling FASST with the WRF-ARW may improve icing forecasts because of its sophisticated approach to handling processes such as meltwater, freezing, thawing, and others that would affect the water and energy budget and in turn affect icing forecasts. Several transformations had to take place in order for the FASST land-surface model and WRF-ARW to work together as fully coupled models. Changes had to be made to the WRF-ARW build mechanisms (Chapter 1, section a) so that FASST would be recognized as a new option that could be chosen through the namelist and compiled with other modules. Similarly, FASST had to be altered to no longer read meteorological data from a file, but accept input from WRF-ARW at each time step in a way that did not alter the integrity or run-time processes of the model. Several icing events were available to test the newly coupled model as well as the performance of other available land-surface models from the WRF-ARW. A variation of event intensities and durations from these events were chosen to give a broader view of the land-surface models' abilities to accurately predict icing in complex terrain. Non- icing events were also used in testing to ensure the land-surface models were not predicting ice in the events where none occurred. When compared to the other land-surface models and observations FASST showed a warm bias in several regions. As the forecasts progressed, FASST appeared to attempt to correct this bias and performed similarly to the other land-surface models and at times better than these land-surface models in areas of the domain not affected by this bias. To correct this warm bias, future investigation should be conducted into the reasoning behind this warm bias, including but not limited to: FASST operation and elevation modeling, WRF-ARW variables and forecasting methods, as well as allowing for spin-up prior to forecast times. Following the correction to the warm bias, FASST can be parallelized to allow for operational forecast performance and included in the WRF-ARW forecasting suite for future software releases. (Abstract shortened by UMI.).

A Dirichlet process model for classifying and forecasting epidemic curves.

PubMed

Nsoesie, Elaine O; Leman, Scotland C; Marathe, Madhav V

2014-01-09

A forecast can be defined as an endeavor to quantitatively estimate a future event or probabilities assigned to a future occurrence. Forecasting stochastic processes such as epidemics is challenging since there are several biological, behavioral, and environmental factors that influence the number of cases observed at each point during an epidemic. However, accurate forecasts of epidemics would impact timely and effective implementation of public health interventions. In this study, we introduce a Dirichlet process (DP) model for classifying and forecasting influenza epidemic curves. The DP model is a nonparametric Bayesian approach that enables the matching of current influenza activity to simulated and historical patterns, identifies epidemic curves different from those observed in the past and enables prediction of the expected epidemic peak time. The method was validated using simulated influenza epidemics from an individual-based model and the accuracy was compared to that of the tree-based classification technique, Random Forest (RF), which has been shown to achieve high accuracy in the early prediction of epidemic curves using a classification approach. We also applied the method to forecasting influenza outbreaks in the United States from 1997-2013 using influenza-like illness (ILI) data from the Centers for Disease Control and Prevention (CDC). We made the following observations. First, the DP model performed as well as RF in identifying several of the simulated epidemics. Second, the DP model correctly forecasted the peak time several days in advance for most of the simulated epidemics. Third, the accuracy of identifying epidemics different from those already observed improved with additional data, as expected. Fourth, both methods correctly classified epidemics with higher reproduction numbers (R) with a higher accuracy compared to epidemics with lower R values. Lastly, in the classification of seasonal influenza epidemics based on ILI data from the CDC, the methods' performance was comparable. Although RF requires less computational time compared to the DP model, the algorithm is fully supervised implying that epidemic curves different from those previously observed will always be misclassified. In contrast, the DP model can be unsupervised, semi-supervised or fully supervised. Since both methods have their relative merits, an approach that uses both RF and the DP model could be beneficial.

Spatiotemporal drought forecasting using nonlinear models

NASA Astrophysics Data System (ADS)

Vasiliades, Lampros; Loukas, Athanasios

2010-05-01

Spatiotemporal data mining is the extraction of unknown and implicit knowledge, structures, spatiotemporal relationships, or patterns not explicitly stored in spatiotemporal databases. As one of data mining techniques, forecasting is widely used to predict the unknown future based upon the patterns hidden in the current and past data. In order to achieve spatiotemporal forecasting, some mature analysis tools, e.g., time series and spatial statistics are extended to the spatial dimension and the temporal dimension, respectively. Drought forecasting plays an important role in the planning and management of natural resources and water resource systems in a river basin. Early and timelines forecasting of a drought event can help to take proactive measures and set out drought mitigation strategies to alleviate the impacts of drought. Despite the widespread application of nonlinear mathematical models, comparative studies on spatiotemporal drought forecasting using different models are still a huge task for modellers. This study uses a promising approach, the Gamma Test (GT), to select the input variables and the training data length, so that the trial and error workload could be greatly reduced. The GT enables to quickly evaluate and estimate the best mean squared error that can be achieved by a smooth model on any unseen data for a given selection of inputs, prior to model construction. The GT is applied to forecast droughts using monthly Standardized Precipitation Index (SPI) timeseries at multiple timescales in several precipitation stations at Pinios river basin in Thessaly region, Greece. Several nonlinear models have been developed efficiently, with the aid of the GT, for 1-month up to 12-month ahead forecasting. Several temporal and spatial statistical indices were considered for the performance evaluation of the models. The predicted results show reasonably good agreement with the actual data for short lead times, whereas the forecasting accuracy decreases with increase in lead time. Finally, the developed nonlinear models could be used in an early warning system for risk and decision analyses at the study area.

How do I know if I’ve improved my continental scale flood early warning system?

NASA Astrophysics Data System (ADS)

Cloke, Hannah L.; Pappenberger, Florian; Smith, Paul J.; Wetterhall, Fredrik

2017-04-01

Flood early warning systems mitigate damages and loss of life and are an economically efficient way of enhancing disaster resilience. The use of continental scale flood early warning systems is rapidly growing. The European Flood Awareness System (EFAS) is a pan-European flood early warning system forced by a multi-model ensemble of numerical weather predictions. Responses to scientific and technical changes can be complex in these computationally expensive continental scale systems, and improvements need to be tested by evaluating runs of the whole system. It is demonstrated here that forecast skill is not correlated with the value of warnings. In order to tell if the system has been improved an evaluation strategy is required that considers both forecast skill and warning value. The combination of a multi-forcing ensemble of EFAS flood forecasts is evaluated with a new skill-value strategy. The full multi-forcing ensemble is recommended for operational forecasting, but, there are spatial variations in the optimal forecast combination. Results indicate that optimizing forecasts based on value rather than skill alters the optimal forcing combination and the forecast performance. Also indicated is that model diversity and ensemble size are both important in achieving best overall performance. The use of several evaluation measures that consider both skill and value is strongly recommended when considering improvements to early warning systems.

Introduction on the operational storm surge forecasting system in Korea Operational Oceanographic System (KOOS)

NASA Astrophysics Data System (ADS)

Kwon, Jae-Il; Park, Kwang-Soon; Choi, Jung-Woon; Lee, Jong-Chan; Heo, Ki-Young; Kim, Sang-Ik

2017-04-01

During last more than 50 years, 258 typhoons passed and affected the Korean peninsula in terms of high winds, storm surges and extreme waves. In this study we explored the performance of the operational storm surge forecasting system in the Korea Operational Oceanographic System (KOOS) with 8 typhoons from 2010 to 2016. The operation storm surge forecasting system for the typhoon in KOOS is based on 2D depth averaged model with tides and CE (U.S. Army Corps of Engineers) wind model. Two key parameters of CE wind model, the locations of typhoon center and its central atmospheric pressure are based from Korea Meteorological administrative (KMA)'s typhoon information provided from 1 day to 3 hour intervals with the approach of typhoon through the KMA's web-site. For 8 typhoons cases, the overall errors, other performances and analysis such as peak time and surge duration are presented in each case. The most important factor in the storm surge errors in the operational forecasting system is the accuracy of typhoon passage prediction.

The Incorporation and Initialization of Cloud Water/ice in AN Operational Forecast Model

NASA Astrophysics Data System (ADS)

Zhao, Qingyun

Quantitative precipitation forecasts have been one of the weakest aspects of numerical weather prediction models. Theoretical studies show that the errors in precipitation calculation can arise from three sources: errors in the large-scale forecasts of primary variables, errors in the crude treatment of condensation/evaporation and precipitation processes, and errors in the model initial conditions. A new precipitation parameterization scheme has been developed to investigate the forecast value of improved precipitation physics via the introduction of cloud water and cloud ice into a numerical prediction model. The main feature of this scheme is the explicit calculation of cloud water and cloud ice in both the convective and stratiform precipitation parameterization. This scheme has been applied to the eta model at the National Meteorological Center. Four extensive tests have been performed. The statistical results showed a significant improvement in the model precipitation forecasts. Diagnostic studies suggest that the inclusion of cloud ice is important in transferring water vapor to precipitation and in the enhancement of latent heat release; the latter subsequently affects the vertical motion field significantly. Since three-dimensional cloud data is absent from the analysis/assimilation system for most numerical models, a method has been proposed to incorporate observed precipitation and nephanalysis data into the data assimilation system to obtain the initial cloud field for the eta model. In this scheme, the initial moisture and vertical motion fields are also improved at the same time as cloud initialization. The physical initialization is performed in a dynamical initialization framework that uses the Newtonian dynamical relaxation method to nudge the model's wind and mass fields toward analyses during a 12-hour data assimilation period. Results from a case study showed that a realistic cloud field was produced by this method at the end of the data assimilation period. Precipitation forecasts have been significantly improved as a result of the improved initial cloud, moisture and vertical motion fields.

Testing the Predictive Power of Coulomb Stress on Aftershock Sequences

NASA Astrophysics Data System (ADS)

Woessner, J.; Lombardi, A.; Werner, M. J.; Marzocchi, W.

2009-12-01

Empirical and statistical models of clustered seismicity are usually strongly stochastic and perceived to be uninformative in their forecasts, since only marginal distributions are used, such as the Omori-Utsu and Gutenberg-Richter laws. In contrast, so-called physics-based aftershock models, based on seismic rate changes calculated from Coulomb stress changes and rate-and-state friction, make more specific predictions: anisotropic stress shadows and multiplicative rate changes. We test the predictive power of models based on Coulomb stress changes against statistical models, including the popular Short Term Earthquake Probabilities and Epidemic-Type Aftershock Sequences models: We score and compare retrospective forecasts on the aftershock sequences of the 1992 Landers, USA, the 1997 Colfiorito, Italy, and the 2008 Selfoss, Iceland, earthquakes. To quantify predictability, we use likelihood-based metrics that test the consistency of the forecasts with the data, including modified and existing tests used in prospective forecast experiments within the Collaboratory for the Study of Earthquake Predictability (CSEP). Our results indicate that a statistical model performs best. Moreover, two Coulomb model classes seem unable to compete: Models based on deterministic Coulomb stress changes calculated from a given fault-slip model, and those based on fixed receiver faults. One model of Coulomb stress changes does perform well and sometimes outperforms the statistical models, but its predictive information is diluted, because of uncertainties included in the fault-slip model. Our results suggest that models based on Coulomb stress changes need to incorporate stochastic features that represent model and data uncertainty.

An experimental seasonal hydrological forecasting system over the Yellow River basin – Part 2: The added value from climate forecast models

DOE PAGES

Yuan, Xing

2016-06-22

This is the second paper of a two-part series on introducing an experimental seasonal hydrological forecasting system over the Yellow River basin in northern China. While the natural hydrological predictability in terms of initial hydrological conditions (ICs) is investigated in a companion paper, the added value from eight North American Multimodel Ensemble (NMME) climate forecast models with a grand ensemble of 99 members is assessed in this paper, with an implicit consideration of human-induced uncertainty in the hydrological models through a post-processing procedure. The forecast skill in terms of anomaly correlation (AC) for 2 m air temperature and precipitation does not necessarily decrease overmore » leads but is dependent on the target month due to a strong seasonality for the climate over the Yellow River basin. As there is more diversity in the model performance for the temperature forecasts than the precipitation forecasts, the grand NMME ensemble mean forecast has consistently higher skill than the best single model up to 6 months for the temperature but up to 2 months for the precipitation. The NMME climate predictions are downscaled to drive the variable infiltration capacity (VIC) land surface hydrological model and a global routing model regionalized over the Yellow River basin to produce forecasts of soil moisture, runoff and streamflow. And the NMME/VIC forecasts are compared with the Ensemble Streamflow Prediction method (ESP/VIC) through 6-month hindcast experiments for each calendar month during 1982–2010. As verified by the VIC offline simulations, the NMME/VIC is comparable to the ESP/VIC for the soil moisture forecasts, and the former has higher skill than the latter only for the forecasts at long leads and for those initialized in the rainy season. The forecast skill for runoff is lower for both forecast approaches, but the added value from NMME/VIC is more obvious, with an increase of the average AC by 0.08–0.2. To compare with the observed streamflow, both the hindcasts from NMME/VIC and ESP/VIC are post-processed through a linear regression model fitted by using VIC offline-simulated streamflow. The post-processed NMME/VIC reduces the root mean squared error (RMSE) from the post-processed ESP/VIC by 5–15 %. And the reduction occurs mostly during the transition from wet to dry seasons. As a result, with the consideration of the uncertainty in the hydrological models, the added value from climate forecast models is decreased especially at short leads, suggesting the necessity of improving the large-scale hydrological models in human-intervened river basins.« less

An experimental seasonal hydrological forecasting system over the Yellow River basin – Part 2: The added value from climate forecast models

DOE Office of Scientific and Technical Information (OSTI.GOV)

Yuan, Xing

This is the second paper of a two-part series on introducing an experimental seasonal hydrological forecasting system over the Yellow River basin in northern China. While the natural hydrological predictability in terms of initial hydrological conditions (ICs) is investigated in a companion paper, the added value from eight North American Multimodel Ensemble (NMME) climate forecast models with a grand ensemble of 99 members is assessed in this paper, with an implicit consideration of human-induced uncertainty in the hydrological models through a post-processing procedure. The forecast skill in terms of anomaly correlation (AC) for 2 m air temperature and precipitation does not necessarily decrease overmore » leads but is dependent on the target month due to a strong seasonality for the climate over the Yellow River basin. As there is more diversity in the model performance for the temperature forecasts than the precipitation forecasts, the grand NMME ensemble mean forecast has consistently higher skill than the best single model up to 6 months for the temperature but up to 2 months for the precipitation. The NMME climate predictions are downscaled to drive the variable infiltration capacity (VIC) land surface hydrological model and a global routing model regionalized over the Yellow River basin to produce forecasts of soil moisture, runoff and streamflow. And the NMME/VIC forecasts are compared with the Ensemble Streamflow Prediction method (ESP/VIC) through 6-month hindcast experiments for each calendar month during 1982–2010. As verified by the VIC offline simulations, the NMME/VIC is comparable to the ESP/VIC for the soil moisture forecasts, and the former has higher skill than the latter only for the forecasts at long leads and for those initialized in the rainy season. The forecast skill for runoff is lower for both forecast approaches, but the added value from NMME/VIC is more obvious, with an increase of the average AC by 0.08–0.2. To compare with the observed streamflow, both the hindcasts from NMME/VIC and ESP/VIC are post-processed through a linear regression model fitted by using VIC offline-simulated streamflow. The post-processed NMME/VIC reduces the root mean squared error (RMSE) from the post-processed ESP/VIC by 5–15 %. And the reduction occurs mostly during the transition from wet to dry seasons. As a result, with the consideration of the uncertainty in the hydrological models, the added value from climate forecast models is decreased especially at short leads, suggesting the necessity of improving the large-scale hydrological models in human-intervened river basins.« less

The MSFC Solar Activity Future Estimation (MSAFE) Model

NASA Technical Reports Server (NTRS)

Suggs, Ron

2017-01-01

The MSAFE model provides forecasts for the solar indices SSN, F10.7, and Ap. These solar indices are used as inputs to space environment models used in orbital spacecraft operations and space mission analysis. Forecasts from the MSAFE model are provided on the MSFC Natural Environments Branch's solar web page and are updated as new monthly observations become available. The MSAFE prediction routine employs a statistical technique that calculates deviations of past solar cycles from the mean cycle and performs a regression analysis to calculate the deviation from the mean cycle of the solar index at the next future time interval. The forecasts are initiated for a given cycle after about 8 to 9 monthly observations from the start of the cycle are collected. A forecast made at the beginning of cycle 24 using the MSAFE program captured the cycle fairly well with some difficulty in discerning the double peak that occurred at solar cycle maximum.

The analysis and application of a new hybrid pollutants forecasting model using modified Kolmogorov-Zurbenko filter.

PubMed

Li, Peizhi; Wang, Yong; Dong, Qingli

2017-04-01

Cities in China suffer from severe smog and haze, and a forecasting system with high accuracy is of great importance to foresee the concentrations of the airborne particles. Compared with chemical transport models, the growing artificial intelligence models can simulate nonlinearities and interactive relationships and getting more accurate results. In this paper, the Kolmogorov-Zurbenko (KZ) filter is modified and firstly applied to construct the model using an artificial intelligence method. The concentration of inhalable particles and fine particulate matter in Dalian are used to analyze the filtered components and test the forecasting accuracy. Besides, an extended experiment is made by implementing a comprehensive comparison and a stability test using data in three other cities in China. Results testify the excellent performance of the developed hybrid models, which can be utilized to better understand the temporal features of pollutants and to perform a better air pollution control and management. Copyright © 2017 Elsevier B.V. All rights reserved.

Vegetation Exposure to Ozone over the Continental United States: Assessment of Exposure Indices by the Eta-CMAQ Air Quality Forecast Model

EPA Science Inventory

This study presents the first evaluation of the performance of the Eta-CMAQ air quality forecast model to predict a variety of widely used seasonal mean and cumulative O₃ exposure indices associated with vegetation using the U.S. AIRNow O₃ observations.

A STUDY OF PROCESS CONTRIBUTIONS TO OZONE FORMATION DURING THE 2004 ICARTT PERIOD USING THE ETA-CMAQ FORECAST MODEL OVER THE NORTHEASTERN U.S.

EPA Science Inventory

First, this study evaluates the Eta-CMAQ forecast model performances for O₃, and related chemical species with the observational data from the aircraft (NOAA P-3 and NASA DC-8) flights, Lidar and ozonesonde during the 2004 International Consortium for Atmospheric Resea...

A STUDY OF PROCESS CONTRIBUTIONS TO OZONE FORMATION DURING THE 2004 ICARTT PERIOD USING THE ETA- CMAQ FORECAST MODEL OVER THE NORTHEASTERN U.S.

EPA Science Inventory

First, this study evaluates the Eta-CMAQ forecast model performances for O₃, and related chemical species with the observational data from the aircraft (NOAA P-3 and NASA DC-8) flights, Lidar and ozonesonde during the 2004 International Consortium for Atmospheric Resea...

Using modified fruit fly optimisation algorithm to perform the function test and case studies

NASA Astrophysics Data System (ADS)

Pan, Wen-Tsao

2013-06-01

Evolutionary computation is a computing mode established by practically simulating natural evolutionary processes based on the concept of Darwinian Theory, and it is a common research method. The main contribution of this paper was to reinforce the function of searching for the optimised solution using the fruit fly optimization algorithm (FOA), in order to avoid the acquisition of local extremum solutions. The evolutionary computation has grown to include the concepts of animal foraging behaviour and group behaviour. This study discussed three common evolutionary computation methods and compared them with the modified fruit fly optimization algorithm (MFOA). It further investigated the ability of the three mathematical functions in computing extreme values, as well as the algorithm execution speed and the forecast ability of the forecasting model built using the optimised general regression neural network (GRNN) parameters. The findings indicated that there was no obvious difference between particle swarm optimization and the MFOA in regards to the ability to compute extreme values; however, they were both better than the artificial fish swarm algorithm and FOA. In addition, the MFOA performed better than the particle swarm optimization in regards to the algorithm execution speed, and the forecast ability of the forecasting model built using the MFOA's GRNN parameters was better than that of the other three forecasting models.

Individual versus superensemble forecasts of seasonal influenza outbreaks in the United States.

PubMed

Yamana, Teresa K; Kandula, Sasikiran; Shaman, Jeffrey

2017-11-01

Recent research has produced a number of methods for forecasting seasonal influenza outbreaks. However, differences among the predicted outcomes of competing forecast methods can limit their use in decision-making. Here, we present a method for reconciling these differences using Bayesian model averaging. We generated retrospective forecasts of peak timing, peak incidence, and total incidence for seasonal influenza outbreaks in 48 states and 95 cities using 21 distinct forecast methods, and combined these individual forecasts to create weighted-average superensemble forecasts. We compared the relative performance of these individual and superensemble forecast methods by geographic location, timing of forecast, and influenza season. We find that, overall, the superensemble forecasts are more accurate than any individual forecast method and less prone to producing a poor forecast. Furthermore, we find that these advantages increase when the superensemble weights are stratified according to the characteristics of the forecast or geographic location. These findings indicate that different competing influenza prediction systems can be combined into a single more accurate forecast product for operational delivery in real time.

Individual versus superensemble forecasts of seasonal influenza outbreaks in the United States

PubMed Central

Kandula, Sasikiran; Shaman, Jeffrey

2017-01-01

Recent research has produced a number of methods for forecasting seasonal influenza outbreaks. However, differences among the predicted outcomes of competing forecast methods can limit their use in decision-making. Here, we present a method for reconciling these differences using Bayesian model averaging. We generated retrospective forecasts of peak timing, peak incidence, and total incidence for seasonal influenza outbreaks in 48 states and 95 cities using 21 distinct forecast methods, and combined these individual forecasts to create weighted-average superensemble forecasts. We compared the relative performance of these individual and superensemble forecast methods by geographic location, timing of forecast, and influenza season. We find that, overall, the superensemble forecasts are more accurate than any individual forecast method and less prone to producing a poor forecast. Furthermore, we find that these advantages increase when the superensemble weights are stratified according to the characteristics of the forecast or geographic location. These findings indicate that different competing influenza prediction systems can be combined into a single more accurate forecast product for operational delivery in real time. PMID:29107987

Using Google Flu Trends data in forecasting influenza-like-illness related ED visits in Omaha, Nebraska.

