Investigating Surface Bias Errors in the Weather Research and Forecasting (WRF) Model using a Geographic Information System (GIS)

DTIC Science & Technology

2015-02-01

WRF ) Model using a Geographic Information System (GIS) by Jeffrey A Smith, Theresa A Foley, John W Raby, and Brian Reen...ARL-TR-7212 ● FEB 2015 US Army Research Laboratory Investigating Surface Bias Errors in the Weather Research and Forecasting ( WRF ) Model...SUBTITLE Investigating surface bias errors in the Weather Research and Forecasting ( WRF ) Model using a Geographic Information System (GIS) 5a

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).

The development rainfall forecasting using kalman filter

NASA Astrophysics Data System (ADS)

Zulfi, Mohammad; Hasan, Moh.; Dwidja Purnomo, Kosala

2018-04-01

Rainfall forecasting is very interesting for agricultural planing. Rainfall information is useful to make decisions about the plan planting certain commodities. In this studies, the rainfall forecasting by ARIMA and Kalman Filter method. Kalman Filter method is used to declare a time series model of which is shown in the form of linear state space to determine the future forecast. This method used a recursive solution to minimize error. The rainfall data in this research clustered by K-means clustering. Implementation of Kalman Filter method is for modelling and forecasting rainfall in each cluster. We used ARIMA (p,d,q) to construct a state space for KalmanFilter model. So, we have four group of the data and one model in each group. In conclusions, Kalman Filter method is better than ARIMA model for rainfall forecasting in each group. It can be showed from error of Kalman Filter method that smaller than error of ARIMA model.

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.

An ensemble-ANFIS based uncertainty assessment model for forecasting multi-scalar standardized precipitation index

NASA Astrophysics Data System (ADS)

Ali, Mumtaz; Deo, Ravinesh C.; Downs, Nathan J.; Maraseni, Tek

2018-07-01

Forecasting drought by means of the World Meteorological Organization-approved Standardized Precipitation Index (SPI) is considered to be a fundamental task to support socio-economic initiatives and effectively mitigating the climate-risk. This study aims to develop a robust drought modelling strategy to forecast multi-scalar SPI in drought-rich regions of Pakistan where statistically significant lagged combinations of antecedent SPI are used to forecast future SPI. With ensemble-Adaptive Neuro Fuzzy Inference System ('ensemble-ANFIS') executed via a 10-fold cross-validation procedure, a model is constructed by randomly partitioned input-target data. Resulting in 10-member ensemble-ANFIS outputs, judged by mean square error and correlation coefficient in the training period, the optimal forecasts are attained by the averaged simulations, and the model is benchmarked with M5 Model Tree and Minimax Probability Machine Regression (MPMR). The results show the proposed ensemble-ANFIS model's preciseness was notably better (in terms of the root mean square and mean absolute error including the Willmott's, Nash-Sutcliffe and Legates McCabe's index) for the 6- and 12- month compared to the 3-month forecasts as verified by the largest error proportions that registered in smallest error band. Applying 10-member simulations, ensemble-ANFIS model was validated for its ability to forecast severity (S), duration (D) and intensity (I) of drought (including the error bound). This enabled uncertainty between multi-models to be rationalized more efficiently, leading to a reduction in forecast error caused by stochasticity in drought behaviours. Through cross-validations at diverse sites, a geographic signature in modelled uncertainties was also calculated. Considering the superiority of ensemble-ANFIS approach and its ability to generate uncertainty-based information, the study advocates the versatility of a multi-model approach for drought-risk forecasting and its prime importance for estimating drought properties over confidence intervals to generate better information for strategic decision-making.

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.

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.

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.

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.

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.

The impact of satellite temperature soundings on the forecasts of a small national meteorological service

NASA Technical Reports Server (NTRS)

Wolfson, N.; Thomasell, A.; Alperson, Z.; Brodrick, H.; Chang, J. T.; Gruber, A.; Ohring, G.

1984-01-01

The impact of introducing satellite temperature sounding data on a numerical weather prediction model of a national weather service is evaluated. A dry five level, primitive equation model which covers most of the Northern Hemisphere, is used for these experiments. Series of parallel forecast runs out to 48 hours are made with three different sets of initial conditions: (1) NOSAT runs, only conventional surface and upper air observations are used; (2) SAT runs, satellite soundings are added to the conventional data over oceanic regions and North Africa; and (3) ALLSAT runs, the conventional upper air observations are replaced by satellite soundings over the entire model domain. The impact on the forecasts is evaluated by three verification methods: the RMS errors in sea level pressure forecasts, systematic errors in sea level pressure forecasts, and errors in subjective forecasts of significant weather elements for a selected portion of the model domain. For the relatively short range of the present forecasts, the major beneficial impacts on the sea level pressure forecasts are found precisely in those areas where the satellite sounding are inserted and where conventional upper air observations are sparse. The RMS and systematic errors are reduced in these regions. The subjective forecasts of significant weather elements are improved with the use of the satellite data. It is found that the ALLSAT forecasts are of a quality comparable to the SAR forecasts.

The Influence of Observation Errors on Analysis Error and Forecast Skill Investigated with an Observing System Simulation Experiment

NASA Technical Reports Server (NTRS)

Prive, N. C.; Errico, R. M.; Tai, K.-S.

2013-01-01

The Global Modeling and Assimilation Office (GMAO) observing system simulation experiment (OSSE) framework is used to explore the response of analysis error and forecast skill to observation quality. In an OSSE, synthetic observations may be created that have much smaller error than real observations, and precisely quantified error may be applied to these synthetic observations. Three experiments are performed in which synthetic observations with magnitudes of applied observation error that vary from zero to twice the estimated realistic error are ingested into the Goddard Earth Observing System Model (GEOS-5) with Gridpoint Statistical Interpolation (GSI) data assimilation for a one-month period representing July. The analysis increment and observation innovation are strongly impacted by observation error, with much larger variances for increased observation error. The analysis quality is degraded by increased observation error, but the change in root-mean-square error of the analysis state is small relative to the total analysis error. Surprisingly, in the 120 hour forecast increased observation error only yields a slight decline in forecast skill in the extratropics, and no discernable degradation of forecast skill in the tropics.

Using Analysis Increments (AI) to Estimate and Correct Systematic Errors in the Global Forecast System (GFS) Online

NASA Astrophysics Data System (ADS)

Bhargava, K.; Kalnay, E.; Carton, J.; Yang, F.

2017-12-01

Systematic forecast errors, arising from model deficiencies, form a significant portion of the total forecast error in weather prediction models like the Global Forecast System (GFS). While much effort has been expended to improve models, substantial model error remains. The aim here is to (i) estimate the model deficiencies in the GFS that lead to systematic forecast errors, (ii) implement an online correction (i.e., within the model) scheme to correct GFS following the methodology of Danforth et al. [2007] and Danforth and Kalnay [2008, GRL]. Analysis Increments represent the corrections that new observations make on, in this case, the 6-hr forecast in the analysis cycle. Model bias corrections are estimated from the time average of the analysis increments divided by 6-hr, assuming that initial model errors grow linearly and first ignoring the impact of observation bias. During 2012-2016, seasonal means of the 6-hr model bias are generally robust despite changes in model resolution and data assimilation systems, and their broad continental scales explain their insensitivity to model resolution. The daily bias dominates the sub-monthly analysis increments and consists primarily of diurnal and semidiurnal components, also requiring a low dimensional correction. Analysis increments in 2015 and 2016 are reduced over oceans, which is attributed to improvements in the specification of the SSTs. These results encourage application of online correction, as suggested by Danforth and Kalnay, for mean, seasonal and diurnal and semidiurnal model biases in GFS to reduce both systematic and random errors. As the error growth in the short-term is still linear, estimated model bias corrections can be added as a forcing term in the model tendency equation to correct online. Preliminary experiments with GFS, correcting temperature and specific humidity online show reduction in model bias in 6-hr forecast. This approach can then be used to guide and optimize the design of sub-grid scale physical parameterizations, more accurate discretization of the model dynamics, boundary conditions, radiative transfer codes, and other potential model improvements which can then replace the empirical correction scheme. The analysis increments also provide guidance in testing new physical parameterizations.

Accuracy of short‐term sea ice drift forecasts using a coupled ice‐ocean model

PubMed Central

Zhang, Jinlun

2015-01-01

Abstract Arctic sea ice drift forecasts of 6 h–9 days for the summer of 2014 are generated using the Marginal Ice Zone Modeling and Assimilation System (MIZMAS); the model is driven by 6 h atmospheric forecasts from the Climate Forecast System (CFSv2). Forecast ice drift speed is compared to drifting buoys and other observational platforms. Forecast positions are compared with actual positions 24 h–8 days since forecast. Forecast results are further compared to those from the forecasts generated using an ice velocity climatology driven by multiyear integrations of the same model. The results are presented in the context of scheduling the acquisition of high‐resolution images that need to follow buoys or scientific research platforms. RMS errors for ice speed are on the order of 5 km/d for 24–48 h since forecast using the sea ice model compared with 9 km/d using climatology. Predicted buoy position RMS errors are 6.3 km for 24 h and 14 km for 72 h since forecast. Model biases in ice speed and direction can be reduced by adjusting the air drag coefficient and water turning angle, but the adjustments do not affect verification statistics. This suggests that improved atmospheric forecast forcing may further reduce the forecast errors. The model remains skillful for 8 days. Using the forecast model increases the probability of tracking a target drifting in sea ice with a 10 km × 10 km image from 60 to 95% for a 24 h forecast and from 27 to 73% for a 48 h forecast. PMID:27818852

Assessing Variability and Errors in Historical Runoff Forecasting with Physical Models and Alternative Data Sources

NASA Astrophysics Data System (ADS)

Penn, C. A.; Clow, D. W.; Sexstone, G. A.

2017-12-01

Water supply forecasts are an important tool for water resource managers in areas where surface water is relied on for irrigating agricultural lands and for municipal water supplies. Forecast errors, which correspond to inaccurate predictions of total surface water volume, can lead to mis-allocated water and productivity loss, thus costing stakeholders millions of dollars. The objective of this investigation is to provide water resource managers with an improved understanding of factors contributing to forecast error, and to help increase the accuracy of future forecasts. In many watersheds of the western United States, snowmelt contributes 50-75% of annual surface water flow and controls both the timing and volume of peak flow. Water supply forecasts from the Natural Resources Conservation Service (NRCS), National Weather Service, and similar cooperators use precipitation and snowpack measurements to provide water resource managers with an estimate of seasonal runoff volume. The accuracy of these forecasts can be limited by available snowpack and meteorological data. In the headwaters of the Rio Grande, NRCS produces January through June monthly Water Supply Outlook Reports. This study evaluates the accuracy of these forecasts since 1990, and examines what factors may contribute to forecast error. The Rio Grande headwaters has experienced recent changes in land cover from bark beetle infestation and a large wildfire, which can affect hydrological processes within the watershed. To investigate trends and possible contributing factors in forecast error, a semi-distributed hydrological model was calibrated and run to simulate daily streamflow for the period 1990-2015. Annual and seasonal watershed and sub-watershed water balance properties were compared with seasonal water supply forecasts. Gridded meteorological datasets were used to assess changes in the timing and volume of spring precipitation events that may contribute to forecast error. Additionally, a spatially-distributed physics-based snow model was used to assess possible effects of land cover change on snowpack properties. Trends in forecasted error are variable while baseline model results show a consistent under-prediction in the recent decade, highlighting possible compounding effects of climate and land cover changes.

Distortion Representation of Forecast Errors for Model Skill Assessment and Objective Analysis. Revision 1.12

NASA Technical Reports Server (NTRS)

Hoffman, Ross N.; Nehrkorn, Thomas; Grassotti, Christopher

1997-01-01

We proposed a novel characterization of errors for numerical weather predictions. In its simplest form we decompose the error into a part attributable to phase errors and a remainder. The phase error is represented in the same fashion as a velocity field and is required to vary slowly and smoothly with position. A general distortion representation allows for the displacement and amplification or bias correction of forecast anomalies. Characterizing and decomposing forecast error in this way has two important applications, which we term the assessment application and the objective analysis application. For the assessment application, our approach results in new objective measures of forecast skill which are more in line with subjective measures of forecast skill and which are useful in validating models and diagnosing their shortcomings. With regard to the objective analysis application, meteorological analysis schemes balance forecast error and observational error to obtain an optimal analysis. Presently, representations of the error covariance matrix used to measure the forecast error are severely limited. For the objective analysis application our approach will improve analyses by providing a more realistic measure of the forecast error. We expect, a priori, that our approach should greatly improve the utility of remotely sensed data which have relatively high horizontal resolution, but which are indirectly related to the conventional atmospheric variables. In this project, we are initially focusing on the assessment application, restricted to a realistic but univariate 2-dimensional situation. Specifically, we study the forecast errors of the sea level pressure (SLP) and 500 hPa geopotential height fields for forecasts of the short and medium range. Since the forecasts are generated by the GEOS (Goddard Earth Observing System) data assimilation system with and without ERS 1 scatterometer data, these preliminary studies serve several purposes. They (1) provide a testbed for the use of the distortion representation of forecast errors, (2) act as one means of validating the GEOS data assimilation system and (3) help to describe the impact of the ERS 1 scatterometer data.

Decomposition of Sources of Errors in Seasonal Streamflow Forecasting over the U.S. Sunbelt

NASA Technical Reports Server (NTRS)

Mazrooei, Amirhossein; Sinah, Tusshar; Sankarasubramanian, A.; Kumar, Sujay V.; Peters-Lidard, Christa D.

2015-01-01

Seasonal streamflow forecasts, contingent on climate information, can be utilized to ensure water supply for multiple uses including municipal demands, hydroelectric power generation, and for planning agricultural operations. However, uncertainties in the streamflow forecasts pose significant challenges in their utilization in real-time operations. In this study, we systematically decompose various sources of errors in developing seasonal streamflow forecasts from two Land Surface Models (LSMs) (Noah3.2 and CLM2), which are forced with downscaled and disaggregated climate forecasts. In particular, the study quantifies the relative contributions of the sources of errors from LSMs, climate forecasts, and downscaling/disaggregation techniques in developing seasonal streamflow forecast. For this purpose, three month ahead seasonal precipitation forecasts from the ECHAM4.5 general circulation model (GCM) were statistically downscaled from 2.8deg to 1/8deg spatial resolution using principal component regression (PCR) and then temporally disaggregated from monthly to daily time step using kernel-nearest neighbor (K-NN) approach. For other climatic forcings, excluding precipitation, we considered the North American Land Data Assimilation System version 2 (NLDAS-2) hourly climatology over the years 1979 to 2010. Then the selected LSMs were forced with precipitation forecasts and NLDAS-2 hourly climatology to develop retrospective seasonal streamflow forecasts over a period of 20 years (1991-2010). Finally, the performance of LSMs in forecasting streamflow under different schemes was analyzed to quantify the relative contribution of various sources of errors in developing seasonal streamflow forecast. Our results indicate that the most dominant source of errors during winter and fall seasons is the errors due to ECHAM4.5 precipitation forecasts, while temporal disaggregation scheme contributes to maximum errors during summer season.

Development of a drought forecasting model for the Asia-Pacific region using remote sensing and climate data: Focusing on Indonesia

NASA Astrophysics Data System (ADS)

Rhee, Jinyoung; Kim, Gayoung; Im, Jungho

2017-04-01

Three regions of Indonesia with different rainfall characteristics were chosen to develop drought forecast models based on machine learning. The 6-month Standardized Precipitation Index (SPI6) was selected as the target variable. The models' forecast skill was compared to the skill of long-range climate forecast models in terms of drought accuracy and regression mean absolute error (MAE). Indonesian droughts are known to be related to El Nino Southern Oscillation (ENSO) variability despite of regional differences as well as monsoon, local sea surface temperature (SST), other large-scale atmosphere-ocean interactions such as Indian Ocean Dipole (IOD) and Southern Pacific Convergence Zone (SPCZ), and local factors including topography and elevation. Machine learning models are thus to enhance drought forecast skill by combining local and remote SST and remote sensing information reflecting initial drought conditions to the long-range climate forecast model results. A total of 126 machine learning models were developed for the three regions of West Java (JB), West Sumatra (SB), and Gorontalo (GO) and six long-range climate forecast models of MSC_CanCM3, MSC_CanCM4, NCEP, NASA, PNU, POAMA as well as one climatology model based on remote sensing precipitation data, and 1 to 6-month lead times. When compared the results between the machine learning models and the long-range climate forecast models, West Java and Gorontalo regions showed similar characteristics in terms of drought accuracy. Drought accuracy of the long-range climate forecast models were generally higher than the machine learning models with short lead times but the opposite appeared for longer lead times. For West Sumatra, however, the machine learning models and the long-range climate forecast models showed similar drought accuracy. The machine learning models showed smaller regression errors for all three regions especially with longer lead times. Among the three regions, the machine learning models developed for Gorontalo showed the highest drought accuracy and the lowest regression error. West Java showed higher drought accuracy compared to West Sumatra, while West Sumatra showed lower regression error compared to West Java. The lower error in West Sumatra may be because of the smaller sample size used for training and evaluation for the region. Regional differences of forecast skill are determined by the effect of ENSO and the following forecast skill of the long-range climate forecast models. While shown somewhat high in West Sumatra, relative importance of remote sensing variables was mostly low in most cases. High importance of the variables based on long-range climate forecast models indicates that the forecast skill of the machine learning models are mostly determined by the forecast skill of the climate models.

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.

Semiparametric modeling: Correcting low-dimensional model error in parametric models

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

Berry, Tyrus, E-mail: thb11@psu.edu; Harlim, John, E-mail: jharlim@psu.edu; Department of Meteorology, the Pennsylvania State University, 503 Walker Building, University Park, PA 16802-5013

2016-03-01

In this paper, a semiparametric modeling approach is introduced as a paradigm for addressing model error arising from unresolved physical phenomena. Our approach compensates for model error by learning an auxiliary dynamical model for the unknown parameters. Practically, the proposed approach consists of the following steps. Given a physics-based model and a noisy data set of historical observations, a Bayesian filtering algorithm is used to extract a time-series of the parameter values. Subsequently, the diffusion forecast algorithm is applied to the retrieved time-series in order to construct the auxiliary model for the time evolving parameters. The semiparametric forecasting algorithm consistsmore » of integrating the existing physics-based model with an ensemble of parameters sampled from the probability density function of the diffusion forecast. To specify initial conditions for the diffusion forecast, a Bayesian semiparametric filtering method that extends the Kalman-based filtering framework is introduced. In difficult test examples, which introduce chaotically and stochastically evolving hidden parameters into the Lorenz-96 model, we show that our approach can effectively compensate for model error, with forecasting skill comparable to that of the perfect model.« less

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.

Randomly correcting model errors in the ARPEGE-Climate v6.1 component of CNRM-CM: applications for seasonal forecasts

NASA Astrophysics Data System (ADS)

Batté, Lauriane; Déqué, Michel

2016-06-01

Stochastic methods are increasingly used in global coupled model climate forecasting systems to account for model uncertainties. In this paper, we describe in more detail the stochastic dynamics technique introduced by Batté and Déqué (2012) in the ARPEGE-Climate atmospheric model. We present new results with an updated version of CNRM-CM using ARPEGE-Climate v6.1, and show that the technique can be used both as a means of analyzing model error statistics and accounting for model inadequacies in a seasonal forecasting framework.The perturbations are designed as corrections of model drift errors estimated from a preliminary weakly nudged re-forecast run over an extended reference period of 34 boreal winter seasons. A detailed statistical analysis of these corrections is provided, and shows that they are mainly made of intra-month variance, thereby justifying their use as in-run perturbations of the model in seasonal forecasts. However, the interannual and systematic error correction terms cannot be neglected. Time correlation of the errors is limited, but some consistency is found between the errors of up to 3 consecutive days.These findings encourage us to test several settings of the random draws of perturbations in seasonal forecast mode. Perturbations are drawn randomly but consistently for all three prognostic variables perturbed. We explore the impact of using monthly mean perturbations throughout a given forecast month in a first ensemble re-forecast (SMM, for stochastic monthly means), and test the use of 5-day sequences of perturbations in a second ensemble re-forecast (S5D, for stochastic 5-day sequences). Both experiments are compared in the light of a REF reference ensemble with initial perturbations only. Results in terms of forecast quality are contrasted depending on the region and variable of interest, but very few areas exhibit a clear degradation of forecasting skill with the introduction of stochastic dynamics. We highlight some positive impacts of the method, mainly on Northern Hemisphere extra-tropics. The 500 hPa geopotential height bias is reduced, and improvements project onto the representation of North Atlantic weather regimes. A modest impact on ensemble spread is found over most regions, which suggests that this method could be complemented by other stochastic perturbation techniques in seasonal forecasting mode.

Forecasting in foodservice: model development, testing, and evaluation.

PubMed

Miller, J L; Thompson, P A; Orabella, M M

1991-05-01

This study was designed to develop, test, and evaluate mathematical models appropriate for forecasting menu-item production demand in foodservice. Data were collected from residence and dining hall foodservices at Ohio State University. Objectives of the study were to collect, code, and analyze the data; develop and test models using actual operation data; and compare forecasting results with current methods in use. Customer count was forecast using deseasonalized simple exponential smoothing. Menu-item demand was forecast by multiplying the count forecast by a predicted preference statistic. Forecasting models were evaluated using mean squared error, mean absolute deviation, and mean absolute percentage error techniques. All models were more accurate than current methods. A broad spectrum of forecasting techniques could be used by foodservice managers with access to a personal computer and spread-sheet and database-management software. The findings indicate that mathematical forecasting techniques may be effective in foodservice operations to control costs, increase productivity, and maximize profits.

Combining empirical approaches and error modelling to enhance predictive uncertainty estimation in extrapolation for operational flood forecasting. Tests on flood events on the Loire basin, France.

NASA Astrophysics Data System (ADS)

Berthet, Lionel; Marty, Renaud; Bourgin, François; Viatgé, Julie; Piotte, Olivier; Perrin, Charles

2017-04-01

An increasing number of operational flood forecasting centres assess the predictive uncertainty associated with their forecasts and communicate it to the end users. This information can match the end-users needs (i.e. prove to be useful for an efficient crisis management) only if it is reliable: reliability is therefore a key quality for operational flood forecasts. In 2015, the French flood forecasting national and regional services (Vigicrues network; www.vigicrues.gouv.fr) implemented a framework to compute quantitative discharge and water level forecasts and to assess the predictive uncertainty. Among the possible technical options to achieve this goal, a statistical analysis of past forecasting errors of deterministic models has been selected (QUOIQUE method, Bourgin, 2014). It is a data-based and non-parametric approach based on as few assumptions as possible about the forecasting error mathematical structure. In particular, a very simple assumption is made regarding the predictive uncertainty distributions for large events outside the range of the calibration data: the multiplicative error distribution is assumed to be constant, whatever the magnitude of the flood. Indeed, the predictive distributions may not be reliable in extrapolation. However, estimating the predictive uncertainty for these rare events is crucial when major floods are of concern. In order to improve the forecasts reliability for major floods, an attempt at combining the operational strength of the empirical statistical analysis and a simple error modelling is done. Since the heteroscedasticity of forecast errors can considerably weaken the predictive reliability for large floods, this error modelling is based on the log-sinh transformation which proved to reduce significantly the heteroscedasticity of the transformed error in a simulation context, even for flood peaks (Wang et al., 2012). Exploratory tests on some operational forecasts issued during the recent floods experienced in France (major spring floods in June 2016 on the Loire river tributaries and flash floods in fall 2016) will be shown and discussed. References Bourgin, F. (2014). How to assess the predictive uncertainty in hydrological modelling? An exploratory work on a large sample of watersheds, AgroParisTech Wang, Q. J., Shrestha, D. L., Robertson, D. E. and Pokhrel, P (2012). A log-sinh transformation for data normalization and variance stabilization. Water Resources Research, , W05514, doi:10.1029/2011WR010973

Stratospheric Assimilation of Chemical Tracer Observations Using a Kalman Filter. Pt. 2; Chi-Square Validated Results and Analysis of Variance and Correlation Dynamics

NASA Technical Reports Server (NTRS)

Menard, Richard; Chang, Lang-Ping

1998-01-01

A Kalman filter system designed for the assimilation of limb-sounding observations of stratospheric chemical tracers, which has four tunable covariance parameters, was developed in Part I (Menard et al. 1998) The assimilation results of CH4 observations from the Cryogenic Limb Array Etalon Sounder instrument (CLAES) and the Halogen Observation Experiment instrument (HALOE) on board of the Upper Atmosphere Research Satellite are described in this paper. A robust (chi)(sup 2) criterion, which provides a statistical validation of the forecast and observational error covariances, was used to estimate the tunable variance parameters of the system. In particular, an estimate of the model error variance was obtained. The effect of model error on the forecast error variance became critical after only three days of assimilation of CLAES observations, although it took 14 days of forecast to double the initial error variance. We further found that the model error due to numerical discretization as arising in the standard Kalman filter algorithm, is comparable in size to the physical model error due to wind and transport modeling errors together. Separate assimilations of CLAES and HALOE observations were compared to validate the state estimate away from the observed locations. A wave-breaking event that took place several thousands of kilometers away from the HALOE observation locations was well captured by the Kalman filter due to highly anisotropic forecast error correlations. The forecast error correlation in the assimilation of the CLAES observations was found to have a structure similar to that in pure forecast mode except for smaller length scales. Finally, we have conducted an analysis of the variance and correlation dynamics to determine their relative importance in chemical tracer assimilation problems. Results show that the optimality of a tracer assimilation system depends, for the most part, on having flow-dependent error correlation rather than on evolving the error variance.

Impact of Representing Model Error in a Hybrid Ensemble-Variational Data Assimilation System for Track Forecast of Tropical Cyclones over the Bay of Bengal

NASA Astrophysics Data System (ADS)

Kutty, Govindan; Muraleedharan, Rohit; Kesarkar, Amit P.

2018-03-01

Uncertainties in the numerical weather prediction models are generally not well-represented in ensemble-based data assimilation (DA) systems. The performance of an ensemble-based DA system becomes suboptimal, if the sources of error are undersampled in the forecast system. The present study examines the effect of accounting for model error treatments in the hybrid ensemble transform Kalman filter—three-dimensional variational (3DVAR) DA system (hybrid) in the track forecast of two tropical cyclones viz. Hudhud and Thane, formed over the Bay of Bengal, using Advanced Research Weather Research and Forecasting (ARW-WRF) model. We investigated the effect of two types of model error treatment schemes and their combination on the hybrid DA system; (i) multiphysics approach, which uses different combination of cumulus, microphysics and planetary boundary layer schemes, (ii) stochastic kinetic energy backscatter (SKEB) scheme, which perturbs the horizontal wind and potential temperature tendencies, (iii) a combination of both multiphysics and SKEB scheme. Substantial improvements are noticed in the track positions of both the cyclones, when flow-dependent ensemble covariance is used in 3DVAR framework. Explicit model error representation is found to be beneficial in treating the underdispersive ensembles. Among the model error schemes used in this study, a combination of multiphysics and SKEB schemes has outperformed the other two schemes with improved track forecast for both the tropical cyclones.

Implementation of bayesian model averaging on the weather data forecasting applications utilizing open weather map

NASA Astrophysics Data System (ADS)

Rahmat, R. F.; Nasution, F. R.; Seniman; Syahputra, M. F.; Sitompul, O. S.

2018-02-01

Weather is condition of air in a certain region at a relatively short period of time, measured with various parameters such as; temperature, air preasure, wind velocity, humidity and another phenomenons in the atmosphere. In fact, extreme weather due to global warming would lead to drought, flood, hurricane and other forms of weather occasion, which directly affects social andeconomic activities. Hence, a forecasting technique is to predict weather with distinctive output, particullary mapping process based on GIS with information about current weather status in certain cordinates of each region with capability to forecast for seven days afterward. Data used in this research are retrieved in real time from the server openweathermap and BMKG. In order to obtain a low error rate and high accuracy of forecasting, the authors use Bayesian Model Averaging (BMA) method. The result shows that the BMA method has good accuracy. Forecasting error value is calculated by mean square error shows (MSE). The error value emerges at minumum temperature rated at 0.28 and maximum temperature rated at 0.15. Meanwhile, the error value of minimum humidity rates at 0.38 and the error value of maximum humidity rates at 0.04. Afterall, the forecasting error rate of wind speed is at 0.076. The lower the forecasting error rate, the more optimized the accuracy is.

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.

Evaluation of the product ratio coherent model in forecasting mortality rates and life expectancy at births by States

NASA Astrophysics Data System (ADS)

Shair, Syazreen Niza; Yusof, Aida Yuzi; Asmuni, Nurin Haniah

2017-05-01

Coherent mortality forecasting models have recently received increasing attention particularly in their application to sub-populations. The advantage of coherent models over independent models is the ability to forecast a non-divergent mortality for two or more sub-populations. One of the coherent models was recently developed by [1] known as the product-ratio model. This model is an extension version of the functional independent model from [2]. The product-ratio model has been applied in a developed country, Australia [1] and has been extended in a developing nation, Malaysia [3]. While [3] accounted for coherency of mortality rates between gender and ethnic group, the coherency between states in Malaysia has never been explored. This paper will forecast the mortality rates of Malaysian sub-populations according to states using the product ratio coherent model and its independent version— the functional independent model. The forecast accuracies of two different models are evaluated using the out-of-sample error measurements— the mean absolute forecast error (MAFE) for age-specific death rates and the mean forecast error (MFE) for the life expectancy at birth. We employ Malaysian mortality time series data from 1991 to 2014, segregated by age, gender and states.

Verification of different forecasts of Hungarian Meteorological Service

NASA Astrophysics Data System (ADS)

Feher, B.

2009-09-01

In this paper I show the results of the forecasts made by the Hungarian Meteorological Service. I focus on the general short- and medium-range forecasts, which contains cloudiness, precipitation, wind speed and temperature for six regions of Hungary. I would like to show the results of some special forecasts as well, such as precipitation predictions which are made for the catchment area of Danube and Tisza rivers, and daily mean temperature predictions used by Hungarian energy companies. The product received by the user is made by the general forecaster, but these predictions are based on the ALADIN and ECMWF outputs. Because of these, the product of the forecaster and the models were also verified. Method like this is able to show us, which weather elements are more difficult to forecast or which regions have higher errors. During the verification procedure the basic errors (mean error, mean absolute error) are calculated. Precipitation amount is classified into five categories, and scores like POD, TS, PC,…etc. were defined by contingency table determined by these categories. The procedure runs fully automatically, all the things forecasters have to do is to print the daily result each morning. Beside the daily result, verification is also made for longer periods like week, month or year. Analyzing the results of longer periods we can say that the best predictions are made for the first few days, and precipitation forecasts are less good for mountainous areas, even, the scores of the forecasters sometimes are higher than the errors of the models. Since forecaster receive results next day, it can helps him/her to reduce mistakes and learn the weakness of the models. This paper contains the verification scores, their trends, the method by which these scores are calculated, and some case studies on worse forecasts.

Effect of Streamflow Forecast Uncertainty on Real-Time Reservoir Operation

NASA Astrophysics Data System (ADS)

Zhao, T.; Cai, X.; Yang, D.

2010-12-01

Various hydrological forecast products have been applied to real-time reservoir operation, including deterministic streamflow forecast (DSF), DSF-based probabilistic streamflow forecast (DPSF), and ensemble streamflow forecast (ESF), which represent forecast uncertainty in the form of deterministic forecast error, deterministic forecast error-based uncertainty distribution, and ensemble forecast errors, respectively. Compared to previous studies that treat these forecast products as ad hoc inputs for reservoir operation models, this paper attempts to model the uncertainties involved in the various forecast products and explores their effect on real-time reservoir operation decisions. In hydrology, there are various indices reflecting the magnitude of streamflow forecast uncertainty; meanwhile, few models illustrate the forecast uncertainty evolution process. This research introduces Martingale Model of Forecast Evolution (MMFE) from supply chain management and justifies its assumptions for quantifying the evolution of uncertainty in streamflow forecast as time progresses. Based on MMFE, this research simulates the evolution of forecast uncertainty in DSF, DPSF, and ESF, and applies the reservoir operation models (dynamic programming, DP; stochastic dynamic programming, SDP; and standard operation policy, SOP) to assess the effect of different forms of forecast uncertainty on real-time reservoir operation. Through a hypothetical single-objective real-time reservoir operation model, the results illustrate that forecast uncertainty exerts significant effects. Reservoir operation efficiency, as measured by a utility function, decreases as the forecast uncertainty increases. Meanwhile, these effects also depend on the type of forecast product being used. In general, the utility of reservoir operation with ESF is nearly as high as the utility obtained with a perfect forecast; the utilities of DSF and DPSF are similar to each other but not as efficient as ESF. Moreover, streamflow variability and reservoir capacity can change the magnitude of the effects of forecast uncertainty, but not the relative merit of DSF, DPSF, and ESF. Schematic diagram of the increase in forecast uncertainty with forecast lead-time and the dynamic updating property of real-time streamflow forecast

Sensitivity of mesoscale-model forecast skill to some initial-data characteristics, data density, data position, analysis procedure and measurement error

NASA Technical Reports Server (NTRS)

Warner, Thomas T.; Key, Lawrence E.; Lario, Annette M.

1989-01-01

The effects of horizontal and vertical data resolution, data density, data location, different objective analysis algorithms, and measurement error on mesoscale-forecast accuracy are studied with observing-system simulation experiments. Domain-averaged errors are shown to generally decrease with time. It is found that the vertical distribution of error growth depends on the initial vertical distribution of the error itself. Larger gravity-inertia wave noise is produced in forecasts with coarser vertical data resolution. The use of a low vertical resolution observing system with three data levels leads to more forecast errors than moderate and high vertical resolution observing systems with 8 and 14 data levels. Also, with poor vertical resolution in soundings, the initial and forecast errors are not affected by the horizontal data resolution.

Joint Seasonal ARMA Approach for Modeling of Load Forecast Errors in Planning Studies

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

Hafen, Ryan P.; Samaan, Nader A.; Makarov, Yuri V.

2014-04-14

To make informed and robust decisions in the probabilistic power system operation and planning process, it is critical to conduct multiple simulations of the generated combinations of wind and load parameters and their forecast errors to handle the variability and uncertainty of these time series. In order for the simulation results to be trustworthy, the simulated series must preserve the salient statistical characteristics of the real series. In this paper, we analyze day-ahead load forecast error data from multiple balancing authority locations and characterize statistical properties such as mean, standard deviation, autocorrelation, correlation between series, time-of-day bias, and time-of-day autocorrelation.more » We then construct and validate a seasonal autoregressive moving average (ARMA) model to model these characteristics, and use the model to jointly simulate day-ahead load forecast error series for all BAs.« less

[Application of wavelet neural networks model to forecast incidence of syphilis].

PubMed

Zhou, Xian-Feng; Feng, Zi-Jian; Yang, Wei-Zhong; Li, Xiao-Song

2011-07-01

To apply Wavelet Neural Networks (WNN) model to forecast incidence of Syphilis. Back Propagation Neural Network (BPNN) and WNN were developed based on the monthly incidence of Syphilis in Sichuan province from 2004 to 2008. The accuracy of forecast was compared between the two models. In the training approximation, the mean absolute error (MAE), rooted mean square error (RMSE) and mean absolute percentage error (MAPE) were 0.0719, 0.0862 and 11.52% respectively for WNN, and 0.0892, 0.1183 and 14.87% respectively for BPNN. The three indexes for generalization of models were 0.0497, 0.0513 and 4.60% for WNN, and 0.0816, 0.1119 and 7.25% for BPNN. WNN is a better model for short-term forecasting of Syphilis.

Towards an Australian ensemble streamflow forecasting system for flood prediction and water management

NASA Astrophysics Data System (ADS)

Bennett, J.; David, R. E.; Wang, Q.; Li, M.; Shrestha, D. L.

2016-12-01

Flood forecasting in Australia has historically relied on deterministic forecasting models run only when floods are imminent, with considerable forecaster input and interpretation. These now co-existed with a continually available 7-day streamflow forecasting service (also deterministic) aimed at operational water management applications such as environmental flow releases. The 7-day service is not optimised for flood prediction. We describe progress on developing a system for ensemble streamflow forecasting that is suitable for both flood prediction and water management applications. Precipitation uncertainty is handled through post-processing of Numerical Weather Prediction (NWP) output with a Bayesian rainfall post-processor (RPP). The RPP corrects biases, downscales NWP output, and produces reliable ensemble spread. Ensemble precipitation forecasts are used to force a semi-distributed conceptual rainfall-runoff model. Uncertainty in precipitation forecasts is insufficient to reliably describe streamflow forecast uncertainty, particularly at shorter lead-times. We characterise hydrological prediction uncertainty separately with a 4-stage error model. The error model relies on data transformation to ensure residuals are homoscedastic and symmetrically distributed. To ensure streamflow forecasts are accurate and reliable, the residuals are modelled using a mixture-Gaussian distribution with distinct parameters for the rising and falling limbs of the forecast hydrograph. In a case study of the Murray River in south-eastern Australia, we show ensemble predictions of floods generally have lower errors than deterministic forecasting methods. We also discuss some of the challenges in operationalising short-term ensemble streamflow forecasts in Australia, including meeting the needs for accurate predictions across all flow ranges and comparing forecasts generated by event and continuous hydrological models.

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.

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

An Extended Objective Evaluation of the 29-km Eta Model for Weather Support to the United States Space Program

NASA Technical Reports Server (NTRS)

Nutter, Paul; Manobianco, John

1998-01-01

This report describes the Applied Meteorology Unit's objective verification of the National Centers for Environmental Prediction 29-km eta model during separate warm and cool season periods from May 1996 through January 1998. The verification of surface and upper-air point forecasts was performed at three selected stations important for 45th Weather Squadron, Spaceflight Meteorology Group, and National Weather Service, Melbourne operational weather concerns. The statistical evaluation identified model biases that may result from inadequate parameterization of physical processes. Since model biases are relatively small compared to the random error component, most of the total model error results from day-to-day variability in the forecasts and/or observations. To some extent, these nonsystematic errors reflect the variability in point observations that sample spatial and temporal scales of atmospheric phenomena that cannot be resolved by the model. On average, Meso-Eta point forecasts provide useful guidance for predicting the evolution of the larger scale environment. A more substantial challenge facing model users in real time is the discrimination of nonsystematic errors that tend to inflate the total forecast error. It is important that model users maintain awareness of ongoing model changes. Such changes are likely to modify the basic error characteristics, particularly near the surface.

A channel dynamics model for real-time flood forecasting

USGS Publications Warehouse

Hoos, Anne B.; Koussis, Antonis D.; Beale, Guy O.

1989-01-01

A new channel dynamics scheme (alternative system predictor in real time (ASPIRE)), designed specifically for real-time river flow forecasting, is introduced to reduce uncertainty in the forecast. ASPIRE is a storage routing model that limits the influence of catchment model forecast errors to the downstream station closest to the catchment. Comparisons with the Muskingum routing scheme in field tests suggest that the ASPIRE scheme can provide more accurate forecasts, probably because discharge observations are used to a maximum advantage and routing reaches (and model errors in each reach) are uncoupled. Using ASPIRE in conjunction with the Kalman filter did not improve forecast accuracy relative to a deterministic updating procedure. Theoretical analysis suggests that this is due to a large process noise to measurement noise ratio.

Performance of univariate forecasting on seasonal diseases: the case of tuberculosis.

PubMed

Permanasari, Adhistya Erna; Rambli, Dayang Rohaya Awang; Dominic, P Dhanapal Durai

2011-01-01

The annual disease incident worldwide is desirable to be predicted for taking appropriate policy to prevent disease outbreak. This chapter considers the performance of different forecasting method to predict the future number of disease incidence, especially for seasonal disease. Six forecasting methods, namely linear regression, moving average, decomposition, Holt-Winter's, ARIMA, and artificial neural network (ANN), were used for disease forecasting on tuberculosis monthly data. The model derived met the requirement of time series with seasonality pattern and downward trend. The forecasting performance was compared using similar error measure in the base of the last 5 years forecast result. The findings indicate that ARIMA model was the most appropriate model since it obtained the less relatively error than the other model.

Dispersion Modeling Using Ensemble Forecasts Compared to ETEX Measurements.

NASA Astrophysics Data System (ADS)

Straume, Anne Grete; N'dri Koffi, Ernest; Nodop, Katrin

1998-11-01

Numerous numerical models are developed to predict long-range transport of hazardous air pollution in connection with accidental releases. When evaluating and improving such a model, it is important to detect uncertainties connected to the meteorological input data. A Lagrangian dispersion model, the Severe Nuclear Accident Program, is used here to investigate the effect of errors in the meteorological input data due to analysis error. An ensemble forecast, produced at the European Centre for Medium-Range Weather Forecasts, is then used as model input. The ensemble forecast members are generated by perturbing the initial meteorological fields of the weather forecast. The perturbations are calculated from singular vectors meant to represent possible forecast developments generated by instabilities in the atmospheric flow during the early part of the forecast. The instabilities are generated by errors in the analyzed fields. Puff predictions from the dispersion model, using ensemble forecast input, are compared, and a large spread in the predicted puff evolutions is found. This shows that the quality of the meteorological input data is important for the success of the dispersion model. In order to evaluate the dispersion model, the calculations are compared with measurements from the European Tracer Experiment. The model manages to predict the measured puff evolution concerning shape and time of arrival to a fairly high extent, up to 60 h after the start of the release. The modeled puff is still too narrow in the advection direction.

Predictability of Solar Radiation for Photovoltaics systems over Europe: from short-term to seasonal time-scales

NASA Astrophysics Data System (ADS)

De Felice, Matteo; Petitta, Marcello; Ruti, Paolo

2014-05-01

Photovoltaic diffusion is steadily growing on Europe, passing from a capacity of almost 14 GWp in 2011 to 21.5 GWp in 2012 [1]. Having accurate forecast is needed for planning and operational purposes, with the possibility to model and predict solar variability at different time-scales. This study examines the predictability of daily surface solar radiation comparing ECMWF operational forecasts with CM-SAF satellite measurements on the Meteosat (MSG) full disk domain. Operational forecasts used are the IFS system up to 10 days and the System4 seasonal forecast up to three months. Forecast are analysed considering average and variance of errors, showing error maps and average on specific domains with respect to prediction lead times. In all the cases, forecasts are compared with predictions obtained using persistence and state-of-art time-series models. We can observe a wide range of errors, with the performance of forecasts dramatically affected by orography and season. Lower errors are on southern Italy and Spain, with errors on some areas consistently under 10% up to ten days during summer (JJA). Finally, we conclude the study with some insight on how to "translate" the error on solar radiation to error on solar power production using available production data from solar power plants. [1] EurObserver, "Baromètre Photovoltaïque, Le journal des énergies renouvables, April 2012."

Distortion Representation of Forecast Errors for Model Skill Assessment and Objective Analysis

NASA Technical Reports Server (NTRS)

Hoffman, Ross N.; Nehrkorn, Thomas; Grassotti, Christopher

1998-01-01

We proposed a novel characterization of errors for numerical weather predictions. A general distortion representation allows for the displacement and amplification or bias correction of forecast anomalies. Characterizing and decomposing forecast error in this way has several important applications, including the model assessment application and the objective analysis application. In this project, we have focused on the assessment application, restricted to a realistic but univariate 2-dimensional situation. Specifically, we study the forecast errors of the sea level pressure (SLP), the 500 hPa geopotential height, and the 315 K potential vorticity fields for forecasts of the short and medium range. The forecasts are generated by the Goddard Earth Observing System (GEOS) data assimilation system with and without ERS-1 scatterometer data. A great deal of novel work has been accomplished under the current contract. In broad terms, we have developed and tested an efficient algorithm for determining distortions. The algorithm and constraints are now ready for application to larger data sets to be used to determine the statistics of the distortion as outlined above, and to be applied in data analysis by using GEOS water vapor imagery to correct short-term forecast errors.

The GEOS Ozone Data Assimilation System: Specification of Error Statistics

NASA Technical Reports Server (NTRS)

Stajner, Ivanka; Riishojgaard, Lars Peter; Rood, Richard B.

2000-01-01

A global three-dimensional ozone data assimilation system has been developed at the Data Assimilation Office of the NASA/Goddard Space Flight Center. The Total Ozone Mapping Spectrometer (TOMS) total ozone and the Solar Backscatter Ultraviolet (SBUV) or (SBUV/2) partial ozone profile observations are assimilated. The assimilation, into an off-line ozone transport model, is done using the global Physical-space Statistical Analysis Scheme (PSAS). This system became operational in December 1999. A detailed description of the statistical analysis scheme, and in particular, the forecast and observation error covariance models is given. A new global anisotropic horizontal forecast error correlation model accounts for a varying distribution of observations with latitude. Correlations are largest in the zonal direction in the tropics where data is sparse. Forecast error variance model is proportional to the ozone field. The forecast error covariance parameters were determined by maximum likelihood estimation. The error covariance models are validated using x squared statistics. The analyzed ozone fields in the winter 1992 are validated against independent observations from ozone sondes and HALOE. There is better than 10% agreement between mean Halogen Occultation Experiment (HALOE) and analysis fields between 70 and 0.2 hPa. The global root-mean-square (RMS) difference between TOMS observed and forecast values is less than 4%. The global RMS difference between SBUV observed and analyzed ozone between 50 and 3 hPa is less than 15%.

Status of the NASA GMAO Observing System Simulation Experiment

NASA Technical Reports Server (NTRS)

Prive, Nikki C.; Errico, Ronald M.

2014-01-01

An Observing System Simulation Experiment (OSSE) is a pure modeling study used when actual observations are too expensive or difficult to obtain. OSSEs are valuable tools for determining the potential impact of new observing systems on numerical weather forecasts and for evaluation of data assimilation systems (DAS). An OSSE has been developed at the NASA Global Modeling and Assimilation Office (GMAO, Errico et al 2013). The GMAO OSSE uses a 13-month integration of the European Centre for Medium- Range Weather Forecasts 2005 operational model at T511/L91 resolution for the Nature Run (NR). Synthetic observations have been updated so that they are based on real observations during the summer of 2013. The emulated observation types include AMSU-A, MHS, IASI, AIRS, and HIRS4 radiance data, GPS-RO, and conventional types including aircraft, rawinsonde, profiler, surface, and satellite winds. The synthetic satellite wind observations are colocated with the NR cloud fields, and the rawinsondes are advected during ascent using the NR wind fields. Data counts for the synthetic observations are matched as closely as possible to real data counts, as shown in Figure 2. Errors are added to the synthetic observations to emulate representativeness and instrument errors. The synthetic errors are calibrated so that the statistics of observation innovation and analysis increments in the OSSE are similar to the same statistics for assimilation of real observations, in an iterative method described by Errico et al (2013). The standard deviations of observation minus forecast (xo-H(xb)) are compared for the OSSE and real data in Figure 3. The synthetic errors include both random, uncorrelated errors, and an additional correlated error component for some observational types. Vertically correlated errors are included for conventional sounding data and GPS-RO, and channel correlated errors are introduced to AIRS and IASI (Figure 4). HIRS, AMSU-A, and MHS have a component of horizontally correlated error. The forecast model used by the GMAO OSSE is the Goddard Earth Observing System Model, Version 5 (GEOS-5) with Gridpoint Statistical Interpolation (GSI) DAS. The model version has been updated to v. 5.13.3, corresponding to the current operational model. Forecasts are run on a cube-sphere grid with 180 points along each edge of the cube (approximately 0.5 degree horizontal resolution) with 72 vertical levels. The DAS is cycled at 6-hour intervals, with 240 hour forecasts launched daily at 0000 UTC. Evaluation of the forecasting skill for July and August is currently underway. Prior versions of the GMAO OSSE have been found to have greater forecasting skill than real world forecasts. It is anticipated that similar forecast skill will be found in the updated OSSE.

Quantifying model uncertainty in seasonal Arctic sea-ice forecasts

NASA Astrophysics Data System (ADS)

Blanchard-Wrigglesworth, Edward; Barthélemy, Antoine; Chevallier, Matthieu; Cullather, Richard; Fučkar, Neven; Massonnet, François; Posey, Pamela; Wang, Wanqiu; Zhang, Jinlun; Ardilouze, Constantin; Bitz, Cecilia; Vernieres, Guillaume; Wallcraft, Alan; Wang, Muyin

2017-04-01

Dynamical model forecasts in the Sea Ice Outlook (SIO) of September Arctic sea-ice extent over the last decade have shown lower skill than that found in both idealized model experiments and hindcasts of previous decades. Additionally, it is unclear how different model physics, initial conditions or post-processing techniques contribute to SIO forecast uncertainty. In this work, we have produced a seasonal forecast of 2015 Arctic summer sea ice using SIO dynamical models initialized with identical sea-ice thickness in the central Arctic. Our goals are to calculate the relative contribution of model uncertainty and irreducible error growth to forecast uncertainty and assess the importance of post-processing, and to contrast pan-Arctic forecast uncertainty with regional forecast uncertainty. We find that prior to forecast post-processing, model uncertainty is the main contributor to forecast uncertainty, whereas after forecast post-processing forecast uncertainty is reduced overall, model uncertainty is reduced by an order of magnitude, and irreducible error growth becomes the main contributor to forecast uncertainty. While all models generally agree in their post-processed forecasts of September sea-ice volume and extent, this is not the case for sea-ice concentration. Additionally, forecast uncertainty of sea-ice thickness grows at a much higher rate along Arctic coastlines relative to the central Arctic ocean. Potential ways of offering spatial forecast information based on the timescale over which the forecast signal beats the noise are also explored.

Growth of Errors and Uncertainties in Medium Range Ensemble Forecasts of U.S. East Coast Cool Season Extratropical Cyclones

NASA Astrophysics Data System (ADS)

Zheng, Minghua

Cool-season extratropical cyclones near the U.S. East Coast often have significant impacts on the safety, health, environment and economy of this most densely populated region. Hence it is of vital importance to forecast these high-impact winter storm events as accurately as possible by numerical weather prediction (NWP), including in the medium-range. Ensemble forecasts are appealing to operational forecasters when forecasting such events because they can provide an envelope of likely solutions to serve user communities. However, it is generally accepted that ensemble outputs are not used efficiently in NWS operations mainly due to the lack of simple and quantitative tools to communicate forecast uncertainties and ensemble verification to assess model errors and biases. Ensemble sensitivity analysis (ESA), which employs a linear correlation and regression between a chosen forecast metric and the forecast state vector, can be used to analyze the forecast uncertainty development for both short- and medium-range forecasts. The application of ESA to a high-impact winter storm in December 2010 demonstrated that the sensitivity signals based on different forecast metrics are robust. In particular, the ESA based on the leading two EOF PCs can separate sensitive regions associated with cyclone amplitude and intensity uncertainties, respectively. The sensitivity signals were verified using the leave-one-out cross validation (LOOCV) method based on a multi-model ensemble from CMC, ECMWF, and NCEP. The climatology of ensemble sensitivities for the leading two EOF PCs based on 3-day and 6-day forecasts of historical cyclone cases was presented. It was found that the EOF1 pattern often represents the intensity variations while the EOF2 pattern represents the track variations along west-southwest and east-northeast direction. For PC1, the upper-level trough associated with the East Coast cyclone and its downstream ridge are important to the forecast uncertainty in cyclone strength. The initial differences in forecasting the ridge along the west coast of North America impact the EOF1 pattern most. For PC2, it was shown that the shift of the tri-polar structure is most significantly related to the cyclone track forecasts. The EOF/fuzzy clustering tool was applied to diagnose the scenarios in operational ensemble forecast of East Coast winter storms. It was shown that the clustering method could efficiently separate the forecast scenarios associated with East Coast storms based on the 90-member multi-model ensemble. A scenario-based ensemble verification method has been proposed and applied it to examine the capability of different EPSs in capturing the analysis scenarios for historical East Coast cyclone cases at lead times of 1-9 days. The results suggest that the NCEP model performs better in short-range forecasts in capturing the analysis scenario although it is under-dispersed. The ECMWF ensemble shows the best performance in the medium range. The CMC model is found to show the smallest percentage of members in the analysis group and a relatively high missing rate, suggesting that it is less reliable regarding capturing the analysis scenario when compared with the other two EPSs. A combination of NCEP and CMC models has been found to reduce the missing rate and improve the error-spread skill in medium- to extended-range forecasts. Based on the orthogonal features of the EOF patterns, the model errors for 1-6-day forecasts have been decomposed for the leading two EOF patterns. The results for error decomposition show that the NCEP model tends to better represent both EOF1 and EOF2 patterns by showing less intensity and displacement errors during 1-3 days. The ECMWF model is found to have the smallest errors in both EOF1 and EOF2 patterns during 4-6 days. We have also found that East Coast cyclones in the ECMWF forecast tend to be towards the southwest of the other two models in representing the EOF2 pattern, which is associated with the southwest-northeast shifting of the cyclone. This result suggests that ECMWF model may have a tendency to show a closer-to-shore solution in forecasting East Coast winter storms. The downstream impacts of Rossby wave packets (RWPs) on the predictability of winter storms are investigated to explore the source of ensemble uncertainties. The composited RWPA anomalies show that there are enhanced RWPs propagating across the Pacific in both large-error and large-spread cases over the verification regions. There are also indications that the errors might propagate with a speed comparable with the group velocity of RWPs. Based on the composite results as well as our observations of the operation daily RWPA, a conceptual model of errors/uncertainty development associated with RWPs has been proposed to serve as a practical tool to understand the evolution of forecast errors and uncertainties associated with the coherent RWPs originating from upstream as far as western Pacific. (Abstract shortened by ProQuest.).

ERM model analysis for adaptation to hydrological model errors

NASA Astrophysics Data System (ADS)

Baymani-Nezhad, M.; Han, D.

2018-05-01

Hydrological conditions are changed continuously and these phenomenons generate errors on flood forecasting models and will lead to get unrealistic results. Therefore, to overcome these difficulties, a concept called model updating is proposed in hydrological studies. Real-time model updating is one of the challenging processes in hydrological sciences and has not been entirely solved due to lack of knowledge about the future state of the catchment under study. Basically, in terms of flood forecasting process, errors propagated from the rainfall-runoff model are enumerated as the main source of uncertainty in the forecasting model. Hence, to dominate the exciting errors, several methods have been proposed by researchers to update the rainfall-runoff models such as parameter updating, model state updating, and correction on input data. The current study focuses on investigations about the ability of rainfall-runoff model parameters to cope with three types of existing errors, timing, shape and volume as the common errors in hydrological modelling. The new lumped model, the ERM model, has been selected for this study to evaluate its parameters for its use in model updating to cope with the stated errors. Investigation about ten events proves that the ERM model parameters can be updated to cope with the errors without the need to recalibrate the model.

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.

How to improve an un-alterable model forecast? A sequential data assimilation based error updating approach

NASA Astrophysics Data System (ADS)

Gragne, A. S.; Sharma, A.; Mehrotra, R.; Alfredsen, K. T.

2012-12-01

Accuracy of reservoir inflow forecasts is instrumental for maximizing value of water resources and influences operation of hydropower reservoirs significantly. Improving hourly reservoir inflow forecasts over a 24 hours lead-time is considered with the day-ahead (Elspot) market of the Nordic exchange market in perspectives. The procedure presented comprises of an error model added on top of an un-alterable constant parameter conceptual model, and a sequential data assimilation routine. The structure of the error model was investigated using freely available software for detecting mathematical relationships in a given dataset (EUREQA) and adopted to contain minimum complexity for computational reasons. As new streamflow data become available the extra information manifested in the discrepancies between measurements and conceptual model outputs are extracted and assimilated into the forecasting system recursively using Sequential Monte Carlo technique. Besides improving forecast skills significantly, the probabilistic inflow forecasts provided by the present approach entrains suitable information for reducing uncertainty in decision making processes related to hydropower systems operation. The potential of the current procedure for improving accuracy of inflow forecasts at lead-times unto 24 hours and its reliability in different seasons of the year will be illustrated and discussed thoroughly.

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.

Multi-model seasonal forecast of Arctic sea-ice: forecast uncertainty at pan-Arctic and regional scales

NASA Astrophysics Data System (ADS)

Blanchard-Wrigglesworth, E.; Barthélemy, A.; Chevallier, M.; Cullather, R.; Fučkar, N.; Massonnet, F.; Posey, P.; Wang, W.; Zhang, J.; Ardilouze, C.; Bitz, C. M.; Vernieres, G.; Wallcraft, A.; Wang, M.

2017-08-01

Dynamical model forecasts in the Sea Ice Outlook (SIO) of September Arctic sea-ice extent over the last decade have shown lower skill than that found in both idealized model experiments and hindcasts of previous decades. Additionally, it is unclear how different model physics, initial conditions or forecast post-processing (bias correction) techniques contribute to SIO forecast uncertainty. In this work, we have produced a seasonal forecast of 2015 Arctic summer sea ice using SIO dynamical models initialized with identical sea-ice thickness in the central Arctic. Our goals are to calculate the relative contribution of model uncertainty and irreducible error growth to forecast uncertainty and assess the importance of post-processing, and to contrast pan-Arctic forecast uncertainty with regional forecast uncertainty. We find that prior to forecast post-processing, model uncertainty is the main contributor to forecast uncertainty, whereas after forecast post-processing forecast uncertainty is reduced overall, model uncertainty is reduced by an order of magnitude, and irreducible error growth becomes the main contributor to forecast uncertainty. While all models generally agree in their post-processed forecasts of September sea-ice volume and extent, this is not the case for sea-ice concentration. Additionally, forecast uncertainty of sea-ice thickness grows at a much higher rate along Arctic coastlines relative to the central Arctic ocean. Potential ways of offering spatial forecast information based on the timescale over which the forecast signal beats the noise are also explored.

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.

A hybrid ARIMA and neural network model applied to forecast catch volumes of Selar crumenophthalmus

NASA Astrophysics Data System (ADS)

Aquino, Ronald L.; Alcantara, Nialle Loui Mar T.; Addawe, Rizavel C.

2017-11-01

The Selar crumenophthalmus with the English name big-eyed scad fish, locally known as matang-baka, is one of the fishes commonly caught along the waters of La Union, Philippines. The study deals with the forecasting of catch volumes of big-eyed scad fish for commercial consumption. The data used are quarterly caught volumes of big-eyed scad fish from 2002 to first quarter of 2017. This actual data is available from the open stat database published by the Philippine Statistics Authority (PSA)whose task is to collect, compiles, analyzes and publish information concerning different aspects of the Philippine setting. Autoregressive Integrated Moving Average (ARIMA) models, Artificial Neural Network (ANN) model and the Hybrid model consisting of ARIMA and ANN were developed to forecast catch volumes of big-eyed scad fish. Statistical errors such as Mean Absolute Errors (MAE) and Root Mean Square Errors (RMSE) were computed and compared to choose the most suitable model for forecasting the catch volume for the next few quarters. A comparison of the results of each model and corresponding statistical errors reveals that the hybrid model, ARIMA-ANN (2,1,2)(6:3:1), is the most suitable model to forecast the catch volumes of the big-eyed scad fish for the next few quarters.

Forecasting Container Throughput at the Doraleh Port in Djibouti through Time Series Analysis

NASA Astrophysics Data System (ADS)

Mohamed Ismael, Hawa; Vandyck, George Kobina

The Doraleh Container Terminal (DCT) located in Djibouti has been noted as the most technologically advanced container terminal on the African continent. DCT's strategic location at the crossroads of the main shipping lanes connecting Asia, Africa and Europe put it in a unique position to provide important shipping services to vessels plying that route. This paper aims to forecast container throughput through the Doraleh Container Port in Djibouti by Time Series Analysis. A selection of univariate forecasting models has been used, namely Triple Exponential Smoothing Model, Grey Model and Linear Regression Model. By utilizing the above three models and their combination, the forecast of container throughput through the Doraleh port was realized. A comparison of the different forecasting results of the three models, in addition to the combination forecast is then undertaken, based on commonly used evaluation criteria Mean Absolute Deviation (MAD) and Mean Absolute Percentage Error (MAPE). The study found that the Linear Regression forecasting Model was the best prediction method for forecasting the container throughput, since its forecast error was the least. Based on the regression model, a ten (10) year forecast for container throughput at DCT has been made.

Forecasting of monsoon heavy rains: challenges in NWP

NASA Astrophysics Data System (ADS)

Sharma, Kuldeep; Ashrit, Raghavendra; Iyengar, Gopal; Bhatla, R.; Rajagopal, E. N.

2016-05-01

Last decade has seen a tremendous improvement in the forecasting skill of numerical weather prediction (NWP) models. This is attributed to increased sophistication in NWP models, which resolve complex physical processes, advanced data assimilation, increased grid resolution and satellite observations. However, prediction of heavy rains is still a challenge since the models exhibit large error in amounts as well as spatial and temporal distribution. Two state-of-art NWP models have been investigated over the Indian monsoon region to assess their ability in predicting the heavy rainfall events. The unified model operational at National Center for Medium Range Weather Forecasting (NCUM) and the unified model operational at the Australian Bureau of Meteorology (Australian Community Climate and Earth-System Simulator -- Global (ACCESS-G)) are used in this study. The recent (JJAS 2015) Indian monsoon season witnessed 6 depressions and 2 cyclonic storms which resulted in heavy rains and flooding. The CRA method of verification allows the decomposition of forecast errors in terms of error in the rainfall volume, pattern and location. The case by case study using CRA technique shows that contribution to the rainfall errors come from pattern and displacement is large while contribution due to error in predicted rainfall volume is least.

Bayesian Hierarchical Models to Augment the Mediterranean Forecast System

DTIC Science & Technology

2010-09-30

In part 2 (Bonazzi et al., 2010), the impact of the ensemble forecast methodology based on MFS-Wind-BHM perturbations is documented. Forecast...absence of dt data stage inputs, the forecast impact of MFS-Error-BHM is neutral. Experiments are underway now to introduce dt back into the MFS-Error...BHM and quantify forecast impacts at MFS. MFS-SuperEnsemble-BHM We have assembled all needed datasets and completed algorithmic development

Short-Term Global Horizontal Irradiance Forecasting Based on Sky Imaging and Pattern Recognition

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

Hodge, Brian S; Feng, Cong; Cui, Mingjian

Accurate short-term forecasting is crucial for solar integration in the power grid. In this paper, a classification forecasting framework based on pattern recognition is developed for 1-hour-ahead global horizontal irradiance (GHI) forecasting. Three sets of models in the forecasting framework are trained by the data partitioned from the preprocessing analysis. The first two sets of models forecast GHI for the first four daylight hours of each day. Then the GHI values in the remaining hours are forecasted by an optimal machine learning model determined based on a weather pattern classification model in the third model set. The weather pattern ismore » determined by a support vector machine (SVM) classifier. The developed framework is validated by the GHI and sky imaging data from the National Renewable Energy Laboratory (NREL). Results show that the developed short-term forecasting framework outperforms the persistence benchmark by 16% in terms of the normalized mean absolute error and 25% in terms of the normalized root mean square error.« less

A global perspective of the limits of prediction skill based on the ECMWF ensemble

NASA Astrophysics Data System (ADS)

Zagar, Nedjeljka

2016-04-01

In this talk presents a new model of the global forecast error growth applied to the forecast errors simulated by the ensemble prediction system (ENS) of the ECMWF. The proxy for forecast errors is the total spread of the ECMWF operational ensemble forecasts obtained by the decomposition of the wind and geopotential fields in the normal-mode functions. In this way, the ensemble spread can be quantified separately for the balanced and inertio-gravity (IG) modes for every forecast range. Ensemble reliability is defined for the balanced and IG modes comparing the ensemble spread with the control analysis in each scale. The results show that initial uncertainties in the ECMWF ENS are largest in the tropical large-scale modes and their spatial distribution is similar to the distribution of the short-range forecast errors. Initially the ensemble spread grows most in the smallest scales and in the synoptic range of the IG modes but the overall growth is dominated by the increase of spread in balanced modes in synoptic and planetary scales in the midlatitudes. During the forecasts, the distribution of spread in the balanced and IG modes grows towards the climatological spread distribution characteristic of the analyses. The ENS system is found to be somewhat under-dispersive which is associated with the lack of tropical variability, primarily the Kelvin waves. The new model of the forecast error growth has three fitting parameters to parameterize the initial fast growth and a more slow exponential error growth later on. The asymptotic values of forecast errors are independent of the exponential growth rate. It is found that the asymptotic values of the errors due to unbalanced dynamics are around 10 days while the balanced and total errors saturate in 3 to 4 weeks. Reference: Žagar, N., R. Buizza, and J. Tribbia, 2015: A three-dimensional multivariate modal analysis of atmospheric predictability with application to the ECMWF ensemble. J. Atmos. Sci., 72, 4423-4444.