PubMed

Araz, Ozgur M; Bentley, Dan; Muelleman, Robert L

2014-09-01

Emergency department (ED) visits increase during the influenza seasons. It is essential to identify statistically significant correlates in order to develop an accurate forecasting model for ED visits. Forecasting influenza-like-illness (ILI)-related ED visits can significantly help in developing robust resource management strategies at the EDs. We first performed correlation analyses to understand temporal correlations between several predictors of ILI-related ED visits. We used the data available for Douglas County, the biggest county in Nebraska, for Omaha, the biggest city in the state, and for a major hospital in Omaha. The data set included total and positive influenza test results from the hospital (ie, Antigen rapid (Ag) and Respiratory Syncytial Virus Infection (RSV) tests); an Internet-based influenza surveillance system data, that is, Google Flu Trends, for both Nebraska and Omaha; total ED visits in Douglas County attributable to ILI; and ILI surveillance network data for Douglas County and Nebraska as the predictors and data for the hospital's ILI-related ED visits as the dependent variable. We used Seasonal Autoregressive Integrated Moving Average and Holt Winters methods with3 linear regression models to forecast ILI-related ED visits at the hospital and evaluated model performances by comparing the root means square errors (RMSEs). Because of strong positive correlations with ILI-related ED visits between 2008 and 2012, we validated the use of Google Flu Trends data as a predictor in an ED influenza surveillance tool. Of the 5 forecasting models we have tested, linear regression models performed significantly better when Google Flu Trends data were included as a predictor. Regression models including Google Flu Trends data as a predictor variable have lower RMSE, and the lowest is achieved when all other variables are also included in the model in our forecasting experiments for the first 5 weeks of 2013 (with RMSE = 57.61). Google Flu Trends data statistically improve the performance of predicting ILI-related ED visits in Douglas County, and this result can be generalized to other communities. Timely and accurate estimates of ED volume during the influenza season, as well as during pandemic outbreaks, can help hospitals plan their ED resources accordingly and lower their costs by optimizing supplies and staffing and can improve service quality by decreasing ED wait times and overcrowding. Copyright © 2014 Elsevier Inc. All rights reserved.

A statistical forecast model using the time-scale decomposition technique to predict rainfall during flood period over the middle and lower reaches of the Yangtze River Valley

NASA Astrophysics Data System (ADS)

Hu, Yijia; Zhong, Zhong; Zhu, Yimin; Ha, Yao

2018-04-01

In this paper, a statistical forecast model using the time-scale decomposition method is established to do the seasonal prediction of the rainfall during flood period (FPR) over the middle and lower reaches of the Yangtze River Valley (MLYRV). This method decomposites the rainfall over the MLYRV into three time-scale components, namely, the interannual component with the period less than 8 years, the interdecadal component with the period from 8 to 30 years, and the interdecadal component with the period larger than 30 years. Then, the predictors are selected for the three time-scale components of FPR through the correlation analysis. At last, a statistical forecast model is established using the multiple linear regression technique to predict the three time-scale components of the FPR, respectively. The results show that this forecast model can capture the interannual and interdecadal variation of FPR. The hindcast of FPR during 14 years from 2001 to 2014 shows that the FPR can be predicted successfully in 11 out of the 14 years. This forecast model performs better than the model using traditional scheme without time-scale decomposition. Therefore, the statistical forecast model using the time-scale decomposition technique has good skills and application value in the operational prediction of FPR over the MLYRV.

Improving flood forecasting capability of physically based distributed hydrological models by parameter optimization

NASA Astrophysics Data System (ADS)

Chen, Y.; Li, J.; Xu, H.

2016-01-01

Physically based distributed hydrological models (hereafter referred to as PBDHMs) divide the terrain of the whole catchment into a number of grid cells at fine resolution and assimilate different terrain data and precipitation to different cells. They are regarded to have the potential to improve the catchment hydrological process simulation and prediction capability. In the early stage, physically based distributed hydrological models are assumed to derive model parameters from the terrain properties directly, so there is no need to calibrate model parameters. However, unfortunately the uncertainties associated with this model derivation are very high, which impacted their application in flood forecasting, so parameter optimization may also be necessary. There are two main purposes for this study: the first is to propose a parameter optimization method for physically based distributed hydrological models in catchment flood forecasting by using particle swarm optimization (PSO) algorithm and to test its competence and to improve its performances; the second is to explore the possibility of improving physically based distributed hydrological model capability in catchment flood forecasting by parameter optimization. In this paper, based on the scalar concept, a general framework for parameter optimization of the PBDHMs for catchment flood forecasting is first proposed that could be used for all PBDHMs. Then, with the Liuxihe model as the study model, which is a physically based distributed hydrological model proposed for catchment flood forecasting, the improved PSO algorithm is developed for the parameter optimization of the Liuxihe model in catchment flood forecasting. The improvements include adoption of the linearly decreasing inertia weight strategy to change the inertia weight and the arccosine function strategy to adjust the acceleration coefficients. This method has been tested in two catchments in southern China with different sizes, and the results show that the improved PSO algorithm could be used for the Liuxihe model parameter optimization effectively and could improve the model capability largely in catchment flood forecasting, thus proving that parameter optimization is necessary to improve the flood forecasting capability of physically based distributed hydrological models. It also has been found that the appropriate particle number and the maximum evolution number of PSO algorithm used for the Liuxihe model catchment flood forecasting are 20 and 30 respectively.

Equation-free mechanistic ecosystem forecasting using empirical dynamic modeling

PubMed Central

Ye, Hao; Beamish, Richard J.; Glaser, Sarah M.; Grant, Sue C. H.; Hsieh, Chih-hao; Richards, Laura J.; Schnute, Jon T.; Sugihara, George

2015-01-01

It is well known that current equilibrium-based models fall short as predictive descriptions of natural ecosystems, and particularly of fisheries systems that exhibit nonlinear dynamics. For example, model parameters assumed to be fixed constants may actually vary in time, models may fit well to existing data but lack out-of-sample predictive skill, and key driving variables may be misidentified due to transient (mirage) correlations that are common in nonlinear systems. With these frailties, it is somewhat surprising that static equilibrium models continue to be widely used. Here, we examine empirical dynamic modeling (EDM) as an alternative to imposed model equations and that accommodates both nonequilibrium dynamics and nonlinearity. Using time series from nine stocks of sockeye salmon (Oncorhynchus nerka) from the Fraser River system in British Columbia, Canada, we perform, for the the first time to our knowledge, real-data comparison of contemporary fisheries models with equivalent EDM formulations that explicitly use spawning stock and environmental variables to forecast recruitment. We find that EDM models produce more accurate and precise forecasts, and unlike extensions of the classic Ricker spawner–recruit equation, they show significant improvements when environmental factors are included. Our analysis demonstrates the strategic utility of EDM for incorporating environmental influences into fisheries forecasts and, more generally, for providing insight into how environmental factors can operate in forecast models, thus paving the way for equation-free mechanistic forecasting to be applied in management contexts. PMID:25733874

Study on the influence of ground and satellite observations on the numerical air-quality for PM10 over Romanian territory

NASA Astrophysics Data System (ADS)

Dumitrache, Rodica Claudia; Iriza, Amalia; Maco, Bogdan Alexandru; Barbu, Cosmin Danut; Hirtl, Marcus; Mantovani, Simone; Nicola, Oana; Irimescu, Anisoara; Craciunescu, Vasile; Ristea, Alina; Diamandi, Andrei

2016-10-01

The numerical forecast of particulate matter concentrations in general, and PM10 in particular is a theme of high socio-economic relevance. The aim of this study was to investigate the impact of ground and satellite data assimilation of PM10 observations into the Weather Research and Forecasting model coupled with Chemistry (WRF-CHEM) numerical air quality model for Romanian territory. This is the first initiative of the kind for this domain of interest. Assimilation of satellite information - e.g. AOT's in air quality models is of interest due to the vast spatial coverage of the observations. Support Vector Regression (SVR) techniques are used to estimate the PM content from heterogeneous data sources, including EO products (Aerosol Optical Thickness), ground measurements and numerical model data (temperature, humidity, wind, etc.). In this study we describe the modeling framework employed and present the evaluation of the impact from the data assimilation of PM10 observations on the forecast of the WRF-CHEM model. Integrations of the WRF-CHEM model in data assimilation enabled/disabled configurations allowed the evaluation of satellite and ground data assimilation impact on the PM10 forecast performance for the Romanian territory. The model integration and evaluation were performed for two months, one in winter conditions (January 2013) and one in summer conditions (June 2013).

Comparison of CMAQ Modeling Study with Discover-AQ 2014 Aircraft Measurements over Colorado

NASA Astrophysics Data System (ADS)

Tang, Y.; Pan, L.; Lee, P.; Tong, D.; Kim, H. C.; Artz, R. S.

2014-12-01

NASA and NCAR jointly led a recent multiple platform-based (space, air and ground) measurement intensive to study air quality and to validate satellite data. The Discover-AQ/FRAPPE field experiment took place along the Colorado Front Range in July and August, 2014. The air quality modeling team of the NOAA Air Resources Laboratory was one of the three teams that provided real-time air quality forecasting for the campaign. The U.S. EPA Community Multi-scale Air Quality (CMAQ) Model was used with emission inventories based on the data set used by the NOAA National Air Quality Forecasting Capacity (NAQFC). By analyzing the forecast results calculated using aircraft measurements, it was found that CO emissions tended to be overestimated, while ethane emissions were underestimated. Biogenic VOCs were also underpredicted. Due to their relatively high altitude, ozone concentrations in Denver and the surrounding areas are affected by both local emissions and transported ozone. The modeled ozone was highly dependent on the meteorological predictions over this region. The complex terrain over the Rocky Mountains also contributed to the model uncertainty. This study discussed the causes of model biases, the forecast performance under different meteorology, and results from using different model grid resolutions. Several data assimilation techniques were further tested to improve the "post-analysis" performance of the modeling system for the period.

Two-Step Forecast of Geomagnetic Storm Using Coronal Mass Ejection and Solar Wind Condition

NASA Technical Reports Server (NTRS)

Kim, R.-S.; Moon, Y.-J.; Gopalswamy, N.; Park, Y.-D.; Kim, Y.-H.

2014-01-01

To forecast geomagnetic storms, we had examined initially observed parameters of coronal mass ejections (CMEs) and introduced an empirical storm forecast model in a previous study. Now we suggest a two-step forecast considering not only CME parameters observed in the solar vicinity but also solar wind conditions near Earth to improve the forecast capability. We consider the empirical solar wind criteria derived in this study (Bz = -5 nT or Ey = 3 mV/m for t = 2 h for moderate storms with minimum Dst less than -50 nT) (i.e. Magnetic Field Magnitude, B (sub z) less than or equal to -5 nanoTeslas or duskward Electrical Field, E (sub y) greater than or equal to 3 millivolts per meter for time greater than or equal to 2 hours for moderate storms with Minimum Disturbance Storm Time, Dst less than -50 nanoTeslas) and a Dst model developed by Temerin and Li (2002, 2006) (TL [i.e. Temerin Li] model). Using 55 CME-Dst pairs during 1997 to 2003, our solar wind criteria produce slightly better forecasts for 31 storm events (90 percent) than the forecasts based on the TL model (87 percent). However, the latter produces better forecasts for 24 nonstorm events (88 percent), while the former correctly forecasts only 71 percent of them. We then performed the two-step forecast. The results are as follows: (i) for 15 events that are incorrectly forecasted using CME parameters, 12 cases (80 percent) can be properly predicted based on solar wind conditions; (ii) if we forecast a storm when both CME and solar wind conditions are satisfied (n, i.e. cap operator - the intersection set that is comprised of all the elements that are common to both), the critical success index becomes higher than that from the forecast using CME parameters alone, however, only 25 storm events (81 percent) are correctly forecasted; and (iii) if we forecast a storm when either set of these conditions is satisfied (?, i.e. cup operator - the union set that is comprised of all the elements of either or both), all geomagnetic storms are correctly forecasted.

Promoting inclusive water governance and forecasting the structure of water consumption based on compositional data: A case study of Beijing.

PubMed

Wei, Yigang; Wang, Zhichao; Wang, Huiwen; Yao, Tang; Li, Yan

2018-09-01

Water is centrally important for agricultural security, environment, people's livelihoods, and socio-economic development, particularly in the face of extreme climate changes. Due to water shortages in many cities, the conflicts between various stakeholders and sectors over water use and allocation are becoming more common and intense. Effective inclusive governance of water use is critical for relieving water use conflicts. In addition, reliable forecasting of the structure of water usage among different sectors is a basic need for effective water governance planning. Although a large number of studies have attempted to forecast water use, little is known about the forecasted structure and trends of water use in the future. This paper aims to develop a forecasting model for the structure of water usage based on compositional data. Compositional data analysis is an effective approach for investigating the internal structure of a system. A host of data transformation methods and forecasting models were adopted and compared in order to derive the best-performing model. According to mean absolute percent error for compositional data (CoMAPE), a hyperspherical-transformation-based vector autoregression model for compositional data (VAR-DRHT) is the best-performing model. The proportions of the agricultural, industrial, domestic and environmental water will be 6.11%, 5.01%, 37.48% and 51.4% by 2020. Several recommendations for water inclusive development are provided to give a better account for the optimization of the water use structure, alleviation of water shortages, and improving stake holders' wellbeing. Overall, although we focus on groundwater, this study presents a powerful framework broadly applicable to resource management. Copyright © 2018 Elsevier B.V. All rights reserved.

Global Positioning System (GPS) Precipitable Water in Forecasting Lightning at Spaceport Canaveral

NASA Technical Reports Server (NTRS)

Kehrer, Kristen; Graf, Brian G.; Roeder, William

2005-01-01

Using meteorology data, focusing on precipitable water (PW), obtained during the 2000-2003 thunderstorm seasons in Central Florida, this paper will, one, assess the skill and accuracy measurements of the current Mazany forecasting tool and, two, provide additional forecasting tools that can be used in predicting lightning. Kennedy Space Center (KSC) and Cape Canaveral Air Force Station (CCAFS) are located in east Central Florida. KSC and CCAFS process and launch manned (NASA Space Shuttle) and unmanned (NASA and Air Force Expendable Launch Vehicles) space vehicles. One of the biggest cost impacts is unplanned launch scrubs due to inclement weather conditions such as thunderstorms. Each launch delay/scrub costs over a quarter million dollars, and the need to land the Shuttle at another landing site and return to KSC costs approximately $ 1M. Given the amount of time lost and costs incurred, the ability to accurately forecast (predict) when lightning will occur can result in significant cost and time savings. All lightning prediction models were developed using binary logistic regression. Lightning is the dependent variable and is binary. The independent variables are the Precipitable Water (PW) value for a given time of the day, the change in PW up to 12 hours, the electric field mill value, and the K-index value. In comparing the Mazany model results for the 1999 period B against actual observations for the 2000-2003 thunderstorm seasons, differences were found in the False Alarm Rate (FAR), Probability of Detection (POD) and Hit Rate (H). On average, the False Alarm Rate (FAR) increased by 58%, the Probability of Detection (POD) decreased by 31% and the Hit Rate decreased by 20%. In comparing the performance of the 6 hour forecast period to the performance of the 1.5 hour forecast period for the Mazany model, the FAR was lower by 15% and the Hit Rate was higher by 7%. However, the POD for the 6 hour forecast period was lower by 16% as compared to the POD of the 1.5 hour forecast period. Neither forecast period performed at the accuracy measures expected. A 2-Hr Forecasting Tool was developed to support a Phase I Lightning Advisory, which requires a 30-minute lead time for predicting lightning.

An Experimental High-Resolution Forecast System During the Vancouver 2010 Winter Olympic and Paralympic Games

NASA Astrophysics Data System (ADS)

Mailhot, J.; Milbrandt, J. A.; Giguère, A.; McTaggart-Cowan, R.; Erfani, A.; Denis, B.; Glazer, A.; Vallée, M.

2014-01-01

Environment Canada ran an experimental numerical weather prediction (NWP) system during the Vancouver 2010 Winter Olympic and Paralympic Games, consisting of nested high-resolution (down to 1-km horizontal grid-spacing) configurations of the GEM-LAM model, with improved geophysical fields, cloud microphysics and radiative transfer schemes, and several new diagnostic products such as density of falling snow, visibility, and peak wind gust strength. The performance of this experimental NWP system has been evaluated in these winter conditions over complex terrain using the enhanced mesoscale observing network in place during the Olympics. As compared to the forecasts from the operational regional 15-km GEM model, objective verification generally indicated significant added value of the higher-resolution models for near-surface meteorological variables (wind speed, air temperature, and dewpoint temperature) with the 1-km model providing the best forecast accuracy. Appreciable errors were noted in all models for the forecasts of wind direction and humidity near the surface. Subjective assessment of several cases also indicated that the experimental Olympic system was skillful at forecasting meteorological phenomena at high-resolution, both spatially and temporally, and provided enhanced guidance to the Olympic forecasters in terms of better timing of precipitation phase change, squall line passage, wind flow channeling, and visibility reduction due to fog and snow.

A prospective earthquake forecast experiment in the western Pacific

NASA Astrophysics Data System (ADS)

Eberhard, David A. J.; Zechar, J. Douglas; Wiemer, Stefan

2012-09-01

Since the beginning of 2009, the Collaboratory for the Study of Earthquake Predictability (CSEP) has been conducting an earthquake forecast experiment in the western Pacific. This experiment is an extension of the Kagan-Jackson experiments begun 15 years earlier and is a prototype for future global earthquake predictability experiments. At the beginning of each year, seismicity models make a spatially gridded forecast of the number of Mw≥ 5.8 earthquakes expected in the next year. For the three participating statistical models, we analyse the first two years of this experiment. We use likelihood-based metrics to evaluate the consistency of the forecasts with the observed target earthquakes and we apply measures based on Student's t-test and the Wilcoxon signed-rank test to compare the forecasts. Overall, a simple smoothed seismicity model (TripleS) performs the best, but there are some exceptions that indicate continued experiments are vital to fully understand the stability of these models, the robustness of model selection and, more generally, earthquake predictability in this region. We also estimate uncertainties in our results that are caused by uncertainties in earthquake location and seismic moment. Our uncertainty estimates are relatively small and suggest that the evaluation metrics are relatively robust. Finally, we consider the implications of our results for a global earthquake forecast experiment.

First Assessment of Itaipu Dam Ensemble Inflow Forecasting System

NASA Astrophysics Data System (ADS)

Mainardi Fan, Fernando; Machado Vieira Lisboa, Auder; Gomes Villa Trinidad, Giovanni; Rógenes Monteiro Pontes, Paulo; Collischonn, Walter; Tucci, Carlos; Costa Buarque, Diogo

2017-04-01

Inflow forecasting for Hydropower Plants (HPP) Dams is one of the prominent uses for hydrological forecasts. A very important HPP in terms of energy generation for South America is the Itaipu Dam, located in the Paraná River, between Brazil and Paraguay countries, with a drainage area of 820.000km2. In this work, we present the development of an ensemble forecasting system for Itaipu, operational since November 2015. The system is based in the MGB-IPH hydrological model, includes hydrodynamics simulations of the main river, and is run every day morning forced by seven different rainfall forecasts: (i) CPTEC-ETA 15km; (ii) CPTEC-BRAMS 5km; (iii) SIMEPAR WRF Ferrier; (iv) SIMEPAR WRF Lin; (v) SIMEPAR WRF Morrison; (vi) SIMEPAR WRF WDM6; (vii) SIMEPAR MEDIAN. The last one (vii) corresponds to the median value of SIMEPAR WRF model versions (iii to vi) rainfall forecasts. Besides the developed system, the "traditional" method for inflow forecasting generation for the Itaipu Dam is also run every day. This traditional method consists in the approximation of the future inflow based on the discharge tendency of upstream telemetric gauges. Nowadays, after all the forecasts are run, the hydrology team of Itaipu develop a consensus forecast, based on all obtained results, which is the one used for the Itaipu HPP Dam operation. After one year of operation a first evaluation of the Ensemble Forecasting System was conducted. Results show that the system performs satisfactory for rising flows up to five days lead time. However, some false alarms were also issued by most ensemble members in some cases. And not in all cases the system performed better than the traditional method, especially during hydrograph recessions. In terms of meteorological forecasts, some members usage are being discontinued. In terms of the hydrodynamics representation, it seems that a better information of rivers cross section could improve hydrographs recession curves forecasts. Those opportunities for improvements are currently being addressed in the system next update.

Evaluation of Flood Forecast and Warning in Elbe river basin - Impact of Forecaster's Strategy

NASA Astrophysics Data System (ADS)

Danhelka, Jan; Vlasak, Tomas

2010-05-01

Czech Hydrometeorological Institute (CHMI) is responsible for flood forecasting and warning in the Czech Republic. To meet that issue CHMI operates hydrological forecasting systems and publish flow forecast in selected profiles. Flood forecast and warning is an output of system that links observation (flow and atmosphere), data processing, weather forecast (especially NWP's QPF), hydrological modeling and modeled outputs evaluation and interpretation by forecaster. Forecast users are interested in final output without separating uncertainties of separate steps of described process. Therefore an evaluation of final operational forecasts was done for profiles within Elbe river basin produced by AquaLog forecasting system during period 2002 to 2008. Effects of uncertainties of observation, data processing and especially meteorological forecasts were not accounted separately. Forecast of flood levels exceedance (peak over the threshold) during forecasting period was the main criterion as flow increase forecast is of the highest importance. Other evaluation criteria included peak flow and volume difference. In addition Nash-Sutcliffe was computed separately for each time step (1 to 48 h) of forecasting period to identify its change with the lead time. Textual flood warnings are issued for administrative regions to initiate flood protection actions in danger of flood. Flood warning hit rate was evaluated at regions level and national level. Evaluation found significant differences of model forecast skill between forecasting profiles, particularly less skill was evaluated at small headwater basins due to domination of QPF uncertainty in these basins. The average hit rate was 0.34 (miss rate = 0.33, false alarm rate = 0.32). However its explored spatial difference is likely to be influenced also by different fit of parameters sets (due to different basin characteristics) and importantly by different impact of human factor. Results suggest that the practice of interactive model operation, experience and forecasting strategy differs between responsible forecasting offices. Warning is based on model outputs interpretation by hydrologists-forecaster. Warning hit rate reached 0.60 for threshold set to lowest flood stage of which 0.11 was underestimation of flood degree (miss 0.22, false alarm 0.28). Critical success index of model forecast was 0.34, while the same criteria for warning reached 0.55. We assume that the increase accounts not only to change of scale from single forecasting point to region for warning, but partly also to forecaster's added value. There is no official warning strategy preferred in the Czech Republic (f.e. tolerance towards higher false alarm rate). Therefore forecaster decision and personal strategy is of great importance. Results show quite successful warning for 1st flood level exceedance, over-warning for 2nd flood level, but under-warning for 3rd (highest) flood level. That suggests general forecaster's preference of medium level warning (2nd flood level is legally determined to be the start of the flood and flood protection activities). In conclusion human forecaster's experience and analysis skill increases flood warning performance notably. However society preference should be specifically addressed in the warning strategy definition to support forecaster's decision making.