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.

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

Fuzzy State Transition and Kalman Filter Applied in Short-Term Traffic Flow Forecasting

PubMed Central

Ming-jun, Deng; Shi-ru, Qu

2015-01-01

Traffic flow is widely recognized as an important parameter for road traffic state forecasting. Fuzzy state transform and Kalman filter (KF) have been applied in this field separately. But the studies show that the former method has good performance on the trend forecasting of traffic state variation but always involves several numerical errors. The latter model is good at numerical forecasting but is deficient in the expression of time hysteretically. This paper proposed an approach that combining fuzzy state transform and KF forecasting model. In considering the advantage of the two models, a weight combination model is proposed. The minimum of the sum forecasting error squared is regarded as a goal in optimizing the combined weight dynamically. Real detection data are used to test the efficiency. Results indicate that the method has a good performance in terms of short-term traffic forecasting. PMID:26779258

Fuzzy State Transition and Kalman Filter Applied in Short-Term Traffic Flow Forecasting.

PubMed

Deng, Ming-jun; Qu, Shi-ru

2015-01-01

Traffic flow is widely recognized as an important parameter for road traffic state forecasting. Fuzzy state transform and Kalman filter (KF) have been applied in this field separately. But the studies show that the former method has good performance on the trend forecasting of traffic state variation but always involves several numerical errors. The latter model is good at numerical forecasting but is deficient in the expression of time hysteretically. This paper proposed an approach that combining fuzzy state transform and KF forecasting model. In considering the advantage of the two models, a weight combination model is proposed. The minimum of the sum forecasting error squared is regarded as a goal in optimizing the combined weight dynamically. Real detection data are used to test the efficiency. Results indicate that the method has a good performance in terms of short-term traffic forecasting.

Compensated Box-Jenkins transfer function for short term load forecast

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

Breipohl, A.; Yu, Z.; Lee, F.N.

In the past years, the Box-Jenkins ARIMA method and the Box-Jenkins transfer function method (BJTF) have been among the most commonly used methods for short term electrical load forecasting. But when there exists a sudden change in the temperature, both methods tend to exhibit larger errors in the forecast. This paper demonstrates that the load forecasting errors resulting from either the BJ ARIMA model or the BJTF model are not simply white noise, but rather well-patterned noise, and the patterns in the noise can be used to improve the forecasts. Thus a compensated Box-Jenkins transfer method (CBJTF) is proposed tomore » improve the accuracy of the load prediction. Some case studies have been made which result in about a 14-33% reduction of the root mean square (RMS) errors of the forecasts, depending on the compensation time period as well as the compensation method used.« less

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.

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.

Possible sources of forecast errors generated by the global/regional assimilation and prediction system for landfalling tropical cyclones. Part I: Initial uncertainties

NASA Astrophysics Data System (ADS)

Zhou, Feifan; Yamaguchi, Munehiko; Qin, Xiaohao

2016-07-01

This paper investigates the possible sources of errors associated with tropical cyclone (TC) tracks forecasted using the Global/Regional Assimilation and Prediction System (GRAPES). The GRAPES forecasts were made for 16 landfalling TCs in the western North Pacific basin during the 2008 and 2009 seasons, with a forecast length of 72 hours, and using the default initial conditions ("initials", hereafter), which are from the NCEP-FNL dataset, as well as ECMWF initials. The forecasts are compared with ECMWF forecasts. The results show that in most TCs, the GRAPES forecasts are improved when using the ECMWF initials compared with the default initials. Compared with the ECMWF initials, the default initials produce lower intensity TCs and a lower intensity subtropical high, but a higher intensity South Asia high and monsoon trough, as well as a higher temperature but lower specific humidity at the TC center. Replacement of the geopotential height and wind fields with the ECMWF initials in and around the TC center at the initial time was found to be the most efficient way to improve the forecasts. In addition, TCs that showed the greatest improvement in forecast accuracy usually had the largest initial uncertainties in TC intensity and were usually in the intensifying phase. The results demonstrate the importance of the initial intensity for TC track forecasts made using GRAPES, and indicate the model is better in describing the intensifying phase than the decaying phase of TCs. Finally, the limit of the improvement indicates that the model error associated with GRAPES forecasts may be the main cause of poor forecasts of landfalling TCs. Thus, further examinations of the model errors are required.

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.

Machine Learning Based Multi-Physical-Model Blending for Enhancing Renewable Energy Forecast -- Improvement via Situation Dependent Error Correction

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

Lu, Siyuan; Hwang, Youngdeok; Khabibrakhmanov, Ildar

With increasing penetration of solar and wind energy to the total energy supply mix, the pressing need for accurate energy forecasting has become well-recognized. Here we report the development of a machine-learning based model blending approach for statistically combining multiple meteorological models for improving the accuracy of solar/wind power forecast. Importantly, we demonstrate that in addition to parameters to be predicted (such as solar irradiance and power), including additional atmospheric state parameters which collectively define weather situations as machine learning input provides further enhanced accuracy for the blended result. Functional analysis of variance shows that the error of individual modelmore » has substantial dependence on the weather situation. The machine-learning approach effectively reduces such situation dependent error thus produces more accurate results compared to conventional multi-model ensemble approaches based on simplistic equally or unequally weighted model averaging. Validation over an extended period of time results show over 30% improvement in solar irradiance/power forecast accuracy compared to forecasts based on the best individual model.« less

Predictability of Bristol Bay, Alaska, sockeye salmon returns one to four years in the future

USGS Publications Warehouse

Adkison, Milo D.; Peterson, R.M.

2000-01-01

Historically, forecast error for returns of sockeye salmon Oncorhynchus nerka to Bristol Bay, Alaska, has been large. Using cross-validation forecast error as our criterion, we selected forecast models for each of the nine principal Bristol Bay drainages. Competing forecast models included stock-recruitment relationships, environmental variables, prior returns of siblings, or combinations of these predictors. For most stocks, we found prior returns of siblings to be the best single predictor of returns; however, forecast accuracy was low even when multiple predictors were considered. For a typical drainage, an 80% confidence interval ranged from one half to double the point forecast. These confidence intervals appeared to be appropriately wide.

Influence of Forecast Accuracy of Photovoltaic Power Output on Facility Planning and Operation of Microgrid under 30 min Power Balancing Control

NASA Astrophysics Data System (ADS)

Kato, Takeyoshi; Sone, Akihito; Shimakage, Toyonari; Suzuoki, Yasuo

A microgrid (MG) is one of the measures for enhancing the high penetration of renewable energy (RE)-based distributed generators (DGs). For constructing a MG economically, the capacity optimization of controllable DGs against RE-based DGs is essential. By using a numerical simulation model developed based on the demonstrative studies on a MG using PAFC and NaS battery as controllable DGs and photovoltaic power generation system (PVS) as a RE-based DG, this study discusses the influence of forecast accuracy of PVS output on the capacity optimization and daily operation evaluated with the cost. The main results are as follows. The required capacity of NaS battery must be increased by 10-40% against the ideal situation without the forecast error of PVS power output. The influence of forecast error on the received grid electricity would not be so significant on annual basis because the positive and negative forecast error varies with days. The annual total cost of facility and operation increases by 2-7% due to the forecast error applied in this study. The impact of forecast error on the facility optimization and operation optimization is almost the same each other at a few percentages, implying that the forecast accuracy should be improved in terms of both the number of times with large forecast error and the average error.

Predicting areas of sustainable error growth in quasigeostrophic flows using perturbation alignment properties

NASA Astrophysics Data System (ADS)

Rivière, G.; Hua, B. L.

2004-10-01

A new perturbation initialization method is used to quantify error growth due to inaccuracies of the forecast model initial conditions in a quasigeostrophic box ocean model describing a wind-driven double gyre circulation. This method is based on recent analytical results on Lagrangian alignment dynamics of the perturbation velocity vector in quasigeostrophic flows. More specifically, it consists in initializing a unique perturbation from the sole knowledge of the control flow properties at the initial time of the forecast and whose velocity vector orientation satisfies a Lagrangian equilibrium criterion. This Alignment-based Initialization method is hereafter denoted as the AI method.In terms of spatial distribution of the errors, we have compared favorably the AI error forecast with the mean error obtained with a Monte-Carlo ensemble prediction. It is shown that the AI forecast is on average as efficient as the error forecast initialized with the leading singular vector for the palenstrophy norm, and significantly more efficient than that for total energy and enstrophy norms. Furthermore, a more precise examination shows that the AI forecast is systematically relevant for all control flows whereas the palenstrophy singular vector forecast leads sometimes to very good scores and sometimes to very bad ones.A principal component analysis at the final time of the forecast shows that the AI mode spatial structure is comparable to that of the first eigenvector of the error covariance matrix for a "bred mode" ensemble. Furthermore, the kinetic energy of the AI mode grows at the same constant rate as that of the "bred modes" from the initial time to the final time of the forecast and is therefore characterized by a sustained phase of error growth. In this sense, the AI mode based on Lagrangian dynamics of the perturbation velocity orientation provides a rationale of the "bred mode" behavior.

Simplification of the Kalman filter for meteorological data assimilation

NASA Technical Reports Server (NTRS)

Dee, Dick P.

1991-01-01

The paper proposes a new statistical method of data assimilation that is based on a simplification of the Kalman filter equations. The forecast error covariance evolution is approximated simply by advecting the mass-error covariance field, deriving the remaining covariances geostrophically, and accounting for external model-error forcing only at the end of each forecast cycle. This greatly reduces the cost of computation of the forecast error covariance. In simulations with a linear, one-dimensional shallow-water model and data generated artificially, the performance of the simplified filter is compared with that of the Kalman filter and the optimal interpolation (OI) method. The simplified filter produces analyses that are nearly optimal, and represents a significant improvement over OI.

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.

Soil Moisture Initialization Error and Subgrid Variability of Precipitation in Seasonal Streamflow Forecasting

NASA Technical Reports Server (NTRS)

Koster, Randal D.; Walker, Gregory K.; Mahanama, Sarith P.; Reichle, Rolf H.

2013-01-01

Offline simulations over the conterminous United States (CONUS) with a land surface model are used to address two issues relevant to the forecasting of large-scale seasonal streamflow: (i) the extent to which errors in soil moisture initialization degrade streamflow forecasts, and (ii) the extent to which a realistic increase in the spatial resolution of forecasted precipitation would improve streamflow forecasts. The addition of error to a soil moisture initialization field is found to lead to a nearly proportional reduction in streamflow forecast skill. The linearity of the response allows the determination of a lower bound for the increase in streamflow forecast skill achievable through improved soil moisture estimation, e.g., through satellite-based soil moisture measurements. An increase in the resolution of precipitation is found to have an impact on large-scale streamflow forecasts only when evaporation variance is significant relative to the precipitation variance. This condition is met only in the western half of the CONUS domain. Taken together, the two studies demonstrate the utility of a continental-scale land surface modeling system as a tool for addressing the science of hydrological prediction.

Operational hydrological forecasting in Bavaria. Part I: Forecast uncertainty

NASA Astrophysics Data System (ADS)

Ehret, U.; Vogelbacher, A.; Moritz, K.; Laurent, S.; Meyer, I.; Haag, I.

2009-04-01

In Bavaria, operational flood forecasting has been established since the disastrous flood of 1999. Nowadays, forecasts based on rainfall information from about 700 raingauges and 600 rivergauges are calculated and issued for nearly 100 rivergauges. With the added experience of the 2002 and 2005 floods, awareness grew that the standard deterministic forecast, neglecting the uncertainty associated with each forecast is misleading, creating a false feeling of unambiguousness. As a consequence, a system to identify, quantify and communicate the sources and magnitude of forecast uncertainty has been developed, which will be presented in part I of this study. In this system, the use of ensemble meteorological forecasts plays a key role which will be presented in part II. Developing the system, several constraints stemming from the range of hydrological regimes and operational requirements had to be met: Firstly, operational time constraints obviate the variation of all components of the modeling chain as would be done in a full Monte Carlo simulation. Therefore, an approach was chosen where only the most relevant sources of uncertainty were dynamically considered while the others were jointly accounted for by static error distributions from offline analysis. Secondly, the dominant sources of uncertainty vary over the wide range of forecasted catchments: In alpine headwater catchments, typically of a few hundred square kilometers in size, rainfall forecast uncertainty is the key factor for forecast uncertainty, with a magnitude dynamically changing with the prevailing predictability of the atmosphere. In lowland catchments encompassing several thousands of square kilometers, forecast uncertainty in the desired range (usually up to two days) is mainly dependent on upstream gauge observation quality, routing and unpredictable human impact such as reservoir operation. The determination of forecast uncertainty comprised the following steps: a) From comparison of gauge observations and several years of archived forecasts, overall empirical error distributions termed 'overall error' were for each gauge derived for a range of relevant forecast lead times. b) The error distributions vary strongly with the hydrometeorological situation, therefore a subdivision into the hydrological cases 'low flow, 'rising flood', 'flood', flood recession' was introduced. c) For the sake of numerical compression, theoretical distributions were fitted to the empirical distributions using the method of moments. Here, the normal distribution was generally best suited. d) Further data compression was achieved by representing the distribution parameters as a function (second-order polynome) of lead time. In general, the 'overall error' obtained from the above procedure is most useful in regions where large human impact occurs and where the influence of the meteorological forecast is limited. In upstream regions however, forecast uncertainty is strongly dependent on the current predictability of the atmosphere, which is contained in the spread of an ensemble forecast. Including this dynamically in the hydrological forecast uncertainty estimation requires prior elimination of the contribution of the weather forecast to the 'overall error'. This was achieved by calculating long series of hydrometeorological forecast tests, where rainfall observations were used instead of forecasts. The resulting error distribution is termed 'model error' and can be applied on hydrological ensemble forecasts, where ensemble rainfall forecasts are used as forcing. The concept will be illustrated by examples (good and bad ones) covering a wide range of catchment sizes, hydrometeorological regimes and quality of hydrological model calibration. The methodology to combine the static and dynamic shares of uncertainty will be presented in part II of this study.

Demand forecasting of electricity in Indonesia with limited historical data

NASA Astrophysics Data System (ADS)

Dwi Kartikasari, Mujiati; Rohmad Prayogi, Arif

2018-03-01

Demand forecasting of electricity is an important activity for electrical agents to know the description of electricity demand in future. Prediction of demand electricity can be done using time series models. In this paper, double moving average model, Holt’s exponential smoothing model, and grey model GM(1,1) are used to predict electricity demand in Indonesia under the condition of limited historical data. The result shows that grey model GM(1,1) has the smallest value of MAE (mean absolute error), MSE (mean squared error), and MAPE (mean absolute percentage error).

Stochastic Forcing for High-Resolution Regional and Global Ocean and Atmosphere-Ocean Coupled Ensemble Forecast System

NASA Astrophysics Data System (ADS)

Rowley, C. D.; Hogan, P. J.; Martin, P.; Thoppil, P.; Wei, M.

2017-12-01

An extended range ensemble forecast system is being developed in the US Navy Earth System Prediction Capability (ESPC), and a global ocean ensemble generation capability to represent uncertainty in the ocean initial conditions has been developed. At extended forecast times, the uncertainty due to the model error overtakes the initial condition as the primary source of forecast uncertainty. Recently, stochastic parameterization or stochastic forcing techniques have been applied to represent the model error in research and operational atmospheric, ocean, and coupled ensemble forecasts. A simple stochastic forcing technique has been developed for application to US Navy high resolution regional and global ocean models, for use in ocean-only and coupled atmosphere-ocean-ice-wave ensemble forecast systems. Perturbation forcing is added to the tendency equations for state variables, with the forcing defined by random 3- or 4-dimensional fields with horizontal, vertical, and temporal correlations specified to characterize different possible kinds of error. Here, we demonstrate the stochastic forcing in regional and global ensemble forecasts with varying perturbation amplitudes and length and time scales, and assess the change in ensemble skill measured by a range of deterministic and probabilistic metrics.

A Comparison Study of Return Ratio-Based Academic Enrollment Forecasting Models. Professional File. Article 129, Spring 2013

ERIC Educational Resources Information Center

Zan, Xinxing Anna; Yoon, Sang Won; Khasawneh, Mohammad; Srihari, Krishnaswami

2013-01-01

In an effort to develop a low-cost and user-friendly forecasting model to minimize forecasting error, we have applied average and exponentially weighted return ratios to project undergraduate student enrollment. We tested the proposed forecasting models with different sets of historical enrollment data, such as university-, school-, and…

Forecasting longitudinal changes in oropharyngeal tumor morphology throughout the course of head and neck radiation therapy

PubMed Central

Yock, Adam D.; Rao, Arvind; Dong, Lei; Beadle, Beth M.; Garden, Adam S.; Kudchadker, Rajat J.; Court, Laurence E.

2014-01-01

Purpose: To create models that forecast longitudinal trends in changing tumor morphology and to evaluate and compare their predictive potential throughout the course of radiation therapy. Methods: Two morphology feature vectors were used to describe 35 gross tumor volumes (GTVs) throughout the course of intensity-modulated radiation therapy for oropharyngeal tumors. The feature vectors comprised the coordinates of the GTV centroids and a description of GTV shape using either interlandmark distances or a spherical harmonic decomposition of these distances. The change in the morphology feature vector observed at 33 time points throughout the course of treatment was described using static, linear, and mean models. Models were adjusted at 0, 1, 2, 3, or 5 different time points (adjustment points) to improve prediction accuracy. The potential of these models to forecast GTV morphology was evaluated using leave-one-out cross-validation, and the accuracy of the models was compared using Wilcoxon signed-rank tests. Results: Adding a single adjustment point to the static model without any adjustment points decreased the median error in forecasting the position of GTV surface landmarks by the largest amount (1.2 mm). Additional adjustment points further decreased the forecast error by about 0.4 mm each. Selection of the linear model decreased the forecast error for both the distance-based and spherical harmonic morphology descriptors (0.2 mm), while the mean model decreased the forecast error for the distance-based descriptor only (0.2 mm). The magnitude and statistical significance of these improvements decreased with each additional adjustment point, and the effect from model selection was not as large as that from adding the initial points. Conclusions: The authors present models that anticipate longitudinal changes in tumor morphology using various models and model adjustment schemes. The accuracy of these models depended on their form, and the utility of these models includes the characterization of patient-specific response with implications for treatment management and research study design. PMID:25086518

Forecasting longitudinal changes in oropharyngeal tumor morphology throughout the course of head and neck radiation therapy

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

Yock, Adam D.; Kudchadker, Rajat J.; Rao, Arvind

2014-08-15

Purpose: To create models that forecast longitudinal trends in changing tumor morphology and to evaluate and compare their predictive potential throughout the course of radiation therapy. Methods: Two morphology feature vectors were used to describe 35 gross tumor volumes (GTVs) throughout the course of intensity-modulated radiation therapy for oropharyngeal tumors. The feature vectors comprised the coordinates of the GTV centroids and a description of GTV shape using either interlandmark distances or a spherical harmonic decomposition of these distances. The change in the morphology feature vector observed at 33 time points throughout the course of treatment was described using static, linear,more » and mean models. Models were adjusted at 0, 1, 2, 3, or 5 different time points (adjustment points) to improve prediction accuracy. The potential of these models to forecast GTV morphology was evaluated using leave-one-out cross-validation, and the accuracy of the models was compared using Wilcoxon signed-rank tests. Results: Adding a single adjustment point to the static model without any adjustment points decreased the median error in forecasting the position of GTV surface landmarks by the largest amount (1.2 mm). Additional adjustment points further decreased the forecast error by about 0.4 mm each. Selection of the linear model decreased the forecast error for both the distance-based and spherical harmonic morphology descriptors (0.2 mm), while the mean model decreased the forecast error for the distance-based descriptor only (0.2 mm). The magnitude and statistical significance of these improvements decreased with each additional adjustment point, and the effect from model selection was not as large as that from adding the initial points. Conclusions: The authors present models that anticipate longitudinal changes in tumor morphology using various models and model adjustment schemes. The accuracy of these models depended on their form, and the utility of these models includes the characterization of patient-specific response with implications for treatment management and research study design.« less

Convective Weather Forecast Accuracy Analysis at Center and Sector Levels

NASA Technical Reports Server (NTRS)

Wang, Yao; Sridhar, Banavar

2010-01-01

This paper presents a detailed convective forecast accuracy analysis at center and sector levels. The study is aimed to provide more meaningful forecast verification measures to aviation community, as well as to obtain useful information leading to the improvements in the weather translation capacity models. In general, the vast majority of forecast verification efforts over past decades have been on the calculation of traditional standard verification measure scores over forecast and observation data analyses onto grids. These verification measures based on the binary classification have been applied in quality assurance of weather forecast products at the national level for many years. Our research focuses on the forecast at the center and sector levels. We calculate the standard forecast verification measure scores for en-route air traffic centers and sectors first, followed by conducting the forecast validation analysis and related verification measures for weather intensities and locations at centers and sectors levels. An approach to improve the prediction of sector weather coverage by multiple sector forecasts is then developed. The weather severe intensity assessment was carried out by using the correlations between forecast and actual weather observation airspace coverage. The weather forecast accuracy on horizontal location was assessed by examining the forecast errors. The improvement in prediction of weather coverage was determined by the correlation between actual sector weather coverage and prediction. observed and forecasted Convective Weather Avoidance Model (CWAM) data collected from June to September in 2007. CWAM zero-minute forecast data with aircraft avoidance probability of 60% and 80% are used as the actual weather observation. All forecast measurements are based on 30-minute, 60- minute, 90-minute, and 120-minute forecasts with the same avoidance probabilities. The forecast accuracy analysis for times under one-hour showed that the errors in intensity and location for center forecast are relatively low. For example, 1-hour forecast intensity and horizontal location errors for ZDC center were about 0.12 and 0.13. However, the correlation between sector 1-hour forecast and actual weather coverage was weak, for sector ZDC32, about 32% of the total variation of observation weather intensity was unexplained by forecast; the sector horizontal location error was about 0.10. The paper also introduces an approach to estimate the sector three-dimensional actual weather coverage by using multiple sector forecasts, which turned out to produce better predictions. Using Multiple Linear Regression (MLR) model for this approach, the correlations between actual observation and the multiple sector forecast model prediction improved by several percents at 95% confidence level in comparison with single sector forecast.

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.

A Hybrid Model for Predicting the Prevalence of Schistosomiasis in Humans of Qianjiang City, China

PubMed Central

Wang, Ying; Lu, Zhouqin; Tian, Lihong; Tan, Li; Shi, Yun; Nie, Shaofa; Liu, Li

2014-01-01

Backgrounds/Objective Schistosomiasis is still a major public health problem in China, despite the fact that the government has implemented a series of strategies to prevent and control the spread of the parasitic disease. Advanced warning and reliable forecasting can help policymakers to adjust and implement strategies more effectively, which will lead to the control and elimination of schistosomiasis. Our aim is to explore the application of a hybrid forecasting model to track the trends of the prevalence of schistosomiasis in humans, which provides a methodological basis for predicting and detecting schistosomiasis infection in endemic areas. Methods A hybrid approach combining the autoregressive integrated moving average (ARIMA) model and the nonlinear autoregressive neural network (NARNN) model to forecast the prevalence of schistosomiasis in the future four years. Forecasting performance was compared between the hybrid ARIMA-NARNN model, and the single ARIMA or the single NARNN model. Results The modelling mean square error (MSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) of the ARIMA-NARNN model was 0.1869×10−4, 0.0029, 0.0419 with a corresponding testing error of 0.9375×10−4, 0.0081, 0.9064, respectively. These error values generated with the hybrid model were all lower than those obtained from the single ARIMA or NARNN model. The forecasting values were 0.75%, 0.80%, 0.76% and 0.77% in the future four years, which demonstrated a no-downward trend. Conclusion The hybrid model has high quality prediction accuracy in the prevalence of schistosomiasis, which provides a methodological basis for future schistosomiasis monitoring and control strategies in the study area. It is worth attempting to utilize the hybrid detection scheme in other schistosomiasis-endemic areas including other infectious diseases. PMID:25119882

Insights on the impact of systematic model errors on data assimilation performance in changing catchments

NASA Astrophysics Data System (ADS)

Pathiraja, S.; Anghileri, D.; Burlando, P.; Sharma, A.; Marshall, L.; Moradkhani, H.

2018-03-01

The global prevalence of rapid and extensive land use change necessitates hydrologic modelling methodologies capable of handling non-stationarity. This is particularly true in the context of Hydrologic Forecasting using Data Assimilation. Data Assimilation has been shown to dramatically improve forecast skill in hydrologic and meteorological applications, although such improvements are conditional on using bias-free observations and model simulations. A hydrologic model calibrated to a particular set of land cover conditions has the potential to produce biased simulations when the catchment is disturbed. This paper sheds new light on the impacts of bias or systematic errors in hydrologic data assimilation, in the context of forecasting in catchments with changing land surface conditions and a model calibrated to pre-change conditions. We posit that in such cases, the impact of systematic model errors on assimilation or forecast quality is dependent on the inherent prediction uncertainty that persists even in pre-change conditions. Through experiments on a range of catchments, we develop a conceptual relationship between total prediction uncertainty and the impacts of land cover changes on the hydrologic regime to demonstrate how forecast quality is affected when using state estimation Data Assimilation with no modifications to account for land cover changes. This work shows that systematic model errors as a result of changing or changed catchment conditions do not always necessitate adjustments to the modelling or assimilation methodology, for instance through re-calibration of the hydrologic model, time varying model parameters or revised offline/online bias estimation.

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.

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.

Mitigating the Impacts of Climate Nonstationarity on Seasonal Streamflow Predictability in the U.S. Southwest

NASA Astrophysics Data System (ADS)

Lehner, Flavio; Wood, Andrew W.; Llewellyn, Dagmar; Blatchford, Douglas B.; Goodbody, Angus G.; Pappenberger, Florian

2017-12-01

Seasonal streamflow predictions provide a critical management tool for water managers in the American Southwest. In recent decades, persistent prediction errors for spring and summer runoff volumes have been observed in a number of watersheds in the American Southwest. While mostly driven by decadal precipitation trends, these errors also relate to the influence of increasing temperature on streamflow in these basins. Here we show that incorporating seasonal temperature forecasts from operational global climate prediction models into streamflow forecasting models adds prediction skill for watersheds in the headwaters of the Colorado and Rio Grande River basins. Current dynamical seasonal temperature forecasts now show sufficient skill to reduce streamflow forecast errors in snowmelt-driven regions. Such predictions can increase the resilience of streamflow forecasting and water management systems in the face of continuing warming as well as decadal-scale temperature variability and thus help to mitigate the impacts of climate nonstationarity on streamflow predictability.

Medium-range reference evapotranspiration forecasts for the contiguous United States based on multi-model numerical weather predictions

NASA Astrophysics Data System (ADS)

Medina, Hanoi; Tian, Di; Srivastava, Puneet; Pelosi, Anna; Chirico, Giovanni B.

2018-07-01

Reference evapotranspiration (ET0) plays a fundamental role in agronomic, forestry, and water resources management. Estimating and forecasting ET0 have long been recognized as a major challenge for researchers and practitioners in these communities. This work explored the potential of multiple leading numerical weather predictions (NWPs) for estimating and forecasting summer ET0 at 101 U.S. Regional Climate Reference Network stations over nine climate regions across the contiguous United States (CONUS). Three leading global NWP model forecasts from THORPEX Interactive Grand Global Ensemble (TIGGE) dataset were used in this study, including the single model ensemble forecasts from the European Centre for Medium-Range Weather Forecasts (EC), the National Centers for Environmental Prediction Global Forecast System (NCEP), and the United Kingdom Meteorological Office forecasts (MO), as well as multi-model ensemble forecasts from the combinations of these NWP models. A regression calibration was employed to bias correct the ET0 forecasts. Impact of individual forecast variables on ET0 forecasts were also evaluated. The results showed that the EC forecasts provided the least error and highest skill and reliability, followed by the MO and NCEP forecasts. The multi-model ensembles constructed from the combination of EC and MO forecasts provided slightly better performance than the single model EC forecasts. The regression process greatly improved ET0 forecast performances, particularly for the regions involving stations near the coast, or with a complex orography. The performance of EC forecasts was only slightly influenced by the size of the ensemble members, particularly at short lead times. Even with less ensemble members, EC still performed better than the other two NWPs. Errors in the radiation forecasts, followed by those in the wind, had the most detrimental effects on the ET0 forecast performances.

A Unified Data Assimilation Strategy for Regional Coupled Atmosphere-Ocean Prediction Systems

NASA Astrophysics Data System (ADS)

Xie, Lian; Liu, Bin; Zhang, Fuqing; Weng, Yonghui

2014-05-01

Improving tropical cyclone (TC) forecasts is a top priority in weather forecasting. Assimilating various observational data to produce better initial conditions for numerical models using advanced data assimilation techniques has been shown to benefit TC intensity forecasts, whereas assimilating large-scale environmental circulation into regional models by spectral nudging or Scale-Selective Data Assimilation (SSDA) has been demonstrated to improve TC track forecasts. Meanwhile, taking into account various air-sea interaction processes by high-resolution coupled air-sea modelling systems has also been shown to improve TC intensity forecasts. Despite the advances in data assimilation and air-sea coupled models, large errors in TC intensity and track forecasting remain. For example, Hurricane Nate (2011) has brought considerable challenge for the TC operational forecasting community, with very large intensity forecast errors (27, 25, and 40 kts for 48, 72, and 96 h, respectively) for the official forecasts. Considering the slow-moving nature of Hurricane Nate, it is reasonable to hypothesize that air-sea interaction processes played a critical role in the intensity change of the storm, and accurate representation of the upper ocean dynamics and thermodynamics is necessary to quantitatively describe the air-sea interaction processes. Currently, data assimilation techniques are generally only applied to hurricane forecasting in stand-alone atmospheric or oceanic model. In fact, most of the regional hurricane forecasting models only included data assimilation techniques for improving the initial condition of the atmospheric model. In such a situation, the benefit of adjustments in one model (atmospheric or oceanic) by assimilating observational data can be compromised by errors from the other model. Thus, unified data assimilation techniques for coupled air-sea modelling systems, which not only simultaneously assimilate atmospheric and oceanic observations into the coupled air-sea modelling system, but also nudging the large-scale environmental flow in the regional model towards global model forecasts are of increasing necessity. In this presentation, we will outline a strategy for an integrated approach in air-sea coupled data assimilation and discuss its benefits and feasibility from incremental results for select historical hurricane cases.

Time series forecasting of future claims amount of SOCSO's employment injury scheme (EIS)

NASA Astrophysics Data System (ADS)

Zulkifli, Faiz; Ismail, Isma Liana; Chek, Mohd Zaki Awang; Jamal, Nur Faezah; Ridzwan, Ahmad Nur Azam Ahmad; Jelas, Imran Md; Noor, Syamsul Ikram Mohd; Ahmad, Abu Bakar

2012-09-01

The Employment Injury Scheme (EIS) provides protection to employees who are injured due to accidents whilst working, commuting from home to the work place or during employee takes a break during an authorized recess time or while travelling that is related with his work. The main purpose of this study is to forecast value on claims amount of EIS for the year 2011 until 2015 by using appropriate models. These models were tested on the actual EIS data from year 1972 until year 2010. Three different forecasting models are chosen for comparisons. These are the Naïve with Trend Model, Average Percent Change Model and Double Exponential Smoothing Model. The best model is selected based on the smallest value of error measures using the Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE). From the result, the best model that best fit the forecast for the EIS is the Average Percent Change Model. Furthermore, the result also shows the claims amount of EIS for the year 2011 to year 2015 continue to trend upwards from year 2010.

The capability of radial basis function to forecast the volume fractions of the annular three-phase flow of gas-oil-water.

PubMed

Roshani, G H; Karami, A; Salehizadeh, A; Nazemi, E

2017-11-01

The problem of how to precisely measure the volume fractions of oil-gas-water mixtures in a pipeline remains as one of the main challenges in the petroleum industry. This paper reports the capability of Radial Basis Function (RBF) in forecasting the volume fractions in a gas-oil-water multiphase system. Indeed, in the present research, the volume fractions in the annular three-phase flow are measured based on a dual energy metering system including the 152 Eu and 137 Cs and one NaI detector, and then modeled by a RBF model. Since the summation of volume fractions are constant (equal to 100%), therefore it is enough for the RBF model to forecast only two volume fractions. In this investigation, three RBF models are employed. The first model is used to forecast the oil and water volume fractions. The next one is utilized to forecast the water and gas volume fractions, and the last one to forecast the gas and oil volume fractions. In the next stage, the numerical data obtained from MCNP-X code must be introduced to the RBF models. Then, the average errors of these three models are calculated and compared. The model which has the least error is picked up as the best predictive model. Based on the results, the best RBF model, forecasts the oil and water volume fractions with the mean relative error of less than 0.5%, which indicates that the RBF model introduced in this study ensures an effective enough mechanism to forecast the results. Copyright © 2017 Elsevier Ltd. All rights reserved.

Satellite-based Calibration of Heat Flux at the Ocean Surface

NASA Astrophysics Data System (ADS)

Barron, C. N.; Dastugue, J. M.; May, J. C.; Rowley, C. D.; Smith, S. R.; Spence, P. L.; Gremes-Cordero, S.

2016-02-01

Model forecasts of upper ocean heat content and variability on diurnal to daily scales are highly dependent on estimates of heat flux through the air-sea interface. Satellite remote sensing is applied to not only inform the initial ocean state but also to mitigate errors in surface heat flux and model representations affecting the distribution of heat in the upper ocean. Traditional assimilation of sea surface temperature (SST) observations re-centers ocean models at the start of each forecast cycle. Subsequent evolution depends on estimates of surface heat fluxes and upper-ocean processes over the forecast period. The COFFEE project (Calibration of Ocean Forcing with satellite Flux Estimates) endeavors to correct ocean forecast bias through a responsive error partition among surface heat flux and ocean dynamics sources. A suite of experiments in the southern California Current demonstrates a range of COFFEE capabilities, showing the impact on forecast error relative to a baseline three-dimensional variational (3DVAR) assimilation using Navy operational global or regional atmospheric forcing. COFFEE addresses satellite-calibration of surface fluxes to estimate surface error covariances and links these to the ocean interior. Experiment cases combine different levels of flux calibration with different assimilation alternatives. The cases may use the original fluxes, apply full satellite corrections during the forecast period, or extend hindcast corrections into the forecast period. Assimilation is either baseline 3DVAR or standard strong-constraint 4DVAR, with work proceeding to add a 4DVAR expanded to include a weak constraint treatment of the surface flux errors. Covariance of flux errors is estimated from the recent time series of forecast and calibrated flux terms. While the California Current examples are shown, the approach is equally applicable to other regions. These approaches within a 3DVAR application are anticipated to be useful for global and larger regional domains where a full 4DVAR methodology may be cost-prohibitive.

Potential Technologies for Assessing Risk Associated with a Mesoscale Forecast

DTIC Science & Technology

2015-10-01

American GFS models, and informally applied on the Weather Research and Forecasting ( WRF ) model. The current CI equation is as follows...Reen B, Penc R. Investigating surface bias errors in the Weather Research and Forecasting ( WRF ) model using a Geographic Information System (GIS). J...Forecast model ( WRF -ARW) with extensions that might include finer terrain resolutions and more detailed representations of the underlying atmospheric

The application of a Grey Markov Model to forecasting annual maximum water levels at hydrological stations

NASA Astrophysics Data System (ADS)

Dong, Sheng; Chi, Kun; Zhang, Qiyi; Zhang, Xiangdong

2012-03-01

Compared with traditional real-time forecasting, this paper proposes a Grey Markov Model (GMM) to forecast the maximum water levels at hydrological stations in the estuary area. The GMM combines the Grey System and Markov theory into a higher precision model. The GMM takes advantage of the Grey System to predict the trend values and uses the Markov theory to forecast fluctuation values, and thus gives forecast results involving two aspects of information. The procedure for forecasting annul maximum water levels with the GMM contains five main steps: 1) establish the GM (1, 1) model based on the data series; 2) estimate the trend values; 3) establish a Markov Model based on relative error series; 4) modify the relative errors caused in step 2, and then obtain the relative errors of the second order estimation; 5) compare the results with measured data and estimate the accuracy. The historical water level records (from 1960 to 1992) at Yuqiao Hydrological Station in the estuary area of the Haihe River near Tianjin, China are utilized to calibrate and verify the proposed model according to the above steps. Every 25 years' data are regarded as a hydro-sequence. Eight groups of simulated results show reasonable agreement between the predicted values and the measured data. The GMM is also applied to the 10 other hydrological stations in the same estuary. The forecast results for all of the hydrological stations are good or acceptable. The feasibility and effectiveness of this new forecasting model have been proved in this paper.

Calibration of Ocean Forcing with satellite Flux Estimates (COFFEE)

NASA Astrophysics Data System (ADS)

Barron, Charlie; Jan, Dastugue; Jackie, May; Rowley, Clark; Smith, Scott; Spence, Peter; Gremes-Cordero, Silvia

2016-04-01

Predicting the evolution of ocean temperature in regional ocean models depends on estimates of surface heat fluxes and upper-ocean processes over the forecast period. Within the COFFEE project (Calibration of Ocean Forcing with satellite Flux Estimates, real-time satellite observations are used to estimate shortwave, longwave, sensible, and latent air-sea heat flux corrections to a background estimate from the prior day's regional or global model forecast. These satellite-corrected fluxes are used to prepare a corrected ocean hindcast and to estimate flux error covariances to project the heat flux corrections for a 3-5 day forecast. In this way, satellite remote sensing is applied to not only inform the initial ocean state but also to mitigate errors in surface heat flux and model representations affecting the distribution of heat in the upper ocean. While traditional assimilation of sea surface temperature (SST) observations re-centers ocean models at the start of each forecast cycle, COFFEE endeavors to appropriately partition and reduce among various surface heat flux and ocean dynamics sources. A suite of experiments in the southern California Current demonstrates a range of COFFEE capabilities, showing the impact on forecast error relative to a baseline three-dimensional variational (3DVAR) assimilation using operational global or regional atmospheric forcing. Experiment cases combine different levels of flux calibration with assimilation alternatives. The cases use the original fluxes, apply full satellite corrections during the forecast period, or extend hindcast corrections into the forecast period. Assimilation is either baseline 3DVAR or standard strong-constraint 4DVAR, with work proceeding to add a 4DVAR expanded to include a weak constraint treatment of the surface flux errors. Covariance of flux errors is estimated from the recent time series of forecast and calibrated flux terms. While the California Current examples are shown, the approach is equally applicable to other regions. These approaches within a 3DVAR application are anticipated to be useful for global and larger regional domains where a full 4DVAR methodology may be cost-prohibitive.

Parametric decadal climate forecast recalibration (DeFoReSt 1.0)

NASA Astrophysics Data System (ADS)

Pasternack, Alexander; Bhend, Jonas; Liniger, Mark A.; Rust, Henning W.; Müller, Wolfgang A.; Ulbrich, Uwe

2018-01-01

Near-term climate predictions such as decadal climate forecasts are increasingly being used to guide adaptation measures. For near-term probabilistic predictions to be useful, systematic errors of the forecasting systems have to be corrected. While methods for the calibration of probabilistic forecasts are readily available, these have to be adapted to the specifics of decadal climate forecasts including the long time horizon of decadal climate forecasts, lead-time-dependent systematic errors (drift) and the errors in the representation of long-term changes and variability. These features are compounded by small ensemble sizes to describe forecast uncertainty and a relatively short period for which typically pairs of reforecasts and observations are available to estimate calibration parameters. We introduce the Decadal Climate Forecast Recalibration Strategy (DeFoReSt), a parametric approach to recalibrate decadal ensemble forecasts that takes the above specifics into account. DeFoReSt optimizes forecast quality as measured by the continuous ranked probability score (CRPS). Using a toy model to generate synthetic forecast observation pairs, we demonstrate the positive effect on forecast quality in situations with pronounced and limited predictability. Finally, we apply DeFoReSt to decadal surface temperature forecasts from the MiKlip prototype system and find consistent, and sometimes considerable, improvements in forecast quality compared with a simple calibration of the lead-time-dependent systematic errors.

Predictability of the Arctic sea ice edge

NASA Astrophysics Data System (ADS)

Goessling, H. F.; Tietsche, S.; Day, J. J.; Hawkins, E.; Jung, T.

2016-02-01

Skillful sea ice forecasts from days to years ahead are becoming increasingly important for the operation and planning of human activities in the Arctic. Here we analyze the potential predictability of the Arctic sea ice edge in six climate models. We introduce the integrated ice-edge error (IIEE), a user-relevant verification metric defined as the area where the forecast and the "truth" disagree on the ice concentration being above or below 15%. The IIEE lends itself to decomposition into an absolute extent error, corresponding to the common sea ice extent error, and a misplacement error. We find that the often-neglected misplacement error makes up more than half of the climatological IIEE. In idealized forecast ensembles initialized on 1 July, the IIEE grows faster than the absolute extent error. This means that the Arctic sea ice edge is less predictable than sea ice extent, particularly in September, with implications for the potential skill of end-user relevant forecasts.

Influence of Forecast Accuracy of Photovoltaic Power Output on Capacity Optimization of Microgrid Composition under 30 min Power Balancing Control

NASA Astrophysics Data System (ADS)

Sone, Akihito; Kato, Takeyoshi; Shimakage, Toyonari; Suzuoki, Yasuo

A microgrid (MG) is one of the measures for enhancing the high penetration of renewable energy (RE)-based distributed generators (DGs). If a number of MGs are controlled to maintain the predetermined electricity demand including RE-based DGs as negative demand, they would contribute to supply-demand balancing of whole electric power system. For constructing a MG economically, the capacity optimization of controllable DGs against RE-based DGs is essential. By using a numerical simulation model developed based on a demonstrative study on a MG using PAFC and NaS battery as controllable DGs and photovoltaic power generation system (PVS) as a RE-based DG, this study discusses the influence of forecast accuracy of PVS output on the capacity optimization. Three forecast cases with different accuracy are compared. The main results are as follows. Even with no forecast error during every 30 min. as the ideal forecast method, the required capacity of NaS battery reaches about 40% of PVS capacity for mitigating the instantaneous forecast error within 30 min. The required capacity to compensate for the forecast error is doubled with the actual forecast method. The influence of forecast error can be reduced by adjusting the scheduled power output of controllable DGs according to the weather forecast. Besides, the required capacity can be reduced significantly if the error of balancing control in a MG is acceptable for a few percentages of periods, because the total periods of large forecast error is not so often.

Ability of matrix models to explain the past and predict the future of plant populations.

USGS Publications Warehouse

McEachern, Kathryn; Crone, Elizabeth E.; Ellis, Martha M.; Morris, William F.; Stanley, Amanda; Bell, Timothy; Bierzychudek, Paulette; Ehrlen, Johan; Kaye, Thomas N.; Knight, Tiffany M.; Lesica, Peter; Oostermeijer, Gerard; Quintana-Ascencio, Pedro F.; Ticktin, Tamara; Valverde, Teresa; Williams, Jennifer I.; Doak, Daniel F.; Ganesan, Rengaian; Thorpe, Andrea S.; Menges, Eric S.

2013-01-01

Uncertainty associated with ecological forecasts has long been recognized, but forecast accuracy is rarely quantified. We evaluated how well data on 82 populations of 20 species of plants spanning 3 continents explained and predicted plant population dynamics. We parameterized stage-based matrix models with demographic data from individually marked plants and determined how well these models forecast population sizes observed at least 5 years into the future. Simple demographic models forecasted population dynamics poorly; only 40% of observed population sizes fell within our forecasts' 95% confidence limits. However, these models explained population dynamics during the years in which data were collected; observed changes in population size during the data-collection period were strongly positively correlated with population growth rate. Thus, these models are at least a sound way to quantify population status. Poor forecasts were not associated with the number of individual plants or years of data. We tested whether vital rates were density dependent and found both positive and negative density dependence. However, density dependence was not associated with forecast error. Forecast error was significantly associated with environmental differences between the data collection and forecast periods. To forecast population fates, more detailed models, such as those that project how environments are likely to change and how these changes will affect population dynamics, may be needed. Such detailed models are not always feasible. Thus, it may be wiser to make risk-averse decisions than to expect precise forecasts from models.

Ability of matrix models to explain the past and predict the future of plant populations.

PubMed

Crone, Elizabeth E; Ellis, Martha M; Morris, William F; Stanley, Amanda; Bell, Timothy; Bierzychudek, Paulette; Ehrlén, Johan; Kaye, Thomas N; Knight, Tiffany M; Lesica, Peter; Oostermeijer, Gerard; Quintana-Ascencio, Pedro F; Ticktin, Tamara; Valverde, Teresa; Williams, Jennifer L; Doak, Daniel F; Ganesan, Rengaian; McEachern, Kathyrn; Thorpe, Andrea S; Menges, Eric S

2013-10-01

Uncertainty associated with ecological forecasts has long been recognized, but forecast accuracy is rarely quantified. We evaluated how well data on 82 populations of 20 species of plants spanning 3 continents explained and predicted plant population dynamics. We parameterized stage-based matrix models with demographic data from individually marked plants and determined how well these models forecast population sizes observed at least 5 years into the future. Simple demographic models forecasted population dynamics poorly; only 40% of observed population sizes fell within our forecasts' 95% confidence limits. However, these models explained population dynamics during the years in which data were collected; observed changes in population size during the data-collection period were strongly positively correlated with population growth rate. Thus, these models are at least a sound way to quantify population status. Poor forecasts were not associated with the number of individual plants or years of data. We tested whether vital rates were density dependent and found both positive and negative density dependence. However, density dependence was not associated with forecast error. Forecast error was significantly associated with environmental differences between the data collection and forecast periods. To forecast population fates, more detailed models, such as those that project how environments are likely to change and how these changes will affect population dynamics, may be needed. Such detailed models are not always feasible. Thus, it may be wiser to make risk-averse decisions than to expect precise forecasts from models. © 2013 Society for Conservation Biology.

Error sensitivity analysis in 10-30-day extended range forecasting by using a nonlinear cross-prediction error model

NASA Astrophysics Data System (ADS)

Xia, Zhiye; Xu, Lisheng; Chen, Hongbin; Wang, Yongqian; Liu, Jinbao; Feng, Wenlan

2017-06-01

Extended range forecasting of 10-30 days, which lies between medium-term and climate prediction in terms of timescale, plays a significant role in decision-making processes for the prevention and mitigation of disastrous meteorological events. The sensitivity of initial error, model parameter error, and random error in a nonlinear crossprediction error (NCPE) model, and their stability in the prediction validity period in 10-30-day extended range forecasting, are analyzed quantitatively. The associated sensitivity of precipitable water, temperature, and geopotential height during cases of heavy rain and hurricane is also discussed. The results are summarized as follows. First, the initial error and random error interact. When the ratio of random error to initial error is small (10-6-10-2), minor variation in random error cannot significantly change the dynamic features of a chaotic system, and therefore random error has minimal effect on the prediction. When the ratio is in the range of 10-1-2 (i.e., random error dominates), attention should be paid to the random error instead of only the initial error. When the ratio is around 10-2-10-1, both influences must be considered. Their mutual effects may bring considerable uncertainty to extended range forecasting, and de-noising is therefore necessary. Second, in terms of model parameter error, the embedding dimension m should be determined by the factual nonlinear time series. The dynamic features of a chaotic system cannot be depicted because of the incomplete structure of the attractor when m is small. When m is large, prediction indicators can vanish because of the scarcity of phase points in phase space. A method for overcoming the cut-off effect ( m > 4) is proposed. Third, for heavy rains, precipitable water is more sensitive to the prediction validity period than temperature or geopotential height; however, for hurricanes, geopotential height is most sensitive, followed by precipitable water.

Improving inflow forecasting into hydropower reservoirs through a complementary modelling framework

NASA Astrophysics Data System (ADS)

Gragne, A. S.; Sharma, A.; Mehrotra, R.; Alfredsen, K.

2014-10-01

Accuracy of reservoir inflow forecasts is instrumental for maximizing the value of water resources and benefits gained through hydropower generation. Improving hourly reservoir inflow forecasts over a 24 h lead-time is considered within the day-ahead (Elspot) market of the Nordic exchange market. We present here a new approach for issuing hourly reservoir inflow forecasts that aims to improve on existing forecasting models that are in place operationally, without needing to modify the pre-existing approach, but instead formulating an additive or complementary model that is independent and captures the structure the existing model may be missing. Besides improving forecast skills of operational models, the approach estimates the uncertainty in the complementary model structure and produces probabilistic inflow forecasts that entrain suitable information for reducing uncertainty in the decision-making processes in hydropower systems operation. The procedure presented comprises an error model added on top of an un-alterable constant parameter conceptual model, the models being demonstrated with reference to the 207 km2 Krinsvatn catchment in central Norway. The structure of the error model is established based on attributes of the residual time series from the conceptual model. Deterministic and probabilistic evaluations revealed an overall significant improvement in forecast accuracy for lead-times up to 17 h. Season based evaluations indicated that the improvement in inflow forecasts varies across seasons and inflow forecasts in autumn and spring are less successful with the 95% prediction interval bracketing less than 95% of the observations for lead-times beyond 17 h.

Seasonal prediction of Indian summer monsoon rainfall in NCEP CFSv2: forecast and predictability error

NASA Astrophysics Data System (ADS)

Pokhrel, Samir; Saha, Subodh Kumar; Dhakate, Ashish; Rahman, Hasibur; Chaudhari, Hemantkumar S.; Salunke, Kiran; Hazra, Anupam; Sujith, K.; Sikka, D. R.

2016-04-01

A detailed analysis of sensitivity to the initial condition for the simulation of the Indian summer monsoon using retrospective forecast by the latest version of the Climate Forecast System version-2 (CFSv2) is carried out. This study primarily focuses on the tropical region of Indian and Pacific Ocean basin, with special emphasis on the Indian land region. The simulated seasonal mean and the inter-annual standard deviations of rainfall, upper and lower level atmospheric circulations and Sea Surface Temperature (SST) tend to be more skillful as the lead forecast time decreases (5 month lead to 0 month lead time i.e. L5-L0). In general spatial correlation (bias) increases (decreases) as forecast lead time decreases. This is further substantiated by their averaged value over the selected study regions over the Indian and Pacific Ocean basins. The tendency of increase (decrease) of model bias with increasing (decreasing) forecast lead time also indicates the dynamical drift of the model. Large scale lower level circulation (850 hPa) shows enhancement of anomalous westerlies (easterlies) over the tropical region of the Indian Ocean (Western Pacific Ocean), which indicates the enhancement of model error with the decrease in lead time. At the upper level circulation (200 hPa) biases in both tropical easterly jet and subtropical westerlies jet tend to decrease as the lead time decreases. Despite enhancement of the prediction skill, mean SST bias seems to be insensitive to the initialization. All these biases are significant and together they make CFSv2 vulnerable to seasonal uncertainties in all the lead times. Overall the zeroth lead (L0) seems to have the best skill, however, in case of Indian summer monsoon rainfall (ISMR), the 3 month lead forecast time (L3) has the maximum ISMR prediction skill. This is valid using different independent datasets, wherein these maximum skill scores are 0.64, 0.42 and 0.57 with respect to the Global Precipitation Climatology Project, CPC Merged Analysis of Precipitation and the India Meteorological Department precipitation dataset respectively for L3. Despite significant El-Niño Southern Oscillation (ENSO) spring predictability barrier at L3, the ISMR skill score is highest at L3. Further, large scale zonal wind shear (Webster-Yang index) and SST over Niño3.4 region is best at L1 and L0. This implies that predictability aspect of ISMR is controlled by factors other than ENSO and Indian Ocean Dipole. Also, the model error (forecast error) outruns the error acquired by the inadequacies in the initial conditions (predictability error). Thus model deficiency is having more serious consequences as compared to the initial condition error for the seasonal forecast. All the model parameters show the increase in the predictability error as the lead decreases over the equatorial eastern Pacific basin and peaks at L2, then it further decreases. The dynamical consistency of both the forecast and the predictability error among all the variables indicates that these biases are purely systematic in nature and improvement of the physical processes in the CFSv2 may enhance the overall predictability.

Distortion Representation of Forecast Errors for Model Skill Assessment and Objective Analysis

NASA Technical Reports Server (NTRS)

Hoffman, Ross N.; Nehrkorn, Thomas; Grassotti, Christopher

1996-01-01

We study a novel characterization of errors for numerical weather predictions. In its simplest form we decompose the error into a part attributable to phase errors and a remainder. The phase error is represented in the same fashion as a velocity field and will be required to vary slowly and smoothly with position. A general distortion representation allows for the displacement and a bias correction of forecast anomalies. In brief, the distortion is determined by minimizing the objective function by varying the displacement and bias correction fields. In the present project we use a global or hemispheric domain, and spherical harmonics to represent these fields. In this project we are initially focusing on the assessment application, restricted to a realistic but univariate 2-dimensional situation. Specifically we study the forecast errors of the 500 hPa geopotential height field for forecasts of the short and medium range. The forecasts are those of the Goddard Earth Observing System data assimilation system. Results presented show that the methodology works, that a large part of the total error may be explained by a distortion limited to triangular truncation at wavenumber 10, and that the remaining residual error contains mostly small spatial scales.

The impact of different background errors in the assimilation of satellite radiances and in-situ observational data using WRFDA for three rainfall events over Iran

NASA Astrophysics Data System (ADS)

Zakeri, Zeinab; Azadi, Majid; Ghader, Sarmad

2018-01-01

Satellite radiances and in-situ observations are assimilated through Weather Research and Forecasting Data Assimilation (WRFDA) system into Advanced Research WRF (ARW) model over Iran and its neighboring area. Domain specific background error based on x and y components of wind speed (UV) control variables is calculated for WRFDA system and some sensitivity experiments are carried out to compare the impact of global background error and the domain specific background errors, both on the precipitation and 2-m temperature forecasts over Iran. Three precipitation events that occurred over the country during January, September and October 2014 are simulated in three different experiments and the results for precipitation and 2-m temperature are verified against the verifying surface observations. Results show that using domain specific background error improves 2-m temperature and 24-h accumulated precipitation forecasts consistently, while global background error may even degrade the forecasts compared to the experiments without data assimilation. The improvement in 2-m temperature is more evident during the first forecast hours and decreases significantly as the forecast length increases.

Why Is Rainfall Error Analysis Requisite for Data Assimilation and Climate Modeling?

NASA Technical Reports Server (NTRS)

Hou, Arthur Y.; Zhang, Sara Q.

2004-01-01

Given the large temporal and spatial variability of precipitation processes, errors in rainfall observations are difficult to quantify yet crucial to making effective use of rainfall data for improving atmospheric analysis, weather forecasting, and climate modeling. We highlight the need for developing a quantitative understanding of systematic and random errors in precipitation observations by examining explicit examples of how each type of errors can affect forecasts and analyses in global data assimilation. We characterize the error information needed from the precipitation measurement community and how it may be used to improve data usage within the general framework of analysis techniques, as well as accuracy requirements from the perspective of climate modeling and global data assimilation.

Error quantification of abnormal extreme high waves in Operational Oceanographic System in Korea

NASA Astrophysics Data System (ADS)

Jeong, Sang-Hun; Kim, Jinah; Heo, Ki-Young; Park, Kwang-Soon

2017-04-01

In winter season, large-height swell-like waves have occurred on the East coast of Korea, causing property damages and loss of human life. It is known that those waves are generated by a local strong wind made by temperate cyclone moving to eastward in the East Sea of Korean peninsula. Because the waves are often occurred in the clear weather, in particular, the damages are to be maximized. Therefore, it is necessary to predict and forecast large-height swell-like waves to prevent and correspond to the coastal damages. In Korea, an operational oceanographic system (KOOS) has been developed by the Korea institute of ocean science and technology (KIOST) and KOOS provides daily basis 72-hours' ocean forecasts such as wind, water elevation, sea currents, water temperature, salinity, and waves which are computed from not only meteorological and hydrodynamic model (WRF, ROMS, MOM, and MOHID) but also wave models (WW-III and SWAN). In order to evaluate the model performance and guarantee a certain level of accuracy of ocean forecasts, a Skill Assessment (SA) system was established as a one of module in KOOS. It has been performed through comparison of model results with in-situ observation data and model errors have been quantified with skill scores. Statistics which are used in skill assessment are including a measure of both errors and correlations such as root-mean-square-error (RMSE), root-mean-square-error percentage (RMSE%), mean bias (MB), correlation coefficient (R), scatter index (SI), circular correlation (CC) and central frequency (CF) that is a frequency with which errors lie within acceptable error criteria. It should be utilized simultaneously not only to quantify an error but also to improve an accuracy of forecasts by providing a feedback interactively. However, in an abnormal phenomena such as high-height swell-like waves in the East coast of Korea, it requires more advanced and optimized error quantification method that allows to predict the abnormal waves well and to improve the accuracy of forecasts by supporting modification of physics and numeric on numerical models through sensitivity test. In this study, we proposed an appropriate method of error quantification especially on abnormal high waves which are occurred by local weather condition. Furthermore, we introduced that how the quantification errors are contributed to improve wind-wave modeling by applying data assimilation and utilizing reanalysis data.

Testing for ontological errors in probabilistic forecasting models of natural systems

PubMed Central

Marzocchi, Warner; Jordan, Thomas H.

2014-01-01

Probabilistic forecasting models describe the aleatory variability of natural systems as well as our epistemic uncertainty about how the systems work. Testing a model against observations exposes ontological errors in the representation of a system and its uncertainties. We clarify several conceptual issues regarding the testing of probabilistic forecasting models for ontological errors: the ambiguity of the aleatory/epistemic dichotomy, the quantification of uncertainties as degrees of belief, the interplay between Bayesian and frequentist methods, and the scientific pathway for capturing predictability. We show that testability of the ontological null hypothesis derives from an experimental concept, external to the model, that identifies collections of data, observed and not yet observed, that are judged to be exchangeable when conditioned on a set of explanatory variables. These conditional exchangeability judgments specify observations with well-defined frequencies. Any model predicting these behaviors can thus be tested for ontological error by frequentist methods; e.g., using P values. In the forecasting problem, prior predictive model checking, rather than posterior predictive checking, is desirable because it provides more severe tests. We illustrate experimental concepts using examples from probabilistic seismic hazard analysis. Severe testing of a model under an appropriate set of experimental concepts is the key to model validation, in which we seek to know whether a model replicates the data-generating process well enough to be sufficiently reliable for some useful purpose, such as long-term seismic forecasting. Pessimistic views of system predictability fail to recognize the power of this methodology in separating predictable behaviors from those that are not. PMID:25097265

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 of the 29-km Eta Model for Weather Support to the United States Space Program

NASA Technical Reports Server (NTRS)

Manobianco, John; Nutter, Paul

1997-01-01

The Applied Meteorology Unit (AMU) conducted a year-long evaluation of NCEP's 29-km mesoscale Eta (meso-eta) weather prediction model in order to identify added value to forecast operations in support of the United States space program. The evaluation was stratified over warm and cool seasons and considered both objective and subjective verification methodologies. Objective verification results generally indicate that meso-eta model point forecasts at selected stations exhibit minimal error growth in terms of RMS errors and are reasonably unbiased. Conversely, results from the subjective verification demonstrate that model forecasts of developing weather events such as thunderstorms, sea breezes, and cold fronts, are not always as accurate as implied by the seasonal error statistics. Sea-breeze case studies reveal that the model generates a dynamically-consistent thermally direct circulation over the Florida peninsula, although at a larger scale than observed. Thunderstorm verification reveals that the meso-eta model is capable of predicting areas of organized convection, particularly during the late afternoon hours but is not capable of forecasting individual thunderstorms. Verification of cold fronts during the cool season reveals that the model is capable of forecasting a majority of cold frontal passages through east central Florida to within +1-h of observed frontal passage.

A Study on Mutil-Scale Background Error Covariances in 3D-Var Data Assimilation

NASA Astrophysics Data System (ADS)

Zhang, Xubin; Tan, Zhe-Min

2017-04-01

The construction of background error covariances is a key component of three-dimensional variational data assimilation. There are different scale background errors and interactions among them in the numerical weather Prediction. However, the influence of these errors and their interactions cannot be represented in the background error covariances statistics when estimated by the leading methods. So, it is necessary to construct background error covariances influenced by multi-scale interactions among errors. With the NMC method, this article firstly estimates the background error covariances at given model-resolution scales. And then the information of errors whose scales are larger and smaller than the given ones is introduced respectively, using different nesting techniques, to estimate the corresponding covariances. The comparisons of three background error covariances statistics influenced by information of errors at different scales reveal that, the background error variances enhance particularly at large scales and higher levels when introducing the information of larger-scale errors by the lateral boundary condition provided by a lower-resolution model. On the other hand, the variances reduce at medium scales at the higher levels, while those show slight improvement at lower levels in the nested domain, especially at medium and small scales, when introducing the information of smaller-scale errors by nesting a higher-resolution model. In addition, the introduction of information of larger- (smaller-) scale errors leads to larger (smaller) horizontal and vertical correlation scales of background errors. Considering the multivariate correlations, the Ekman coupling increases (decreases) with the information of larger- (smaller-) scale errors included, whereas the geostrophic coupling in free atmosphere weakens in both situations. The three covariances obtained in above work are used in a data assimilation and model forecast system respectively, and then the analysis-forecast cycles for a period of 1 month are conducted. Through the comparison of both analyses and forecasts from this system, it is found that the trends for variation in analysis increments with information of different scale errors introduced are consistent with those for variation in variances and correlations of background errors. In particular, introduction of smaller-scale errors leads to larger amplitude of analysis increments for winds at medium scales at the height of both high- and low- level jet. And analysis increments for both temperature and humidity are greater at the corresponding scales at middle and upper levels under this circumstance. These analysis increments improve the intensity of jet-convection system which includes jets at different levels and coupling between them associated with latent heat release, and these changes in analyses contribute to the better forecasts for winds and temperature in the corresponding areas. When smaller-scale errors are included, analysis increments for humidity enhance significantly at large scales at lower levels to moisten southern analyses. This humidification devotes to correcting dry bias there and eventually improves forecast skill of humidity. Moreover, inclusion of larger- (smaller-) scale errors is beneficial for forecast quality of heavy (light) precipitation at large (small) scales due to the amplification (diminution) of intensity and area in precipitation forecasts but tends to overestimate (underestimate) light (heavy) precipitation .

Arima model and exponential smoothing method: A comparison

NASA Astrophysics Data System (ADS)

Wan Ahmad, Wan Kamarul Ariffin; Ahmad, Sabri

2013-04-01

This study shows the comparison between Autoregressive Moving Average (ARIMA) model and Exponential Smoothing Method in making a prediction. The comparison is focused on the ability of both methods in making the forecasts with the different number of data sources and the different length of forecasting period. For this purpose, the data from The Price of Crude Palm Oil (RM/tonne), Exchange Rates of Ringgit Malaysia (RM) in comparison to Great Britain Pound (GBP) and also The Price of SMR 20 Rubber Type (cents/kg) with three different time series are used in the comparison process. Then, forecasting accuracy of each model is measured by examinethe prediction error that producedby using Mean Squared Error (MSE), Mean Absolute Percentage Error (MAPE), and Mean Absolute deviation (MAD). The study shows that the ARIMA model can produce a better prediction for the long-term forecasting with limited data sources, butcannot produce a better prediction for time series with a narrow range of one point to another as in the time series for Exchange Rates. On the contrary, Exponential Smoothing Method can produce a better forecasting for Exchange Rates that has a narrow range of one point to another for its time series, while itcannot produce a better prediction for a longer forecasting period.