A short-term and high-resolution distribution system load forecasting approach using support vector regression with hybrid parameters optimization

DOE PAGES

Jiang, Huaiguang; Zhang, Yingchen; Muljadi, Eduard; ...

2016-01-01

This paper proposes an approach for distribution system load forecasting, which aims to provide highly accurate short-term load forecasting with high resolution utilizing a support vector regression (SVR) based forecaster and a two-step hybrid parameters optimization method. Specifically, because the load profiles in distribution systems contain abrupt deviations, a data normalization is designed as the pretreatment for the collected historical load data. Then an SVR model is trained by the load data to forecast the future load. For better performance of SVR, a two-step hybrid optimization algorithm is proposed to determine the best parameters. In the first step of themore » hybrid optimization algorithm, a designed grid traverse algorithm (GTA) is used to narrow the parameters searching area from a global to local space. In the second step, based on the result of the GTA, particle swarm optimization (PSO) is used to determine the best parameters in the local parameter space. After the best parameters are determined, the SVR model is used to forecast the short-term load deviation in the distribution system. The performance of the proposed approach is compared to some classic methods in later sections of the paper.« less

Evaluation of Real-Time Convection-Permitting Precipitation Forecasts in China During the 2013-2014 Summer Season

NASA Astrophysics Data System (ADS)

Zhu, Kefeng; Xue, Ming; Zhou, Bowen; Zhao, Kun; Sun, Zhengqi; Fu, Peiling; Zheng, Yongguang; Zhang, Xiaoling; Meng, Qingtao

2018-01-01

Forecasts at a 4 km convection-permitting resolution over China during the summer season have been produced with the Weather Research and Forecasting model at Nanjing University since 2013. Precipitation forecasts from 2013 to 2014 are evaluated with dense rain gauge observations and compared with operational global model forecasts. Overall, the 4 km forecasts show very good agreement with observations over most parts of China, outperforming global forecasts in terms of spatial distribution, intensity, and diurnal variation. Quantitative evaluations with the Gilbert skill score further confirm the better performance of the 4 km forecasts over global forecasts for heavy precipitation, especially for the thresholds of 100 and 150 mm d-1. Besides bulk characteristics, the representations of some unique features of summer precipitation in China under the influence of the East Asian summer monsoon are further evaluated. These include the northward progression and southward retreat of the main rainband through the summer season, the diurnal variations of precipitation, and the meridional and zonal propagation of precipitation episodes associated with background synoptic flow and the embedded mesoscale convective systems. The 4 km forecast is able to faithfully reproduce most of the features while overprediction of afternoon convection near the southern China coast is found to be a main deficiency that requires further investigations.

Sensitivity analysis of a data assimilation technique for hindcasting and forecasting hydrodynamics of a complex coastal water body

NASA Astrophysics Data System (ADS)

Ren, Lei; Hartnett, Michael

2017-02-01

Accurate forecasting of coastal surface currents is of great economic importance due to marine activities such as marine renewable energy and fish farms in coastal regions in recent twenty years. Advanced oceanographic observation systems such as satellites and radars can provide many parameters of interest, such as surface currents and waves, with fine spatial resolution in near real time. To enhance modelling capability, data assimilation (DA) techniques which combine the available measurements with the hydrodynamic models have been used since the 1990s in oceanography. Assimilating measurements into hydrodynamic models makes the original model background states follow the observation trajectory, then uses it to provide more accurate forecasting information. Galway Bay is an open, wind dominated water body on which two coastal radars are deployed. An efficient and easy to implement sequential DA algorithm named Optimal Interpolation (OI) was used to blend radar surface current data into a three-dimensional Environmental Fluid Dynamics Code (EFDC) model. Two empirical parameters, horizontal correlation length and DA cycle length (CL), are inherent within OI. No guidance has previously been published regarding selection of appropriate values of these parameters or how sensitive OI DA is to variations in their values. Detailed sensitivity analysis has been performed on both of these parameters and results presented. Appropriate value of DA CL was examined and determined on producing the minimum Root-Mean-Square-Error (RMSE) between radar data and model background states. Analysis was performed to evaluate assimilation index (AI) of using an OI DA algorithm in the model. AI of the half-day forecasting mean vectors' directions was over 50% in the best assimilation model. The ability of using OI to improve model forecasts was also assessed and is reported upon.

Using phenomenological models for forecasting the 2015 Ebola challenge.

PubMed

Pell, Bruce; Kuang, Yang; Viboud, Cecile; Chowell, Gerardo

2018-03-01

The rising number of novel pathogens threatening the human population has motivated the application of mathematical modeling for forecasting the trajectory and size of epidemics. We summarize the real-time forecasting results of the logistic equation during the 2015 Ebola challenge focused on predicting synthetic data derived from a detailed individual-based model of Ebola transmission dynamics and control. We also carry out a post-challenge comparison of two simple phenomenological models. In particular, we systematically compare the logistic growth model and a recently introduced generalized Richards model (GRM) that captures a range of early epidemic growth profiles ranging from sub-exponential to exponential growth. Specifically, we assess the performance of each model for estimating the reproduction number, generate short-term forecasts of the epidemic trajectory, and predict the final epidemic size. During the challenge the logistic equation consistently underestimated the final epidemic size, peak timing and the number of cases at peak timing with an average mean absolute percentage error (MAPE) of 0.49, 0.36 and 0.40, respectively. Post-challenge, the GRM which has the flexibility to reproduce a range of epidemic growth profiles ranging from early sub-exponential to exponential growth dynamics outperformed the logistic growth model in ascertaining the final epidemic size as more incidence data was made available, while the logistic model underestimated the final epidemic even with an increasing amount of data of the evolving epidemic. Incidence forecasts provided by the generalized Richards model performed better across all scenarios and time points than the logistic growth model with mean RMS decreasing from 78.00 (logistic) to 60.80 (GRM). Both models provided reasonable predictions of the effective reproduction number, but the GRM slightly outperformed the logistic growth model with a MAPE of 0.08 compared to 0.10, averaged across all scenarios and time points. Our findings further support the consideration of transmission models that incorporate flexible early epidemic growth profiles in the forecasting toolkit. Such models are particularly useful for quickly evaluating a developing infectious disease outbreak using only case incidence time series of the early phase of an infectious disease outbreak. Copyright © 2016 The Authors. Published by Elsevier B.V. All rights reserved.

High resolution statistical downscaling of the EUROSIP seasonal prediction. Application for southeastern Romania

NASA Astrophysics Data System (ADS)

Busuioc, Aristita; Dumitrescu, Alexandru; Dumitrache, Rodica; Iriza, Amalia

2017-04-01

Seasonal climate forecasts in Europe are currently issued at the European Centre for Medium-Range Weather Forecasts (ECMWF) in the form of multi-model ensemble predictions available within the "EUROSIP" system. Different statistical techniques to calibrate, downscale and combine the EUROSIP direct model output are used to optimize the quality of the final probabilistic forecasts. In this study, a statistical downscaling model (SDM) based on canonical correlation analysis (CCA) is used to downscale the EUROSIP seasonal forecast at a spatial resolution of 1km x 1km over the Movila farm placed in southeastern Romania. This application is achieved in the framework of the H2020 MOSES project (http://www.moses-project.eu). The combination between monthly standardized values of three climate variables (maximum/minimum temperatures-Tmax/Tmin, total precipitation-Prec) is used as predictand while combinations of various large-scale predictors are tested in terms of their availability as outputs in the seasonal EUROSIP probabilistic forecasting (sea level pressure, temperature at 850 hPa and geopotential height at 500 hPa). The predictors are taken from the ECMWF system considering 15 members of the ensemble, for which the hindcasts since 1991 until present are available. The model was calibrated over the period 1991-2014 and predictions for summers 2015 and 2016 were achieved. The calibration was made for the ensemble average as well as for each ensemble member. The model was developed for each lead time: one month anticipation for June, two months anticipation for July and three months anticipation for August. The main conclusions from these preliminary results are: best predictions (in terms of the anomaly sign) for Tmax (July-2 months anticipation, August-3 months anticipation) for both years (2015, 2016); for Tmin - good predictions only for August (3 months anticipation ) for both years; for precipitation, good predictions for July (2 months anticipation) in 2015 and August (3 months anticipation) in 2016; failed prediction for June (1-month anticipation) for all parameters. To see if the results obtained for 2015 and 2016 summers are in agreement with the general ECMWF model performance in forecast of the three predictors used in the CCA SDM calibration, the mean bias and root mean square errors (RMSE) calculated over the entire period in each grid point, for each ensemble member and ensemble average were computed. The obtained results are confirmed, showing highest ECMWF performance in forecasting of the three predictors for 3 months anticipation (August) and lowest performance for one month anticipation (June). The added value of the CCA SDM in forecasting local Tmax/Tmin and total precipitation was compared to the ECMWF performance using nearest grid point method. Comparisons were performed for the 1991-2014 period, taking into account the forecast made in May for July. An important improvement was found for the CCA SDM predictions in terms of the RMSE value (computed against observations) for Tmax/Tmin and less for precipitation. The tests are in progress for the other summer months (June, July).

Time-Series Approaches for Forecasting the Number of Hospital Daily Discharged Inpatients.

PubMed

Ting Zhu; Li Luo; Xinli Zhang; Yingkang Shi; Wenwu Shen

2017-03-01

For hospitals where decisions regarding acceptable rates of elective admissions are made in advance based on expected available bed capacity and emergency requests, accurate predictions of inpatient bed capacity are especially useful for capacity reservation purposes. As given, the remaining unoccupied beds at the end of each day, bed capacity of the next day can be obtained by examining the forecasts of the number of discharged patients during the next day. The features of fluctuations in daily discharges like trend, seasonal cycles, special-day effects, and autocorrelation complicate decision optimizing, while time-series models can capture these features well. This research compares three models: a model combining seasonal regression and ARIMA, a multiplicative seasonal ARIMA (MSARIMA) model, and a combinatorial model based on MSARIMA and weighted Markov Chain models in generating forecasts of daily discharges. The models are applied to three years of discharge data of an entire hospital. Several performance measures like the direction of the symmetry value, normalized mean squared error, and mean absolute percentage error are utilized to capture the under- and overprediction in model selection. The findings indicate that daily discharges can be forecast by using the proposed models. A number of important practical implications are discussed, such as the use of accurate forecasts in discharge planning, admission scheduling, and capacity reservation.

Coastal and Riverine Flood Forecast Model powered by ADCIRC

NASA Astrophysics Data System (ADS)

Khalid, A.; Ferreira, C.

2017-12-01

Coastal flooding is becoming a major threat to increased population in the coastal areas. To protect coastal communities from tropical storms & hurricane damages, early warning systems are being developed. These systems have the capability of real time flood forecasting to identify hazardous coastal areas and aid coastal communities in rescue operations. State of the art hydrodynamic models forced by atmospheric forcing have given modelers the ability to forecast storm surge, water levels and currents. This helps to identify the areas threatened by intense storms. Study on Chesapeake Bay area has gained national importance because of its combined riverine and coastal phenomenon, which leads to greater uncertainty in flood predictions. This study presents an automated flood forecast system developed by following Advanced Circulation (ADCIRC) Surge Guidance System (ASGS) guidelines and tailored to take in riverine and coastal boundary forcing, thus includes all the hydrodynamic processes to forecast total water in the Potomac River. As studies on tidal and riverine flow interaction are very scarce in number, our forecast system would be a scientific tool to examine such area and fill the gaps with precise prediction for Potomac River. Real-time observations from National Oceanic and Atmospheric Administration (NOAA) and field measurements have been used as model boundary feeding. The model performance has been validated by using major historical riverine and coastal flooding events. Hydrodynamic model ADCIRC produced promising predictions for flood inundation areas. As better forecasts can be achieved by using coupled models, this system is developed to take boundary conditions from Global WaveWatchIII for the research purposes. Wave and swell propagation will be fed through Global WavewatchIII model to take into account the effects of swells and currents. This automated forecast system is currently undergoing rigorous testing to include any missing parameters which might provide better and more reliable forecast for the flood affected communities.

Drought forecasting in Luanhe River basin involving climatic indices

NASA Astrophysics Data System (ADS)

Ren, Weinan; Wang, Yixuan; Li, Jianzhu; Feng, Ping; Smith, Ronald J.

2017-11-01

Drought is regarded as one of the most severe natural disasters globally. This is especially the case in Tianjin City, Northern China, where drought can affect economic development and people's livelihoods. Drought forecasting, the basis of drought management, is an important mitigation strategy. In this paper, we evolve a probabilistic forecasting model, which forecasts transition probabilities from a current Standardized Precipitation Index (SPI) value to a future SPI class, based on conditional distribution of multivariate normal distribution to involve two large-scale climatic indices at the same time, and apply the forecasting model to 26 rain gauges in the Luanhe River basin in North China. The establishment of the model and the derivation of the SPI are based on the hypothesis of aggregated monthly precipitation that is normally distributed. Pearson correlation and Shapiro-Wilk normality tests are used to select appropriate SPI time scale and large-scale climatic indices. Findings indicated that longer-term aggregated monthly precipitation, in general, was more likely to be considered normally distributed and forecasting models should be applied to each gauge, respectively, rather than to the whole basin. Taking Liying Gauge as an example, we illustrate the impact of the SPI time scale and lead time on transition probabilities. Then, the controlled climatic indices of every gauge are selected by Pearson correlation test and the multivariate normality of SPI, corresponding climatic indices for current month and SPI 1, 2, and 3 months later are demonstrated using Shapiro-Wilk normality test. Subsequently, we illustrate the impact of large-scale oceanic-atmospheric circulation patterns on transition probabilities. Finally, we use a score method to evaluate and compare the performance of the three forecasting models and compare them with two traditional models which forecast transition probabilities from a current to a future SPI class. The results show that the three proposed models outperform the two traditional models and involving large-scale climatic indices can improve the forecasting accuracy.

Multivariate time series modeling of short-term system scale irrigation demand

NASA Astrophysics Data System (ADS)

Perera, Kushan C.; Western, Andrew W.; George, Biju; Nawarathna, Bandara

2015-12-01

Travel time limits the ability of irrigation system operators to react to short-term irrigation demand fluctuations that result from variations in weather, including very hot periods and rainfall events, as well as the various other pressures and opportunities that farmers face. Short-term system-wide irrigation demand forecasts can assist in system operation. Here we developed a multivariate time series (ARMAX) model to forecast irrigation demands with respect to aggregated service points flows (IDCGi, ASP) and off take regulator flows (IDCGi, OTR) based across 5 command areas, which included area covered under four irrigation channels and the study area. These command area specific ARMAX models forecast 1-5 days ahead daily IDCGi, ASP and IDCGi, OTR using the real time flow data recorded at the service points and the uppermost regulators and observed meteorological data collected from automatic weather stations. The model efficiency and the predictive performance were quantified using the root mean squared error (RMSE), Nash-Sutcliffe model efficiency coefficient (NSE), anomaly correlation coefficient (ACC) and mean square skill score (MSSS). During the evaluation period, NSE for IDCGi, ASP and IDCGi, OTR across 5 command areas were ranged 0.98-0.78. These models were capable of generating skillful forecasts (MSSS ⩾ 0.5 and ACC ⩾ 0.6) of IDCGi, ASP and IDCGi, OTR for all 5 lead days and IDCGi, ASP and IDCGi, OTR forecasts were better than using the long term monthly mean irrigation demand. Overall these predictive performance from the ARMAX time series models were higher than almost all the previous studies we are aware. Further, IDCGi, ASP and IDCGi, OTR forecasts have improved the operators' ability to react for near future irrigation demand fluctuations as the developed ARMAX time series models were self-adaptive to reflect the short-term changes in the irrigation demand with respect to various pressures and opportunities that farmers' face, such as changing water policy, continued development of water markets, drought and changing technology.

Ensemble Data Assimilation for Streamflow Forecasting: Experiments with Ensemble Kalman Filter and Particle Filter

NASA Astrophysics Data System (ADS)

Hirpa, F. A.; Gebremichael, M.; Hopson, T. M.; Wojick, R.

2011-12-01

We present results of data assimilation of ground discharge observation and remotely sensed soil moisture observations into Sacramento Soil Moisture Accounting (SACSMA) model in a small watershed (1593 km2) in Minnesota, the Unites States. Specifically, we perform assimilation experiments with Ensemble Kalman Filter (EnKF) and Particle Filter (PF) in order to improve streamflow forecast accuracy at six hourly time step. The EnKF updates the soil moisture states in the SACSMA from the relative errors of the model and observations, while the PF adjust the weights of the state ensemble members based on the likelihood of the forecast. Results of the improvements of each filter over the reference model (without data assimilation) will be presented. Finally, the EnKF and PF are coupled together to further improve the streamflow forecast accuracy.

A deep learning framework for financial time series using stacked autoencoders and long-short term memory.

PubMed

Bao, Wei; Yue, Jun; Rao, Yulei

2017-01-01

The application of deep learning approaches to finance has received a great deal of attention from both investors and researchers. This study presents a novel deep learning framework where wavelet transforms (WT), stacked autoencoders (SAEs) and long-short term memory (LSTM) are combined for stock price forecasting. The SAEs for hierarchically extracted deep features is introduced into stock price forecasting for the first time. The deep learning framework comprises three stages. First, the stock price time series is decomposed by WT to eliminate noise. Second, SAEs is applied to generate deep high-level features for predicting the stock price. Third, high-level denoising features are fed into LSTM to forecast the next day's closing price. Six market indices and their corresponding index futures are chosen to examine the performance of the proposed model. Results show that the proposed model outperforms other similar models in both predictive accuracy and profitability performance.

Trends in the predictive performance of raw ensemble weather forecasts

NASA Astrophysics Data System (ADS)

Hemri, Stephan; Scheuerer, Michael; Pappenberger, Florian; Bogner, Konrad; Haiden, Thomas

2015-04-01

Over the last two decades the paradigm in weather forecasting has shifted from being deterministic to probabilistic. Accordingly, numerical weather prediction (NWP) models have been run increasingly as ensemble forecasting systems. The goal of such ensemble forecasts is to approximate the forecast probability distribution by a finite sample of scenarios. Global ensemble forecast systems, like the European Centre for Medium-Range Weather Forecasts (ECMWF) ensemble, are prone to probabilistic biases, and are therefore not reliable. They particularly tend to be underdispersive for surface weather parameters. Hence, statistical post-processing is required in order to obtain reliable and sharp forecasts. In this study we apply statistical post-processing to ensemble forecasts of near-surface temperature, 24-hour precipitation totals, and near-surface wind speed from the global ECMWF model. Our main objective is to evaluate the evolution of the difference in skill between the raw ensemble and the post-processed forecasts. The ECMWF ensemble is under continuous development, and hence its forecast skill improves over time. Parts of these improvements may be due to a reduction of probabilistic bias. Thus, we first hypothesize that the gain by post-processing decreases over time. Based on ECMWF forecasts from January 2002 to March 2014 and corresponding observations from globally distributed stations we generate post-processed forecasts by ensemble model output statistics (EMOS) for each station and variable. Parameter estimates are obtained by minimizing the Continuous Ranked Probability Score (CRPS) over rolling training periods that consist of the n days preceding the initialization dates. Given the higher average skill in terms of CRPS of the post-processed forecasts for all three variables, we analyze the evolution of the difference in skill between raw ensemble and EMOS forecasts. The fact that the gap in skill remains almost constant over time, especially for near-surface wind speed, suggests that improvements to the atmospheric model have an effect quite different from what calibration by statistical post-processing is doing. That is, they are increasing potential skill. Thus this study indicates that (a) further model development is important even if one is just interested in point forecasts, and (b) statistical post-processing is important because it will keep adding skill in the foreseeable future.

The SPoRT-WRF: Evaluating the Impact of NASA Datasets on Convective Forecasts

NASA Technical Reports Server (NTRS)

Zavodsky, Bradley; Case, Jonathan; Kozlowski, Danielle; Molthan, Andrew

2012-01-01

The Short-term Prediction Research and Transition Center (SPoRT) is a collaborative partnership between NASA and operational forecasting entities, including a number of National Weather Service offices. SPoRT transitions real-time NASA products and capabilities to its partners to address specific operational forecast challenges. One challenge that forecasters face is applying convection-allowing numerical models to predict mesoscale convective weather. In order to address this specific forecast challenge, SPoRT produces real-time mesoscale model forecasts using the Weather Research and Forecasting (WRF) model that includes unique NASA products and capabilities. Currently, the SPoRT configuration of the WRF model (SPoRT-WRF) incorporates the 4-km Land Information System (LIS) land surface data, 1-km SPoRT sea surface temperature analysis and 1-km Moderate resolution Imaging Spectroradiometer (MODIS) greenness vegetation fraction (GVF) analysis, and retrieved thermodynamic profiles from the Atmospheric Infrared Sounder (AIRS). The LIS, SST, and GVF data are all integrated into the SPoRT-WRF through adjustments to the initial and boundary conditions, and the AIRS data are assimilated into a 9-hour SPoRT WRF forecast each day at 0900 UTC. This study dissects the overall impact of the NASA datasets and the individual surface and atmospheric component datasets on daily mesoscale forecasts. A case study covering the super tornado outbreak across the Ce ntral and Southeastern United States during 25-27 April 2011 is examined. Three different forecasts are analyzed including the SPoRT-WRF (NASA surface and atmospheric data), the SPoRT WRF without AIRS (NASA surface data only), and the operational National Severe Storms Laboratory (NSSL) WRF (control with no NASA data). The forecasts are compared qualitatively by examining simulated versus observed radar reflectivity. Differences between the simulated reflectivity are further investigated using convective parameters along with model soundings to determine the impacts of the various NASA datasets. Additionally, quantitative evaluation of select meteorological parameters is performed using the Meteorological Evaluation Tools model verification package to compare forecasts to in situ surface and upper air observations.