A Case Study of the Impact of AIRS Temperature Retrievals on Numerical Weather Prediction

NASA Technical Reports Server (NTRS)

Reale, O.; Atlas, R.; Jusem, J. C.

2004-01-01

Large errors in numerical weather prediction are often associated with explosive cyclogenesis. Most studes focus on the under-forecasting error, i.e. cases of rapidly developing cyclones which are poorly predicted in numerical models. However, the over-forecasting error (i.e., to predict an explosively developing cyclone which does not occur in reality) is a very common error that severely impacts the forecasting skill of all models and may also present economic costs if associated with operational forecasting. Unnecessary precautions taken by marine activities can result in severe economic loss. Moreover, frequent occurrence of over-forecasting can undermine the reliance on operational weather forecasting. Therefore, it is important to understand and reduce the prdctions of extreme weather associated with explosive cyclones which do not actually develop. In this study we choose a very prominent case of over-forecasting error in the northwestern Pacific. A 960 hPa cyclone develops in less than 24 hour in the 5-day forecast, with a deepening rate of about 30 hPa in one day. The cyclone is not versed in the analyses and is thus a case of severe over-forecasting. By assimilating AIRS data, the error is largely eliminated. By following the propagation of the anomaly that generates the spurious cyclone, it is found that a small mid-tropospheric geopotential height negative anomaly over the northern part of the Indian subcontinent in the initial conditions, propagates westward, is amplified by orography, and generates a very intense jet streak in the subtropical jet stream, with consequent explosive cyclogenesis over the Pacific. The AIRS assimilation eliminates this anomaly that may have been caused by erroneous upper-air data, and represents the jet stream more correctly. The energy associated with the jet is distributed over a much broader area and as a consequence a multiple, but much more moderate cyclogenesis is observed.

Evaluation of the 29-km Eta Model. Part 1; Objective Verification at Three Selected Stations

NASA Technical Reports Server (NTRS)

Nutter, Paul A.; Manobianco, John; Merceret, Francis J. (Technical Monitor)

1998-01-01

This paper describes an objective verification of the National Centers for Environmental Prediction (NCEP) 29-km eta model from May 1996 through January 1998. The evaluation was designed to assess the model's surface and upper-air point forecast accuracy at three selected locations during separate warm (May - August) and cool (October - January) season periods. In order to enhance sample sizes available for statistical calculations, the objective verification includes two consecutive warm and cool season periods. Systematic model deficiencies comprise the larger portion of the total error in most of the surface forecast variables that were evaluated. The error characteristics for both surface and upper-air forecasts vary widely by parameter, season, and station location. At upper levels, a few characteristic biases are identified. Overall however, the upper-level errors are more nonsystematic in nature and could be explained partly by observational measurement uncertainty. With a few exceptions, the upper-air results also indicate that 24-h model error growth is not statistically significant. In February and August 1997, NCEP implemented upgrades to the eta model's physical parameterizations that were designed to change some of the model's error characteristics near the surface. The results shown in this paper indicate that these upgrades led to identifiable and statistically significant changes in forecast accuracy for selected surface parameters. While some of the changes were expected, others were not consistent with the intent of the model updates and further emphasize the need for ongoing sensitivity studies and localized statistical verification efforts. Objective verification of point forecasts is a stringent measure of model performance, but when used alone, is not enough to quantify the overall value that model guidance may add to the forecast process. Therefore, results from a subjective verification of the meso-eta model over the Florida peninsula are discussed in the companion paper by Manobianco and Nutter. Overall verification results presented here and in part two should establish a reasonable benchmark from which model users and developers may pursue the ongoing eta model verification strategies in the future.

Forecasting malaria in a highly endemic country using environmental and clinical predictors.

PubMed

Zinszer, Kate; Kigozi, Ruth; Charland, Katia; Dorsey, Grant; Brewer, Timothy F; Brownstein, John S; Kamya, Moses R; Buckeridge, David L

2015-06-18

Malaria thrives in poor tropical and subtropical countries where local resources are limited. Accurate disease forecasts can provide public and clinical health services with the information needed to implement targeted approaches for malaria control that make effective use of limited resources. The objective of this study was to determine the relevance of environmental and clinical predictors of malaria across different settings in Uganda. Forecasting models were based on health facility data collected by the Uganda Malaria Surveillance Project and satellite-derived rainfall, temperature, and vegetation estimates from 2006 to 2013. Facility-specific forecasting models of confirmed malaria were developed using multivariate autoregressive integrated moving average models and produced weekly forecast horizons over a 52-week forecasting period. The model with the most accurate forecasts varied by site and by forecast horizon. Clinical predictors were retained in the models with the highest predictive power for all facility sites. The average error over the 52 forecasting horizons ranged from 26 to 128% whereas the cumulative burden forecast error ranged from 2 to 22%. Clinical data, such as drug treatment, could be used to improve the accuracy of malaria predictions in endemic settings when coupled with environmental predictors. Further exploration of malaria forecasting is necessary to improve its accuracy and value in practice, including examining other environmental and intervention predictors, including insecticide-treated nets.

Modeled Forecasts of Dengue Fever in San Juan, Puerto Rico Using NASA Satellite Enhanced Weather Forecasts

NASA Astrophysics Data System (ADS)

Morin, C.; Quattrochi, D. A.; Zavodsky, B.; Case, J.

2015-12-01

Dengue fever (DF) is an important mosquito transmitted disease that is strongly influenced by meteorological and environmental conditions. Recent research has focused on forecasting DF case numbers based on meteorological data. However, these forecasting tools have generally relied on empirical models that require long DF time series to train. Additionally, their accuracy has been tested retrospectively, using past meteorological data. Consequently, the operational utility of the forecasts are still in question because the error associated with weather and climate forecasts are not reflected in the results. Using up-to-date weekly dengue case numbers for model parameterization and weather forecast data as meteorological input, we produced weekly forecasts of DF cases in San Juan, Puerto Rico. Each week, the past weeks' case counts were used to re-parameterize a process-based DF model driven with updated weather forecast data to generate forecasts of DF case numbers. Real-time weather forecast data was produced using the Weather Research and Forecasting (WRF) numerical weather prediction (NWP) system enhanced using additional high-resolution NASA satellite data. This methodology was conducted in a weekly iterative process with each DF forecast being evaluated using county-level DF cases reported by the Puerto Rico Department of Health. The one week DF forecasts were accurate especially considering the two sources of model error. First, weather forecasts were sometimes inaccurate and generally produced lower than observed temperatures. Second, the DF model was often overly influenced by the previous weeks DF case numbers, though this phenomenon could be lessened by increasing the number of simulations included in the forecast. Although these results are promising, we would like to develop a methodology to produce longer range forecasts so that public health workers can better prepare for dengue epidemics.

Verification of an ensemble prediction system for storm surge forecast in the Adriatic Sea

NASA Astrophysics Data System (ADS)

Mel, Riccardo; Lionello, Piero

2014-12-01

In the Adriatic Sea, storm surges present a significant threat to Venice and to the flat coastal areas of the northern coast of the basin. Sea level forecast is of paramount importance for the management of daily activities and for operating the movable barriers that are presently being built for the protection of the city. In this paper, an EPS (ensemble prediction system) for operational forecasting of storm surge in the northern Adriatic Sea is presented and applied to a 3-month-long period (October-December 2010). The sea level EPS is based on the HYPSE (hydrostatic Padua Sea elevation) model, which is a standard single-layer nonlinear shallow water model, whose forcings (mean sea level pressure and surface wind fields) are provided by the ensemble members of the ECMWF (European Center for Medium-Range Weather Forecasts) EPS. Results are verified against observations at five tide gauges located along the Croatian and Italian coasts of the Adriatic Sea. Forecast uncertainty increases with the predicted value of the storm surge and with the forecast lead time. The EMF (ensemble mean forecast) provided by the EPS has a rms (root mean square) error lower than the DF (deterministic forecast), especially for short (up to 3 days) lead times. Uncertainty for short lead times of the forecast and for small storm surges is mainly caused by uncertainty of the initial condition of the hydrodynamical model. Uncertainty for large lead times and large storm surges is mainly caused by uncertainty in the meteorological forcings. The EPS spread increases with the rms error of the forecast. For large lead times the EPS spread and the forecast error substantially coincide. However, the EPS spread in this study, which does not account for uncertainty in the initial condition, underestimates the error during the early part of the forecast and for small storm surge values. On the contrary, it overestimates the rms error for large surge values. The PF (probability forecast) of the EPS has a clear skill in predicting the actual probability distribution of sea level, and it outperforms simple "dressed" PF methods. A probability estimate based on the single DF is shown to be inadequate. However, a PF obtained with a prescribed Gaussian distribution and centered on the DF value performs very similarly to the EPS-based PF.

Extended Range Prediction of Indian Summer Monsoon: Current status

NASA Astrophysics Data System (ADS)

Sahai, A. K.; Abhilash, S.; Borah, N.; Joseph, S.; Chattopadhyay, R.; S, S.; Rajeevan, M.; Mandal, R.; Dey, A.

2014-12-01

The main focus of this study is to develop forecast consensus in the extended range prediction (ERP) of monsoon Intraseasonal oscillations using a suit of different variants of Climate Forecast system (CFS) model. In this CFS based Grand MME prediction system (CGMME), the ensemble members are generated by perturbing the initial condition and using different configurations of CFSv2. This is to address the role of different physical mechanisms known to have control on the error growth in the ERP in the 15-20 day time scale. The final formulation of CGMME is based on 21 ensembles of the standalone Global Forecast System (GFS) forced with bias corrected forecasted SST from CFS, 11 low resolution CFST126 and 11 high resolution CFST382. Thus, we develop the multi-model consensus forecast for the ERP of Indian summer monsoon (ISM) using a suite of different variants of CFS model. This coordinated international effort lead towards the development of specific tailor made regional forecast products over Indian region. Skill of deterministic and probabilistic categorical rainfall forecast as well the verification of large-scale low frequency monsoon intraseasonal oscillations has been carried out using hindcast from 2001-2012 during the monsoon season in which all models are initialized at every five days starting from 16May to 28 September. The skill of deterministic forecast from CGMME is better than the best participating single model ensemble configuration (SME). The CGMME approach is believed to quantify the uncertainty in both initial conditions and model formulation. Main improvement is attained in probabilistic forecast which is because of an increase in the ensemble spread, thereby reducing the error due to over-confident ensembles in a single model configuration. For probabilistic forecast, three tercile ranges are determined by ranking method based on the percentage of ensemble members from all the participating models falls in those three categories. CGMME further added value to both deterministic and probability forecast compared to raw SME's and this better skill is probably flows from large spread and improved spread-error relationship. CGMME system is currently capable of generating ER prediction in real time and successfully delivering its experimental operational ER forecast of ISM for the last few years.

Evaluation of Satellite and Model Precipitation Products Over Turkey

NASA Astrophysics Data System (ADS)

Yilmaz, M. T.; Amjad, M.

2017-12-01

Satellite-based remote sensing, gauge stations, and models are the three major platforms to acquire precipitation dataset. Among them satellites and models have the advantage of retrieving spatially and temporally continuous and consistent datasets, while the uncertainty estimates of these retrievals are often required for many hydrological studies to understand the source and the magnitude of the uncertainty in hydrological response parameters. In this study, satellite and model precipitation data products are validated over various temporal scales (daily, 3-daily, 7-daily, 10-daily and monthly) using in-situ measured precipitation observations from a network of 733 gauges from all over the Turkey. Tropical Rainfall Measurement Mission (TRMM) Multi-satellite Precipitation Analysis (TMPA) 3B42 version 7 and European Center of Medium-Range Weather Forecast (ECMWF) model estimates (daily, 3-daily, 7-daily and 10-daily accumulated forecast) are used in this study. Retrievals are evaluated for their mean and standard deviation and their accuracies are evaluated via bias, root mean square error, error standard deviation and correlation coefficient statistics. Intensity vs frequency analysis and some contingency table statistics like percent correct, probability of detection, false alarm ratio and critical success index are determined using daily time-series. Both ECMWF forecasts and TRMM observations, on average, overestimate the precipitation compared to gauge estimates; wet biases are 10.26 mm/month and 8.65 mm/month, respectively for ECMWF and TRMM. RMSE values of ECMWF forecasts and TRMM estimates are 39.69 mm/month and 41.55 mm/month, respectively. Monthly correlations between Gauges-ECMWF, Gauges-TRMM and ECMWF-TRMM are 0.76, 0.73 and 0.81, respectively. The model and the satellite error statistics are further compared against the gauges error statistics based on inverse distance weighting (IWD) analysis. Both the model and satellite data have less IWD errors (14.72 mm/month and 10.75 mm/month, respectively) compared to gauges IWD error (21.58 mm/month). These results show that, on average, ECMWF forecast data have higher skill than TRMM observations. Overall, both ECMWF forecast data and TRMM observations show good potential for catchment scale hydrological analysis.

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.

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.

Forecasting daily streamflow using online sequential extreme learning machines

NASA Astrophysics Data System (ADS)

Lima, Aranildo R.; Cannon, Alex J.; Hsieh, William W.

2016-06-01

While nonlinear machine methods have been widely used in environmental forecasting, in situations where new data arrive continually, the need to make frequent model updates can become cumbersome and computationally costly. To alleviate this problem, an online sequential learning algorithm for single hidden layer feedforward neural networks - the online sequential extreme learning machine (OSELM) - is automatically updated inexpensively as new data arrive (and the new data can then be discarded). OSELM was applied to forecast daily streamflow at two small watersheds in British Columbia, Canada, at lead times of 1-3 days. Predictors used were weather forecast data generated by the NOAA Global Ensemble Forecasting System (GEFS), and local hydro-meteorological observations. OSELM forecasts were tested with daily, monthly or yearly model updates. More frequent updating gave smaller forecast errors, including errors for data above the 90th percentile. Larger datasets used in the initial training of OSELM helped to find better parameters (number of hidden nodes) for the model, yielding better predictions. With the online sequential multiple linear regression (OSMLR) as benchmark, we concluded that OSELM is an attractive approach as it easily outperformed OSMLR in forecast accuracy.

A review of multimodel superensemble forecasting for weather, seasonal climate, and hurricanes

NASA Astrophysics Data System (ADS)

Krishnamurti, T. N.; Kumar, V.; Simon, A.; Bhardwaj, A.; Ghosh, T.; Ross, R.

2016-06-01

This review provides a summary of work in the area of ensemble forecasts for weather, climate, oceans, and hurricanes. This includes a combination of multiple forecast model results that does not dwell on the ensemble mean but uses a unique collective bias reduction procedure. A theoretical framework for this procedure is provided, utilizing a suite of models that is constructed from the well-known Lorenz low-order nonlinear system. A tutorial that includes a walk-through table and illustrates the inner workings of the multimodel superensemble's principle is provided. Systematic errors in a single deterministic model arise from a host of features that range from the model's initial state (data assimilation), resolution, representation of physics, dynamics, and ocean processes, local aspects of orography, water bodies, and details of the land surface. Models, in their diversity of representation of such features, end up leaving unique signatures of systematic errors. The multimodel superensemble utilizes as many as 10 million weights to take into account the bias errors arising from these diverse features of multimodels. The design of a single deterministic forecast models that utilizes multiple features from the use of the large volume of weights is provided here. This has led to a better understanding of the error growths and the collective bias reductions for several of the physical parameterizations within diverse models, such as cumulus convection, planetary boundary layer physics, and radiative transfer. A number of examples for weather, seasonal climate, hurricanes and sub surface oceanic forecast skills of member models, the ensemble mean, and the superensemble are provided.

Post-processing through linear regression

NASA Astrophysics Data System (ADS)

van Schaeybroeck, B.; Vannitsem, S.

2011-03-01

Various post-processing techniques are compared for both deterministic and ensemble forecasts, all based on linear regression between forecast data and observations. In order to evaluate the quality of the regression methods, three criteria are proposed, related to the effective correction of forecast error, the optimal variability of the corrected forecast and multicollinearity. The regression schemes under consideration include the ordinary least-square (OLS) method, a new time-dependent Tikhonov regularization (TDTR) method, the total least-square method, a new geometric-mean regression (GM), a recently introduced error-in-variables (EVMOS) method and, finally, a "best member" OLS method. The advantages and drawbacks of each method are clarified. These techniques are applied in the context of the 63 Lorenz system, whose model version is affected by both initial condition and model errors. For short forecast lead times, the number and choice of predictors plays an important role. Contrarily to the other techniques, GM degrades when the number of predictors increases. At intermediate lead times, linear regression is unable to provide corrections to the forecast and can sometimes degrade the performance (GM and the best member OLS with noise). At long lead times the regression schemes (EVMOS, TDTR) which yield the correct variability and the largest correlation between ensemble error and spread, should be preferred.

Moisture Forecast Bias Correction in GEOS DAS

NASA Technical Reports Server (NTRS)

Dee, D.

1999-01-01

Data assimilation methods rely on numerous assumptions about the errors involved in measuring and forecasting atmospheric fields. One of the more disturbing of these is that short-term model forecasts are assumed to be unbiased. In case of atmospheric moisture, for example, observational evidence shows that the systematic component of errors in forecasts and analyses is often of the same order of magnitude as the random component. we have implemented a sequential algorithm for estimating forecast moisture bias from rawinsonde data in the Goddard Earth Observing System Data Assimilation System (GEOS DAS). The algorithm is designed to remove the systematic component of analysis errors and can be easily incorporated in an existing statistical data assimilation system. We will present results of initial experiments that show a significant reduction of bias in the GEOS DAS moisture analyses.

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.

Improving real-time inflow forecasting into hydropower reservoirs through a complementary modelling framework

NASA Astrophysics Data System (ADS)

Gragne, A. S.; Sharma, A.; Mehrotra, R.; Alfredsen, K.

2015-08-01

Accuracy of reservoir inflow forecasts is instrumental for maximizing the value of water resources and benefits gained through hydropower generation. Improving hourly reservoir inflow forecasts over a 24 h lead time is considered within the day-ahead (Elspot) market of the Nordic exchange market. A complementary modelling framework presents an approach for improving real-time forecasting without needing to modify the pre-existing forecasting model, but instead formulating an independent additive or complementary model that captures the structure the existing operational model may be missing. We present here the application of this principle for issuing improved hourly inflow forecasts into hydropower reservoirs over extended lead times, and the parameter estimation procedure reformulated to deal with bias, persistence and heteroscedasticity. The procedure presented comprises an error model added on top of an unalterable constant parameter conceptual model. This procedure is applied in the 207 km2 Krinsvatn catchment in central Norway. The structure of the error model is established based on attributes of the residual time series from the conceptual model. Besides improving forecast skills of operational models, the approach estimates the uncertainty in the complementary model structure and produces probabilistic inflow forecasts that entrain suitable information for reducing uncertainty in the decision-making processes in hydropower systems operation. Deterministic and probabilistic evaluations revealed an overall significant improvement in forecast accuracy for lead times up to 17 h. Evaluation of the percentage of observations bracketed in the forecasted 95 % confidence interval indicated that the degree of success in containing 95 % of the observations varies across seasons and hydrologic years.

Volcanic ash modeling with the NMMB-MONARCH-ASH model: quantification of offline modeling errors

NASA Astrophysics Data System (ADS)

Marti, Alejandro; Folch, Arnau

2018-03-01

Volcanic ash modeling systems are used to simulate the atmospheric dispersion of volcanic ash and to generate forecasts that quantify the impacts from volcanic eruptions on infrastructures, air quality, aviation, and climate. The efficiency of response and mitigation actions is directly associated with the accuracy of the volcanic ash cloud detection and modeling systems. Operational forecasts build on offline coupled modeling systems in which meteorological variables are updated at the specified coupling intervals. Despite the concerns from other communities regarding the accuracy of this strategy, the quantification of the systematic errors and shortcomings associated with the offline modeling systems has received no attention. This paper employs the NMMB-MONARCH-ASH model to quantify these errors by employing different quantitative and categorical evaluation scores. The skills of the offline coupling strategy are compared against those from an online forecast considered to be the best estimate of the true outcome. Case studies are considered for a synthetic eruption with constant eruption source parameters and for two historical events, which suitably illustrate the severe aviation disruptive effects of European (2010 Eyjafjallajökull) and South American (2011 Cordón Caulle) volcanic eruptions. Evaluation scores indicate that systematic errors due to the offline modeling are of the same order of magnitude as those associated with the source term uncertainties. In particular, traditional offline forecasts employed in operational model setups can result in significant uncertainties, failing to reproduce, in the worst cases, up to 45-70 % of the ash cloud of an online forecast. These inconsistencies are anticipated to be even more relevant in scenarios in which the meteorological conditions change rapidly in time. The outcome of this paper encourages operational groups responsible for real-time advisories for aviation to consider employing computationally efficient online dispersal models.

Analog-Based Postprocessing of Navigation-Related Hydrological Ensemble Forecasts

NASA Astrophysics Data System (ADS)

Hemri, S.; Klein, B.

2017-11-01

Inland waterway transport benefits from probabilistic forecasts of water levels as they allow to optimize the ship load and, hence, to minimize the transport costs. Probabilistic state-of-the-art hydrologic ensemble forecasts inherit biases and dispersion errors from the atmospheric ensemble forecasts they are driven with. The use of statistical postprocessing techniques like ensemble model output statistics (EMOS) allows for a reduction of these systematic errors by fitting a statistical model based on training data. In this study, training periods for EMOS are selected based on forecast analogs, i.e., historical forecasts that are similar to the forecast to be verified. Due to the strong autocorrelation of water levels, forecast analogs have to be selected based on entire forecast hydrographs in order to guarantee similar hydrograph shapes. Custom-tailored measures of similarity for forecast hydrographs comprise hydrological series distance (SD), the hydrological matching algorithm (HMA), and dynamic time warping (DTW). Verification against observations reveals that EMOS forecasts for water level at three gauges along the river Rhine with training periods selected based on SD, HMA, and DTW compare favorably with reference EMOS forecasts, which are based on either seasonal training periods or on training periods obtained by dividing the hydrological forecast trajectories into runoff regimes.

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.

Implementation and Research on the Operational Use of the Mesoscale Prediction Model COAMPS in Poland

DTIC Science & Technology

2007-09-30

COAMPS model. Bogumil Jakubiak, University of Warsaw – participated in EGU General Assembly , Vienna Austria 15-20 April 2007 giving one oral and two...conditional forecast (background) error probability density function using an ensemble of the model forecast to generate background error statistics...COAMPS system on ICM machines at Warsaw University for the purpose of providing operational support to the general public using the ICM meteorological

MO-C-17A-04: Forecasting Longitudinal Changes in Oropharyngeal Tumor Morphology Throughout the Course of Head and Neck Radiation Therapy

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

Yock, A; UT Graduate School of Biomedical Sciences, Houston, TX; Rao, A

2014-06-15

Purpose: To generate, evaluate, and compare models that predict longitudinal changes in tumor morphology throughout the course of radiation therapy. Methods: Two morphology feature vectors were used to describe the size, shape, and position of 35 oropharyngeal GTVs at each treatment fraction during intensity-modulated radiation therapy. The feature vectors comprised the coordinates of the GTV centroids and one of two shape descriptors. One shape descriptor was based on radial distances between the GTV centroid and 614 GTV surface landmarks. The other was based on a spherical harmonic decomposition of these distances. Feature vectors over the course of therapy were describedmore » using static, linear, and mean models. The error of these models in forecasting GTV morphology was evaluated with leave-one-out cross-validation, and their accuracy was compared using Wilcoxon signed-rank tests. The effect of adjusting model parameters at 1, 2, 3, or 5 time points (adjustment points) was also evaluated. Results: The addition of a single adjustment point to the static model decreased the median error in forecasting the position of GTV surface landmarks by 1.2 mm (p<0.001). Additional adjustment points further decreased forecast error by about 0.4 mm each. The linear model decreased forecast error compared to the static model for feature vectors based on both shape descriptors (0.2 mm), while the mean model did so only for those based on the inter-landmark distances (0.2 mm). The decrease in forecast error due to adding adjustment points was greater than that due to model selection. Both effects diminished with subsequent adjustment points. Conclusion: Models of tumor morphology that include information from prior patients and/or prior treatment fractions are able to predict the tumor surface at each treatment fraction during radiation therapy. The predicted tumor morphology can be compared with patient anatomy or dose distributions, opening the possibility of anticipatory re-planning. American Legion Auxiliary Fellowship; The University of Texas Graduate School of Biomedical Sciences at Houston.« less

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.

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.

Sensitivity of the model error parameter specification in weak-constraint four-dimensional variational data assimilation

NASA Astrophysics Data System (ADS)

Shaw, Jeremy A.; Daescu, Dacian N.

2017-08-01

This article presents the mathematical framework to evaluate the sensitivity of a forecast error aspect to the input parameters of a weak-constraint four-dimensional variational data assimilation system (w4D-Var DAS), extending the established theory from strong-constraint 4D-Var. Emphasis is placed on the derivation of the equations for evaluating the forecast sensitivity to parameters in the DAS representation of the model error statistics, including bias, standard deviation, and correlation structure. A novel adjoint-based procedure for adaptive tuning of the specified model error covariance matrix is introduced. Results from numerical convergence tests establish the validity of the model error sensitivity equations. Preliminary experiments providing a proof-of-concept are performed using the Lorenz multi-scale model to illustrate the theoretical concepts and potential benefits for practical applications.

On the Likely Utility of Hybrid Weights Optimized for Variances in Hybrid Error Covariance Models

NASA Astrophysics Data System (ADS)

Satterfield, E.; Hodyss, D.; Kuhl, D.; Bishop, C. H.

2017-12-01

Because of imperfections in ensemble data assimilation schemes, one cannot assume that the ensemble covariance is equal to the true error covariance of a forecast. Previous work demonstrated how information about the distribution of true error variances given an ensemble sample variance can be revealed from an archive of (observation-minus-forecast, ensemble-variance) data pairs. Here, we derive a simple and intuitively compelling formula to obtain the mean of this distribution of true error variances given an ensemble sample variance from (observation-minus-forecast, ensemble-variance) data pairs produced by a single run of a data assimilation system. This formula takes the form of a Hybrid weighted average of the climatological forecast error variance and the ensemble sample variance. Here, we test the extent to which these readily obtainable weights can be used to rapidly optimize the covariance weights used in Hybrid data assimilation systems that employ weighted averages of static covariance models and flow-dependent ensemble based covariance models. Univariate data assimilation and multi-variate cycling ensemble data assimilation are considered. In both cases, it is found that our computationally efficient formula gives Hybrid weights that closely approximate the optimal weights found through the simple but computationally expensive process of testing every plausible combination of weights.

A GLM Post-processor to Adjust Ensemble Forecast Traces

NASA Astrophysics Data System (ADS)

Thiemann, M.; Day, G. N.; Schaake, J. C.; Draijer, S.; Wang, L.

2011-12-01

The skill of hydrologic ensemble forecasts has improved in the last years through a better understanding of climate variability, better climate forecasts and new data assimilation techniques. Having been extensively utilized for probabilistic water supply forecasting, interest is developing to utilize these forecasts in operational decision making. Hydrologic ensemble forecast members typically have inherent biases in flow timing and volume caused by (1) structural errors in the models used, (2) systematic errors in the data used to calibrate those models, (3) uncertain initial hydrologic conditions, and (4) uncertainties in the forcing datasets. Furthermore, hydrologic models have often not been developed for operational decision points and ensemble forecasts are thus not always available where needed. A statistical post-processor can be used to address these issues. The post-processor should (1) correct for systematic biases in flow timing and volume, (2) preserve the skill of the available raw forecasts, (3) preserve spatial and temporal correlation as well as the uncertainty in the forecasted flow data, (4) produce adjusted forecast ensembles that represent the variability of the observed hydrograph to be predicted, and (5) preserve individual forecast traces as equally likely. The post-processor should also allow for the translation of available ensemble forecasts to hydrologically similar locations where forecasts are not available. This paper introduces an ensemble post-processor (EPP) developed in support of New York City water supply operations. The EPP employs a general linear model (GLM) to (1) adjust available ensemble forecast traces and (2) create new ensembles for (nearby) locations where only historic flow observations are available. The EPP is calibrated by developing daily and aggregated statistical relationships form historical flow observations and model simulations. These are then used in operation to obtain the conditional probability density function (PDF) of the observations to be predicted, thus jointly adjusting individual ensemble members. These steps are executed in a normalized transformed space ('z'-space) to account for the strong non-linearity in the flow observations involved. A data window centered on each calibration date is used to minimize impacts from sampling errors and data noise. Testing on datasets from California and New York suggests that the EPP can successfully minimize biases in ensemble forecasts, while preserving the raw forecast skill in a 'days to weeks' forecast horizon and reproducing the variability of climatology for 'weeks to years' forecast horizons.

Energy Storage Sizing Taking Into Account Forecast Uncertainties and Receding Horizon Operation

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

Baker, Kyri; Hug, Gabriela; Li, Xin

Energy storage systems (ESS) have the potential to be very beneficial for applications such as reducing the ramping of generators, peak shaving, and balancing not only the variability introduced by renewable energy sources, but also the uncertainty introduced by errors in their forecasts. Optimal usage of storage may result in reduced generation costs and an increased use of renewable energy. However, optimally sizing these devices is a challenging problem. This paper aims to provide the tools to optimally size an ESS under the assumption that it will be operated under a model predictive control scheme and that the forecast ofmore » the renewable energy resources include prediction errors. A two-stage stochastic model predictive control is formulated and solved, where the optimal usage of the storage is simultaneously determined along with the optimal generation outputs and size of the storage. Wind forecast errors are taken into account in the optimization problem via probabilistic constraints for which an analytical form is derived. This allows for the stochastic optimization problem to be solved directly, without using sampling-based approaches, and sizing the storage to account not only for a wide range of potential scenarios, but also for a wide range of potential forecast errors. In the proposed formulation, we account for the fact that errors in the forecast affect how the device is operated later in the horizon and that a receding horizon scheme is used in operation to optimally use the available storage.« less

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

Forecast models for suicide: Time-series analysis with data from Italy.

PubMed

Preti, Antonio; Lentini, Gianluca

2016-01-01

The prediction of suicidal behavior is a complex task. To fine-tune targeted preventative interventions, predictive analytics (i.e. forecasting future risk of suicide) is more important than exploratory data analysis (pattern recognition, e.g. detection of seasonality in suicide time series). This study sets out to investigate the accuracy of forecasting models of suicide for men and women. A total of 101 499 male suicides and of 39 681 female suicides - occurred in Italy from 1969 to 2003 - were investigated. In order to apply the forecasting model and test its accuracy, the time series were split into a training set (1969 to 1996; 336 months) and a test set (1997 to 2003; 84 months). The main outcome was the accuracy of forecasting models on the monthly number of suicides. These measures of accuracy were used: mean absolute error; root mean squared error; mean absolute percentage error; mean absolute scaled error. In both male and female suicides a change in the trend pattern was observed, with an increase from 1969 onwards to reach a maximum around 1990 and decrease thereafter. The variances attributable to the seasonal and trend components were, respectively, 24% and 64% in male suicides, and 28% and 41% in female ones. Both annual and seasonal historical trends of monthly data contributed to forecast future trends of suicide with a margin of error around 10%. The finding is clearer in male than in female time series of suicide. The main conclusion of the study is that models taking seasonality into account seem to be able to derive information on deviation from the mean when this occurs as a zenith, but they fail to reproduce it when it occurs as a nadir. Preventative efforts should concentrate on the factors that influence the occurrence of increases above the main trend in both seasonal and cyclic patterns of suicides.