Do quantitative decadal forecasts from GCMs provide decision relevant skill?

NASA Astrophysics Data System (ADS)

Suckling, E. B.; Smith, L. A.

2012-04-01

It is widely held that only physics-based simulation models can capture the dynamics required to provide decision-relevant probabilistic climate predictions. This fact in itself provides no evidence that predictions from today's GCMs are fit for purpose. Empirical (data-based) models are employed to make probability forecasts on decadal timescales, where it is argued that these 'physics free' forecasts provide a quantitative 'zero skill' target for the evaluation of forecasts based on more complicated models. It is demonstrated that these zero skill models are competitive with GCMs on decadal scales for probability forecasts evaluated over the last 50 years. Complications of statistical interpretation due to the 'hindcast' nature of this experiment, and the likely relevance of arguments that the lack of hindcast skill is irrelevant as the signal will soon 'come out of the noise' are discussed. A lack of decision relevant quantiative skill does not bring the science-based insights of anthropogenic warming into doubt, but it does call for a clear quantification of limits, as a function of lead time, for spatial and temporal scales on which decisions based on such model output are expected to prove maladaptive. Failing to do so may risk the credibility of science in support of policy in the long term. The performance amongst a collection of simulation models is evaluated, having transformed ensembles of point forecasts into probability distributions through the kernel dressing procedure [1], according to a selection of proper skill scores [2] and contrasted with purely data-based empirical models. Data-based models are unlikely to yield realistic forecasts for future climate change if the Earth system moves away from the conditions observed in the past, upon which the models are constructed; in this sense the empirical model defines zero skill. When should a decision relevant simulation model be expected to significantly outperform such empirical models? Probability forecasts up to ten years ahead (decadal forecasts) are considered, both on global and regional spatial scales for surface air temperature. Such decadal forecasts are not only important in terms of providing information on the impacts of near-term climate change, but also from the perspective of climate model validation, as hindcast experiments and a sufficient database of historical observations allow standard forecast verification methods to be used. Simulation models from the ENSEMBLES hindcast experiment [3] are evaluated and contrasted with static forecasts of the observed climatology, persistence forecasts and against simple statistical models, called dynamic climatology (DC). It is argued that DC is a more apropriate benchmark in the case of a non-stationary climate. It is found that the ENSEMBLES models do not demonstrate a significant increase in skill relative to the empirical models even at global scales over any lead time up to a decade ahead. It is suggested that the contsruction and co-evaluation with the data-based models become a regular component of the reporting of large simulation model forecasts. The methodology presented may easily be adapted to other forecasting experiments and is expected to influence the design of future experiments. The inclusion of comparisons with dynamic climatology and other data-based approaches provide important information to both scientists and decision makers on which aspects of state-of-the-art simulation forecasts are likely to be fit for purpose. [1] J. Bröcker and L. A. Smith. From ensemble forecasts to predictive distributions, Tellus A, 60(4), 663-678 (2007). [2] J. Bröcker and L. A. Smith. Scoring probabilistic forecasts: The importance of being proper, Weather and Forecasting, 22, 382-388 (2006). [3] F. J. Doblas-Reyes, A. Weisheimer, T. N. Palmer, J. M. Murphy and D. Smith. Forecast quality asessment of the ENSEMBLES seasonal-to-decadal stream 2 hindcasts, ECMWF Technical Memorandum, 621 (2010).

Entropy Econometrics for combining regional economic forecasts: A Data-Weighted Prior Estimator

NASA Astrophysics Data System (ADS)

Fernández-Vázquez, Esteban; Moreno, Blanca

2017-10-01

Forecast combination has been studied in econometrics for a long time, and the literature has shown the superior performance of forecast combination over individual predictions. However, there is still controversy on which is the best procedure to specify the forecast weights. This paper explores the possibility of using a procedure based on Entropy Econometrics, which allows setting the weights for the individual forecasts as a mixture of different alternatives. In particular, we examine the ability of the Data-Weighted Prior Estimator proposed by Golan (J Econom 101(1):165-193, 2001) to combine forecasting models in a context of small sample sizes, a relative common scenario when dealing with time series for regional economies. We test the validity of the proposed approach using a simulation exercise and a real-world example that aims at predicting gross regional product growth rates for a regional economy. The forecasting performance of the Data-Weighted Prior Estimator proposed is compared with other combining methods. The simulation results indicate that in scenarios of heavily ill-conditioned datasets the approach suggested dominates other forecast combination strategies. The empirical results are consistent with the conclusions found in the numerical experiment.

Crude oil price forecasting based on hybridizing wavelet multiple linear regression model, particle swarm optimization techniques, and principal component analysis.

PubMed

Shabri, Ani; Samsudin, Ruhaidah

2014-01-01

Crude oil prices do play significant role in the global economy and are a key input into option pricing formulas, portfolio allocation, and risk measurement. In this paper, a hybrid model integrating wavelet and multiple linear regressions (MLR) is proposed for crude oil price forecasting. In this model, Mallat wavelet transform is first selected to decompose an original time series into several subseries with different scale. Then, the principal component analysis (PCA) is used in processing subseries data in MLR for crude oil price forecasting. The particle swarm optimization (PSO) is used to adopt the optimal parameters of the MLR model. To assess the effectiveness of this model, daily crude oil market, West Texas Intermediate (WTI), has been used as the case study. Time series prediction capability performance of the WMLR model is compared with the MLR, ARIMA, and GARCH models using various statistics measures. The experimental results show that the proposed model outperforms the individual models in forecasting of the crude oil prices series.

Crude Oil Price Forecasting Based on Hybridizing Wavelet Multiple Linear Regression Model, Particle Swarm Optimization Techniques, and Principal Component Analysis

PubMed Central

Shabri, Ani; Samsudin, Ruhaidah

2014-01-01

Crude oil prices do play significant role in the global economy and are a key input into option pricing formulas, portfolio allocation, and risk measurement. In this paper, a hybrid model integrating wavelet and multiple linear regressions (MLR) is proposed for crude oil price forecasting. In this model, Mallat wavelet transform is first selected to decompose an original time series into several subseries with different scale. Then, the principal component analysis (PCA) is used in processing subseries data in MLR for crude oil price forecasting. The particle swarm optimization (PSO) is used to adopt the optimal parameters of the MLR model. To assess the effectiveness of this model, daily crude oil market, West Texas Intermediate (WTI), has been used as the case study. Time series prediction capability performance of the WMLR model is compared with the MLR, ARIMA, and GARCH models using various statistics measures. The experimental results show that the proposed model outperforms the individual models in forecasting of the crude oil prices series. PMID:24895666

Cloud Impacts on Pavement Temperature in Energy Balance Models

NASA Astrophysics Data System (ADS)

Walker, C. L.

2013-12-01

Forecast systems provide decision support for end-users ranging from the solar energy industry to municipalities concerned with road safety. Pavement temperature is an important variable when considering vehicle response to various weather conditions. A complex, yet direct relationship exists between tire and pavement temperatures. Literature has shown that as tire temperature increases, friction decreases which affects vehicle performance. Many forecast systems suffer from inaccurate radiation forecasts resulting in part from the inability to model different types of clouds and their influence on radiation. This research focused on forecast improvement by determining how cloud type impacts the amount of shortwave radiation reaching the surface and subsequent pavement temperatures. The study region was the Great Plains where surface solar radiation data were obtained from the High Plains Regional Climate Center's Automated Weather Data Network stations. Road pavement temperature data were obtained from the Meteorological Assimilation Data Ingest System. Cloud properties and radiative transfer quantities were obtained from the Clouds and Earth's Radiant Energy System mission via Aqua and Terra Moderate Resolution Imaging Spectroradiometer satellite products. An additional cloud data set was incorporated from the Naval Research Laboratory Cloud Classification algorithm. Statistical analyses using a modified nearest neighbor approach were first performed relating shortwave radiation variability with road pavement temperature fluctuations. Then statistical associations were determined between the shortwave radiation and cloud property data sets. Preliminary results suggest that substantial pavement forecasting improvement is possible with the inclusion of cloud-specific information. Future model sensitivity testing seeks to quantify the magnitude of forecast improvement.

Forecasting wetting and drying of post-wildfire soils in response to precipitation: A time series optimization approach

NASA Astrophysics Data System (ADS)

Basak, A.; Kulkarni, C.; Schmidt, K. M.; Mengshoel, O. J.

2015-12-01

Volumetric water content (VWC) in soils is critical for forecasting thresholds for runoff-driven erosion caused by rainfall. Even though theoretical relations (e.g., Richards equation) have been developed to quantify VWC in unsaturated granular soils, site-specific field conditions and hysteresis of suction and VWC in soil preclude their direct use. Although attempts have previously been made to forecast VWC using various time-series models (e.g., autoregressive integrated moving average or ARIMA), these approaches lack hydrologic foundations and perform poorly when used to forecast VWC over time periods longer than 24 hours. In this work, we extend an existing Antecedent Water Index (AWI) based model to express VWC as a function of time and rainfall. AWI models typically overfit data and cannot be used for forecast VWC over long time periods. We developed a new model to overcome this limitation, which accumulates rainfall over a time window and fits a diverse range of wetting and drying curves. Hydraulic redistribution parameters in this model bear resemblance to hydrologic processes driven by gravity and suction. This model reasonably forecasts VWC using only initial VWC values and rainfall forecasts. Experimental VWC data were collected from steep gradient post-wildfire sites in southern California. Rapid landscape change was observed in response to small to moderate rain storms. We formulated a mean-squared error minimization problem over the model parameters and optimized using genetic algorithms. We found that our model fits VWC data for 3 distinct soil textures, each occurring at 3 different depths below the ground surface (5 cm, 15 cm, and 30 cm). Our model successfully forecasts VWC trends, such as drying and wetting rate. To a certain extent, our model achieves spatial and seasonal generalizability. Our accumulative rainfall model is also applicable to continuous predictions, where VWC values are repeatedly used to predict future ones within a 12-hr time frame.

Portals for Real-Time Earthquake Data and Forecasting: Challenge and Promise (Invited)

NASA Astrophysics Data System (ADS)

Rundle, J. B.; Holliday, J. R.; Graves, W. R.; Feltstykket, R.; Donnellan, A.; Glasscoe, M. T.

2013-12-01

Earthquake forecasts have been computed by a variety of countries world-wide for over two decades. For the most part, forecasts have been computed for insurance, reinsurance and underwriters of catastrophe bonds. However, recent events clearly demonstrate that mitigating personal risk is becoming the responsibility of individual members of the public. Open access to a variety of web-based forecasts, tools, utilities and information is therefore required. Portals for data and forecasts present particular challenges, and require the development of both apps and the client/server architecture to deliver the basic information in real time. The basic forecast model we consider is the Natural Time Weibull (NTW) method (JBR et al., Phys. Rev. E, 86, 021106, 2012). This model uses small earthquakes (';seismicity-based models') to forecast the occurrence of large earthquakes, via data-mining algorithms combined with the ANSS earthquake catalog. This method computes large earthquake probabilities using the number of small earthquakes that have occurred in a region since the last large earthquake. Localizing these forecasts in space so that global forecasts can be computed in real time presents special algorithmic challenges, which we describe in this talk. Using 25 years of data from the ANSS California-Nevada catalog of earthquakes, we compute real-time global forecasts at a grid scale of 0.1o. We analyze and monitor the performance of these models using the standard tests, which include the Reliability/Attributes and Receiver Operating Characteristic (ROC) tests. It is clear from much of the analysis that data quality is a major limitation on the accurate computation of earthquake probabilities. We discuss the challenges of serving up these datasets over the web on web-based platforms such as those at www.quakesim.org , www.e-decider.org , and www.openhazards.com.

Impact of scatterometer wind (ASCAT-A/B) data assimilation on semi real-time forecast system at KIAPS

NASA Astrophysics Data System (ADS)

Han, H. J.; Kang, J. H.

2016-12-01

Since Jul. 2015, KIAPS (Korea Institute of Atmospheric Prediction Systems) has been performing the semi real-time forecast system to assess the performance of their forecast system as a NWP model. KPOP (KIAPS Protocol for Observation Processing) is a part of KIAPS data assimilation system and has been performing well in KIAPS semi real-time forecast system. In this study, due to the fact that KPOP would be able to treat the scatterometer wind data, we analyze the effect of scatterometer wind (ASCAT-A/B) on KIAPS semi real-time forecast system. O-B global distribution and statistics of scatterometer wind give use two information which are the difference between background field and observation is not too large and KPOP processed the scatterometer wind data well. The changes of analysis increment because of O-B global distribution appear remarkably at the bottom of atmospheric field. It also shows that scatterometer wind data cover wide ocean where data would be able to short. Performance of scatterometer wind data can be checked through the vertical error reduction against IFS between background and analysis field and vertical statistics of O-A. By these analysis result, we can notice that scatterometer wind data will influence the positive effect on lower level performance of semi real-time forecast system at KIAPS. After, long-term result based on effect of scatterometer wind data will be analyzed.

Evaluation of the CFSv2 CMIP5 decadal predictions

NASA Astrophysics Data System (ADS)

Bombardi, Rodrigo J.; Zhu, Jieshun; Marx, Lawrence; Huang, Bohua; Chen, Hua; Lu, Jian; Krishnamurthy, Lakshmi; Krishnamurthy, V.; Colfescu, Ioana; Kinter, James L.; Kumar, Arun; Hu, Zeng-Zhen; Moorthi, Shrinivas; Tripp, Patrick; Wu, Xingren; Schneider, Edwin K.

2015-01-01

Retrospective decadal forecasts were undertaken using the Climate Forecast System version 2 (CFSv2) as part of Coupled Model Intercomparison Project 5. Decadal forecasts were performed separately by the National Center for Environmental Prediction (NCEP) and by the Center for Ocean-Land-Atmosphere Studies (COLA), with the centers using two different analyses for the ocean initial conditions the NCEP Climate Forecast System Reanalysis (CFSR) and the NEMOVAR-COMBINE analysis. COLA also examined the sensitivity to the inclusion of forcing by specified volcanic aerosols. Biases in the CFSv2 for both sets of initial conditions include cold midlatitude sea surface temperatures, and rapid melting of sea ice associated with warm polar oceans. Forecasts from the NEMOVAR-COMBINE analysis showed strong weakening of the Atlantic Meridional Overturning Circulation (AMOC), eventually approaching the weaker AMOC associated with CFSR. The decadal forecasts showed high predictive skill over the Indian, the western Pacific, and the Atlantic Oceans and low skill over the central and eastern Pacific. The volcanic forcing shows only small regional differences in predictability of surface temperature at 2m (T2m) in comparison to forecasts without volcanic forcing, especially over the Indian Ocean. An ocean heat content (OHC) budget analysis showed that the OHC has substantial memory, indicating potential for the decadal predictability of T2m; however, the model has a systematic drift in global mean OHC. The results suggest that the reduction of model biases may be the most productive path towards improving the model's decadal forecasts.

Japanese earthquake predictability experiment with multiple runs before and after the 2011 Tohoku-oki earthquake

NASA Astrophysics Data System (ADS)

Hirata, N.; Tsuruoka, H.; Yokoi, S.

2011-12-01

The current Japanese national earthquake prediction program emphasizes the importance of modeling as well as monitoring for a sound scientific development of earthquake prediction research. One major focus of the current program is to move toward creating testable earthquake forecast models. For this purpose, in 2009 we joined the Collaboratory for the Study of Earthquake Predictability (CSEP) and installed, through an international collaboration, the CSEP Testing Centre, an infrastructure to encourage researchers to develop testable models for Japan. We started Japanese earthquake predictability experiment on November 1, 2009. The experiment consists of 12 categories, with 4 testing classes with different time spans (1 day, 3 months, 1 year and 3 years) and 3 testing regions called 'All Japan,' 'Mainland,' and 'Kanto.' A total of 160 models, as of August 2013, were submitted, and are currently under the CSEP official suite of tests for evaluating the performance of forecasts. We will present results of prospective forecast and testing for periods before and after the 2011 Tohoku-oki earthquake. Because a seismic activity has changed dramatically since the 2011 event, performances of models have been affected very much. In addition, as there is the problem of authorized catalogue related to the completeness magnitude, most models did not pass the CSEP consistency tests. Also, we will discuss the retrospective earthquake forecast experiments for aftershocks of the 2011 Tohoku-oki earthquake. Our aim is to describe what has turned out to be the first occasion for setting up a research environment for rigorous earthquake forecasting in Japan.

Japanese earthquake predictability experiment with multiple runs before and after the 2011 Tohoku-oki earthquake

NASA Astrophysics Data System (ADS)

Hirata, N.; Tsuruoka, H.; Yokoi, S.

2013-12-01

The current Japanese national earthquake prediction program emphasizes the importance of modeling as well as monitoring for a sound scientific development of earthquake prediction research. One major focus of the current program is to move toward creating testable earthquake forecast models. For this purpose, in 2009 we joined the Collaboratory for the Study of Earthquake Predictability (CSEP) and installed, through an international collaboration, the CSEP Testing Centre, an infrastructure to encourage researchers to develop testable models for Japan. We started Japanese earthquake predictability experiment on November 1, 2009. The experiment consists of 12 categories, with 4 testing classes with different time spans (1 day, 3 months, 1 year and 3 years) and 3 testing regions called 'All Japan,' 'Mainland,' and 'Kanto.' A total of 160 models, as of August 2013, were submitted, and are currently under the CSEP official suite of tests for evaluating the performance of forecasts. We will present results of prospective forecast and testing for periods before and after the 2011 Tohoku-oki earthquake. Because a seismic activity has changed dramatically since the 2011 event, performances of models have been affected very much. In addition, as there is the problem of authorized catalogue related to the completeness magnitude, most models did not pass the CSEP consistency tests. Also, we will discuss the retrospective earthquake forecast experiments for aftershocks of the 2011 Tohoku-oki earthquake. Our aim is to describe what has turned out to be the first occasion for setting up a research environment for rigorous earthquake forecasting in Japan.

An interdisciplinary approach for earthquake modelling and forecasting

NASA Astrophysics Data System (ADS)

Han, P.; Zhuang, J.; Hattori, K.; Ogata, Y.

2016-12-01

Earthquake is one of the most serious disasters, which may cause heavy casualties and economic losses. Especially in the past two decades, huge/mega earthquakes have hit many countries. Effective earthquake forecasting (including time, location, and magnitude) becomes extremely important and urgent. To date, various heuristically derived algorithms have been developed for forecasting earthquakes. Generally, they can be classified into two types: catalog-based approaches and non-catalog-based approaches. Thanks to the rapid development of statistical seismology in the past 30 years, now we are able to evaluate the performances of these earthquake forecast approaches quantitatively. Although a certain amount of precursory information is available in both earthquake catalogs and non-catalog observations, the earthquake forecast is still far from satisfactory. In most case, the precursory phenomena were studied individually. An earthquake model that combines self-exciting and mutually exciting elements was developed by Ogata and Utsu from the Hawkes process. The core idea of this combined model is that the status of the event at present is controlled by the event itself (self-exciting) and all the external factors (mutually exciting) in the past. In essence, the conditional intensity function is a time-varying Poisson process with rate λ(t), which is composed of the background rate, the self-exciting term (the information from past seismic events), and the external excitation term (the information from past non-seismic observations). This model shows us a way to integrate the catalog-based forecast and non-catalog-based forecast. Against this background, we are trying to develop a new earthquake forecast model which combines catalog-based and non-catalog-based approaches.

Obtaining high-resolution stage forecasts by coupling large-scale hydrologic models with sensor data

NASA Astrophysics Data System (ADS)

Fries, K. J.; Kerkez, B.

2017-12-01

We investigate how "big" quantities of distributed sensor data can be coupled with a large-scale hydrologic model, in particular the National Water Model (NWM), to obtain hyper-resolution forecasts. The recent launch of the NWM provides a great example of how growing computational capacity is enabling a new generation of massive hydrologic models. While the NWM spans an unprecedented spatial extent, there remain many questions about how to improve forecast at the street-level, the resolution at which many stakeholders make critical decisions. Further, the NWM runs on supercomputers, so water managers who may have access to their own high-resolution measurements may not readily be able to assimilate them into the model. To that end, we ask the question: how can the advances of the large-scale NWM be coupled with new local observations to enable hyper-resolution hydrologic forecasts? A methodology is proposed whereby the flow forecasts of the NWM are directly mapped to high-resolution stream levels using Dynamical System Identification. We apply the methodology across a sensor network of 182 gages in Iowa. Of these sites, approximately one third have shown to perform well in high-resolution flood forecasting when coupled with the outputs of the NWM. The quality of these forecasts is characterized using Principal Component Analysis and Random Forests to identify where the NWM may benefit from new sources of local observations. We also discuss how this approach can help municipalities identify where they should place low-cost sensors to most benefit from flood forecasts of the NWM.

Adaptive use of research aircraft data sets for hurricane forecasts

NASA Astrophysics Data System (ADS)

Biswas, M. K.; Krishnamurti, T. N.

2008-02-01

This study uses an adaptive observational strategy for hurricane forecasting. It shows the impacts of Lidar Atmospheric Sensing Experiment (LASE) and dropsonde data sets from Convection and Moisture Experiment (CAMEX) field campaigns on hurricane track and intensity forecasts. The following cases are used in this study: Bonnie, Danielle and Georges of 1998 and Erin, Gabrielle and Humberto of 2001. A single model run for each storm is carried out using the Florida State University Global Spectral Model (FSUGSM) with the European Center for Medium Range Weather Forecasts (ECMWF) analysis as initial conditions, in addition to 50 other model runs where the analysis is randomly perturbed for each storm. The centers of maximum variance of the DLM heights are located from the forecast error variance fields at the 84-hr forecast. Back correlations are then performed using the centers of these maximum variances and the fields at the 36-hr forecast. The regions having the highest correlations in the vicinity of the hurricanes are indicative of regions from where the error growth emanates and suggests the need for additional observations. Data sets are next assimilated in those areas that contain high correlations. Forecasts are computed using the new initial conditions for the storm cases, and track and intensity skills are then examined with respect to the control forecast. The adaptive strategy is capable of identifying sensitive areas where additional observations can help in reducing the hurricane track forecast errors. A reduction of position error by approximately 52% for day 3 of forecast (averaged over 7 storm cases) over the control runs is observed. The intensity forecast shows only a slight positive impact due to the model’s coarse resolution.

Performance of Trajectory Models with Wind Uncertainty

NASA Technical Reports Server (NTRS)

Lee, Alan G.; Weygandt, Stephen S.; Schwartz, Barry; Murphy, James R.

2009-01-01

Typical aircraft trajectory predictors use wind forecasts but do not account for the forecast uncertainty. A method for generating estimates of wind prediction uncertainty is described and its effect on aircraft trajectory prediction uncertainty is investigated. The procedure for estimating the wind prediction uncertainty relies uses a time-lagged ensemble of weather model forecasts from the hourly updated Rapid Update Cycle (RUC) weather prediction system. Forecast uncertainty is estimated using measures of the spread amongst various RUC time-lagged ensemble forecasts. This proof of concept study illustrates the estimated uncertainty and the actual wind errors, and documents the validity of the assumed ensemble-forecast accuracy relationship. Aircraft trajectory predictions are made using RUC winds with provision for the estimated uncertainty. Results for a set of simulated flights indicate this simple approach effectively translates the wind uncertainty estimate into an aircraft trajectory uncertainty. A key strength of the method is the ability to relate uncertainty to specific weather phenomena (contained in the various ensemble members) allowing identification of regional variations in uncertainty.

A research model--forecasting incident rates from optimized safety program intervention strategies.

PubMed

Iyer, P S; Haight, J M; Del Castillo, E; Tink, B W; Hawkins, P W

2005-01-01

INTRODUCTION/PROBLEM: Property damage incidents, workplace injuries, and safety programs designed to prevent them, are expensive aspects of doing business in contemporary industry. The National Safety Council (2002) estimated that workplace injuries cost $146.6 billion per year. Because companies are resource limited, optimizing intervention strategies to decrease incidents with less costly programs can contribute to improved productivity. Systematic data collection methods were employed and the forecasting ability of a time-lag relationship between interventions and incident rates was studied using various statistical methods (an intervention is not expected to have an immediate nor infinitely lasting effect on the incident rate). As a follow up to the initial work, researchers developed two models designed to forecast incident rates. One is based on past incident rate performance and the other on the configuration and level of effort applied to the safety and health program. Researchers compared actual incident performance to the prediction capability of each model over 18 months in the forestry operations at an electricity distribution company and found the models to allow accurate prediction of incident rates. These models potentially have powerful implications as a business-planning tool for human resource allocation and for designing an optimized safety and health intervention program to minimize incidents. Depending on the mathematical relationship, one can determine what interventions, where and how much to apply them, and when to increase or reduce human resource input as determined by the forecasted performance.

The FireWork air quality forecast system with near-real-time biomass burning emissions: Recent developments and evaluation of performance for the 2015 North American wildfire season.

PubMed

Pavlovic, Radenko; Chen, Jack; Anderson, Kerry; Moran, Michael D; Beaulieu, Paul-André; Davignon, Didier; Cousineau, Sophie

2016-09-01

Environment and Climate Change Canada's FireWork air quality (AQ) forecast system for North America with near-real-time biomass burning emissions has been running experimentally during the Canadian wildfire season since 2013. The system runs twice per day with model initializations at 00 UTC and 12 UTC, and produces numerical AQ forecast guidance with 48-hr lead time. In this work we describe the FireWork system, which incorporates near-real-time biomass burning emissions based on the Canadian Wildland Fire Information System (CWFIS) as an input to the operational Regional Air Quality Deterministic Prediction System (RAQDPS). To demonstrate the capability of the system we analyzed two forecast periods in 2015 (June 2-July 15, and August 15-31) when fire activity was high, and observed fire-smoke-impacted areas in western Canada and the western United States. Modeled PM2.5 surface concentrations were compared with surface measurements and benchmarked with results from the operational RAQDPS, which did not consider near-real-time biomass burning emissions. Model performance statistics showed that FireWork outperformed RAQDPS with improvements in forecast hourly PM2.5 across the region; the results were especially significant for stations near the path of fire plume trajectories. Although the hourly PM2.5 concentrations predicted by FireWork still displayed bias for areas with active fires for these two periods (mean bias [MB] of -7.3 µg m(-3) and 3.1 µg m(-3)), it showed better forecast skill than the RAQDPS (MB of -11.7 µg m(-3) and -5.8 µg m(-3)) and demonstrated a greater ability to capture temporal variability of episodic PM2.5 events (correlation coefficient values of 0.50 and 0.69 for FireWork compared to 0.03 and 0.11 for RAQDPS). A categorical forecast comparison based on an hourly PM2.5 threshold of 30 µg m(-3) also showed improved scores for probability of detection (POD), critical success index (CSI), and false alarm rate (FAR). Smoke from wildfires can have a large impact on regional air quality (AQ) and can expose populations to elevated pollution levels. Environment and Climate Change Canada has been producing operational air quality forecasts for all of Canada since 2009 and is now working to include near-real-time wildfire emissions (NRTWE) in its operational AQ forecasting system. An experimental forecast system named FireWork, which includes NRTWE, has been undergoing testing and evaluation since 2013. A performance analysis of FireWork forecasts for the 2015 wildfire season shows that FireWork provides significant improvements to surface PM2.5 forecasts and valuable guidance to regional forecasters and first responders.

Validation of Volcanic Ash Forecasting Performed by the Washington Volcanic Ash Advisory Center

NASA Astrophysics Data System (ADS)

Salemi, A.; Hanna, J.

2009-12-01

In support of NOAA’s mission to protect life and property, the Satellite Analysis Branch (SAB) uses satellite imagery to monitor volcanic eruptions and track volcanic ash. The Washington Volcanic Ash Advisory Center (VAAC) was established in late 1997 through an agreement with the International Civil Aviation Organization (ICAO). A volcanic ash advisory (VAA) is issued every 6 hours while an eruption is occurring. Information about the current location and height of the volcanic ash as well as any pertinent meteorological information is contained within the VAA. In addition, when ash is detected in satellite imagery, 6-, 12- and 18-hour forecasts of ash height and location are provided. This information is garnered from many sources including Meteorological Watch Offices (MWOs), pilot reports (PIREPs), model forecast winds, radiosondes and volcano observatories. The Washington VAAC has performed a validation of their 6, 12 and 18 hour airborne volcanic ash forecasts issued since October, 2007. The volcanic ash forecasts are viewed dichotomously (yes/no) with the frequency of yes and no events placed into a contingency table. A large variety of categorical statistics useful in describing forecast performance are then computed from the resulting contingency table.

Time series modeling and forecasting using memetic algorithms for regime-switching models.

PubMed

Bergmeir, Christoph; Triguero, Isaac; Molina, Daniel; Aznarte, José Luis; Benitez, José Manuel

2012-11-01

In this brief, we present a novel model fitting procedure for the neuro-coefficient smooth transition autoregressive model (NCSTAR), as presented by Medeiros and Veiga. The model is endowed with a statistically founded iterative building procedure and can be interpreted in terms of fuzzy rule-based systems. The interpretability of the generated models and a mathematically sound building procedure are two very important properties of forecasting models. The model fitting procedure employed by the original NCSTAR is a combination of initial parameter estimation by a grid search procedure with a traditional local search algorithm. We propose a different fitting procedure, using a memetic algorithm, in order to obtain more accurate models. An empirical evaluation of the method is performed, applying it to various real-world time series originating from three forecasting competitions. The results indicate that we can significantly enhance the accuracy of the models, making them competitive to models commonly used in the field.

Short-Term Forecasts Using NU-WRF for the Winter Olympics 2018

NASA Technical Reports Server (NTRS)

Srikishen, Jayanthi; Case, Jonathan L.; Petersen, Walter A.; Iguchi, Takamichi; Tao, Wei-Kuo; Zavodsky, Bradley T.; Molthan, Andrew

2017-01-01

The NASA Unified-Weather Research and Forecasting model (NU-WRF) will be included for testing and evaluation in the forecast demonstration project (FDP) of the International Collaborative Experiment -PyeongChang 2018 Olympic and Paralympic (ICE-POP) Winter Games. An international array of radar and supporting ground based observations together with various forecast and now-cast models will be operational during ICE-POP. In conjunction with personnel from NASA's Goddard Space Flight Center, the NASA Short-term Prediction Research and Transition (SPoRT) Center is developing benchmark simulations for a real-time NU-WRF configuration to run during the FDP. ICE-POP observational datasets will be used to validate model simulations and investigate improved model physics and performance for prediction of snow events during the research phase (RDP) of the project The NU-WRF model simulations will also support NASA Global Precipitation Measurement (GPM) Mission ground-validation physical and direct validation activities in relation to verifying, testing and improving satellite-based snowfall retrieval algorithms over complex terrain.

Forecasting currency circulation data of Bank Indonesia by using hybrid ARIMAX-ANN model

NASA Astrophysics Data System (ADS)

Prayoga, I. Gede Surya Adi; Suhartono, Rahayu, Santi Puteri

2017-05-01

The purpose of this study is to forecast currency inflow and outflow data of Bank Indonesia. Currency circulation in Indonesia is highly influenced by the presence of Eid al-Fitr. One way to forecast the data with Eid al-Fitr effect is using autoregressive integrated moving average with exogenous input (ARIMAX) model. However, ARIMAX is a linear model, which cannot handle nonlinear correlation structures of the data. In the field of forecasting, inaccurate predictions can be considered caused by the existence of nonlinear components that are uncaptured by the model. In this paper, we propose a hybrid model of ARIMAX and artificial neural networks (ANN) that can handle both linear and nonlinear correlation. This method was applied for 46 series of currency inflow and 46 series of currency outflow. The results showed that based on out-of-sample root mean squared error (RMSE), the hybrid models are up to10.26 and 10.65 percent better than ARIMAX for inflow and outflow series, respectively. It means that ANN performs well in modeling nonlinear correlation of the data and can increase the accuracy of linear model.

Reservoir inflow forecasting with a modified coactive neuro-fuzzy inference system: a case study for a semi-arid region

NASA Astrophysics Data System (ADS)

Allawi, Mohammed Falah; Jaafar, Othman; Mohamad Hamzah, Firdaus; Mohd, Nuruol Syuhadaa; Deo, Ravinesh C.; El-Shafie, Ahmed

2017-10-01

Existing forecast models applied for reservoir inflow forecasting encounter several drawbacks, due to the difficulty of the underlying mathematical procedures being to cope with and to mimic the naturalization and stochasticity of the inflow data patterns. In this study, appropriate adjustments to the conventional coactive neuro-fuzzy inference system (CANFIS) method are proposed to improve the mathematical procedure, thus enabling a better detection of the high nonlinearity patterns found in the reservoir inflow training data. This modification includes the updating of the back propagation algorithm, leading to a consequent update of the membership rules and the induction of the centre-weighted set rather than the global weighted set used in feature extraction. The modification also aids in constructing an integrated model that is able to not only detect the nonlinearity in the training data but also the wide range of features within the training data records used to simulate the forecasting model. To demonstrate the model's efficacy, the proposed CANFIS method has been applied to forecast monthly inflow data at Aswan High Dam (AHD), located in southern Egypt. Comparative analyses of the forecasting skill of the modified CANFIS and the conventional ANFIS model are carried out with statistical score indicators to assess the reliability of the developed method. The statistical metrics support the better performance of the developed CANFIS model, which significantly outperforms the ANFIS model to attain a low relative error value (23%), mean absolute error (1.4 BCM month-1), root mean square error (1.14 BCM month-1), and a relative large coefficient of determination (0.94). The present study ascertains the better utility of the modified CANFIS model in respect to the traditional ANFIS model applied in reservoir inflow forecasting for a semi-arid region.

The assisted prediction modelling frame with hybridisation and ensemble for business risk forecasting and an implementation

NASA Astrophysics Data System (ADS)

Li, Hui; Hong, Lu-Yao; Zhou, Qing; Yu, Hai-Jie

2015-08-01

The business failure of numerous companies results in financial crises. The high social costs associated with such crises have made people to search for effective tools for business risk prediction, among which, support vector machine is very effective. Several modelling means, including single-technique modelling, hybrid modelling, and ensemble modelling, have been suggested in forecasting business risk with support vector machine. However, existing literature seldom focuses on the general modelling frame for business risk prediction, and seldom investigates performance differences among different modelling means. We reviewed researches on forecasting business risk with support vector machine, proposed the general assisted prediction modelling frame with hybridisation and ensemble (APMF-WHAE), and finally, investigated the use of principal components analysis, support vector machine, random sampling, and group decision, under the general frame in forecasting business risk. Under the APMF-WHAE frame with support vector machine as the base predictive model, four specific predictive models were produced, namely, pure support vector machine, a hybrid support vector machine involved with principal components analysis, a support vector machine ensemble involved with random sampling and group decision, and an ensemble of hybrid support vector machine using group decision to integrate various hybrid support vector machines on variables produced from principle components analysis and samples from random sampling. The experimental results indicate that hybrid support vector machine and ensemble of hybrid support vector machines were able to produce dominating performance than pure support vector machine and support vector machine ensemble.

Hydrological time series modeling: A comparison between adaptive neuro-fuzzy, neural network and autoregressive techniques

NASA Astrophysics Data System (ADS)

Lohani, A. K.; Kumar, Rakesh; Singh, R. D.

2012-06-01

SummaryTime series modeling is necessary for the planning and management of reservoirs. More recently, the soft computing techniques have been used in hydrological modeling and forecasting. In this study, the potential of artificial neural networks and neuro-fuzzy system in monthly reservoir inflow forecasting are examined by developing and comparing monthly reservoir inflow prediction models, based on autoregressive (AR), artificial neural networks (ANNs) and adaptive neural-based fuzzy inference system (ANFIS). To take care the effect of monthly periodicity in the flow data, cyclic terms are also included in the ANN and ANFIS models. Working with time series flow data of the Sutlej River at Bhakra Dam, India, several ANN and adaptive neuro-fuzzy models are trained with different input vectors. To evaluate the performance of the selected ANN and adaptive neural fuzzy inference system (ANFIS) models, comparison is made with the autoregressive (AR) models. The ANFIS model trained with the input data vector including previous inflows and cyclic terms of monthly periodicity has shown a significant improvement in the forecast accuracy in comparison with the ANFIS models trained with the input vectors considering only previous inflows. In all cases ANFIS gives more accurate forecast than the AR and ANN models. The proposed ANFIS model coupled with the cyclic terms is shown to provide better representation of the monthly inflow forecasting for planning and operation of reservoir.

Developing a dengue forecast model using machine learning: A case study in China

PubMed Central

Zhang, Qin; Wang, Li; Xiao, Jianpeng; Zhang, Qingying; Luo, Ganfeng; Li, Zhihao; He, Jianfeng; Zhang, Yonghui; Ma, Wenjun

2017-01-01

Background In China, dengue remains an important public health issue with expanded areas and increased incidence recently. Accurate and timely forecasts of dengue incidence in China are still lacking. We aimed to use the state-of-the-art machine learning algorithms to develop an accurate predictive model of dengue. Methodology/Principal findings Weekly dengue cases, Baidu search queries and climate factors (mean temperature, relative humidity and rainfall) during 2011–2014 in Guangdong were gathered. A dengue search index was constructed for developing the predictive models in combination with climate factors. The observed year and week were also included in the models to control for the long-term trend and seasonality. Several machine learning algorithms, including the support vector regression (SVR) algorithm, step-down linear regression model, gradient boosted regression tree algorithm (GBM), negative binomial regression model (NBM), least absolute shrinkage and selection operator (LASSO) linear regression model and generalized additive model (GAM), were used as candidate models to predict dengue incidence. Performance and goodness of fit of the models were assessed using the root-mean-square error (RMSE) and R-squared measures. The residuals of the models were examined using the autocorrelation and partial autocorrelation function analyses to check the validity of the models. The models were further validated using dengue surveillance data from five other provinces. The epidemics during the last 12 weeks and the peak of the 2014 large outbreak were accurately forecasted by the SVR model selected by a cross-validation technique. Moreover, the SVR model had the consistently smallest prediction error rates for tracking the dynamics of dengue and forecasting the outbreaks in other areas in China. Conclusion and significance The proposed SVR model achieved a superior performance in comparison with other forecasting techniques assessed in this study. The findings can help the government and community respond early to dengue epidemics. PMID:29036169

Skill assessment of Korea operational oceanographic system (KOOS)

NASA Astrophysics Data System (ADS)

Kim, J.; Park, K.

2016-02-01

For the ocean forecast system in Korea, the Korea operational oceanographic system (KOOS) has been developed and pre-operated since 2009 by the Korea institute of ocean science and technology (KIOST) funded by the Korean government. KOOS provides real time information and forecasts for marine environmental conditions in order to support all kinds of activities in the sea. Furthermore, more significant purpose of the KOOS information is to response and support to maritime problems and accidents such as oil spill, red-tide, shipwreck, extraordinary wave, coastal inundation and so on. Accordingly, it is essential to evaluate prediction accuracy and efforts to improve accuracy. The forecast accuracy should meet or exceed target benchmarks before its products are approved for release to the public.In this paper, we conduct error quantification of the forecasts using skill assessment technique for judgement of the KOOS performance. Skill assessment statistics includes the measures of errors and correlations such as root-mean-square-error (RMSE), mean bias (MB), correlation coefficient (R), and index of agreement (IOA) and the frequency with which errors lie within specified limits termed the central frequency (CF).The KOOS provides 72-hour daily forecast data such as air pressure, wind, water elevation, currents, wave, water temperature, and salinity produced by meteorological and hydrodynamic numerical models of WRF, ROMS, MOM5, WAM, WW3, and MOHID. The skill assessment has been performed through comparison of model results with in-situ observation data (Figure 1) for the period from 1 July, 2010 to 31 March, 2015 in Table 1 and model errors have been quantified with skill scores and CF determined by acceptable criteria depending on predicted variables (Table 2). Moreover, we conducted quantitative evaluation of spatio-temporal pattern correlation between numerical models and observation data such as sea surface temperature (SST) and sea surface current produced by ocean sensor in satellites and high frequency (HF) radar, respectively. Those quantified errors can allow to objective assessment of the KOOS performance and used can reveal different aspects of model inefficiency. Based on these results, various model components are tested and developed in order to improve forecast accuracy.

Short-range quantitative precipitation forecasting using Deep Learning approaches

NASA Astrophysics Data System (ADS)

Akbari Asanjan, A.; Yang, T.; Gao, X.; Hsu, K. L.; Sorooshian, S.

2017-12-01

Predicting short-range quantitative precipitation is very important for flood forecasting, early flood warning and other hydrometeorological purposes. This study aims to improve the precipitation forecasting skills using a recently developed and advanced machine learning technique named Long Short-Term Memory (LSTM). The proposed LSTM learns the changing patterns of clouds from Cloud-Top Brightness Temperature (CTBT) images, retrieved from the infrared channel of Geostationary Operational Environmental Satellite (GOES), using a sophisticated and effective learning method. After learning the dynamics of clouds, the LSTM model predicts the upcoming rainy CTBT events. The proposed model is then merged with a precipitation estimation algorithm termed Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN) to provide precipitation forecasts. The results of merged LSTM with PERSIANN are compared to the results of an Elman-type Recurrent Neural Network (RNN) merged with PERSIANN and Final Analysis of Global Forecast System model over the states of Oklahoma, Florida and Oregon. The performance of each model is investigated during 3 storm events each located over one of the study regions. The results indicate the outperformance of merged LSTM forecasts comparing to the numerical and statistical baselines in terms of Probability of Detection (POD), False Alarm Ratio (FAR), Critical Success Index (CSI), RMSE and correlation coefficient especially in convective systems. The proposed method shows superior capabilities in short-term forecasting over compared methods.

Continental scale data assimilation of discharge and its effect on flow predictions

NASA Astrophysics Data System (ADS)

Weerts, Albrecht; Schellekens, Jaap; van Dijk, Albert

2017-04-01

Floods are the most frequent of natural disasters, affecting millions of people across the globe every year. The anticipation and forecasting of floods at the global scale is crucial to preparing for severe events and providing early awareness where local flood models and warning services may not exist (Emmerton et al., 2016). Current global flood forecasting system heavily rely on forecast forcing (precipitation, temperature, reference potential evaporation) to derive initial state estimates of the hydrological model for the next forecast (e.g. by glueing the first day of subsequent forecast as proxy for the historical observed forcing). It is clear that this approach is not perfect and that data assimilation can help to overcome some of the weaknesses of this approach. So far most hydrologic da studies have focused mostly on catchment scale. Here we conduct a da experiment by assimilating multiple streamflow observations across the contiguous united states (CONUS) and Europe into a global hydrological model (W3RA) and run with and without localization method using OpenDA in the global flood forecasting information system (GLOFFIS). It is shown that assimilation of streamflow holds considerable potential for improving global scale flood forecasting (improving NSE scores from 0 to 0.7 and beyond). Weakness in the model (e.g. structural problems and missing processes) and forcing that influence the performance will be highlighted.

Continental scale data assimilation of discharge and its effect on flow predictions across the contiguous US (CONUS)

NASA Astrophysics Data System (ADS)

Weerts, A.; Schellekens, J.; van Dijk, A.; Molenaar, R.