An Ensemble System Based on Hybrid EGARCH-ANN with Different Distributional Assumptions to Predict S&P 500 Intraday Volatility

NASA Astrophysics Data System (ADS)

Lahmiri, S.; Boukadoum, M.

2015-10-01

Accurate forecasting of stock market volatility is an important issue in portfolio risk management. In this paper, an ensemble system for stock market volatility is presented. It is composed of three different models that hybridize the exponential generalized autoregressive conditional heteroscedasticity (GARCH) process and the artificial neural network trained with the backpropagation algorithm (BPNN) to forecast stock market volatility under normal, t-Student, and generalized error distribution (GED) assumption separately. The goal is to design an ensemble system where each single hybrid model is capable to capture normality, excess skewness, or excess kurtosis in the data to achieve complementarity. The performance of each EGARCH-BPNN and the ensemble system is evaluated by the closeness of the volatility forecasts to realized volatility. Based on mean absolute error and mean of squared errors, the experimental results show that proposed ensemble model used to capture normality, skewness, and kurtosis in data is more accurate than the individual EGARCH-BPNN models in forecasting the S&P 500 intra-day volatility based on one and five-minute time horizons data.

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.

Evaluation of NMME temperature and precipitation bias and forecast skill for South Asia

NASA Astrophysics Data System (ADS)

Cash, Benjamin A.; Manganello, Julia V.; Kinter, James L.

2017-08-01

Systematic error and forecast skill for temperature and precipitation in two regions of Southern Asia are investigated using hindcasts initialized May 1 from the North American Multi-Model Ensemble. We focus on two contiguous but geographically and dynamically diverse regions: the Extended Indian Monsoon Rainfall (70-100E, 10-30 N) and the nearby mountainous area of Pakistan and Afghanistan (60-75E, 23-39 N). Forecast skill is assessed using the Sign test framework, a rigorous statistical method that can be applied to non-Gaussian variables such as precipitation and to different ensemble sizes without introducing bias. We find that models show significant systematic error in both precipitation and temperature for both regions. The multi-model ensemble mean (MMEM) consistently yields the lowest systematic error and the highest forecast skill for both regions and variables. However, we also find that the MMEM consistently provides a statistically significant increase in skill over climatology only in the first month of the forecast. While the MMEM tends to provide higher overall skill than climatology later in the forecast, the differences are not significant at the 95% level. We also find that MMEMs constructed with a relatively small number of ensemble members per model can equal or outperform MMEMs constructed with more members in skill. This suggests some ensemble members either provide no contribution to overall skill or even detract from it.

Ensemble Data Assimilation Without Ensembles: Methodology and Application to Ocean Data Assimilation

NASA Technical Reports Server (NTRS)

Keppenne, Christian L.; Rienecker, Michele M.; Kovach, Robin M.; Vernieres, Guillaume

2013-01-01

Two methods to estimate background error covariances for data assimilation are introduced. While both share properties with the ensemble Kalman filter (EnKF), they differ from it in that they do not require the integration of multiple model trajectories. Instead, all the necessary covariance information is obtained from a single model integration. The first method is referred-to as SAFE (Space Adaptive Forecast error Estimation) because it estimates error covariances from the spatial distribution of model variables within a single state vector. It can thus be thought of as sampling an ensemble in space. The second method, named FAST (Flow Adaptive error Statistics from a Time series), constructs an ensemble sampled from a moving window along a model trajectory. The underlying assumption in these methods is that forecast errors in data assimilation are primarily phase errors in space and/or time.

Constraining Stochastic Parametrisation Schemes Using High-Resolution Model Simulations

NASA Astrophysics Data System (ADS)

Christensen, H. M.; Dawson, A.; Palmer, T.

2017-12-01

Stochastic parametrisations are used in weather and climate models as a physically motivated way to represent model error due to unresolved processes. Designing new stochastic schemes has been the target of much innovative research over the last decade. While a focus has been on developing physically motivated approaches, many successful stochastic parametrisation schemes are very simple, such as the European Centre for Medium-Range Weather Forecasts (ECMWF) multiplicative scheme `Stochastically Perturbed Parametrisation Tendencies' (SPPT). The SPPT scheme improves the skill of probabilistic weather and seasonal forecasts, and so is widely used. However, little work has focused on assessing the physical basis of the SPPT scheme. We address this matter by using high-resolution model simulations to explicitly measure the `error' in the parametrised tendency that SPPT seeks to represent. The high resolution simulations are first coarse-grained to the desired forecast model resolution before they are used to produce initial conditions and forcing data needed to drive the ECMWF Single Column Model (SCM). By comparing SCM forecast tendencies with the evolution of the high resolution model, we can measure the `error' in the forecast tendencies. In this way, we provide justification for the multiplicative nature of SPPT, and for the temporal and spatial scales of the stochastic perturbations. However, we also identify issues with the SPPT scheme. It is therefore hoped these measurements will improve both holistic and process based approaches to stochastic parametrisation. Figure caption: Instantaneous snapshot of the optimal SPPT stochastic perturbation, derived by comparing high-resolution simulations with a low resolution forecast model.

Probabilistic Wind Power Ramp Forecasting Based on a Scenario Generation Method

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

Wang, Qin; Florita, Anthony R; Krishnan, Venkat K

Wind power ramps (WPRs) are particularly important in the management and dispatch of wind power and currently drawing the attention of balancing authorities. With the aim to reduce the impact of WPRs for power system operations, this paper develops a probabilistic ramp forecasting method based on a large number of simulated scenarios. An ensemble machine learning technique is first adopted to forecast the basic wind power forecasting scenario and calculate the historical forecasting errors. A continuous Gaussian mixture model (GMM) is used to fit the probability distribution function (PDF) of forecasting errors. The cumulative distribution function (CDF) is analytically deduced.more » The inverse transform method based on Monte Carlo sampling and the CDF is used to generate a massive number of forecasting error scenarios. An optimized swinging door algorithm is adopted to extract all the WPRs from the complete set of wind power forecasting scenarios. The probabilistic forecasting results of ramp duration and start-time are generated based on all scenarios. Numerical simulations on publicly available wind power data show that within a predefined tolerance level, the developed probabilistic wind power ramp forecasting method is able to predict WPRs with a high level of sharpness and accuracy.« less

Probabilistic Wind Power Ramp Forecasting Based on a Scenario Generation Method: Preprint

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

Wang, Qin; Florita, Anthony R; Krishnan, Venkat K

2017-08-31

Wind power ramps (WPRs) are particularly important in the management and dispatch of wind power, and they are currently drawing the attention of balancing authorities. With the aim to reduce the impact of WPRs for power system operations, this paper develops a probabilistic ramp forecasting method based on a large number of simulated scenarios. An ensemble machine learning technique is first adopted to forecast the basic wind power forecasting scenario and calculate the historical forecasting errors. A continuous Gaussian mixture model (GMM) is used to fit the probability distribution function (PDF) of forecasting errors. The cumulative distribution function (CDF) ismore » analytically deduced. The inverse transform method based on Monte Carlo sampling and the CDF is used to generate a massive number of forecasting error scenarios. An optimized swinging door algorithm is adopted to extract all the WPRs from the complete set of wind power forecasting scenarios. The probabilistic forecasting results of ramp duration and start time are generated based on all scenarios. Numerical simulations on publicly available wind power data show that within a predefined tolerance level, the developed probabilistic wind power ramp forecasting method is able to predict WPRs with a high level of sharpness and accuracy.« less

A Posteriori Correction of Forecast and Observation Error Variances

NASA Technical Reports Server (NTRS)

Rukhovets, Leonid

2005-01-01

Proposed method of total observation and forecast error variance correction is based on the assumption about normal distribution of "observed-minus-forecast" residuals (O-F), where O is an observed value and F is usually a short-term model forecast. This assumption can be accepted for several types of observations (except humidity) which are not grossly in error. Degree of nearness to normal distribution can be estimated by the symmetry or skewness (luck of symmetry) a(sub 3) = mu(sub 3)/sigma(sup 3) and kurtosis a(sub 4) = mu(sub 4)/sigma(sup 4) - 3 Here mu(sub i) = i-order moment, sigma is a standard deviation. It is well known that for normal distribution a(sub 3) = a(sub 4) = 0.

Forecasting impact injuries of unrestrained occupants in railway vehicle passenger compartments.

PubMed

Xie, Suchao; Zhou, Hui

2014-01-01

In order to predict the injury parameters of the occupants corresponding to different experimental parameters and to determine impact injury indices conveniently and efficiently, a model forecasting occupant impact injury was established in this work. The work was based on finite experimental observation values obtained by numerical simulation. First, the various factors influencing the impact injuries caused by the interaction between unrestrained occupants and the compartment's internal structures were collated and the most vulnerable regions of the occupant's body were analyzed. Then, the forecast model was set up based on a genetic algorithm-back propagation (GA-BP) hybrid algorithm, which unified the individual characteristics of the back propagation-artificial neural network (BP-ANN) model and the genetic algorithm (GA). The model was well suited to studies of occupant impact injuries and allowed multiple-parameter forecasts of the occupant impact injuries to be realized assuming values for various influencing factors. Finally, the forecast results for three types of secondary collision were analyzed using forecasting accuracy evaluation methods. All of the results showed the ideal accuracy of the forecast model. When an occupant faced a table, the relative errors between the predicted and experimental values of the respective injury parameters were kept within ± 6.0 percent and the average relative error (ARE) values did not exceed 3.0 percent. When an occupant faced a seat, the relative errors between the predicted and experimental values of the respective injury parameters were kept within ± 5.2 percent and the ARE values did not exceed 3.1 percent. When the occupant faced another occupant, the relative errors between the predicted and experimental values of the respective injury parameters were kept within ± 6.3 percent and the ARE values did not exceed 3.8 percent. The injury forecast model established in this article reduced repeat experiment times and improved the design efficiency of the internal compartment's structure parameters, and it provided a new way for assessing the safety performance of the interior structural parameters in existing, and newly designed, railway vehicle compartments.

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.

Development of extended WRF variational data assimilation system (WRFDA) for WRF non-hydrostatic mesoscale model

NASA Astrophysics Data System (ADS)

Pattanayak, Sujata; Mohanty, U. C.

2018-06-01

The paper intends to present the development of the extended weather research forecasting data assimilation (WRFDA) system in the framework of the non-hydrostatic mesoscale model core of weather research forecasting system (WRF-NMM), as an imperative aspect of numerical modeling studies. Though originally the WRFDA provides improved initial conditions for advanced research WRF, we have successfully developed a unified WRFDA utility that can be used by the WRF-NMM core, as well. After critical evaluation, it has been strategized to develop a code to merge WRFDA framework and WRF-NMM output. In this paper, we have provided a few selected implementations and initial results through single observation test, and background error statistics like eigenvalues, eigenvector and length scale among others, which showcase the successful development of extended WRFDA code for WRF-NMM model. Furthermore, the extended WRFDA system is applied for the forecast of three severe cyclonic storms: Nargis (27 April-3 May 2008), Aila (23-26 May 2009) and Jal (4-8 November 2010) formed over the Bay of Bengal. Model results are compared and contrasted within the analysis fields and later on with high-resolution model forecasts. The mean initial position error is reduced by 33% with WRFDA as compared to GFS analysis. The vector displacement errors in track forecast are reduced by 33, 31, 30 and 20% to 24, 48, 72 and 96 hr forecasts respectively, in data assimilation experiments as compared to control run. The model diagnostics indicates successful implementation of WRFDA within the WRF-NMM system.

Suboptimal schemes for atmospheric data assimilation based on the Kalman filter

NASA Technical Reports Server (NTRS)

Todling, Ricardo; Cohn, Stephen E.

1994-01-01

This work is directed toward approximating the evolution of forecast error covariances for data assimilation. The performance of different algorithms based on simplification of the standard Kalman filter (KF) is studied. These are suboptimal schemes (SOSs) when compared to the KF, which is optimal for linear problems with known statistics. The SOSs considered here are several versions of optimal interpolation (OI), a scheme for height error variance advection, and a simplified KF in which the full height error covariance is advected. To employ a methodology for exact comparison among these schemes, a linear environment is maintained, in which a beta-plane shallow-water model linearized about a constant zonal flow is chosen for the test-bed dynamics. The results show that constructing dynamically balanced forecast error covariances rather than using conventional geostrophically balanced ones is essential for successful performance of any SOS. A posteriori initialization of SOSs to compensate for model - data imbalance sometimes results in poor performance. Instead, properly constructed dynamically balanced forecast error covariances eliminate the need for initialization. When the SOSs studied here make use of dynamically balanced forecast error covariances, the difference among their performances progresses naturally from conventional OI to the KF. In fact, the results suggest that even modest enhancements of OI, such as including an approximate dynamical equation for height error variances while leaving height error correlation structure homogeneous, go a long way toward achieving the performance of the KF, provided that dynamically balanced cross-covariances are constructed and that model errors are accounted for properly. The results indicate that such enhancements are necessary if unconventional data are to have a positive impact.

Thirty Years of Improving the NCEP Global Forecast System

NASA Astrophysics Data System (ADS)

White, G. H.; Manikin, G.; Yang, F.

2014-12-01

Current eight day forecasts by the NCEP Global Forecast System are as accurate as five day forecasts 30 years ago. This revolution in weather forecasting reflects increases in computer power, improvements in the assimilation of observations, especially satellite data, improvements in model physics, improvements in observations and international cooperation and competition. One important component has been and is the diagnosis, evaluation and reduction of systematic errors. The effect of proposed improvements in the GFS on systematic errors is one component of the thorough testing of such improvements by the Global Climate and Weather Modeling Branch. Examples of reductions in systematic errors in zonal mean temperatures and winds and other fields will be presented. One challenge in evaluating systematic errors is uncertainty in what reality is. Model initial states can be regarded as the best overall depiction of the atmosphere, but can be misleading in areas of few observations or for fields not well observed such as humidity or precipitation over the oceans. Verification of model physics is particularly difficult. The Environmental Modeling Center emphasizes the evaluation of systematic biases against observations. Recently EMC has placed greater emphasis on synoptic evaluation and on precipitation, 2-meter temperatures and dew points and 10 meter winds. A weekly EMC map discussion reviews the performance of many models over the United States and has helped diagnose and alleviate significant systematic errors in the GFS, including a near surface summertime evening cold wet bias over the eastern US and a multi-week period when the GFS persistently developed bogus tropical storms off Central America. The GFS exhibits a wet bias for light rain and a dry bias for moderate to heavy rain over the continental United States. Significant changes to the GFS are scheduled to be implemented in the fall of 2014. These include higher resolution, improved physics and improvements to the assimilation. These changes significantly improve the tropospheric flow and reduce a tropical upper tropospheric warm bias. One important error remaining is the failure of the GFS to maintain deep convection over Indonesia and in the tropical west Pacific. This and other current systematic errors will be presented.

Statistical models for estimating daily streamflow in Michigan

USGS Publications Warehouse

Holtschlag, D.J.; Salehi, Habib

1992-01-01

Statistical models for estimating daily streamflow were analyzed for 25 pairs of streamflow-gaging stations in Michigan. Stations were paired by randomly choosing a station operated in 1989 at which 10 or more years of continuous flow data had been collected and at which flow is virtually unregulated; a nearby station was chosen where flow characteristics are similar. Streamflow data from the 25 randomly selected stations were used as the response variables; streamflow data at the nearby stations were used to generate a set of explanatory variables. Ordinary-least squares regression (OLSR) equations, autoregressive integrated moving-average (ARIMA) equations, and transfer function-noise (TFN) equations were developed to estimate the log transform of flow for the 25 randomly selected stations. The precision of each type of equation was evaluated on the basis of the standard deviation of the estimation errors. OLSR equations produce one set of estimation errors; ARIMA and TFN models each produce l sets of estimation errors corresponding to the forecast lead. The lead-l forecast is the estimate of flow l days ahead of the most recent streamflow used as a response variable in the estimation. In this analysis, the standard deviation of lead l ARIMA and TFN forecast errors were generally lower than the standard deviation of OLSR errors for l < 2 days and l < 9 days, respectively. Composite estimates were computed as a weighted average of forecasts based on TFN equations and backcasts (forecasts of the reverse-ordered series) based on ARIMA equations. The standard deviation of composite errors varied throughout the length of the estimation interval and generally was at maximum near the center of the interval. For comparison with OLSR errors, the mean standard deviation of composite errors were computed for intervals of length 1 to 40 days. The mean standard deviation of length-l composite errors were generally less than the standard deviation of the OLSR errors for l < 32 days. In addition, the composite estimates ensure a gradual transition between periods of estimated and measured flows. Model performance among stations of differing model error magnitudes were compared by computing ratios of the mean standard deviation of the length l composite errors to the standard deviation of OLSR errors. The mean error ratio for the set of 25 selected stations was less than 1 for intervals l < 32 days. Considering the frequency characteristics of the length of intervals of estimated record in Michigan, the effective mean error ratio for intervals < 30 days was 0.52. Thus, for intervals of estimation of 1 month or less, the error of the composite estimate is substantially lower than error of the OLSR estimate.

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.

Incorporating Yearly Derived Winter Wheat Maps Into Winter Wheat Yield Forecasting Model

NASA Technical Reports Server (NTRS)

Skakun, S.; Franch, B.; Roger, J.-C.; Vermote, E.; Becker-Reshef, I.; Justice, C.; Santamaría-Artigas, A.

2016-01-01

Wheat is one of the most important cereal crops in the world. Timely and accurate forecast of wheat yield and production at global scale is vital in implementing food security policy. Becker-Reshef et al. (2010) developed a generalized empirical model for forecasting winter wheat production using remote sensing data and official statistics. This model was implemented using static wheat maps. In this paper, we analyze the impact of incorporating yearly wheat masks into the forecasting model. We propose a new approach of producing in season winter wheat maps exploiting satellite data and official statistics on crop area only. Validation on independent data showed that the proposed approach reached 6% to 23% of omission error and 10% to 16% of commission error when mapping winter wheat 2-3 months before harvest. In general, we found a limited impact of using yearly winter wheat masks over a static mask for the study regions.

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

Morley, Steven

The PyForecastTools package provides Python routines for calculating metrics for model validation, forecast verification and model comparison. For continuous predictands the package provides functions for calculating bias (mean error, mean percentage error, median log accuracy, symmetric signed bias), and for calculating accuracy (mean squared error, mean absolute error, mean absolute scaled error, normalized RMSE, median symmetric accuracy). Convenience routines to calculate the component parts (e.g. forecast error, scaled error) of each metric are also provided. To compare models the package provides: generic skill score; percent better. Robust measures of scale including median absolute deviation, robust standard deviation, robust coefficient ofmore » variation and the Sn estimator are all provided by the package. Finally, the package implements Python classes for NxN contingency tables. In the case of a multi-class prediction, accuracy and skill metrics such as proportion correct and the Heidke and Peirce skill scores are provided as object methods. The special case of a 2x2 contingency table inherits from the NxN class and provides many additional metrics for binary classification: probability of detection, probability of false detection, false alarm ration, threat score, equitable threat score, bias. Confidence intervals for many of these quantities can be calculated using either the Wald method or Agresti-Coull intervals.« less

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

NASA Astrophysics Data System (ADS)

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

2011-02-01

The wind power forecasts 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 applied in this study is based on meteorological models that provide 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. For this purpose a training of a Neural Network (NN) to link directly the forecasted meteorological data and the power data has been performed. One wind farm has been examined located in a mountain area in the south of Italy (Sicily). First we compare the performances of a prediction based on meteorological data coming from a single model with those obtained by the combination of models (RAMS, ECMWF deterministic, LAMI). It is shown that the multi models approach reduces the day-ahead normalized RMSE forecast error (normalized by nominal power) of at least 1% compared to the singles models approach. Finally we have focused on the possibility of using the ensemble model system (EPS by 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 three days ahead period.

Targeted observations to improve tropical cyclone track forecasts in the Atlantic and eastern Pacific basins

NASA Astrophysics Data System (ADS)

Aberson, Sim David

In 1997, the National Hurricane Center and the Hurricane Research Division began conducting operational synoptic surveillance missions with the Gulfstream IV-SP jet aircraft to improve operational forecast models. During the first two years, twenty-four missions were conducted around tropical cyclones threatening the continental United States, Puerto Rico, and the Virgin Islands. Global Positioning System dropwindsondes were released from the aircraft at 150--200 km intervals along the flight track in the tropical cyclone environment to obtain wind, temperature, and humidity profiles from flight level (around 150 hPa) to the surface. The observations were processed and formatted aboard the aircraft and transmitted to the National Centers for Environmental Prediction (NCEP). There, they were ingested into the Global Data Assimilation System that subsequently provides initial and time-dependent boundary conditions for numerical models that forecast tropical cyclone track and intensity. Three dynamical models were employed in testing the targeting and sampling strategies. With the assimilation into the numerical guidance of all the observations gathered during the surveillance missions, only the 12-h Geophysical Fluid Dynamics Laboratory Hurricane Model forecast showed statistically significant improvement. Neither the forecasts from the Aviation run of the Global Spectral Model nor the shallow-water VICBAR model were improved with the assimilation of the dropwindsonde data. This mediocre result is found to be due mainly to the difficulty in operationally quantifying the storm-motion vector used to create accurate synthetic data to represent the tropical cyclone vortex in the models. A secondary limit on forecast improvements from the surveillance missions is the limited amount of data provided by the one surveillance aircraft in regular missions. The inability of some surveillance missions to surround the tropical cyclone with dropwindsonde observations is a possible third limit, though the results are inconclusive. Due to limited aircraft resources, optimal observing strategies for these missions must be developed. Since observations in areas of decaying error modes are unlikely to have large impact on subsequent forecasts, such strategies should be based on taking observations in those geographic locations corresponding to the most rapidly growing error modes in the numerical models and on known deficiencies in current data assimilation systems. Here, the most rapidly growing modes are represented by areas of large forecast spread in the NCEP bred-mode global ensemble forecasting system. The sampling strategy requires sampling the entire target region at approximately the same resolution as the North American rawinsonde network to limit the possibly spurious spread of information from dropwindsonde observations into data-sparse regions where errors are likely to grow. When only the subset of data in these fully-sampled target regions is assimilated into the numerical models, statistically significant reduction of the track forecast errors of up to 25% within the critical first two days of the forecast are seen. These model improvements are comparable with the cumulative business-as-usual track forecast model improvements expected over eighteen years.

Simulated forecast error and climate drift resulting from the omission of the upper stratosphere in numerical models

NASA Technical Reports Server (NTRS)

Boville, Byron A.; Baumhefner, David P.

1990-01-01

Using an NCAR community climate model, Version I, the forecast error growth and the climate drift resulting from the omission of the upper stratosphere are investigated. In the experiment, the control simulation is a seasonal integration of a medium horizontal general circulation model with 30 levels extending from the surface to the upper mesosphere, while the main experiment uses an identical model, except that only the bottom 15 levels (below 10 mb) are retained. It is shown that both random and systematic errors develop rapidly in the lower stratosphere with some local propagation into the troposphere in the 10-30-day time range. The random growth rate in the troposphere in the case of the altered upper boundary was found to be slightly faster than that for the initial-condition uncertainty alone. However, this is not likely to make a significant impact in operational forecast models, because the initial-condition uncertainty is very large.

Benefits of an ultra large and multiresolution ensemble for estimating available wind power

NASA Astrophysics Data System (ADS)

Berndt, Jonas; Hoppe, Charlotte; Elbern, Hendrik

2016-04-01

In this study we investigate the benefits of an ultra large ensemble with up to 1000 members including multiple nesting with a target horizontal resolution of 1 km. The ensemble shall be used as a basis to detect events of extreme errors in wind power forecasting. Forecast value is the wind vector at wind turbine hub height (~ 100 m) in the short range (1 to 24 hour). Current wind power forecast systems rest already on NWP ensemble models. However, only calibrated ensembles from meteorological institutions serve as input so far, with limited spatial resolution (˜10 - 80 km) and member number (˜ 50). Perturbations related to the specific merits of wind power production are yet missing. Thus, single extreme error events which are not detected by such ensemble power forecasts occur infrequently. The numerical forecast model used in this study is the Weather Research and Forecasting Model (WRF). Model uncertainties are represented by stochastic parametrization of sub-grid processes via stochastically perturbed parametrization tendencies and in conjunction via the complementary stochastic kinetic-energy backscatter scheme already provided by WRF. We perform continuous ensemble updates by comparing each ensemble member with available observations using a sequential importance resampling filter to improve the model accuracy while maintaining ensemble spread. Additionally, we use different ensemble systems from global models (ECMWF and GFS) as input and boundary conditions to capture different synoptic conditions. Critical weather situations which are connected to extreme error events are located and corresponding perturbation techniques are applied. The demanding computational effort is overcome by utilising the supercomputer JUQUEEN at the Forschungszentrum Juelich.

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.

Forecast of dengue incidence using temperature and rainfall.

PubMed

Hii, Yien Ling; Zhu, Huaiping; Ng, Nawi; Ng, Lee Ching; Rocklöv, Joacim

2012-01-01

An accurate early warning system to predict impending epidemics enhances the effectiveness of preventive measures against dengue fever. The aim of this study was to develop and validate a forecasting model that could predict dengue cases and provide timely early warning in Singapore. We developed a time series Poisson multivariate regression model using weekly mean temperature and cumulative rainfall over the period 2000-2010. Weather data were modeled using piecewise linear spline functions. We analyzed various lag times between dengue and weather variables to identify the optimal dengue forecasting period. Autoregression, seasonality and trend were considered in the model. We validated the model by forecasting dengue cases for week 1 of 2011 up to week 16 of 2012 using weather data alone. Model selection and validation were based on Akaike's Information Criterion, standardized Root Mean Square Error, and residuals diagnoses. A Receiver Operating Characteristics curve was used to analyze the sensitivity of the forecast of epidemics. The optimal period for dengue forecast was 16 weeks. Our model forecasted correctly with errors of 0.3 and 0.32 of the standard deviation of reported cases during the model training and validation periods, respectively. It was sensitive enough to distinguish between outbreak and non-outbreak to a 96% (CI = 93-98%) in 2004-2010 and 98% (CI = 95%-100%) in 2011. The model predicted the outbreak in 2011 accurately with less than 3% possibility of false alarm. We have developed a weather-based dengue forecasting model that allows warning 16 weeks in advance of dengue epidemics with high sensitivity and specificity. We demonstrate that models using temperature and rainfall could be simple, precise, and low cost tools for dengue forecasting which could be used to enhance decision making on the timing, scale of vector control operations, and utilization of limited resources.

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.

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.

How is the weather? Forecasting inpatient glycemic control

PubMed Central

Saulnier, George E; Castro, Janna C; Cook, Curtiss B; Thompson, Bithika M

2017-01-01

Aim: Apply methods of damped trend analysis to forecast inpatient glycemic control. Method: Observed and calculated point-of-care blood glucose data trends were determined over 62 weeks. Mean absolute percent error was used to calculate differences between observed and forecasted values. Comparisons were drawn between model results and linear regression forecasting. Results: The forecasted mean glucose trends observed during the first 24 and 48 weeks of projections compared favorably to the results provided by linear regression forecasting. However, in some scenarios, the damped trend method changed inferences compared with linear regression. In all scenarios, mean absolute percent error values remained below the 10% accepted by demand industries. Conclusion: Results indicate that forecasting methods historically applied within demand industries can project future inpatient glycemic control. Additional study is needed to determine if forecasting is useful in the analyses of other glucometric parameters and, if so, how to apply the techniques to quality improvement. PMID:29134125

Supplier Short Term Load Forecasting Using Support Vector Regression and Exogenous Input

NASA Astrophysics Data System (ADS)

Matijaš, Marin; Vukićcević, Milan; Krajcar, Slavko

2011-09-01

In power systems, task of load forecasting is important for keeping equilibrium between production and consumption. With liberalization of electricity markets, task of load forecasting changed because each market participant has to forecast their own load. Consumption of end-consumers is stochastic in nature. Due to competition, suppliers are not in a position to transfer their costs to end-consumers; therefore it is essential to keep forecasting error as low as possible. Numerous papers are investigating load forecasting from the perspective of the grid or production planning. We research forecasting models from the perspective of a supplier. In this paper, we investigate different combinations of exogenous input on the simulated supplier loads and show that using points of delivery as a feature for Support Vector Regression leads to lower forecasting error, while adding customer number in different datasets does the opposite.

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.

Application Of Multi-grid Method On China Seas' Temperature Forecast

NASA Astrophysics Data System (ADS)

Li, W.; Xie, Y.; He, Z.; Liu, K.; Han, G.; Ma, J.; Li, D.

2006-12-01

Correlation scales have been used in traditional scheme of 3-dimensional variational (3D-Var) data assimilation to estimate the background error covariance for the numerical forecast and reanalysis of atmosphere and ocean for decades. However there are still some drawbacks of this scheme. First, the correlation scales are difficult to be determined accurately. Second, the positive definition of the first-guess error covariance matrix cannot be guaranteed unless the correlation scales are sufficiently small. Xie et al. (2005) indicated that a traditional 3D-Var only corrects some certain wavelength errors and its accuracy depends on the accuracy of the first-guess covariance. And in general, short wavelength error can not be well corrected until long one is corrected and then inaccurate first-guess covariance may mistakenly take long wave error as short wave ones and result in erroneous analysis. For the purpose of quickly minimizing the errors of long and short waves successively, a new 3D-Var data assimilation scheme, called multi-grid data assimilation scheme, is proposed in this paper. By assimilating the shipboard SST and temperature profiles data into a numerical model of China Seas, we applied this scheme in two-month data assimilation and forecast experiment which ended in a favorable result. Comparing with the traditional scheme of 3D-Var, the new scheme has higher forecast accuracy and a lower forecast Root-Mean-Square (RMS) error. Furthermore, this scheme was applied to assimilate the SST of shipboard, AVHRR Pathfinder Version 5.0 SST and temperature profiles at the same time, and a ten-month forecast experiment on sea temperature of China Seas was carried out, in which a successful forecast result was obtained. Particularly, the new scheme is demonstrated a great numerical efficiency in these analyses.

Dynamical Downscaling of Seasonal Climate Prediction over Nordeste Brazil with ECHAM3 and NCEP's Regional Spectral Models at IRI.

NASA Astrophysics Data System (ADS)

Nobre, Paulo; Moura, Antonio D.; Sun, Liqiang

2001-12-01

This study presents an evaluation of a seasonal climate forecast done with the International Research Institute for Climate Prediction (IRI) dynamical forecast system (regional model nested into a general circulation model) over northern South America for January-April 1999, encompassing the rainy season over Brazil's Nordeste. The one-way nesting is one in two tiers: first the NCEP's Regional Spectral Model (RSM) runs with an 80-km grid mesh forced by the ECHAM3 atmospheric general circulation model (AGCM) outputs; then the RSM runs with a finer grid mesh (20 km) forced by the forecasts generated by the RSM-80. An ensemble of three realizations is done. Lower boundary conditions over the oceans for both ECHAM and RSM model runs are sea surface temperature forecasts over the tropical oceans. Soil moisture is initialized by ECHAM's inputs. The rainfall forecasts generated by the regional model are compared with those of the AGCM and observations. It is shown that the regional model at 80-km resolution improves upon the AGCM rainfall forecast, reducing both seasonal bias and root-mean-square error. On the other hand, the RSM-20 forecasts presented larger errors, with spatial patterns that resemble those of local topography. The better forecast of the position and width of the intertropical convergence zone (ITCZ) over the tropical Atlantic by the RSM-80 model is one of the principal reasons for better-forecast scores of the RSM-80 relative to the AGCM. The regional model improved the spatial as well as the temporal details of rainfall distribution, and also presenting the minimum spread among the ensemble members. The statistics of synoptic-scale weather variability on seasonal timescales were best forecast with the regional 80-km model over the Nordeste. The possibility of forecasting the frequency distribution of dry and wet spells within the rainy season is encouraging.