2016-12-01

Floods are the most frequent of natural disasters, affecting millions of people across the globe every year. The anticipation and forecasting of floods at the global scale is crucial to preparing for severe events and providing early awareness where local flood models and warning services may not exist (Emmerton et al., 2016). Current global flood forecasting system heavily rely on forecast forcing (precipitation, temperature, reference potential evaporation) to derive initial state estimates of the hydrological model for the next forecast (e.g. by glueing the first day of subsequent forecast as proxy for the historical observed forcing). It is clear that this approach is not perfect and that data assimilation can help to overcome some of the weaknesses of this approach. So far most hydrologic da studies have focused mostly on catchment scale. Here we conduct a da experiment by assimilating multiple streamflow observations across the contiguous united states (CONUS) into a global hydrological model (W3RA) and run with and without localization method using OpenDA in the global flood forecasting information system (GLOFFIS). It is shown that assimilation of streamflow holds considerable potential for improving global scale flood forecasting (improving NSE scores from 0 to 0.7 and beyond). Weakness in the model (e.g. structural problems and missing processes) and forcing that influence the performance will be highlighted.

Ecosystem Model Skill Assessment. Yes We Can!

PubMed

Olsen, Erik; Fay, Gavin; Gaichas, Sarah; Gamble, Robert; Lucey, Sean; Link, Jason S

2016-01-01

Accelerated changes to global ecosystems call for holistic and integrated analyses of past, present and future states under various pressures to adequately understand current and projected future system states. Ecosystem models can inform management of human activities in a complex and changing environment, but are these models reliable? Ensuring that models are reliable for addressing management questions requires evaluating their skill in representing real-world processes and dynamics. Skill has been evaluated for just a limited set of some biophysical models. A range of skill assessment methods have been reviewed but skill assessment of full marine ecosystem models has not yet been attempted. We assessed the skill of the Northeast U.S. (NEUS) Atlantis marine ecosystem model by comparing 10-year model forecasts with observed data. Model forecast performance was compared to that obtained from a 40-year hindcast. Multiple metrics (average absolute error, root mean squared error, modeling efficiency, and Spearman rank correlation), and a suite of time-series (species biomass, fisheries landings, and ecosystem indicators) were used to adequately measure model skill. Overall, the NEUS model performed above average and thus better than expected for the key species that had been the focus of the model tuning. Model forecast skill was comparable to the hindcast skill, showing that model performance does not degenerate in a 10-year forecast mode, an important characteristic for an end-to-end ecosystem model to be useful for strategic management purposes. We identify best-practice approaches for end-to-end ecosystem model skill assessment that would improve both operational use of other ecosystem models and future model development. We show that it is possible to not only assess the skill of a complicated marine ecosystem model, but that it is necessary do so to instill confidence in model results and encourage their use for strategic management. Our methods are applicable to any type of predictive model, and should be considered for use in fields outside ecology (e.g. economics, climate change, and risk assessment).

Calibration and parameterization of a semi-distributed hydrological model to support sub-daily ensemble flood forecasting; a watershed in southeast Brazil

NASA Astrophysics Data System (ADS)

de Almeida Bressiani, D.; Srinivasan, R.; Mendiondo, E. M.

2013-12-01

The use of distributed or semi-distributed models to represent the processes and dynamics of a watershed in the last few years has increased. These models are important tools to predict and forecast the hydrological responses of the watersheds, and they can subside disaster risk management and planning. However they usually have a lot of parameters, of which, due to the spatial and temporal variability of the processes, are not known, specially in developing countries; therefore a robust and sensible calibration is very important. This study conduced a sub-daily calibration and parameterization of the Soil & Water Assessment Tool (SWAT) for a 12,600 km2 watershed in southeast Brazil, and uses ensemble forecasts to evaluate if the model can be used as a tool for flood forecasting. The Piracicaba Watershed, in São Paulo State, is mainly rural, but has about 4 million of population in highly relevant urban areas, and three cities in the list of critical cities of the National Center for Natural Disasters Monitoring and Alerts. For calibration: the watershed was divided in areas with similar hydrological characteristics, for each of these areas one gauge station was chosen for calibration; this procedure was performed to evaluate the effectiveness of calibrating in fewer places, since areas with the same group of groundwater, soil, land use and slope characteristics should have similar parameters; making calibration a less time-consuming task. The sensibility analysis and calibration were performed on the software SWAT-CUP with the optimization algorithm: Sequential Uncertainly Fitting Version 2 (SUFI-2), which uses Latin hypercube sampling scheme in an iterative process. The performance of the models to evaluate the calibration and validation was done with: Nash-Sutcliffe efficiency coefficient (NSE), determination coefficient (r2), root mean square error (RMSE), and percent bias (PBIAS), with monthly average values of NSE around 0.70, r2 of 0.9, normalized RMSE of 0.01, and PBIAS of 10. Past events were analysed to evaluate the possibility of using the SWAT developed model for Piracicaba watershed as a tool for ensemble flood forecasting. For the ensemble evaluation members from the numerical model Eta were used. Eta is an atmospheric model used for research and operational purposes, with 5km resolution, and is updated twice a day (00 e 12 UTC) for a ten day horizon, with precipitation and weather estimates for each hour. The parameterized SWAT model performed overall well for ensemble flood forecasting.

Nowcasting of deep convective clouds and heavy precipitation: Comparison study between NWP model simulation and extrapolation

NASA Astrophysics Data System (ADS)

Bližňák, Vojtěch; Sokol, Zbyněk; Zacharov, Petr

2017-02-01

An evaluation of convective cloud forecasts performed with the numerical weather prediction (NWP) model COSMO and extrapolation of cloud fields is presented using observed data derived from the geostationary satellite Meteosat Second Generation (MSG). The present study focuses on the nowcasting range (1-5 h) for five severe convective storms in their developing stage that occurred during the warm season in the years 2012-2013. Radar reflectivity and extrapolated radar reflectivity data were assimilated for at least 6 h depending on the time of occurrence of convection. Synthetic satellite imageries were calculated using radiative transfer model RTTOV v10.2, which was implemented into the COSMO model. NWP model simulations of IR10.8 μm and WV06.2 μm brightness temperatures (BTs) with a horizontal resolution of 2.8 km were interpolated into the satellite projection and objectively verified against observations using Root Mean Square Error (RMSE), correlation coefficient (CORR) and Fractions Skill Score (FSS) values. Naturally, the extrapolation of cloud fields yielded an approximately 25% lower RMSE, 20% higher CORR and 15% higher FSS at the beginning of the second forecasted hour compared to the NWP model forecasts. On the other hand, comparable scores were observed for the third hour, whereas the NWP forecasts outperformed the extrapolation by 10% for RMSE, 15% for CORR and up to 15% for FSS during the fourth forecasted hour and 15% for RMSE, 27% for CORR and up to 15% for FSS during the fifth forecasted hour. The analysis was completed by a verification of the precipitation forecasts yielding approximately 8% higher RMSE, 15% higher CORR and up to 45% higher FSS when the NWP model simulation is used compared to the extrapolation for the first hour. Both the methods yielded unsatisfactory level of precipitation forecast accuracy from the fourth forecasted hour onward.

A Kalman filter for a two-dimensional shallow-water model

NASA Technical Reports Server (NTRS)

Parrish, D. F.; Cohn, S. E.

1985-01-01

A two-dimensional Kalman filter is described for data assimilation for making weather forecasts. The filter is regarded as superior to the optimal interpolation method because the filter determines the forecast error covariance matrix exactly instead of using an approximation. A generalized time step is defined which includes expressions for one time step of the forecast model, the error covariance matrix, the gain matrix, and the evolution of the covariance matrix. Subsequent time steps are achieved by quantifying the forecast variables or employing a linear extrapolation from a current variable set, assuming the forecast dynamics are linear. Calculations for the evolution of the error covariance matrix are banded, i.e., are performed only with the elements significantly different from zero. Experimental results are provided from an application of the filter to a shallow-water simulation covering a 6000 x 6000 km grid.

Forecasting urban water demand: A meta-regression analysis.

PubMed

Sebri, Maamar

2016-12-01

Water managers and planners require accurate water demand forecasts over the short-, medium- and long-term for many purposes. These range from assessing water supply needs over spatial and temporal patterns to optimizing future investments and planning future allocations across competing sectors. This study surveys the empirical literature on the urban water demand forecasting using the meta-analytical approach. Specifically, using more than 600 estimates, a meta-regression analysis is conducted to identify explanations of cross-studies variation in accuracy of urban water demand forecasting. Our study finds that accuracy depends significantly on study characteristics, including demand periodicity, modeling method, forecasting horizon, model specification and sample size. The meta-regression results remain robust to different estimators employed as well as to a series of sensitivity checks performed. The importance of these findings lies in the conclusions and implications drawn out for regulators and policymakers and for academics alike. Copyright © 2016. Published by Elsevier Ltd.

A prospective earthquake forecast experiment for Japan

NASA Astrophysics Data System (ADS)

Yokoi, Sayoko; Nanjo, Kazuyoshi; Tsuruoka, Hiroshi; Hirata, Naoshi

2013-04-01

One major focus of the current Japanese earthquake prediction research program (2009-2013) is to move toward creating testable earthquake forecast models. For this purpose we started an experiment of forecasting earthquake activity in Japan under the framework of the Collaboratory for the Study of Earthquake Predictability (CSEP) through an international collaboration. We established the CSEP Testing Centre, an infrastructure to encourage researchers to develop testable models for Japan, and to conduct verifiable prospective tests of their model performance. On 1 November in 2009, we started the 1st earthquake forecast testing experiment for the Japan area. We use the unified JMA catalogue compiled by the Japan Meteorological Agency as authorized catalogue. The experiment consists of 12 categories, with 4 testing classes with different time spans (1 day, 3 months, 1 year, and 3 years) and 3 testing regions called All Japan, Mainland, and Kanto. A total of 91 models were submitted to CSEP-Japan, and are evaluated with the CSEP official suite of tests about forecast performance. In this presentation, we show the results of the experiment of the 3-month testing class for 5 rounds. HIST-ETAS7pa, MARFS and RI10K models corresponding to the All Japan, Mainland and Kanto regions showed the best score based on the total log-likelihood. It is also clarified that time dependency of model parameters is no effective factor to pass the CSEP consistency tests for the 3-month testing class in all regions. Especially, spatial distribution in the All Japan region was too difficult to pass consistency test due to multiple events at a bin. Number of target events for a round in the Mainland region tended to be smaller than model's expectation during all rounds, which resulted in rejections of consistency test because of overestimation. In the Kanto region, pass ratios of consistency tests in each model showed more than 80%, which was associated with good balanced forecasting of event number and spatial distribution. Due to the multiple rounds of the experiment, we are now understanding the stability of models, robustness of model selection and earthquake predictability in each region beyond stochastic fluctuations of seismicity. We plan to use the results for design of 3 dimensional earthquake forecasting model in Kanto region, which is supported by the special project for reducing vulnerability for urban mega earthquake disasters from Ministy of Education, Culture, Sports and Technology of Japan.

Real-time Ensemble Forecasting of Coronal Mass Ejections using the WSA-ENLIL+Cone Model

NASA Astrophysics Data System (ADS)

Mays, M. L.; Taktakishvili, A.; Pulkkinen, A. A.; MacNeice, P. J.; Rastaetter, L.; Kuznetsova, M. M.; Odstrcil, D.

2013-12-01

Ensemble forecasting of coronal mass ejections (CMEs) provides significant information in that it provides an estimation of the spread or uncertainty in CME arrival time predictions due to uncertainties in determining CME input parameters. Ensemble modeling of CME propagation in the heliosphere is performed by forecasters at the Space Weather Research Center (SWRC) using the WSA-ENLIL cone model available at the Community Coordinated Modeling Center (CCMC). SWRC is an in-house research-based operations team at the CCMC which provides interplanetary space weather forecasting for NASA's robotic missions and performs real-time model validation. A distribution of n (routinely n=48) CME input parameters are generated using the CCMC Stereo CME Analysis Tool (StereoCAT) which employs geometrical triangulation techniques. These input parameters are used to perform n different simulations yielding an ensemble of solar wind parameters at various locations of interest (satellites or planets), including a probability distribution of CME shock arrival times (for hits), and geomagnetic storm strength (for Earth-directed hits). Ensemble simulations have been performed experimentally in real-time at the CCMC since January 2013. We present the results of ensemble simulations for a total of 15 CME events, 10 of which were performed in real-time. The observed CME arrival was within the range of ensemble arrival time predictions for 5 out of the 12 ensemble runs containing hits. The average arrival time prediction was computed for each of the twelve ensembles predicting hits and using the actual arrival time an average absolute error of 8.20 hours was found for all twelve ensembles, which is comparable to current forecasting errors. Some considerations for the accuracy of ensemble CME arrival time predictions include the importance of the initial distribution of CME input parameters, particularly the mean and spread. When the observed arrivals are not within the predicted range, this still allows the ruling out of prediction errors caused by tested CME input parameters. Prediction errors can also arise from ambient model parameters such as the accuracy of the solar wind background, and other limitations. Additionally the ensemble modeling setup was used to complete a parametric event case study of the sensitivity of the CME arrival time prediction to free parameters for ambient solar wind model and CME.

Modeling Markov switching ARMA-GARCH neural networks models and an application to forecasting stock returns.

PubMed

Bildirici, Melike; Ersin, Özgür

2014-01-01

The study has two aims. The first aim is to propose a family of nonlinear GARCH models that incorporate fractional integration and asymmetric power properties to MS-GARCH processes. The second purpose of the study is to augment the MS-GARCH type models with artificial neural networks to benefit from the universal approximation properties to achieve improved forecasting accuracy. Therefore, the proposed Markov-switching MS-ARMA-FIGARCH, APGARCH, and FIAPGARCH processes are further augmented with MLP, Recurrent NN, and Hybrid NN type neural networks. The MS-ARMA-GARCH family and MS-ARMA-GARCH-NN family are utilized for modeling the daily stock returns in an emerging market, the Istanbul Stock Index (ISE100). Forecast accuracy is evaluated in terms of MAE, MSE, and RMSE error criteria and Diebold-Mariano equal forecast accuracy tests. The results suggest that the fractionally integrated and asymmetric power counterparts of Gray's MS-GARCH model provided promising results, while the best results are obtained for their neural network based counterparts. Further, among the models analyzed, the models based on the Hybrid-MLP and Recurrent-NN, the MS-ARMA-FIAPGARCH-HybridMLP, and MS-ARMA-FIAPGARCH-RNN provided the best forecast performances over the baseline single regime GARCH models and further, over the Gray's MS-GARCH model. Therefore, the models are promising for various economic applications.

Modeling Markov Switching ARMA-GARCH Neural Networks Models and an Application to Forecasting Stock Returns

PubMed Central

Bildirici, Melike; Ersin, Özgür

2014-01-01

The study has two aims. The first aim is to propose a family of nonlinear GARCH models that incorporate fractional integration and asymmetric power properties to MS-GARCH processes. The second purpose of the study is to augment the MS-GARCH type models with artificial neural networks to benefit from the universal approximation properties to achieve improved forecasting accuracy. Therefore, the proposed Markov-switching MS-ARMA-FIGARCH, APGARCH, and FIAPGARCH processes are further augmented with MLP, Recurrent NN, and Hybrid NN type neural networks. The MS-ARMA-GARCH family and MS-ARMA-GARCH-NN family are utilized for modeling the daily stock returns in an emerging market, the Istanbul Stock Index (ISE100). Forecast accuracy is evaluated in terms of MAE, MSE, and RMSE error criteria and Diebold-Mariano equal forecast accuracy tests. The results suggest that the fractionally integrated and asymmetric power counterparts of Gray's MS-GARCH model provided promising results, while the best results are obtained for their neural network based counterparts. Further, among the models analyzed, the models based on the Hybrid-MLP and Recurrent-NN, the MS-ARMA-FIAPGARCH-HybridMLP, and MS-ARMA-FIAPGARCH-RNN provided the best forecast performances over the baseline single regime GARCH models and further, over the Gray's MS-GARCH model. Therefore, the models are promising for various economic applications. PMID:24977200

Development of On-line Wildfire Emissions for the Operational Canadian Air Quality Forecast System

NASA Astrophysics Data System (ADS)

Pavlovic, R.; Menard, S.; Chen, J.; Anselmo, D.; Paul-Andre, B.; Gravel, S.; Moran, M. D.; Davignon, D.

2013-12-01

An emissions processing system has been developed to incorporate near-real-time emissions from wildfires and large prescribed burns into Environment Canada's real-time GEM-MACH air quality (AQ) forecast system. Since the GEM-MACH forecast domain covers Canada and most of the USA, including Alaska, fire location information is needed for both of these large countries. Near-real-time satellite data are obtained and processed separately for the two countries for organizational reasons. Fire location and fuel consumption data for Canada are provided by the Canadian Forest Service's Canadian Wild Fire Information System (CWFIS) while fire location and emissions data for the U.S. are provided by the SMARTFIRE (Satellite Mapping Automated Reanalysis Tool for Fire Incident Reconciliation) system via the on-line BlueSky Gateway. During AQ model runs, emissions from individual fire sources are injected into elevated model layers based on plume-rise calculations and then transport and chemistry calculations are performed. This 'on the fly' approach to the insertion of emissions provides greater flexibility since on-line meteorology is used and reduces computational overhead in emission pre-processing. An experimental wildfire version of GEM-MACH was run in real-time mode for the summers of 2012 and 2013. 48-hour forecasts were generated every 12 hours (at 00 and 12 UTC). Noticeable improvements in the AQ forecasts for PM2.5 were seen in numerous regions where fire activity was high. Case studies evaluating model performance for specific regions, computed objective scores, and subjective evaluations by AQ forecasters will be included in this presentation. Using the lessons learned from the last two summers, Environment Canada will continue to work towards the goal of incorporating near-real-time intermittent wildfire emissions within the operational air quality forecast system.

Simultaneous Estimation of Model State Variables and Observation and Forecast Biases Using a Two-Stage Hybrid Kalman Filter

NASA Technical Reports Server (NTRS)

Pauwels, V. R. N.; DeLannoy, G. J. M.; Hendricks Franssen, H.-J.; Vereecken, H.

2013-01-01

In this paper, we present a two-stage hybrid Kalman filter to estimate both observation and forecast bias in hydrologic models, in addition to state variables. The biases are estimated using the discrete Kalman filter, and the state variables using the ensemble Kalman filter. A key issue in this multi-component assimilation scheme is the exact partitioning of the difference between observation and forecasts into state, forecast bias and observation bias updates. Here, the error covariances of the forecast bias and the unbiased states are calculated as constant fractions of the biased state error covariance, and the observation bias error covariance is a function of the observation prediction error covariance. In a series of synthetic experiments, focusing on the assimilation of discharge into a rainfall-runoff model, it is shown that both static and dynamic observation and forecast biases can be successfully estimated. The results indicate a strong improvement in the estimation of the state variables and resulting discharge as opposed to the use of a bias-unaware ensemble Kalman filter. Furthermore, minimal code modification in existing data assimilation software is needed to implement the method. The results suggest that a better performance of data assimilation methods should be possible if both forecast and observation biases are taken into account.

Advances in air quality prediction with the use of integrated systems

NASA Astrophysics Data System (ADS)

Dragani, R.; Benedetti, A.; Engelen, R. J.; Peuch, V. H.

2017-12-01

Recent years have seen the rise of global operational atmospheric composition forecasting systems for several applications including climate monitoring, provision of boundary conditions for regional air quality forecasting, energy sector applications, to mention a few. Typically, global forecasts are provided in the medium-range up to five days ahead and are initialized with an analysis based on satellite data. In this work we present the latest advances in data assimilation using the ECMWF's 4D-Var system extended to atmospheric composition which is currently operational under the Copernicus Atmosphere Monitoring Service of the European Commission. The service is based on acquisition of all relevant data available in near-real-time, the processing of these datasets in the assimilation and the subsequent dissemination of global forecasts at ECMWF. The global forecasts are used by the CAMS regional models as boundary conditions for the European forecasts based on a multi-model ensemble. The global forecasts are also used to provide boundary conditions for other parts of the world (e.g., China) and are freely available to all interested entities. Some of the regional models also perform assimilation of satellite and ground-based observations. All products are assessed, validated and made publicly available on https://atmosphere.copernicus.eu/.

Relating anomaly correlation to lead time: Clustering analysis of CFSv2 forecasts of summer precipitation in China

NASA Astrophysics Data System (ADS)

Zhao, Tongtiegang; Liu, Pan; Zhang, Yongyong; Ruan, Chengqing

2017-09-01

Global climate model (GCM) forecasts are an integral part of long-range hydroclimatic forecasting. We propose to use clustering to explore anomaly correlation, which indicates the performance of raw GCM forecasts, in the three-dimensional space of latitude, longitude, and initialization time. Focusing on a certain period of the year, correlations for forecasts initialized at different preceding periods form a vector. The vectors of anomaly correlation across different GCM grid cells are clustered to reveal how GCM forecasts perform as time progresses. Through the case study of Climate Forecast System Version 2 (CFSv2) forecasts of summer precipitation in China, we observe that the correlation at a certain cell oscillates with lead time and can become negative. The use of clustering reveals two meaningful patterns that characterize the relationship between anomaly correlation and lead time. For some grid cells in Central and Southwest China, CFSv2 forecasts exhibit positive correlations with observations and they tend to improve as time progresses. This result suggests that CFSv2 forecasts tend to capture the summer precipitation induced by the East Asian monsoon and the South Asian monsoon. It also indicates that CFSv2 forecasts can potentially be applied to improving hydrological forecasts in these regions. For some other cells, the correlations are generally close to zero at different lead times. This outcome implies that CFSv2 forecasts still have plenty of room for further improvement. The robustness of the patterns has been tested using both hierarchical clustering and k-means clustering and examined with the Silhouette score.

Verify MesoNAM Performance

NASA Technical Reports Server (NTRS)

Bauman, William H., III

2010-01-01

The AMU conducted an objective analysis of the MesoNAM forecasts compared to observed values from sensors at specified KSC/CCAFS wind towers by calculating the following statistics to verify the performance of the model: 1) Bias (mean difference), 2) Standard deviation of Bias, 3) Root Mean Square Error (RMSE), and 4) Hypothesis test for Bias = O. The 45 WS LWOs use the MesoNAM to support launch weather operations. However, the actual performance of the model at KSC and CCAFS had not been measured objectively. The analysis compared the MesoNAM forecast winds, temperature and dew point to the observed values from the sensors on wind towers. The data were stratified by tower sensor, month and onshore/offshore wind direction based on the orientation of the coastline to each tower's location. The model's performance statistics were then calculated for each wind tower based on sensor height and model initialization time. The period of record for the data used in this task was based on the operational start of the current MesoNAM in mid-August 2006 and so the task began with the first full month of data, September 2006, through May 2010. The analysis of model performance indicated: a) The accuracy decreased as the forecast valid time from the model initialization increased, b) There was a diurnal signal in T with a cool bias during the late night and a warm bias during the afternoon, c) There was a diurnal signal in Td with a low bias during the afternoon and a high bias during the late night, and d) The model parameters at each vertical level most closely matched the observed parameters at heights closest to those vertical levels. The AMU developed a GUI that consists of a multi-level drop-down menu written in JavaScript embedded within the HTML code. This tool allows the LWO to easily and efficiently navigate among the charts and spreadsheet files containing the model performance statistics. The objective statistics give the LWOs knowledge of the model's strengths and weaknesses and the GUI allows quick access to the data which will result in improved forecasts for operations.

Temporal patterns and forecast of dengue infection in Northeastern Thailand.

PubMed

Silawan, Tassanee; Singhasivanon, Pratap; Kaewkungwal, Jaranit; Nimmanitya, Suchitra; Suwonkerd, Wanapa

2008-01-01

This study aimed to determine temporal patterns and develop a forecasting model for dengue incidence in northeastern Thailand. Reported cases were obtained from the Thailand national surveillance system. The temporal patterns were displayed by plotting monthly rates, the seasonal-trend decomposition procedure based on loess (STL) was performed using R 2.2.1 software, and the trend was assessed using Poisson regression. The forecasting model for dengue incidence was performed in R 2.2.1 and Intercooled Stata 9.2 using the seasonal Autoregressive Integrated Moving Average (ARIMA) model. The model was evaluated by comparing predicted versus actual rates of dengue for 1996 to 2005 and used to forecast monthly rates during January to December 2006. The results reveal that epidemics occurred every two years, with approximately three years per epidemic, and that the next epidemic will take place in 2006 to 2008. It was found that if a month increased, the rate ratio for dengue infection decreased by a factor 0.9919 for overall region and 0.9776 to 0.9984 for individual provinces. The amplitude of the peak, which was evident in June or July, was 11.32 to 88.08 times greater than the rest of the year. The seasonal ARIMA (2, 1, 0) (0, 1, 1)12 model was model with the best fit for regionwide data of total dengue incidence whereas the models with the best fit varied by province. The forecasted regional monthly rates during January to December 2006 should range from 0.27 to 17.89 per 100,000 population. The peak for 2006 should be much higher than the peak for 2005. The highest peaks in 2006 should be in Loei, Buri Ram, Surin, Nakhon Phanom, and Ubon Ratchathani Provinces.

The CMEMS IBI-MFC Forecasting Service in 2017: Evolution and Novelties associated to the CMEMS service release

NASA Astrophysics Data System (ADS)

Lorente, Pablo; Sotillo, Marcos G.; Gutknecht, Elodie; Dabrowski, Tomasz; Aouf, Lotfi; Toledano, Cristina; Amo-Baladron, Arancha; Aznar, Roland; De Pascual, Alvaro; Levier, Bruno; Bowyer, Peter; Rainaud, Romain; Alvarez-Fanjul, Enrique

2017-04-01

The IBI-MFC (Iberia-Biscay-Ireland Monitoring & Forecasting Centre) has been providing daily ocean model estimates and forecasts of diverse physical parameters for the IBI regional seas since 2011, first in the frame of MyOcean projects and later as part of the Copernicus Marine Environment Monitoring Service (CMEMS). By April 2017, coincident with the V3 CMEMS Service Release, the IBI-MFC will extend their near real time (NRT) forecast capabilities. Two new operational IBI forecast systems will be operationally run to generate high resolution biochemical (BIO) and wave (WAV) products on the IBI area. The IBI-NRT-BIO forecast system, based on a 1/36° NEMO-PISCES model application, is run once a week coupled with the IBI physical forecast solution and nested to the CMEMS GLOBAL-BIO solution. On the other hand, the IBI-NRT-WAV system, based on a MeteoFrance-WAM 10km resolution model application, runs twice a day using ECMWF wind forcing. Among other novelties related to the evolution of the IBI physical (PHY) solution, it is worthwhile mentioning the provision, as part of the IBI-NRT-PHY product daily updated, of three-dimensional hourly data on specific areas within the IBI domain. The delivery of these new hourly data along the whole water column has been achieved after the request from IBI users, in order to foster downscaling approaches by providing coherent open boundary conditions to any potential high-resolution coastal model nested to IBI regional solution. An extensive skill assessment of IBI-NRT forecast products has been conducted through the NARVAL (North Atlantic Regional VALidation) web tool, by means of the automatic computation of statistical metrics and quality indicators. By now, this tool has been focused on the validation of the IBI-NRT-PHY system. Nowadays, NARVAL is facing a significant upgrade to validate the aforementioned new biogeochemical and wave IBI products. To this aim, satellite derived observations of chlorophyll and significant wave height will be used, together with in-situ wave parameters measured by mooring buoys. Within this validation framework, special emphasis has been placed on the intercomparison of different forecast model solutions in overlapping areas in order to evaluate models' performances and prognostic capabilities. This common uncertainty estimates of IBI and other model solution is currently performed by NARVAL using both CMEMS forecast model sources (i.e. GLOBAL-MFC, MED-MFC and NWS-MFC) and non-CMEMS operational forecast solutions (mostly downstream application nested to the IBI solution). With respect to the IBI multi-year (MY) products, it is worth mentioning that the actual biogeochemical and physical reanalysis products will be re-run along year 2017, extending its time coverage backwards until 1992. Based on these IBI-MY products, a variety of climatic indicators related to essential oceanographic processes (i.e. western coastal upwelling or the Mediterranean Outflow Water) are currently being computed.

An Overview of the National Weather Service National Water Model

NASA Astrophysics Data System (ADS)

Cosgrove, B.; Gochis, D.; Clark, E. P.; Cui, Z.; Dugger, A. L.; Feng, X.; Karsten, L. R.; Khan, S.; Kitzmiller, D.; Lee, H. S.; Liu, Y.; McCreight, J. L.; Newman, A. J.; Oubeidillah, A.; Pan, L.; Pham, C.; Salas, F.; Sampson, K. M.; Sood, G.; Wood, A.; Yates, D. N.; Yu, W.

2016-12-01

The National Weather Service (NWS) Office of Water Prediction (OWP), in conjunction with the National Center for Atmospheric Research (NCAR) and the NWS National Centers for Environmental Prediction (NCEP) recently implemented version 1.0 of the National Water Model (NWM) into operations. This model is an hourly cycling uncoupled analysis and forecast system that provides streamflow for 2.7 million river reaches and other hydrologic information on 1km and 250m grids. It will provide complementary hydrologic guidance at current NWS river forecast locations and significantly expand guidance coverage and type in underserved locations. The core of this system is the NCAR-supported community Weather Research and Forecasting (WRF)-Hydro hydrologic model. It ingests forcing from a variety of sources including Multi-Sensor Multi-Radar (MRMS) radar-gauge observed precipitation data and High Resolution Rapid Refresh (HRRR), Rapid Refresh (RAP), Global Forecast System (GFS) and Climate Forecast System (CFS) forecast data. WRF-Hydro is configured to use the Noah-Multi Parameterization (Noah-MP) Land Surface Model (LSM) to simulate land surface processes. Separate water routing modules perform diffusive wave surface routing and saturated subsurface flow routing on a 250m grid, and Muskingum-Cunge channel routing down National Hydrogaphy Dataset Plus V2 (NHDPlusV2) stream reaches. River analyses and forecasts are provided across a domain encompassing the Continental United States (CONUS) and hydrologically contributing areas, while land surface output is available on a larger domain that extends beyond the CONUS into Canada and Mexico (roughly from latitude 19N to 58N). The system includes an analysis and assimilation configuration along with three forecast configurations. These include a short-range 15 hour deterministic forecast, a medium-Range 10 day deterministic forecast and a long-range 30 day 16-member ensemble forecast. United Sates Geologic Survey (USGS) streamflow observations are assimilated into the analysis and assimilation configuration, and all four configurations benefit from the inclusion of 1,260 reservoirs. An overview of the National Water Model will be given, along with information on ongoing evaluation activities and plans for future NWM enhancements.

Automated system for smoke dispersion prediction due to wild fires in Alaska

NASA Astrophysics Data System (ADS)

Kulchitsky, A.; Stuefer, M.; Higbie, L.; Newby, G.

2007-12-01

Community climate models have enabled development of specific environmental forecast systems. The University of Alaska (UAF) smoke group was created to adapt a smoke forecast system to the Alaska region. The US Forest Service (USFS) Missoula Fire Science Lab had developed a smoke forecast system based on the Weather Research and Forecasting (WRF) Model including chemistry (WRF/Chem). Following the successful experience of USFS, which runs their model operationally for the contiguous U.S., we develop a similar system for Alaska in collaboration with scientists from the USFS Missoula Fire Science Lab. Wildfires are a significant source of air pollution in Alaska because the climate and vegetation favor annual summer fires that burn huge areas. Extreme cases occurred in 2004, when an area larger than Maryland (more than 25000~km2) burned. Small smoke particles with a diameter less than 10~μm can penetrate deep into lungs causing health problems. Smoke also creates a severe restriction to air transport and has tremendous economical effect. The smoke dispersion and forecast system for Alaska was developed at the Geophysical Institute (GI) and the Arctic Region Supercomputing Center (ARSC), both at University of Alaska Fairbanks (UAF). They will help the public and plan activities a few days in advance to avoid dangerous smoke exposure. The availability of modern high performance supercomputers at ARSC allows us to create and run high-resolution, WRF-based smoke dispersion forecast for the entire State of Alaska. The core of the system is a Python program that manages the independent pieces. Our adapted Alaska system performs the following steps \\begin{itemize} Calculate the medium-resolution weather forecast using WRF/Met. Adapt the near real-time satellite-derived wildfire location and extent data that are received via direct broadcast from UAF's "Geographic Information Network of Alaska" (GINA) Calculate fuel moisture using WRF forecasts and National Fire Danger Rating System (NFDRS) fuel maps Calculate smoke emission components using a first order fire emission model Model the smoke plume rise yielding a vertically distribution that accounts for one-dimensional (vertical) concentrations of smoke constituents in the atmosphere above the fire Run WRF/Chem at high resolution for the forecast Use standard graphical tools to provide accessible smoke dispersion The system run twice each day at ARSC. The results will be freely available from a dedicated wildfire smoke web portal at ARSC.

Seamless atmospheric modeling across the hydrostatic-nonhydrostatic scales - preliminary results using an unstructured-Voronoi mesh for weather prediction.

NASA Astrophysics Data System (ADS)

Skamarock, W. C.

2015-12-01

One of the major problems in atmospheric model applications is the representation of deep convection within the models; explicit simulation of deep convection on fine meshes performs much better than sub-grid parameterized deep convection on coarse meshes. Unfortunately, the high cost of explicit convective simulation has meant it has only been used to down-scale global simulations in weather prediction and regional climate applications, typically using traditional one-way interactive nesting technology. We have been performing real-time weather forecast tests using a global non-hydrostatic atmospheric model (the Model for Prediction Across Scales, MPAS) that employs a variable-resolution unstructured Voronoi horizontal mesh (nominally hexagons) to span hydrostatic to nonhydrostatic scales. The smoothly varying Voronoi mesh eliminates many downscaling problems encountered using traditional one- or two-way grid nesting. Our test weather forecasts cover two periods - the 2015 Spring Forecast Experiment conducted at the NOAA Storm Prediction Center during the month of May in which we used a 50-3 km mesh, and the PECAN field program examining nocturnal convection over the US during the months of June and July in which we used a 15-3 km mesh. An important aspect of this modeling system is that the model physics be scale-aware, particularly the deep convection parameterization. These MPAS simulations employ the Grell-Freitas scale-aware convection scheme. Our test forecasts show that the scheme produces a gradual transition in the deep convection, from the deep unstable convection being handled entirely by the convection scheme on the coarse mesh regions (dx > 15 km), to the deep convection being almost entirely explicit on the 3 km NA region of the meshes. We will present results illustrating the performance of critical aspects of the MPAS model in these tests.

Sensitivity of WRF precipitation field to assimilation sources in northeastern Spain

NASA Astrophysics Data System (ADS)

Lorenzana, Jesús; Merino, Andrés; García-Ortega, Eduardo; Fernández-González, Sergio; Gascón, Estíbaliz; Hermida, Lucía; Sánchez, José Luis; López, Laura; Marcos, José Luis

2015-04-01

Numerical weather prediction (NWP) of precipitation is a challenge. Models predict precipitation after solving many physical processes. In particular, mesoscale NWP models have different parameterizations, such as microphysics, cumulus or radiation schemes. These facilitate, according to required spatial and temporal resolutions, precipitation fields with increasing reliability. Nevertheless, large uncertainties are inherent to precipitation forecasting. Consequently, assimilation methods are very important. The Atmospheric Physics Group at the University of León in Spain and the Castile and León Supercomputing Center carry out daily weather prediction based on the Weather Research and Forecasting (WRF) model, covering the entire Iberian Peninsula. Forecasts of severe precipitation affecting the Ebro Valley, in the southern Pyrenees range of northeastern Spain, are crucial in the decision-making process for managing reservoirs or initializing runoff models. These actions can avert floods and ensure uninterrupted economic activity in the area. We investigated a set of cases corresponding to intense or severe precipitation patterns, using a rain gauge network. Simulations were performed with a dual objective, i.e., to analyze forecast improvement using a specific assimilation method, and to study the sensitivity of model outputs to different types of assimilation data. A WRF forecast model initialized by an NCEP SST analysis was used as the control run. The assimilation was based on the Meteorological Assimilation Data Ingest System (MADIS) developed by NOAA. The MADIS data used were METAR, maritime, ACARS, radiosonde, and satellite products. The results show forecast improvement using the suggested assimilation method, and differences in the accuracy of forecast precipitation patterns varied with the assimilation data source.

Statistical and Machine Learning forecasting methods: Concerns and ways forward

PubMed Central

Makridakis, Spyros; Assimakopoulos, Vassilios

2018-01-01

Machine Learning (ML) methods have been proposed in the academic literature as alternatives to statistical ones for time series forecasting. Yet, scant evidence is available about their relative performance in terms of accuracy and computational requirements. The purpose of this paper is to evaluate such performance across multiple forecasting horizons using a large subset of 1045 monthly time series used in the M3 Competition. After comparing the post-sample accuracy of popular ML methods with that of eight traditional statistical ones, we found that the former are dominated across both accuracy measures used and for all forecasting horizons examined. Moreover, we observed that their computational requirements are considerably greater than those of statistical methods. The paper discusses the results, explains why the accuracy of ML models is below that of statistical ones and proposes some possible ways forward. The empirical results found in our research stress the need for objective and unbiased ways to test the performance of forecasting methods that can be achieved through sizable and open competitions allowing meaningful comparisons and definite conclusions. PMID:29584784

Comparison between stochastic and machine learning methods for hydrological multi-step ahead forecasting: All forecasts are wrong!

NASA Astrophysics Data System (ADS)

Papacharalampous, Georgia; Tyralis, Hristos; Koutsoyiannis, Demetris

2017-04-01

Machine learning (ML) is considered to be a promising approach to hydrological processes forecasting. We conduct a comparison between several stochastic and ML point estimation methods by performing large-scale computational experiments based on simulations. The purpose is to provide generalized results, while the respective comparisons in the literature are usually based on case studies. The stochastic methods used include simple methods, models from the frequently used families of Autoregressive Moving Average (ARMA), Autoregressive Fractionally Integrated Moving Average (ARFIMA) and Exponential Smoothing models. The ML methods used are Random Forests (RF), Support Vector Machines (SVM) and Neural Networks (NN). The comparison refers to the multi-step ahead forecasting properties of the methods. A total of 20 methods are used, among which 9 are the ML methods. 12 simulation experiments are performed, while each of them uses 2 000 simulated time series of 310 observations. The time series are simulated using stochastic processes from the families of ARMA and ARFIMA models. Each time series is split into a fitting (first 300 observations) and a testing set (last 10 observations). The comparative assessment of the methods is based on 18 metrics, that quantify the methods' performance according to several criteria related to the accurate forecasting of the testing set, the capturing of its variation and the correlation between the testing and forecasted values. The most important outcome of this study is that there is not a uniformly better or worse method. However, there are methods that are regularly better or worse than others with respect to specific metrics. It appears that, although a general ranking of the methods is not possible, their classification based on their similar or contrasting performance in the various metrics is possible to some extent. Another important conclusion is that more sophisticated methods do not necessarily provide better forecasts compared to simpler methods. It is pointed out that the ML methods do not differ dramatically from the stochastic methods, while it is interesting that the NN, RF and SVM algorithms used in this study offer potentially very good performance in terms of accuracy. It should be noted that, although this study focuses on hydrological processes, the results are of general scientific interest. Another important point in this study is the use of several methods and metrics. Using fewer methods and fewer metrics would have led to a very different overall picture, particularly if those fewer metrics corresponded to fewer criteria. For this reason, we consider that the proposed methodology is appropriate for the evaluation of forecasting methods.

Multi-Step Time Series Forecasting with an Ensemble of Varied Length Mixture Models.

PubMed

Ouyang, Yicun; Yin, Hujun

2018-05-01

Many real-world problems require modeling and forecasting of time series, such as weather temperature, electricity demand, stock prices and foreign exchange (FX) rates. Often, the tasks involve predicting over a long-term period, e.g. several weeks or months. Most existing time series models are inheritably for one-step prediction, that is, predicting one time point ahead. Multi-step or long-term prediction is difficult and challenging due to the lack of information and uncertainty or error accumulation. The main existing approaches, iterative and independent, either use one-step model recursively or treat the multi-step task as an independent model. They generally perform poorly in practical applications. In this paper, as an extension of the self-organizing mixture autoregressive (AR) model, the varied length mixture (VLM) models are proposed to model and forecast time series over multi-steps. The key idea is to preserve the dependencies between the time points within the prediction horizon. Training data are segmented to various lengths corresponding to various forecasting horizons, and the VLM models are trained in a self-organizing fashion on these segments to capture these dependencies in its component AR models of various predicting horizons. The VLM models form a probabilistic mixture of these varied length models. A combination of short and long VLM models and an ensemble of them are proposed to further enhance the prediction performance. The effectiveness of the proposed methods and their marked improvements over the existing methods are demonstrated through a number of experiments on synthetic data, real-world FX rates and weather temperatures.

Exploiting teleconnection indices for probabilistic forecasting of drought class transitions in Sicily region (Italy)

NASA Astrophysics Data System (ADS)

Bonaccorso, Brunella; Cancelliere, Antonino

2015-04-01

In the present study two probabilistic models for short-medium term drought forecasting able to include information provided by teleconnection indices are proposed and applied to Sicily region (Italy). Drought conditions are expressed in terms of the Standardized Precipitation-Evapotranspiration Index (SPEI) at different aggregation time scales. More specifically, a multivariate approach based on normal distribution is developed in order to estimate: 1) on the one hand transition probabilities to future SPEI drought classes and 2) on the other hand, SPEI forecasts at a generic time horizon M, as functions of past values of SPEI and the selected teleconnection index. To this end, SPEI series at 3, 4 and 6 aggregation time scales for Sicily region are extracted from the Global SPEI database, SPEIbase , available at Web repository of the Spanish National Research Council (http://sac.csic.es/spei/database.html), and averaged over the study area. In particular, SPEIbase v2.3 with spatial resolution of 0.5° lat/lon and temporal coverage between January 1901 and December 2013 is used. A preliminary correlation analysis is carried out to investigate the link between the drought index and different teleconnection patterns, namely: the North Atlantic Oscillation (NAO), the Scandinavian (SCA) and the East Atlantic-West Russia (EA-WR) patterns. Results of such analysis indicate a strongest influence of NAO on drought conditions in Sicily with respect to other teleconnection indices. Then, the proposed forecasting methodology is applied and the skill in forecasting of the proposed models is quantitatively assessed through the application of a simple score approach and of performance indices. Results indicate that inclusion of NAO index generally enhance model performance thus confirming the suitability of the models for short- medium term forecast of drought conditions.

Inflow forecasting model construction with stochastic time series for coordinated dam operation

NASA Astrophysics Data System (ADS)

Kim, T.; Jung, Y.; Kim, H.; Heo, J. H.

2014-12-01

Dam inflow forecasting is one of the most important tasks in dam operation for an effective water resources management and control. In general, dam inflow forecasting with stochastic time series model is possible to apply when the data is stationary because most of stochastic process based on stationarity. However, recent hydrological data cannot be satisfied the stationarity anymore because of climate change. Therefore a stochastic time series model, which can consider seasonality and trend in the data series, named SARIMAX(Seasonal Autoregressive Integrated Average with eXternal variable) model were constructed in this study. This SARIMAX model could increase the performance of stochastic time series model by considering the nonstationarity components and external variable such as precipitation. For application, the models were constructed for four coordinated dams on Han river in South Korea with monthly time series data. As a result, the models of each dam have similar performance and it would be possible to use the model for coordinated dam operation.Acknowledgement This research was supported by a grant 'Establishing Active Disaster Management System of Flood Control Structures by using 3D BIM Technique' [NEMA-NH-12-57] from the Natural Hazard Mitigation Research Group, National Emergency Management Agency of Korea.

Skilful seasonal forecasts of streamflow over Europe?