Why Don't We Learn to Accurately Forecast Feelings? How Misremembering Our Predictions Blinds Us to Past Forecasting Errors

ERIC Educational Resources Information Center

Meyvis, Tom; Ratner, Rebecca K.; Levav, Jonathan

2010-01-01

Why do affective forecasting errors persist in the face of repeated disconfirming evidence? Five studies demonstrate that people misremember their forecasts as consistent with their experience and thus fail to perceive the extent of their forecasting error. As a result, people do not learn from past forecasting errors and fail to adjust subsequent…

Two Topics in Seasonal Streamflow Forecasting: Soil Moisture Initialization Error and Precipitation Downscaling

NASA Technical Reports Server (NTRS)

Koster, Randal; Walker, Greg; Mahanama, Sarith; Reichle, Rolf

2012-01-01

Continental-scale offline simulations with a land surface model are used to address two important issues in the forecasting of large-scale seasonal streamflow: (i) the extent to which errors in soil moisture initialization degrade streamflow forecasts, and (ii) the extent to which the downscaling of seasonal precipitation forecasts, if it could be done accurately, would improve streamflow forecasts. The reduction in streamflow forecast skill (with forecasted streamflow measured against observations) associated with adding noise to a soil moisture field is found to be, to first order, proportional to the average reduction in the accuracy of the soil moisture field itself. This result has implications for streamflow forecast improvement under satellite-based soil moisture measurement programs. In the second and more idealized ("perfect model") analysis, precipitation downscaling is found to have an impact on large-scale streamflow forecasts only if two conditions are met: (i) evaporation variance is significant relative to the precipitation variance, and (ii) the subgrid spatial variance of precipitation is adequately large. In the large-scale continental region studied (the conterminous United States), these two conditions are met in only a somewhat limited area.

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.

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.

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.

Variational Continuous Assimilation of TMI and SSM/I Rain Rates: Impact on GEOS-3 Hurricane Analyses and Forecasts

NASA Technical Reports Server (NTRS)

Hou, Arthur Y.; Zhang, Sara Q.; Reale, Oreste

2003-01-01

We describe a variational continuous assimilation (VCA) algorithm for assimilating tropical rainfall data using moisture and temperature tendency corrections as the control variable to offset model deficiencies. For rainfall assimilation, model errors are of special concern since model-predicted precipitation is based on parameterized moist physics, which can have substantial systematic errors. This study examines whether a VCA scheme using the forecast model as a weak constraint offers an effective pathway to precipitation assimilation. The particular scheme we exarnine employs a '1+1' dimension precipitation observation operator based on a 6-h integration of a column model of moist physics from the Goddard Earth Observing System (GEOS) global data assimilation system DAS). In earlier studies, we tested a simplified version of this scheme and obtained improved monthly-mean analyses and better short-range forecast skills. This paper describes the full implementation ofthe 1+1D VCA scheme using background and observation error statistics, and examines how it may improve GEOS analyses and forecasts of prominent tropical weather systems such as hurricanes. Parallel assimilation experiments with and without rainfall data for Hurricanes Bonnie and Floyd show that assimilating 6-h TMI and SSM/I surfice rain rates leads to more realistic storm features in the analysis, which, in turn, provide better initial conditions for 5-day storm track prediction and precipitation forecast. These results provide evidence that addressing model deficiencies in moisture tendency may be crucial to making effective use of precipitation information in data assimilation.

Integration of Behind-the-Meter PV Fleet Forecasts into Utility Grid System Operations

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

Hoff, Thomas Hoff; Kankiewicz, Adam

Four major research objectives were completed over the course of this study. Three of the objectives were to evaluate three, new, state-of-the-art solar irradiance forecasting models. The fourth objective was to improve the California Independent System Operator’s (ISO) load forecasts by integrating behind-the-meter (BTM) PV forecasts. The three, new, state-of-the-art solar irradiance forecasting models included: the infrared (IR) satellite-based cloud motion vector (CMV) model; the WRF-SolarCA model and variants; and the Optimized Deep Machine Learning (ODML)-training model. The first two forecasting models targeted known weaknesses in current operational solar forecasts. They were benchmarked against existing operational numerical weather prediction (NWP)more » forecasts, visible satellite CMV forecasts, and measured PV plant power production. IR CMV, WRF-SolarCA, and ODML-training forecasting models all improved the forecast to a significant degree. Improvements varied depending on time of day, cloudiness index, and geographic location. The fourth objective was to demonstrate that the California ISO’s load forecasts could be improved by integrating BTM PV forecasts. This objective represented the project’s most exciting and applicable gains. Operational BTM forecasts consisting of 200,000+ individual rooftop PV forecasts were delivered into the California ISO’s real-time automated load forecasting (ALFS) environment. They were then evaluated side-by-side with operational load forecasts with no BTM-treatment. Overall, ALFS-BTM day-ahead (DA) forecasts performed better than baseline ALFS forecasts when compared to actual load data. Specifically, ALFS-BTM DA forecasts were observed to have the largest reduction of error during the afternoon on cloudy days. Shorter term 30 minute-ahead ALFS-BTM forecasts were shown to have less error under all sky conditions, especially during the morning time periods when traditional load forecasts often experience their largest uncertainties. This work culminated in a GO decision being made by the California ISO to include zonal BTM forecasts into its operational load forecasting system. The California ISO’s Manager of Short Term Forecasting, Jim Blatchford, summarized the research performed in this project with the following quote: “The behind-the-meter (BTM) California ISO region forecasting research performed by Clean Power Research and sponsored by the Department of Energy’s SUNRISE program was an opportunity to verify value and demonstrate improved load forecast capability. In 2016, the California ISO will be incorporating the BTM forecast into the Hour Ahead and Day Ahead load models to look for improvements in the overall load forecast accuracy as BTM PV capacity continues to grow.”« less

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.

Real-time prediction of atmospheric Lagrangian coherent structures based on forecast data: An application and error analysis

NASA Astrophysics Data System (ADS)

BozorgMagham, Amir E.; Ross, Shane D.; Schmale, David G.

2013-09-01

The language of Lagrangian coherent structures (LCSs) provides a new means for studying transport and mixing of passive particles advected by an atmospheric flow field. Recent observations suggest that LCSs govern the large-scale atmospheric motion of airborne microorganisms, paving the way for more efficient models and management strategies for the spread of infectious diseases affecting plants, domestic animals, and humans. In addition, having reliable predictions of the timing of hyperbolic LCSs may contribute to improved aerobiological sampling of microorganisms with unmanned aerial vehicles and LCS-based early warning systems. Chaotic atmospheric dynamics lead to unavoidable forecasting errors in the wind velocity field, which compounds errors in LCS forecasting. In this study, we reveal the cumulative effects of errors of (short-term) wind field forecasts on the finite-time Lyapunov exponent (FTLE) fields and the associated LCSs when realistic forecast plans impose certain limits on the forecasting parameters. Objectives of this paper are to (a) quantify the accuracy of prediction of FTLE-LCS features and (b) determine the sensitivity of such predictions to forecasting parameters. Results indicate that forecasts of attracting LCSs exhibit less divergence from the archive-based LCSs than the repelling features. This result is important since attracting LCSs are the backbone of long-lived features in moving fluids. We also show under what circumstances one can trust the forecast results if one merely wants to know if an LCS passed over a region and does not need to precisely know the passage time.

A study on characteristics of retrospective optimal interpolation with WRF testbed

NASA Astrophysics Data System (ADS)

Kim, S.; Noh, N.; Lim, G.

2012-12-01

This study presents the application of retrospective optimal interpolation (ROI) with Weather Research and Forecasting model (WRF). Song et al. (2009) suggest ROI method which is an optimal interpolation (OI) that gradually assimilates observations over the analysis window for variance-minimum estimate of an atmospheric state at the initial time of the analysis window. Song and Lim (2011) improve the method by incorporating eigen-decomposition and covariance inflation. ROI method assimilates the data at post analysis time using perturbation method (Errico and Raeder, 1999) without adjoint model. In this study, ROI method is applied to WRF model to validate the algorithm and to investigate the capability. The computational costs for ROI can be reduced due to the eigen-decomposition of background error covariance. Using the background error covariance in eigen-space, 1-profile assimilation experiment is performed. The difference between forecast errors with assimilation and without assimilation is obviously increased as time passed, which means the improvement of forecast error by assimilation. The characteristics and strength/weakness of ROI method are investigated by conducting the experiments with other data assimilation method.

Ocean Data Impacts in Global HYCOM

DTIC Science & Technology

2014-08-01

The purpose of assimilation is to reduce the model initial condition error. Improved initial con- ditions should lead to an improved forecast...the determination of locations where forecast errors are sensitive to the initial conditions are essential for improving the data assimilation system...longwave radiation, total (large scale plus convective) precipitation, ground/sea temperature, zonal and me- ridional wind velocities at 10m, mean sea

Stochastic Model of Seasonal Runoff Forecasts

NASA Astrophysics Data System (ADS)

Krzysztofowicz, Roman; Watada, Leslie M.

1986-03-01

Each year the National Weather Service and the Soil Conservation Service issue a monthly sequence of five (or six) categorical forecasts of the seasonal snowmelt runoff volume. To describe uncertainties in these forecasts for the purposes of optimal decision making, a stochastic model is formulated. It is a discrete-time, finite, continuous-space, nonstationary Markov process. Posterior densities of the actual runoff conditional upon a forecast, and transition densities of forecasts are obtained from a Bayesian information processor. Parametric densities are derived for the process with a normal prior density of the runoff and a linear model of the forecast error. The structure of the model and the estimation procedure are motivated by analyses of forecast records from five stations in the Snake River basin, from the period 1971-1983. The advantages of supplementing the current forecasting scheme with a Bayesian analysis are discussed.

Prediction skill of tropical synoptic scale transients from ECMWF and NCEP ensemble prediction systems

DOE PAGES

Taraphdar, S.; Mukhopadhyay, P.; Leung, L. Ruby; ...

2016-12-05

The prediction skill of tropical synoptic scale transients (SSTR) such as monsoon low and depression during the boreal summer of 2007–2009 are assessed using high resolution ECMWF and NCEP TIGGE forecasts data. By analyzing 246 forecasts for lead times up to 10 days, it is found that the models have good skills in forecasting the planetary scale means but the skills of SSTR remain poor, with the latter showing no skill beyond 2 days for the global tropics and Indian region. Consistent forecast skills among precipitation, velocity potential, and vorticity provide evidence that convection is the primary process responsible formore » precipitation. The poor skills of SSTR can be attributed to the larger random error in the models as they fail to predict the locations and timings of SSTR. Strong correlation between the random error and synoptic precipitation suggests that the former starts to develop from regions of convection. As the NCEP model has larger biases of synoptic scale precipitation, it has a tendency to generate more random error that ultimately reduces the prediction skill of synoptic systems in that model. Finally, the larger biases in NCEP may be attributed to the model moist physics and/or coarser horizontal resolution compared to ECMWF.« less

Prediction skill of tropical synoptic scale transients from ECMWF and NCEP ensemble prediction systems

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

Taraphdar, S.; Mukhopadhyay, P.; Leung, L. Ruby

The prediction skill of tropical synoptic scale transients (SSTR) such as monsoon low and depression during the boreal summer of 2007–2009 are assessed using high resolution ECMWF and NCEP TIGGE forecasts data. By analyzing 246 forecasts for lead times up to 10 days, it is found that the models have good skills in forecasting the planetary scale means but the skills of SSTR remain poor, with the latter showing no skill beyond 2 days for the global tropics and Indian region. Consistent forecast skills among precipitation, velocity potential, and vorticity provide evidence that convection is the primary process responsible formore » precipitation. The poor skills of SSTR can be attributed to the larger random error in the models as they fail to predict the locations and timings of SSTR. Strong correlation between the random error and synoptic precipitation suggests that the former starts to develop from regions of convection. As the NCEP model has larger biases of synoptic scale precipitation, it has a tendency to generate more random error that ultimately reduces the prediction skill of synoptic systems in that model. Finally, the larger biases in NCEP may be attributed to the model moist physics and/or coarser horizontal resolution compared to ECMWF.« less

Research on wind field algorithm of wind lidar based on BP neural network and grey prediction

NASA Astrophysics Data System (ADS)

Chen, Yong; Chen, Chun-Li; Luo, Xiong; Zhang, Yan; Yang, Ze-hou; Zhou, Jie; Shi, Xiao-ding; Wang, Lei

2018-01-01

This paper uses the BP neural network and grey algorithm to forecast and study radar wind field. In order to reduce the residual error in the wind field prediction which uses BP neural network and grey algorithm, calculating the minimum value of residual error function, adopting the residuals of the gray algorithm trained by BP neural network, using the trained network model to forecast the residual sequence, using the predicted residual error sequence to modify the forecast sequence of the grey algorithm. The test data show that using the grey algorithm modified by BP neural network can effectively reduce the residual value and improve the prediction precision.

Daily Reservoir Inflow Forecasting using Deep Learning with Downscaled Multi-General Circulation Models (GCMs) Platform

NASA Astrophysics Data System (ADS)

Li, D.; Fang, N. Z.

2017-12-01

Dallas-Fort Worth Metroplex (DFW) has a population of over 7 million depending on many water supply reservoirs. The reservoir inflow plays a vital role in water supply decision making process and long-term strategic planning for the region. This paper demonstrates a method of utilizing deep learning algorithms and multi-general circulation model (GCM) platform to forecast reservoir inflow for three reservoirs within the DFW: Eagle Mountain Lake, Lake Benbrook and Lake Arlington. Ensemble empirical mode decomposition was firstly employed to extract the features, which were then represented by the deep belief networks (DBNs). The first 75 years of the historical data (1940 -2015) were used to train the model, while the last 2 years of the data (2016-2017) were used for the model validation. The weights of each DBN gained from the training process were then applied to establish a neural network (NN) that was able to forecast reservoir inflow. Feature predictors used for the forecasting model were generated from weather forecast results of the downscaled multi-GCM platform for the North Texas region. By comparing root mean square error (RMSE) and mean bias error (MBE) with the observed data, the authors found that the deep learning with downscaled multi-GCM platform is an effective approach in the reservoir inflow forecasting.

Error models for official mortality forecasts.

PubMed

Alho, J M; Spencer, B D

1990-09-01

"The Office of the Actuary, U.S. Social Security Administration, produces alternative forecasts of mortality to reflect uncertainty about the future.... In this article we identify the components and assumptions of the official forecasts and approximate them by stochastic parametric models. We estimate parameters of the models from past data, derive statistical intervals for the forecasts, and compare them with the official high-low intervals. We use the models to evaluate the forecasts rather than to develop different predictions of the future. Analysis of data from 1972 to 1985 shows that the official intervals for mortality forecasts for males or females aged 45-70 have approximately a 95% chance of including the true mortality rate in any year. For other ages the chances are much less than 95%." excerpt

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.

Predictability of CFSv2 in the tropical Indo-Pacific region, at daily and subseasonal time scales

NASA Astrophysics Data System (ADS)

Krishnamurthy, V.

2018-06-01

The predictability of a coupled climate model is evaluated at daily and intraseasonal time scales in the tropical Indo-Pacific region during boreal summer and winter. This study has assessed the daily retrospective forecasts of the Climate Forecast System version 2 from the National Centers of Environmental Prediction for the period 1982-2010. The growth of errors in the forecasts of daily precipitation, monsoon intraseasonal oscillation (MISO) and the Madden-Julian oscillation (MJO) is studied. The seasonal cycle of the daily climatology of precipitation is reasonably well predicted except for the underestimation during the peak of summer. The anomalies follow the typical pattern of error growth in nonlinear systems and show no difference between summer and winter. The initial errors in all the cases are found to be in the nonlinear phase of the error growth. The doubling time of small errors is estimated by applying Lorenz error formula. For summer and winter, the doubling time of the forecast errors is in the range of 4-7 and 5-14 days while the doubling time of the predictability errors is 6-8 and 8-14 days, respectively. The doubling time in MISO during the summer and MJO during the winter is in the range of 12-14 days, indicating higher predictability and providing optimism for long-range prediction. There is no significant difference in the growth of forecasts errors originating from different phases of MISO and MJO, although the prediction of the active phase seems to be slightly better.

Spatial Ensemble Postprocessing of Precipitation Forecasts Using High Resolution Analyses

NASA Astrophysics Data System (ADS)

Lang, Moritz N.; Schicker, Irene; Kann, Alexander; Wang, Yong

2017-04-01

Ensemble prediction systems are designed to account for errors or uncertainties in the initial and boundary conditions, imperfect parameterizations, etc. However, due to sampling errors and underestimation of the model errors, these ensemble forecasts tend to be underdispersive, and to lack both reliability and sharpness. To overcome such limitations, statistical postprocessing methods are commonly applied to these forecasts. In this study, a full-distributional spatial post-processing method is applied to short-range precipitation forecasts over Austria using Standardized Anomaly Model Output Statistics (SAMOS). Following Stauffer et al. (2016), observation and forecast fields are transformed into standardized anomalies by subtracting a site-specific climatological mean and dividing by the climatological standard deviation. Due to the need of fitting only a single regression model for the whole domain, the SAMOS framework provides a computationally inexpensive method to create operationally calibrated probabilistic forecasts for any arbitrary location or for all grid points in the domain simultaneously. Taking advantage of the INCA system (Integrated Nowcasting through Comprehensive Analysis), high resolution analyses are used for the computation of the observed climatology and for model training. The INCA system operationally combines station measurements and remote sensing data into real-time objective analysis fields at 1 km-horizontal resolution and 1 h-temporal resolution. The precipitation forecast used in this study is obtained from a limited area model ensemble prediction system also operated by ZAMG. The so called ALADIN-LAEF provides, by applying a multi-physics approach, a 17-member forecast at a horizontal resolution of 10.9 km and a temporal resolution of 1 hour. The performed SAMOS approach statistically combines the in-house developed high resolution analysis and ensemble prediction system. The station-based validation of 6 hour precipitation sums shows a mean improvement of more than 40% in CRPS when compared to bilinearly interpolated uncalibrated ensemble forecasts. The validation on randomly selected grid points, representing the true height distribution over Austria, still indicates a mean improvement of 35%. The applied statistical model is currently set up for 6-hourly and daily accumulation periods, but will be extended to a temporal resolution of 1-3 hours within a new probabilistic nowcasting system operated by ZAMG.

Physiologically grounded metrics of model skill: a case study estimating heat stress in intertidal populations

PubMed Central

Kish, Nicole E.; Helmuth, Brian; Wethey, David S.

2016-01-01

Models of ecological responses to climate change fundamentally assume that predictor variables, which are often measured at large scales, are to some degree diagnostic of the smaller-scale biological processes that ultimately drive patterns of abundance and distribution. Given that organisms respond physiologically to stressors, such as temperature, in highly non-linear ways, small modelling errors in predictor variables can potentially result in failures to predict mortality or severe stress, especially if an organism exists near its physiological limits. As a result, a central challenge facing ecologists, particularly those attempting to forecast future responses to environmental change, is how to develop metrics of forecast model skill (the ability of a model to predict defined events) that are biologically meaningful and reflective of underlying processes. We quantified the skill of four simple models of body temperature (a primary determinant of physiological stress) of an intertidal mussel, Mytilus californianus, using common metrics of model performance, such as root mean square error, as well as forecast verification skill scores developed by the meteorological community. We used a physiologically grounded framework to assess each model's ability to predict optimal, sub-optimal, sub-lethal and lethal physiological responses. Models diverged in their ability to predict different levels of physiological stress when evaluated using skill scores, even though common metrics, such as root mean square error, indicated similar accuracy overall. Results from this study emphasize the importance of grounding assessments of model skill in the context of an organism's physiology and, especially, of considering the implications of false-positive and false-negative errors when forecasting the ecological effects of environmental change. PMID:27729979

Short-term Wind Forecasting at Wind Farms using WRF-LES and Actuator Disk Model

NASA Astrophysics Data System (ADS)

Kirkil, Gokhan

2017-04-01

Short-term wind forecasts are obtained for a wind farm on a mountainous terrain using WRF-LES. Multi-scale simulations are also performed using different PBL parameterizations. Turbines are parameterized using Actuator Disc Model. LES models improved the forecasts. Statistical error analysis is performed and ramp events are analyzed. Complex topography of the study area affects model performance, especially the accuracy of wind forecasts were poor for cross valley-mountain flows. By means of LES, we gain new knowledge about the sources of spatial and temporal variability of wind fluctuations such as the configuration of wind turbines.

Study on the medical meteorological forecast of the number of hypertension inpatient based on SVR

NASA Astrophysics Data System (ADS)

Zhai, Guangyu; Chai, Guorong; Zhang, Haifeng

2017-06-01

The purpose of this study is to build a hypertension prediction model by discussing the meteorological factors for hypertension incidence. The research method is selecting the standard data of relative humidity, air temperature, visibility, wind speed and air pressure of Lanzhou from 2010 to 2012(calculating the maximum, minimum and average value with 5 days as a unit ) as the input variables of Support Vector Regression(SVR) and the standard data of hypertension incidence of the same period as the output dependent variables to obtain the optimal prediction parameters by cross validation algorithm, then by SVR algorithm learning and training, a SVR forecast model for hypertension incidence is built. The result shows that the hypertension prediction model is composed of 15 input independent variables, the training accuracy is 0.005, the final error is 0.0026389. The forecast accuracy based on SVR model is 97.1429%, which is higher than statistical forecast equation and neural network prediction method. It is concluded that SVR model provides a new method for hypertension prediction with its simple calculation, small error as well as higher historical sample fitting and Independent sample forecast capability.

Multiple regression and Artificial Neural Network for long-term rainfall forecasting using large scale climate modes

NASA Astrophysics Data System (ADS)

Mekanik, F.; Imteaz, M. A.; Gato-Trinidad, S.; Elmahdi, A.

2013-10-01

In this study, the application of Artificial Neural Networks (ANN) and Multiple regression analysis (MR) to forecast long-term seasonal spring rainfall in Victoria, Australia was investigated using lagged El Nino Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) as potential predictors. The use of dual (combined lagged ENSO-IOD) input sets for calibrating and validating ANN and MR Models is proposed to investigate the simultaneous effect of past values of these two major climate modes on long-term spring rainfall prediction. The MR models that did not violate the limits of statistical significance and multicollinearity were selected for future spring rainfall forecast. The ANN was developed in the form of multilayer perceptron using Levenberg-Marquardt algorithm. Both MR and ANN modelling were assessed statistically using mean square error (MSE), mean absolute error (MAE), Pearson correlation (r) and Willmott index of agreement (d). The developed MR and ANN models were tested on out-of-sample test sets; the MR models showed very poor generalisation ability for east Victoria with correlation coefficients of -0.99 to -0.90 compared to ANN with correlation coefficients of 0.42-0.93; ANN models also showed better generalisation ability for central and west Victoria with correlation coefficients of 0.68-0.85 and 0.58-0.97 respectively. The ability of multiple regression models to forecast out-of-sample sets is compatible with ANN for Daylesford in central Victoria and Kaniva in west Victoria (r = 0.92 and 0.67 respectively). The errors of the testing sets for ANN models are generally lower compared to multiple regression models. The statistical analysis suggest the potential of ANN over MR models for rainfall forecasting using large scale climate modes.

Prediction skill of rainstorm events over India in the TIGGE weather prediction models

NASA Astrophysics Data System (ADS)

Karuna Sagar, S.; Rajeevan, M.; Vijaya Bhaskara Rao, S.; Mitra, A. K.

2017-12-01

Extreme rainfall events pose a serious threat of leading to severe floods in many countries worldwide. Therefore, advance prediction of its occurrence and spatial distribution is very essential. In this paper, an analysis has been made to assess the skill of numerical weather prediction models in predicting rainstorms over India. Using gridded daily rainfall data set and objective criteria, 15 rainstorms were identified during the monsoon season (June to September). The analysis was made using three TIGGE (THe Observing System Research and Predictability Experiment (THORPEX) Interactive Grand Global Ensemble) models. The models considered are the European Centre for Medium-Range Weather Forecasts (ECMWF), National Centre for Environmental Prediction (NCEP) and the UK Met Office (UKMO). Verification of the TIGGE models for 43 observed rainstorm days from 15 rainstorm events has been made for the period 2007-2015. The comparison reveals that rainstorm events are predictable up to 5 days in advance, however with a bias in spatial distribution and intensity. The statistical parameters like mean error (ME) or Bias, root mean square error (RMSE) and correlation coefficient (CC) have been computed over the rainstorm region using the multi-model ensemble (MME) mean. The study reveals that the spread is large in ECMWF and UKMO followed by the NCEP model. Though the ensemble spread is quite small in NCEP, the ensemble member averages are not well predicted. The rank histograms suggest that the forecasts are under prediction. The modified Contiguous Rain Area (CRA) technique was used to verify the spatial as well as the quantitative skill of the TIGGE models. Overall, the contribution from the displacement and pattern errors to the total RMSE is found to be more in magnitude. The volume error increases from 24 hr forecast to 48 hr forecast in all the three models.

Diagnosis of North American Multi-Model Ensemble (NMME) skill for predicting floods and droughts over the continental USA

NASA Astrophysics Data System (ADS)

Slater, L. J.; Villarini, G.; Bradley, A.

2015-12-01

Model predictions of precipitation and temperature are crucial to mitigate the impacts of major flood and drought events through informed planning and response. However, the potential value and applicability of these predictions is inescapably linked to their forecast quality. The North-American Multi-Model Ensemble (NMME) is a multi-agency supported forecasting system for intraseasonal to interannual (ISI) climate predictions. Retrospective forecasts and real-time information are provided by each agency free of charge to facilitate collaborative research efforts for predicting future climate conditions as well as extreme weather events such as floods and droughts. Using the PRISM climate mapping system as the reference data, we examine the skill of five General Circulation Models (GCMs) from the NMME project to forecast monthly and seasonal precipitation and temperature over seven sub-regions of the continental United States. For each model, we quantify the seasonal accuracy of the forecast relative to observed precipitation using the mean square error skill score. This score is decomposed to assess the accuracy of the forecast in the absence of biases (potential skill), and in the presence of conditional (slope reliability) and unconditional (standardized mean error) biases. The quantification of these biases allows us to diagnose each model's skill over a full range temporal and spatial scales. Finally, we test each model's forecasting skill by evaluating its ability to predict extended periods of extreme temperature and precipitation that were conducive to 'billion-dollar' historical flood and drought events in different regions of the continental USA. The forecasting skill of the individual climate models is summarized and presented along with a discussion of different multi-model averaging techniques for predicting such events.

The propagation of wind errors through ocean wave hindcasts

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

Holthuijsen, L.H.; Booij, N.; Bertotti, L.

1996-08-01

To estimate uncertainties in wave forecast and hindcasts, computations have been carried out for a location in the Mediterranean Sea using three different analyses of one historic wind field. These computations involve a systematic sensitivity analysis and estimated wind field errors. This technique enables a wave modeler to estimate such uncertainties in other forecasts and hindcasts if only one wind analysis is available.

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.

A study on the characteristics of retrospective optimal interpolation using an Observing System Simulation Experiment

NASA Astrophysics Data System (ADS)

Kim, Shin-Woo; Noh, Nam-Kyu; Lim, Gyu-Ho

2013-04-01

This study presents the introduction of retrospective optimal interpolation (ROI) and its application with Weather Research and Forecasting model (WRF). Song et al. (2009) suggested ROI method which is an optimal interpolation (OI) that gradually assimilates observations over the analysis window for variance-minimum estimate of an atmospheric state at the initial time of the analysis window. The assimilation window of ROI algorithm is gradually increased, similar with that of the quasi-static variational assimilation (QSVA; Pires et al., 1996). Unlike QSVA method, however, ROI method assimilates the data at post analysis time using perturbation method (Verlaan and Heemink, 1997) without adjoint model. Song and Lim (2011) improved this method by incorporating eigen-decomposition and covariance inflation. The computational costs for ROI can be reduced due to the eigen-decomposition of background error covariance which can concentrate ROI analyses on the error variances of governing eigenmodes by transforming the control variables into eigenspace. A total energy norm is used for the normalization of each control variables. In this study, ROI method is applied to WRF model with Observing System Simulation Experiment (OSSE) to validate the algorithm and to investigate the capability. Horizontal wind, pressure, potential temperature, and water vapor mixing ratio are used for control variables and observations. Firstly, 1-profile assimilation experiment is performed. Subsequently, OSSE's are performed using the virtual observing system which consists of synop, ship, and sonde data. The difference between forecast errors with assimilation and without assimilation is obviously increased as time passed, which means the improvement of forecast error with the assimilation by ROI. The characteristics and strength/weakness of ROI method are also investigated by conducting the experiments with 3D-Var (3-dimensional variational) method and 4D-Var (4-dimensional variational) method. In the initial time, ROI produces a larger forecast error than that of 4D-Var. However, the difference between the two experimental results is decreased gradually with time, and the ROI shows apparently better result (i.e., smaller forecast error) than that of 4D-Var after 9-hour forecast.

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.

Tropical forecasting - Predictability perspective

NASA Technical Reports Server (NTRS)

Shukla, J.

1989-01-01

Results are presented of classical predictability studies and forecast experiments with observed initial conditions to show the nature of initial error growth and final error equilibration for the tropics and midlatitudes, separately. It is found that the theoretical upper limit of tropical circulation predictability is far less than for midlatitudes. The error growth for a complete general circulation model is compared to a dry version of the same model in which there is no prognostic equation for moisture, and diabatic heat sources are prescribed. It is found that the growth rate of synoptic-scale errors for the dry model is significantly smaller than for the moist model, suggesting that the interactions between dynamics and moist processes are among the important causes of atmospheric flow predictability degradation. Results are then presented of numerical experiments showing that correct specification of the slowly varying boundary condition of SST produces significant improvement in the prediction of time-averaged circulation and rainfall over the tropics.

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.

Measures of model performance based on the log accuracy ratio

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

Morley, Steven Karl; Brito, Thiago Vasconcelos; Welling, Daniel T.

Quantitative assessment of modeling and forecasting of continuous quantities uses a variety of approaches. We review existing literature describing metrics for forecast accuracy and bias, concentrating on those based on relative errors and percentage errors. Of these accuracy metrics, the mean absolute percentage error (MAPE) is one of the most common across many fields and has been widely applied in recent space science literature and we highlight the benefits and drawbacks of MAPE and proposed alternatives. We then introduce the log accuracy ratio, and derive from it two metrics: the median symmetric accuracy; and the symmetric signed percentage bias. Robustmore » methods for estimating the spread of a multiplicative linear model using the log accuracy ratio are also presented. The developed metrics are shown to be easy to interpret, robust, and to mitigate the key drawbacks of their more widely-used counterparts based on relative errors and percentage errors. Their use is illustrated with radiation belt electron flux modeling examples.« less

Measures of model performance based on the log accuracy ratio

DOE PAGES

Morley, Steven Karl; Brito, Thiago Vasconcelos; Welling, Daniel T.

2018-01-03

Quantitative assessment of modeling and forecasting of continuous quantities uses a variety of approaches. We review existing literature describing metrics for forecast accuracy and bias, concentrating on those based on relative errors and percentage errors. Of these accuracy metrics, the mean absolute percentage error (MAPE) is one of the most common across many fields and has been widely applied in recent space science literature and we highlight the benefits and drawbacks of MAPE and proposed alternatives. We then introduce the log accuracy ratio, and derive from it two metrics: the median symmetric accuracy; and the symmetric signed percentage bias. Robustmore » methods for estimating the spread of a multiplicative linear model using the log accuracy ratio are also presented. The developed metrics are shown to be easy to interpret, robust, and to mitigate the key drawbacks of their more widely-used counterparts based on relative errors and percentage errors. Their use is illustrated with radiation belt electron flux modeling examples.« less

Improving medium-range ensemble streamflow forecasts through statistical post-processing

NASA Astrophysics Data System (ADS)

Mendoza, Pablo; Wood, Andy; Clark, Elizabeth; Nijssen, Bart; Clark, Martyn; Ramos, Maria-Helena; Nowak, Kenneth; Arnold, Jeffrey

2017-04-01

Probabilistic hydrologic forecasts are a powerful source of information for decision-making in water resources operations. A common approach is the hydrologic model-based generation of streamflow forecast ensembles, which can be implemented to account for different sources of uncertainties - e.g., from initial hydrologic conditions (IHCs), weather forecasts, and hydrologic model structure and parameters. In practice, hydrologic ensemble forecasts typically have biases and spread errors stemming from errors in the aforementioned elements, resulting in a degradation of probabilistic properties. In this work, we compare several statistical post-processing techniques applied to medium-range ensemble streamflow forecasts obtained with the System for Hydromet Applications, Research and Prediction (SHARP). SHARP is a fully automated prediction system for the assessment and demonstration of short-term to seasonal streamflow forecasting applications, developed by the National Center for Atmospheric Research, University of Washington, U.S. Army Corps of Engineers, and U.S. Bureau of Reclamation. The suite of post-processing techniques includes linear blending, quantile mapping, extended logistic regression, quantile regression, ensemble analogs, and the generalized linear model post-processor (GLMPP). We assess and compare these techniques using multi-year hindcasts in several river basins in the western US. This presentation discusses preliminary findings about the effectiveness of the techniques for improving probabilistic skill, reliability, discrimination, sharpness and resolution.