NASA Astrophysics Data System (ADS)

Arnal, Louise; Cloke, Hannah L.; Stephens, Elisabeth; Wetterhall, Fredrik; Prudhomme, Christel; Neumann, Jessica; Krzeminski, Blazej; Pappenberger, Florian

2018-04-01

This paper considers whether there is any added value in using seasonal climate forecasts instead of historical meteorological observations for forecasting streamflow on seasonal timescales over Europe. A Europe-wide analysis of the skill of the newly operational EFAS (European Flood Awareness System) seasonal streamflow forecasts (produced by forcing the Lisflood model with the ECMWF System 4 seasonal climate forecasts), benchmarked against the ensemble streamflow prediction (ESP) forecasting approach (produced by forcing the Lisflood model with historical meteorological observations), is undertaken. The results suggest that, on average, the System 4 seasonal climate forecasts improve the streamflow predictability over historical meteorological observations for the first month of lead time only (in terms of hindcast accuracy, sharpness and overall performance). However, the predictability varies in space and time and is greater in winter and autumn. Parts of Europe additionally exhibit a longer predictability, up to 7 months of lead time, for certain months within a season. In terms of hindcast reliability, the EFAS seasonal streamflow hindcasts are on average less skilful than the ESP for all lead times. The results also highlight the potential usefulness of the EFAS seasonal streamflow forecasts for decision-making (measured in terms of the hindcast discrimination for the lower and upper terciles of the simulated streamflow). Although the ESP is the most potentially useful forecasting approach in Europe, the EFAS seasonal streamflow forecasts appear more potentially useful than the ESP in some regions and for certain seasons, especially in winter for almost 40 % of Europe. Patterns in the EFAS seasonal streamflow hindcast skill are however not mirrored in the System 4 seasonal climate hindcasts, hinting at the need for a better understanding of the link between hydrological and meteorological variables on seasonal timescales, with the aim of improving climate-model-based seasonal streamflow forecasting.

A Hybrid Approach on Tourism Demand Forecasting

NASA Astrophysics Data System (ADS)

Nor, M. E.; Nurul, A. I. M.; Rusiman, M. S.

2018-04-01

Tourism has become one of the important industries that contributes to the country’s economy. Tourism demand forecasting gives valuable information to policy makers, decision makers and organizations related to tourism industry in order to make crucial decision and planning. However, it is challenging to produce an accurate forecast since economic data such as the tourism data is affected by social, economic and environmental factors. In this study, an equally-weighted hybrid method, which is a combination of Box-Jenkins and Artificial Neural Networks, was applied to forecast Malaysia’s tourism demand. The forecasting performance was assessed by taking the each individual method as a benchmark. The results showed that this hybrid approach outperformed the other two models

Forecasting Optimal Solar Energy Supply in Jiangsu Province (China): A Systematic Approach Using Hybrid of Weather and Energy Forecast Models

PubMed Central

Zhao, Xiuli; Yiranbon, Ethel

2014-01-01

The idea of aggregating information is clearly recognizable in the daily lives of all entities whether as individuals or as a group, since time immemorial corporate organizations, governments, and individuals as economic agents aggregate information to formulate decisions. Energy planning represents an investment-decision problem where information needs to be aggregated from credible sources to predict both demand and supply of energy. To do this there are varying methods ranging from the use of portfolio theory to managing risk and maximizing portfolio performance under a variety of unpredictable economic outcomes. The future demand for energy and need to use solar energy in order to avoid future energy crisis in Jiangsu province in China require energy planners in the province to abandon their reliance on traditional, “least-cost,” and stand-alone technology cost estimates and instead evaluate conventional and renewable energy supply on the basis of a hybrid of optimization models in order to ensure effective and reliable supply. Our task in this research is to propose measures towards addressing optimal solar energy forecasting by employing a systematic optimization approach based on a hybrid of weather and energy forecast models. After giving an overview of the sustainable energy issues in China, we have reviewed and classified the various models that existing studies have used to predict the influences of the weather influences and the output of solar energy production units. Further, we evaluate the performance of an exemplary ensemble model which combines the forecast output of two popular statistical prediction methods using a dynamic weighting factor. PMID:24511292

Forecasting optimal solar energy supply in Jiangsu Province (China): a systematic approach using hybrid of weather and energy forecast models.

PubMed

Zhao, Xiuli; Asante Antwi, Henry; Yiranbon, Ethel

2014-01-01

The idea of aggregating information is clearly recognizable in the daily lives of all entities whether as individuals or as a group, since time immemorial corporate organizations, governments, and individuals as economic agents aggregate information to formulate decisions. Energy planning represents an investment-decision problem where information needs to be aggregated from credible sources to predict both demand and supply of energy. To do this there are varying methods ranging from the use of portfolio theory to managing risk and maximizing portfolio performance under a variety of unpredictable economic outcomes. The future demand for energy and need to use solar energy in order to avoid future energy crisis in Jiangsu province in China require energy planners in the province to abandon their reliance on traditional, "least-cost," and stand-alone technology cost estimates and instead evaluate conventional and renewable energy supply on the basis of a hybrid of optimization models in order to ensure effective and reliable supply. Our task in this research is to propose measures towards addressing optimal solar energy forecasting by employing a systematic optimization approach based on a hybrid of weather and energy forecast models. After giving an overview of the sustainable energy issues in China, we have reviewed and classified the various models that existing studies have used to predict the influences of the weather influences and the output of solar energy production units. Further, we evaluate the performance of an exemplary ensemble model which combines the forecast output of two popular statistical prediction methods using a dynamic weighting factor.

Real-time short-term forecast of water inflow into Bureyskaya reservoir

NASA Astrophysics Data System (ADS)

Motovilov, Yury

2017-04-01

During several recent years, a methodology for operational optimization in hydrosystems including forecasts of the hydrological situation has been developed on example of Burea reservoir. The forecasts accuracy improvement of the water inflow into the reservoir during planning of water and energy regime was one of the main goals for implemented research. Burea river is the second left largest Amur tributary after Zeya river with its 70.7 thousand square kilometers watershed and 723 km-long river course. A variety of natural conditions - from plains in the southern part to northern mountainous areas determine a significant spatio-temporal variability in runoff generation patterns and river regime. Bureyskaya hydropower plant (HPP) with watershed area 65.2 thousand square kilometers is a key station in the Russian Far Eastern energy system providing its reliable operation. With a spacious reservoir, Bureyskaya HPP makes a significant contribution to the protection of the Amur region from catastrophic floods. A physically-based distributed model of runoff generation based on the ECOMAG (ECOlogical Model for Applied Geophysics) hydrological modeling platform has been developed for the Burea River basin. The model describes processes of interception of rainfall/snowfall by the canopy, snow accumulation and melt, soil freezing and thawing, water infiltration into unfrozen and frozen soil, evapotranspiration, thermal and water regime of soil, overland, subsurface, ground and river flow. The governing model's equations are derived from integration of the basic hydro- and thermodynamics equations of water and heat vertical transfer in snowpack, frozen/unfrozen soil, horizontal water flow under and over catchment slopes, etc. The model setup for Bureya river basin included watershed and river network schematization with GIS module by DEM analysis, meteorological time-series preparation, model calibration and validation against historical observations. The results showed good model performance as compared to observed inflow data into the Bureya reservoir and high diagnostic potential of data-modeling system of the runoff formation. With the use of this system the following flowchart for short-range forecasting inflow into Bureyskoe reservoir and forecast correction technique using continuously updated hydrometeorological data has been developed: 1 - Daily renewal of weather observations and forecasts database via the Internet; 2 - Daily runoff calculation from the beginning of the current year to current date is conducted; 3 - Short-range (up to 7 days) forecast is generated based on weather forecast. The idea underlying the model assimilation of newly obtained hydro meteorological information to adjust short-range hydrological forecasts lies in the assumption of the forecast errors inertia. Then the difference between calculated and observed streamflow at the forecast release date is "scattered" with specific weights to calculated streamflow for the forecast lead time. During 2016 this forecasts method of the inflow into the Bureyskaya reservoir up to 7 days is tested in online mode. Satisfactory evaluated short-range inflow forecast success rate is obtained. Tests of developed method have shown strong sensitivity to the results of short-term precipitation forecasts.

Seasonal forecasting for water resource management: the example of CNR Genissiat dam on the Rhone River in France

NASA Astrophysics Data System (ADS)

Dommanget, Etienne; Bellier, Joseph; Ben Daoud, Aurélien; Graff, Benjamin

2014-05-01

Compagnie Nationale du Rhône (CNR) has been granted the concession to operate the Rhone River from the Swiss border to the Mediterranean Sea since 1933 and carries out three interdependent missions: navigation, irrigation and hydropower production. Nowadays, CNR generates one quarter of France's hydropower electricity. The convergence of public and private interests around optimizing the management of water resources throughout the French Rhone valley led CNR to develop hydrological models dedicated to discharge seasonal forecasting. Indeed, seasonal forecasting is a major issue for CNR and water resource management, in order to optimize long-term investments of the produced electricity, plan dam maintenance operations and anticipate low water period. Seasonal forecasting models have been developed on the Genissiat dam. With an installed capacity of 420MW, Genissiat dam is the first of the 19 CNR's hydropower plants. Discharge forecasting at Genissiat dam is strategic since its inflows contributes to 20% of the total Rhone average discharge and consequently to 40% of the total Rhone hydropower production. Forecasts are based on hydrological statistical models. Discharge on the main Rhone River tributaries upstream Genissiat dam are forecasted from 1 to 6 months ahead thanks to multiple linear regressions. Inputs data of these regressions are identified depending on river hydrological regimes and periods of the year. For the melting season, from spring to summer, snow water equivalent (SWE) data are of major importance. SWE data are calculated from Crocus model (Météo France) and SLF's model (Switzerland). CNR hydro-meteorological forecasters assessed meteorological trends regarding precipitations for the next coming months. These trends are used to generate stochastically precipitation scenarios in order to complement regression data set. This probabilistic approach build a decision-making supports for CNR's water resource management team and provides them with seasonal forecasts and their confidence interval. After a presentation of CNR methodology, results for the years 2011 and 2013 will illustrate CNR's seasonal forecasting models ability. These years are of particular interest regarding water resource management seeing that they are, respectively, unusually dry and snowy. Model performances will be assessed in comparison with historical climatology thanks to CRPS skill score.

A back-fitting algorithm to improve real-time flood forecasting

NASA Astrophysics Data System (ADS)

Zhang, Xiaojing; Liu, Pan; Cheng, Lei; Liu, Zhangjun; Zhao, Yan

2018-07-01

Real-time flood forecasting is important for decision-making with regards to flood control and disaster reduction. The conventional approach involves a postprocessor calibration strategy that first calibrates the hydrological model and then estimates errors. This procedure can simulate streamflow consistent with observations, but obtained parameters are not optimal. Joint calibration strategies address this issue by refining hydrological model parameters jointly with the autoregressive (AR) model. In this study, five alternative schemes are used to forecast floods. Scheme I uses only the hydrological model, while scheme II includes an AR model for error correction. In scheme III, differencing is used to remove non-stationarity in the error series. A joint inference strategy employed in scheme IV calibrates the hydrological and AR models simultaneously. The back-fitting algorithm, a basic approach for training an additive model, is adopted in scheme V to alternately recalibrate hydrological and AR model parameters. The performance of the five schemes is compared with a case study of 15 recorded flood events from China's Baiyunshan reservoir basin. Our results show that (1) schemes IV and V outperform scheme III during the calibration and validation periods and (2) scheme V is inferior to scheme IV in the calibration period, but provides better results in the validation period. Joint calibration strategies can therefore improve the accuracy of flood forecasting. Additionally, the back-fitting recalibration strategy produces weaker overcorrection and a more robust performance compared with the joint inference strategy.

Development of the ClearSky smoke dispersion forecast system for agricultural field burning in the Pacific Northwest

NASA Astrophysics Data System (ADS)

Jain, Rahul; Vaughan, Joseph; Heitkamp, Kyle; Ramos, Charleston; Claiborn, Candis; Schreuder, Maarten; Schaaf, Mark; Lamb, Brian

The post-harvest burning of agricultural fields is commonly used to dispose of crop residue and provide other desired services such as pest control. Despite careful regulation of burning, smoke plumes from field burning in the Pacific Northwest commonly degrade air quality, particularly for rural populations. In this paper, ClearSky, a numerical smoke dispersion forecast system for agricultural field burning that was developed to support smoke management in the Inland Pacific Northwest, is described. ClearSky began operation during the summer through fall burn season of 2002 and continues to the present. ClearSky utilizes Mesoscale Meteorological Model version 5 (MM5v3) forecasts from the University of Washington, data on agricultural fields, a web-based user interface for defining burn scenarios, the Lagrangian CALPUFF dispersion model and web-served animations of plume forecasts. The ClearSky system employs a unique hybrid source configuration, which treats the flaming portion of a field as a buoyant line source and the smoldering portion of the field as a buoyant area source. Limited field observations show that this hybrid approach yields reasonable plume rise estimates using source parameters derived from recent field burning emission field studies. The performance of this modeling system was evaluated for 2003 by comparing forecast meteorology against meteorological observations, and comparing model-predicted hourly averaged PM 2.5 concentrations against observations. Examples from this evaluation illustrate that while the ClearSky system can accurately predict PM 2.5 surface concentrations due to field burning, the overall model performance depends strongly on meteorological forecast error. Statistical evaluation of the meteorological forecast at seven surface stations indicates a strong relationship between topographical complexity near the station and absolute wind direction error with wind direction errors increasing from approximately 20° for sites in open areas to 70° or more for sites in very complex terrain. The analysis also showed some days with good forecast meteorology with absolute mean error in wind direction less than 30° when ClearSky correctly predicted PM 2.5 surface concentrations at receptors affected by field burns. On several other days with similar levels of wind direction error the model did not predict apparent plume impacts. In most of these cases, there were no reported burns in the vicinity of the monitor and, thus, it appeared that other, non-reported burns were responsible for the apparent plume impact at the monitoring site. These cases do not provide information on the performance of the model, but rather indicate that further work is needed to identify all burns and to improve burn reports in an accurate and timely manner. There were also a number of days with wind direction errors exceeding 70° when the forecast system did not correctly predict plume behavior.

Human-model hybrid Korean air quality forecasting system.

PubMed

Chang, Lim-Seok; Cho, Ara; Park, Hyunju; Nam, Kipyo; Kim, Deokrae; Hong, Ji-Hyoung; Song, Chang-Keun

2016-09-01

The Korean national air quality forecasting system, consisting of the Weather Research and Forecasting, the Sparse Matrix Operator Kernel Emissions, and the Community Modeling and Analysis (CMAQ), commenced from August 31, 2013 with target pollutants of particulate matters (PM) and ozone. Factors contributing to PM forecasting accuracy include CMAQ inputs of meteorological field and emissions, forecasters' capacity, and inherent CMAQ limit. Four numerical experiments were conducted including two global meteorological inputs from the Global Forecast System (GFS) and the Unified Model (UM), two emissions from the Model Intercomparison Study Asia (MICS-Asia) and the Intercontinental Chemical Transport Experiment (INTEX-B) for the Northeast Asia with Clear Air Policy Support System (CAPSS) for South Korea, and data assimilation of the Monitoring Atmospheric Composition and Climate (MACC). Significant PM underpredictions by using both emissions were found for PM mass and major components (sulfate and organic carbon). CMAQ predicts PM2.5 much better than PM10 (NMB of PM2.5: -20~-25%, PM10: -43~-47%). Forecasters' error usually occurred at the next day of high PM event. Once CMAQ fails to predict high PM event the day before, forecasters are likely to dismiss the model predictions on the next day which turns out to be true. The best combination of CMAQ inputs is the set of UM global meteorological field, MICS-Asia and CAPSS 2010 emissions with the NMB of -12.3%, the RMSE of 16.6μ/m(3) and the R(2) of 0.68. By using MACC data as an initial and boundary condition, the performance skill of CMAQ would be improved, especially in the case of undefined coarse emission. A variety of methods such as ensemble and data assimilation are considered to improve further the accuracy of air quality forecasting, especially for high PM events to be comparable to for all cases. The growing utilization of the air quality forecast induced the public strongly to demand that the accuracy of the national forecasting be improved. In this study, we investigated the problems in the current forecasting as well as various alternatives to solve the problems. Such efforts to improve the accuracy of the forecast are expected to contribute to the protection of public health by increasing the availability of the forecast system.

Drought Monitoring and Forecasting Using the Princeton/U Washington National Hydrologic Forecasting System

NASA Astrophysics Data System (ADS)

Wood, E. F.; Yuan, X.; Roundy, J. K.; Lettenmaier, D. P.; Mo, K. C.; Xia, Y.; Ek, M. B.

2011-12-01

Extreme hydrologic events in the form of droughts or floods are a significant source of social and economic damage in many parts of the world. Having sufficient warning of extreme events allows managers to prepare for and reduce the severity of their impacts. A hydrologic forecast system can give seasonal predictions that can be used by mangers to make better decisions; however there is still much uncertainty associated with such a system. Therefore it is important to understand the forecast skill of the system before transitioning to operational usage. Seasonal reforecasts (1982 - 2010) from the NCEP Climate Forecast System (both version 1 (CFS) and version 2 (CFSv2), Climate Prediction Center (CPC) outlooks and the European Seasonal Interannual Prediction (EUROSIP) system, are assessed for forecasting skill in drought prediction across the U.S., both singularly and as a multi-model system The Princeton/U Washington national hydrologic monitoring and forecast system is being implemented at NCEP/EMC via their Climate Test Bed as the experimental hydrological forecast system to support U.S. operational drought prediction. Using our system, the seasonal forecasts are biased corrected, downscaled and used to drive the Variable Infiltration Capacity (VIC) land surface model to give seasonal forecasts of hydrologic variables with lead times of up to six months. Results are presented for a number of events, with particular focus on the Apalachicola-Chattahoochee-Flint (ACF) River Basin in the South Eastern United States, which has experienced a number of severe droughts in recent years and is a pilot study basin for the National Integrated Drought Information System (NIDIS). The performance of the VIC land surface model is evaluated using observational forcing when compared to observed streamflow. The effectiveness of the forecast system to predict streamflow and soil moisture is evaluated when compared with observed streamflow and modeled soil moisture driven by observed atmospheric forcing. The forecast skills from the dynamical seasonal models (CFSv1, CFSv2, EUROSIP) and CPC are also compared with forecasts based on the Ensemble Streamflow Prediction (ESP) method, which uses initial conditions and historical forcings to generate seasonal forecasts. The skill of the system to predict drought, drought recovery and related hydrological conditions such as low-flows is assessed, along with quantified uncertainty.

Evaluating the improvements of the BOLAM meteorological model operational at ISPRA: A case study approach - preliminary results

NASA Astrophysics Data System (ADS)

Mariani, S.; Casaioli, M.; Lastoria, B.; Accadia, C.; Flavoni, S.

2009-04-01

The Institute for Environmental Protection and Research - ISPRA (former Agency for Environmental Protection and Technical Services - APAT) runs operationally since 2000 an integrated meteo-marine forecasting chain, named the Hydro-Meteo-Marine Forecasting System (Sistema Idro-Meteo-Mare - SIMM), formed by a cascade of four numerical models, telescoping from the Mediterranean basin to the Venice Lagoon, and initialized by means of analyses and forecasts from the European Centre for Medium-Range Weather Forecasts (ECMWF). The operational integrated system consists of a meteorological model, the parallel verision of BOlogna Limited Area Model (BOLAM), coupled over the Mediterranean sea with a WAve Model (WAM), a high-resolution shallow-water model of the Adriatic and Ionian Sea, namely the Princeton Ocean Model (POM), and a finite-element version of the same model (VL-FEM) on the Venice Lagoon, aimed to forecast the acqua alta events. Recently, the physically based, fully distributed, rainfall-runoff TOPographic Kinematic APproximation and Integration (TOPKAPI) model has been integrated into the system, coupled to BOLAM, over two river basins, located in the central and northeastern part of Italy, respectively. However, at the present time, this latter part of the forecasting chain is not operational and it is used in a research configuration. BOLAM was originally implemented in 2000 onto the Quadrics parallel supercomputer (and for this reason referred to as QBOLAM, as well) and only at the end of 2006 it was ported (together with the other operational marine models of the forecasting chain) onto the Silicon Graphics Inc. (SGI) Altix 8-processor machine. In particular, due to the Quadrics implementation, the Kuo scheme was formerly implemented into QBOLAM for the cumulus convection parameterization. On the contrary, when porting SIMM onto the Altix Linux cluster, it was achievable to implement into QBOLAM the more advanced convection parameterization by Kain and Fritsch. A fully updated serial version of the BOLAM code has been recently acquired. Code improvements include a more precise advection scheme (Weighted Average Flux); explicit advection of five hydrometeors, and state-of-the-art parameterization schemes for radiation, convection, boundary layer turbulence and soil processes (also with possible choice among different available schemes). The operational implementation of the new code into the SIMM model chain, which requires the development of a parallel version, will be achieved during 2009. In view of this goal, the comparative verification of the different model versions' skill represents a fundamental task. On this purpose, it has been decided to evaluate the performance improvement of the new BOLAM code (in the available serial version, hereinafter BOLAM 2007) with respect to the version with the Kain-Fritsch scheme (hereinafter KF version) and to the older one employing the Kuo scheme (hereinafter Kuo version). In the present work, verification of precipitation forecasts from the three BOLAM versions is carried on in a case study approach. The intense rainfall episode occurred on 10th - 17th December 2008 over Italy has been considered. This event produced indeed severe damages in Rome and its surrounding areas. Objective and subjective verification methods have been employed in order to evaluate model performance against an observational dataset including rain gauge observations and satellite imagery. Subjective comparison of observed and forecast precipitation fields is suitable to give an overall description of the forecast quality. Spatial errors (e.g., shifting and pattern errors) and rainfall volume error can be assessed quantitatively by means of object-oriented methods. By comparing satellite images with model forecast fields, it is possible to investigate the differences between the evolution of the observed weather system and the predicted ones, and its sensitivity to the improvements in the model code. Finally, the error in forecasting the cyclone evolution can be tentatively related with the precipitation forecast error.