A data-driven SVR model for long-term runoff prediction and uncertainty analysis based on the Bayesian framework

NASA Astrophysics Data System (ADS)

Liang, Zhongmin; Li, Yujie; Hu, Yiming; Li, Binquan; Wang, Jun

2017-06-01

Accurate and reliable long-term forecasting plays an important role in water resources management and utilization. In this paper, a hybrid model called SVR-HUP is presented to predict long-term runoff and quantify the prediction uncertainty. The model is created based on three steps. First, appropriate predictors are selected according to the correlations between meteorological factors and runoff. Second, a support vector regression (SVR) model is structured and optimized based on the LibSVM toolbox and a genetic algorithm. Finally, using forecasted and observed runoff, a hydrologic uncertainty processor (HUP) based on a Bayesian framework is used to estimate the posterior probability distribution of the simulated values, and the associated uncertainty of prediction was quantitatively analyzed. Six precision evaluation indexes, including the correlation coefficient (CC), relative root mean square error (RRMSE), relative error (RE), mean absolute percentage error (MAPE), Nash-Sutcliffe efficiency (NSE), and qualification rate (QR), are used to measure the prediction accuracy. As a case study, the proposed approach is applied in the Han River basin, South Central China. Three types of SVR models are established to forecast the monthly, flood season and annual runoff volumes. The results indicate that SVR yields satisfactory accuracy and reliability at all three scales. In addition, the results suggest that the HUP cannot only quantify the uncertainty of prediction based on a confidence interval but also provide a more accurate single value prediction than the initial SVR forecasting result. Thus, the SVR-HUP model provides an alternative method for long-term runoff forecasting.

Time-series-based hybrid mathematical modelling method adapted to forecast automotive and medical waste generation: Case study of Lithuania.

PubMed

Karpušenkaitė, Aistė; Ruzgas, Tomas; Denafas, Gintaras

2018-05-01

The aim of the study was to create a hybrid forecasting method that could produce higher accuracy forecasts than previously used 'pure' time series methods. Mentioned methods were already tested with total automotive waste, hazardous automotive waste, and total medical waste generation, but demonstrated at least a 6% error rate in different cases and efforts were made to decrease it even more. Newly developed hybrid models used a random start generation method to incorporate different time-series advantages and it helped to increase the accuracy of forecasts by 3%-4% in hazardous automotive waste and total medical waste generation cases; the new model did not increase the accuracy of total automotive waste generation forecasts. Developed models' abilities to forecast short- and mid-term forecasts were tested using prediction horizon.

KARMA4

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

Khalil, Mohammad; Salloum, Maher; Lee, Jina

2017-07-10

KARMA4 is a C++ library for autoregressive moving average (ARMA) modeling and forecasting of time-series data while incorporating both process and observation error. KARMA4 is designed for fitting and forecasting of time-series data for predictive purposes.

Confidence intervals in Flow Forecasting by using artificial neural networks

NASA Astrophysics Data System (ADS)

Panagoulia, Dionysia; Tsekouras, George

2014-05-01

One of the major inadequacies in implementation of Artificial Neural Networks (ANNs) for flow forecasting is the development of confidence intervals, because the relevant estimation cannot be implemented directly, contrasted to the classical forecasting methods. The variation in the ANN output is a measure of uncertainty in the model predictions based on the training data set. Different methods for uncertainty analysis, such as bootstrap, Bayesian, Monte Carlo, have already proposed for hydrologic and geophysical models, while methods for confidence intervals, such as error output, re-sampling, multi-linear regression adapted to ANN have been used for power load forecasting [1-2]. The aim of this paper is to present the re-sampling method for ANN prediction models and to develop this for flow forecasting of the next day. The re-sampling method is based on the ascending sorting of the errors between real and predicted values for all input vectors. The cumulative sample distribution function of the prediction errors is calculated and the confidence intervals are estimated by keeping the intermediate value, rejecting the extreme values according to the desired confidence levels, and holding the intervals symmetrical in probability. For application of the confidence intervals issue, input vectors are used from the Mesochora catchment in western-central Greece. The ANN's training algorithm is the stochastic training back-propagation process with decreasing functions of learning rate and momentum term, for which an optimization process is conducted regarding the crucial parameters values, such as the number of neurons, the kind of activation functions, the initial values and time parameters of learning rate and momentum term etc. Input variables are historical data of previous days, such as flows, nonlinearly weather related temperatures and nonlinearly weather related rainfalls based on correlation analysis between the under prediction flow and each implicit input variable of different ANN structures [3]. The performance of each ANN structure is evaluated by the voting analysis based on eleven criteria, which are the root mean square error (RMSE), the correlation index (R), the mean absolute percentage error (MAPE), the mean percentage error (MPE), the mean percentage error (ME), the percentage volume in errors (VE), the percentage error in peak (MF), the normalized mean bias error (NMBE), the normalized root mean bias error (NRMSE), the Nash-Sutcliffe model efficiency coefficient (E) and the modified Nash-Sutcliffe model efficiency coefficient (E1). The next day flow for the test set is calculated using the best ANN structure's model. Consequently, the confidence intervals of various confidence levels for training, evaluation and test sets are compared in order to explore the generalisation dynamics of confidence intervals from training and evaluation sets. [1] H.S. Hippert, C.E. Pedreira, R.C. Souza, "Neural networks for short-term load forecasting: A review and evaluation," IEEE Trans. on Power Systems, vol. 16, no. 1, 2001, pp. 44-55. [2] G. J. Tsekouras, N.E. Mastorakis, F.D. Kanellos, V.T. Kontargyri, C.D. Tsirekis, I.S. Karanasiou, Ch.N. Elias, A.D. Salis, P.A. Kontaxis, A.A. Gialketsi: "Short term load forecasting in Greek interconnected power system using ANN: Confidence Interval using a novel re-sampling technique with corrective Factor", WSEAS International Conference on Circuits, Systems, Electronics, Control & Signal Processing, (CSECS '10), Vouliagmeni, Athens, Greece, December 29-31, 2010. [3] D. Panagoulia, I. Trichakis, G. J. Tsekouras: "Flow Forecasting via Artificial Neural Networks - A Study for Input Variables conditioned on atmospheric circulation", European Geosciences Union, General Assembly 2012 (NH1.1 / AS1.16 - Extreme meteorological and hydrological events induced by severe weather and climate change), Vienna, Austria, 22-27 April 2012.

Generating short-term probabilistic wind power scenarios via nonparametric forecast error density estimators: Generating short-term probabilistic wind power scenarios via nonparametric forecast error density estimators

DOE PAGES

Staid, Andrea; Watson, Jean -Paul; Wets, Roger J. -B.; ...

2017-07-11

Forecasts of available wind power are critical in key electric power systems operations planning problems, including economic dispatch and unit commitment. Such forecasts are necessarily uncertain, limiting the reliability and cost effectiveness of operations planning models based on a single deterministic or “point” forecast. A common approach to address this limitation involves the use of a number of probabilistic scenarios, each specifying a possible trajectory of wind power production, with associated probability. We present and analyze a novel method for generating probabilistic wind power scenarios, leveraging available historical information in the form of forecasted and corresponding observed wind power timemore » series. We estimate non-parametric forecast error densities, specifically using epi-spline basis functions, allowing us to capture the skewed and non-parametric nature of error densities observed in real-world data. We then describe a method to generate probabilistic scenarios from these basis functions that allows users to control for the degree to which extreme errors are captured.We compare the performance of our approach to the current state-of-the-art considering publicly available data associated with the Bonneville Power Administration, analyzing aggregate production of a number of wind farms over a large geographic region. Finally, we discuss the advantages of our approach in the context of specific power systems operations planning problems: stochastic unit commitment and economic dispatch. Here, our methodology is embodied in the joint Sandia – University of California Davis Prescient software package for assessing and analyzing stochastic operations strategies.« less

Generating short-term probabilistic wind power scenarios via nonparametric forecast error density estimators: Generating short-term probabilistic wind power scenarios via nonparametric forecast error density estimators

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

Staid, Andrea; Watson, Jean -Paul; Wets, Roger J. -B.

Forecasts of available wind power are critical in key electric power systems operations planning problems, including economic dispatch and unit commitment. Such forecasts are necessarily uncertain, limiting the reliability and cost effectiveness of operations planning models based on a single deterministic or “point” forecast. A common approach to address this limitation involves the use of a number of probabilistic scenarios, each specifying a possible trajectory of wind power production, with associated probability. We present and analyze a novel method for generating probabilistic wind power scenarios, leveraging available historical information in the form of forecasted and corresponding observed wind power timemore » series. We estimate non-parametric forecast error densities, specifically using epi-spline basis functions, allowing us to capture the skewed and non-parametric nature of error densities observed in real-world data. We then describe a method to generate probabilistic scenarios from these basis functions that allows users to control for the degree to which extreme errors are captured.We compare the performance of our approach to the current state-of-the-art considering publicly available data associated with the Bonneville Power Administration, analyzing aggregate production of a number of wind farms over a large geographic region. Finally, we discuss the advantages of our approach in the context of specific power systems operations planning problems: stochastic unit commitment and economic dispatch. Here, our methodology is embodied in the joint Sandia – University of California Davis Prescient software package for assessing and analyzing stochastic operations strategies.« less

A novel hybrid forecasting model for PM₁₀ and SO₂ daily concentrations.

PubMed

Wang, Ping; Liu, Yong; Qin, Zuodong; Zhang, Guisheng

2015-02-01

Air-quality forecasting in urban areas is difficult because of the uncertainties in describing both the emission and meteorological fields. The use of incomplete information in the training phase restricts practical air-quality forecasting. In this paper, we propose a hybrid artificial neural network and a hybrid support vector machine, which effectively enhance the forecasting accuracy of an artificial neural network (ANN) and support vector machine (SVM) by revising the error term of the traditional methods. The hybrid methodology can be described in two stages. First, we applied the ANN or SVM forecasting system with historical data and exogenous parameters, such as meteorological variables. Then, the forecasting target was revised by the Taylor expansion forecasting model using the residual information of the error term in the previous stage. The innovation involved in this approach is that it sufficiently and validly utilizes the useful residual information on an incomplete input variable condition. The proposed method was evaluated by experiments using a 2-year dataset of daily PM₁₀ (particles with a diameter of 10 μm or less) concentrations and SO₂ (sulfur dioxide) concentrations from four air pollution monitoring stations located in Taiyuan, China. The theoretical analysis and experimental results demonstrated that the forecasting accuracy of the proposed model is very promising. Copyright © 2014 Elsevier B.V. All rights reserved.

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.

Use of forecasting signatures to help distinguish periodicity, randomness, and chaos in ripples and other spatial patterns

USGS Publications Warehouse

Rubin, D.M.

1992-01-01

Forecasting of one-dimensional time series previously has been used to help distinguish periodicity, chaos, and noise. This paper presents two-dimensional generalizations for making such distinctions for spatial patterns. The techniques are evaluated using synthetic spatial patterns and then are applied to a natural example: ripples formed in sand by blowing wind. Tests with the synthetic patterns demonstrate that the forecasting techniques can be applied to two-dimensional spatial patterns, with the same utility and limitations as when applied to one-dimensional time series. One limitation is that some combinations of periodicity and randomness exhibit forecasting signatures that mimic those of chaos. For example, sine waves distorted with correlated phase noise have forecasting errors that increase with forecasting distance, errors that, are minimized using nonlinear models at moderate embedding dimensions, and forecasting properties that differ significantly between the original and surrogates. Ripples formed in sand by flowing air or water typically vary in geometry from one to another, even when formed in a flow that is uniform on a large scale; each ripple modifies the local flow or sand-transport field, thereby influencing the geometry of the next ripple downcurrent. Spatial forecasting was used to evaluate the hypothesis that such a deterministic process - rather than randomness or quasiperiodicity - is responsible for the variation between successive ripples. This hypothesis is supported by a forecasting error that increases with forecasting distance, a greater accuracy of nonlinear relative to linear models, and significant differences between forecasts made with the original ripples and those made with surrogate patterns. Forecasting signatures cannot be used to distinguish ripple geometry from sine waves with correlated phase noise, but this kind of structure can be ruled out by two geometric properties of the ripples: Successive ripples are highly correlated in wavelength, and ripple crests display dislocations such as branchings and mergers. ?? 1992 American Institute of Physics.

Improved track forecasting of a typhoon reaching landfall from four-dimensional variational data assimilation of AMSU-A retrieved data

NASA Astrophysics Data System (ADS)

Zhao, Ying; Wang, Bin; Ji, Zhongzhen; Liang, Xudong; Deng, Guo; Zhang, Xin

2005-07-01

In this study, an attempt to improve typhoon forecasts is made by incorporating three-dimensional Advanced Microwave Sounding Unit-A (AMSU-A) retrieved wind and temperature and the central sea level pressure of cyclones from typhoon reports or bogus surface low data into initial conditions, on the basis of the Fifth-Generation National Center for Atmospheric Research/Pennsylvania State University Mesoscale Model (MM5) four-dimensional variational data assimilation (4DVar) system with a full-physics adjoint model. All the above-mentioned data are found to be useful for improvement of typhoon forecasts in this mesoscale data assimilation experiment. The comparison tests showed the following results: (1) The assimilation of the satellite-retrieved data was found to have a positive impact on the typhoon track forecast, but the landing position error is ˜150 km. (2) The assimilation of both the satellite-retrieved data and moving information of the typhoon center dramatically improved the track forecast and captured the recurvature and landfall. The mean track error during the 72-hour forecast is 69 km. The predicted typhoon intensity, however, is much weaker than that from observations. (3) The assimilation of both the satellite-retrieved data and the bogus surface low data improved the intensity and track forecasts more significantly than the assimilation of only bogus surface low data (bogus data assimilation) did. The mean errors during the 72-hour forecast are 2.6 hPa for the minimum sea level pressure and 87 km for track position. However, the forecasted landing time is ˜6 hours earlier than the observed one.

Skill of a global seasonal ensemble streamflow forecasting system

NASA Astrophysics Data System (ADS)

Candogan Yossef, Naze; Winsemius, Hessel; Weerts, Albrecht; van Beek, Rens; Bierkens, Marc

2013-04-01

Forecasting of water availability and scarcity is a prerequisite for managing the risks and opportunities caused by the inter-annual variability of streamflow. Reliable seasonal streamflow forecasts are necessary to prepare for an appropriate response in disaster relief, management of hydropower reservoirs, water supply, agriculture and navigation. Seasonal hydrological forecasting on a global scale could be valuable especially for developing regions of the world, where effective hydrological forecasting systems are scarce. In this study, we investigate the forecasting skill of the global seasonal streamflow forecasting system FEWS-World, using the global hydrological model PCR-GLOBWB. FEWS-World has been setup within the European Commission 7th Framework Programme project Global Water Scarcity Information Service (GLOWASIS). Skill is assessed in historical simulation mode as well as retroactive forecasting mode. The assessment in historical simulation mode used a meteorological forcing based on observations from the Climate Research Unit of the University of East Anglia and the ERA-40 reanalysis of the European Center for Medium-Range Weather Forecasts (ECMWF). We assessed the skill of the global hydrological model PCR-GLOBWB in reproducing past discharge extremes in 20 large rivers of the world. This preliminary assessment concluded that the prospects for seasonal forecasting with PCR-GLOBWB or comparable models are positive. However this assessment did not include actual meteorological forecasts. Thus the meteorological forcing errors were not assessed. Yet, in a forecasting setup, the predictive skill of a hydrological forecasting system is affected by errors due to uncertainty from numerical weather prediction models. For the assessment in retroactive forecasting mode, the model is forced with actual ensemble forecasts from the seasonal forecast archives of ECMWF. Skill is assessed at 78 stations on large river basins across the globe, for all the months of the year and for lead times up to 6 months. The forecasted discharges are compared with observed monthly streamflow records using the ensemble verification measures Brier Skill Score (BSS) and Continuous Ranked Probability Score (CRPS). The eventual goal is to transfer FEWS-World to operational forecasting mode, where the system will use operational seasonal forecasts from ECMWF. The results will be disseminated on the internet, and hopefully provide information that is valuable for users in data and model-poor regions of the world.

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 WRF-ROI system by incorporating eigen-decomposition

NASA Astrophysics Data System (ADS)

Kim, S.; Noh, N.; Song, H.; Lim, G.

2011-12-01

This study presents the development of WRF-ROI system, which is the implementation of Retrospective Optimal Interpolation (ROI) to the Weather Research and Forecasting model (WRF). ROI is a new data assimilation algorithm introduced by Song et al. (2009) and Song and Lim (2009). The formulation of ROI is similar with that of Optimal Interpolation (OI), but ROI iteratively assimilates an observation set at a post analysis time into a prior analysis, possibly providing the high quality reanalysis data. ROI method assimilates the data at post analysis time using perturbation method (Errico and Raeder, 1999) without adjoint model. In previous study, ROI method is applied to Lorenz 40-variable model (Lorenz, 1996) to validate the algorithm and to investigate the capability. It is therefore required to apply this ROI method into a more realistic and complicated model framework such as WRF. In this research, the reduced-rank formulation of ROI is used instead of a reduced-resolution method. The computational costs can be reduced due to the eigen-decomposition of background error covariance in the reduced-rank method. When single profile of observations is assimilated in the WRF-ROI system by incorporating eigen-decomposition, the analysis error tends to be reduced if compared with the background error. The difference between forecast errors with assimilation and without assimilation is obviously increased as time passed, which means the improvement of forecast error by assimilation.

Time series regression and ARIMAX for forecasting currency flow at Bank Indonesia in Sulawesi region

NASA Astrophysics Data System (ADS)

Suharsono, Agus; Suhartono, Masyitha, Aulia; Anuravega, Arum

2015-12-01

The purpose of the study is to forecast the outflow and inflow of currency at Indonesian Central Bank or Bank Indonesia (BI) in Sulawesi Region. The currency outflow and inflow data tend to have a trend pattern which is influenced by calendar variation effects. Therefore, this research focuses to apply some forecasting methods that could handle calendar variation effects, i.e. Time Series Regression (TSR) and ARIMAX models, and compare the forecast accuracy with ARIMA model. The best model is selected based on the lowest of Root Mean Squares Errors (RMSE) at out-sample dataset. The results show that ARIMA is the best model for forecasting the currency outflow and inflow at South Sulawesi. Whereas, the best model for forecasting the currency outflow at Central Sulawesi and Southeast Sulawesi, and for forecasting the currency inflow at South Sulawesi and North Sulawesi is TSR. Additionally, ARIMAX is the best model for forecasting the currency outflow at North Sulawesi. Hence, the results show that more complex models do not neccessary yield more accurate forecast than the simpler one.

An autoregressive integrated moving average model for short-term prediction of hepatitis C virus seropositivity among male volunteer blood donors in Karachi, Pakistan

PubMed Central

Akhtar, Saeed; Rozi, Shafquat

2009-01-01

AIM: To identify the stochastic autoregressive integrated moving average (ARIMA) model for short term forecasting of hepatitis C virus (HCV) seropositivity among volunteer blood donors in Karachi, Pakistan. METHODS: Ninety-six months (1998-2005) data on HCV seropositive cases (1000-1 × month-1) among male volunteer blood donors tested at four major blood banks in Karachi, Pakistan were subjected to ARIMA modeling. Subsequently, a fitted ARIMA model was used to forecast HCV seropositive donors for 91-96 mo to contrast with observed series of the same months. To assess the forecast accuracy, the mean absolute error rate (%) between the observed and predicted HCV seroprevalence was calculated. Finally, a fitted ARIMA model was used for short-term forecasts beyond the observed series. RESULTS: The goodness-of-fit test of the optimum ARIMA (2,1,7) model showed non-significant autocorrelations in the residuals of the model. The forecasts by ARIMA for 91-96 mo closely followed the pattern of observed series for the same months, with mean monthly absolute forecast errors (%) over 6 mo of 6.5%. The short-term forecasts beyond the observed series adequately captured the pattern in the data and showed increasing tendency of HCV seropositivity with a mean ± SD HCV seroprevalence (1000-1 × month-1) of 24.3 ± 1.4 over the forecast interval. CONCLUSION: To curtail HCV spread, public health authorities need to educate communities and health care providers about HCV transmission routes based on known HCV epidemiology in Pakistan and its neighboring countries. Future research may focus on factors associated with hyperendemic levels of HCV infection. PMID:19340903

Short-term ensemble radar rainfall forecasts for hydrological applications

NASA Astrophysics Data System (ADS)

Codo de Oliveira, M.; Rico-Ramirez, M. A.

2016-12-01

Flooding is a very common natural disaster around the world, putting local population and economy at risk. Forecasting floods several hours ahead and issuing warnings are of main importance to permit proper response in emergency situations. However, it is important to know the uncertainties related to the rainfall forecasting in order to produce more reliable forecasts. Nowcasting models (short-term rainfall forecasts) are able to produce high spatial and temporal resolution predictions that are useful in hydrological applications. Nonetheless, they are subject to uncertainties mainly due to the nowcasting model used, errors in radar rainfall estimation, temporal development of the velocity field and to the fact that precipitation processes such as growth and decay are not taken into account. In this study an ensemble generation scheme using rain gauge data as a reference to estimate radars errors is used to produce forecasts with up to 3h lead-time. The ensembles try to assess in a realistic way the residual uncertainties that remain even after correction algorithms are applied in the radar data. The ensembles produced are compered to a stochastic ensemble generator. Furthermore, the rainfall forecast output was used as an input in a hydrodynamic sewer network model and also in hydrological model for catchments of different sizes in north England. A comparative analysis was carried of how was carried out to assess how the radar uncertainties propagate into these models. The first named author is grateful to CAPES - Ciencia sem Fronteiras for funding this PhD research.

Multiple linear regression and regression with time series error models in forecasting PM10 concentrations in Peninsular Malaysia.

PubMed

Ng, Kar Yong; Awang, Norhashidah

2018-01-06

Frequent haze occurrences in Malaysia have made the management of PM 10 (particulate matter with aerodynamic less than 10 μm) pollution a critical task. This requires knowledge on factors associating with PM 10 variation and good forecast of PM 10 concentrations. Hence, this paper demonstrates the prediction of 1-day-ahead daily average PM 10 concentrations based on predictor variables including meteorological parameters and gaseous pollutants. Three different models were built. They were multiple linear regression (MLR) model with lagged predictor variables (MLR1), MLR model with lagged predictor variables and PM 10 concentrations (MLR2) and regression with time series error (RTSE) model. The findings revealed that humidity, temperature, wind speed, wind direction, carbon monoxide and ozone were the main factors explaining the PM 10 variation in Peninsular Malaysia. Comparison among the three models showed that MLR2 model was on a same level with RTSE model in terms of forecasting accuracy, while MLR1 model was the worst.

Impact of Initial Condition Errors and Precipitation Forecast Bias on Drought Simulation and Prediction in the Huaihe River Basin

NASA Astrophysics Data System (ADS)

Xu, H.; Luo, L.; Wu, Z.

2016-12-01

Drought, regarded as one of the major disasters all over the world, is not always easy to detect and forecast. Hydrological models coupled with Numerical Weather Prediction (NWP) has become a relatively effective method for drought monitoring and prediction. The accuracy of hydrological initial condition (IC) and the skill of NWP precipitation forecast can both heavily affect the quality and skill of hydrological forecast. In the study, the Variable Infiltration Capacity (VIC) model and Global Environmental Multi-scale (GEM) model were used to investigate the roles of IC and NWP forecast accuracy on hydrological predictions. A rev-ESP type experiment was conducted for a number of drought events in the Huaihe river basin. The experiment suggests that errors in ICs indeed affect the drought simulations by VIC and thus the drought monitoring. Although errors introduced in the ICs diminish gradually, the influence sometimes can last beyond 12 months. Using the soil moisture anomaly percentage index (SMAPI) as the metric to measure drought severity for the study region, we are able to quantify that time scale of influence from IC ranges. The analysis shows that the time scale is directly related to the magnitude of the introduced IC range and the average precipitation intensity. In order to explore how systematic bias correction in GEM forecasted precipitation can affect precipitation and hydrological forecast, we then both used station and gridded observations to eliminate biases of forecasted data. Meanwhile, different precipitation inputs with corrected data during drought process were conducted by VIC to investigate the changes of drought simulations, thus demonstrated short-term rolling drought prediction using a better performed corrected precipitation forecast. There is a word limit on the length of the abstract. So make sure your abstract fits the requirement. If this version is too long, try to shorten it as much as you can.

Adaptive correction of ensemble forecasts

NASA Astrophysics Data System (ADS)

Pelosi, Anna; Battista Chirico, Giovanni; Van den Bergh, Joris; Vannitsem, Stephane

2017-04-01

Forecasts from numerical weather prediction (NWP) models often suffer from both systematic and non-systematic errors. These are present in both deterministic and ensemble forecasts, and originate from various sources such as model error and subgrid variability. Statistical post-processing techniques can partly remove such errors, which is particularly important when NWP outputs concerning surface weather variables are employed for site specific applications. Many different post-processing techniques have been developed. For deterministic forecasts, adaptive methods such as the Kalman filter are often used, which sequentially post-process the forecasts by continuously updating the correction parameters as new ground observations become available. These methods are especially valuable when long training data sets do not exist. For ensemble forecasts, well-known techniques are ensemble model output statistics (EMOS), and so-called "member-by-member" approaches (MBM). Here, we introduce a new adaptive post-processing technique for ensemble predictions. The proposed method is a sequential Kalman filtering technique that fully exploits the information content of the ensemble. One correction equation is retrieved and applied to all members, however the parameters of the regression equations are retrieved by exploiting the second order statistics of the forecast ensemble. We compare our new method with two other techniques: a simple method that makes use of a running bias correction of the ensemble mean, and an MBM post-processing approach that rescales the ensemble mean and spread, based on minimization of the Continuous Ranked Probability Score (CRPS). We perform a verification study for the region of Campania in southern Italy. We use two years (2014-2015) of daily meteorological observations of 2-meter temperature and 10-meter wind speed from 18 ground-based automatic weather stations distributed across the region, comparing them with the corresponding COSMO-LEPS ensemble forecasts. Deterministic verification scores (e.g., mean absolute error, bias) and probabilistic scores (e.g., CRPS) are used to evaluate the post-processing techniques. We conclude that the new adaptive method outperforms the simpler running bias-correction. The proposed adaptive method often outperforms the MBM method in removing bias. The MBM method has the advantage of correcting the ensemble spread, although it needs more training data.

Time series forecasting using ERNN and QR based on Bayesian model averaging

NASA Astrophysics Data System (ADS)

Pwasong, Augustine; Sathasivam, Saratha

2017-08-01

The Bayesian model averaging technique is a multi-model combination technique. The technique was employed to amalgamate the Elman recurrent neural network (ERNN) technique with the quadratic regression (QR) technique. The amalgamation produced a hybrid technique known as the hybrid ERNN-QR technique. The potentials of forecasting with the hybrid technique are compared with the forecasting capabilities of individual techniques of ERNN and QR. The outcome revealed that the hybrid technique is superior to the individual techniques in the mean square error sense.

A forecasting model for dengue incidence in the District of Gampaha, Sri Lanka.

PubMed

Withanage, Gayan P; Viswakula, Sameera D; Nilmini Silva Gunawardena, Y I; Hapugoda, Menaka D

2018-04-24

Dengue is one of the major health problems in Sri Lanka causing an enormous social and economic burden to the country. An accurate early warning system can enhance the efficiency of preventive measures. The aim of the study was to develop and validate a simple accurate forecasting model for the District of Gampaha, Sri Lanka. Three time-series regression models were developed using monthly rainfall, rainy days, temperature, humidity, wind speed and retrospective dengue incidences over the period January 2012 to November 2015 for the District of Gampaha, Sri Lanka. Various lag times were analyzed to identify optimum forecasting periods including interactions of multiple lags. The models were validated using epidemiological data from December 2015 to November 2017. Prepared models were compared based on Akaike's information criterion, Bayesian information criterion and residual analysis. The selected model forecasted correctly with mean absolute errors of 0.07 and 0.22, and root mean squared errors of 0.09 and 0.28, for training and validation periods, respectively. There were no dengue epidemics observed in the district during the training period and nine outbreaks occurred during the forecasting period. The proposed model captured five outbreaks and correctly rejected 14 within the testing period of 24 months. The Pierce skill score of the model was 0.49, with a receiver operating characteristic of 86% and 92% sensitivity. The developed weather based forecasting model allows warnings of impending dengue outbreaks and epidemics in advance of one month with high accuracy. Depending upon climatic factors, the previous month's dengue cases had a significant effect on the dengue incidences of the current month. The simple, precise and understandable forecasting model developed could be used to manage limited public health resources effectively for patient management, vector surveillance and intervention programmes in the district.

Sampling strategies based on singular vectors for assimilated models in ocean forecasting systems

NASA Astrophysics Data System (ADS)

Fattorini, Maria; Brandini, Carlo; Ortolani, Alberto

2016-04-01

Meteorological and oceanographic models do need observations, not only as a ground truth element to verify the quality of the models, but also to keep model forecast error acceptable: through data assimilation techniques which merge measured and modelled data, natural divergence of numerical solutions from reality can be reduced / controlled and a more reliable solution - called analysis - is computed. Although this concept is valid in general, its application, especially in oceanography, raises many problems due to three main reasons: the difficulties that have ocean models in reaching an acceptable state of equilibrium, the high measurements cost and the difficulties in realizing them. The performances of the data assimilation procedures depend on the particular observation networks in use, well beyond the background quality and the used assimilation method. In this study we will present some results concerning the great impact of the dataset configuration, in particular measurements position, on the evaluation of the overall forecasting reliability of an ocean model. The aim consists in identifying operational criteria to support the design of marine observation networks at regional scale. In order to identify the observation network able to minimize the forecast error, a methodology based on Singular Vectors Decomposition of the tangent linear model is proposed. Such a method can give strong indications on the local error dynamics. In addition, for the purpose of avoiding redundancy of information contained in the data, a minimal distance among data positions has been chosen on the base of a spatial correlation analysis of the hydrodynamic fields under investigation. This methodology has been applied for the choice of data positions starting from simplified models, like an ideal double-gyre model and a quasi-geostrophic one. Model configurations and data assimilation are based on available ROMS routines, where a variational assimilation algorithm (4D-var) is included as part of the code These first applications have provided encouraging results in terms of increased predictability time and reduced forecast error, also improving the quality of the analysis used to recover the real circulation patterns from a first guess quite far from the real state.

Assessment of Global Forecast Ocean Assimilation Model (FOAM) using new satellite SST data

NASA Astrophysics Data System (ADS)

Ascione Kenov, Isabella; Sykes, Peter; Fiedler, Emma; McConnell, Niall; Ryan, Andrew; Maksymczuk, Jan

2016-04-01

There is an increased demand for accurate ocean weather information for applications in the field of marine safety and navigation, water quality, offshore commercial operations, monitoring of oil spills and pollutants, among others. The Met Office, UK, provides ocean forecasts to customers from governmental, commercial and ecological sectors using the Global Forecast Ocean Assimilation Model (FOAM), an operational modelling system which covers the global ocean and runs daily, using the NEMO (Nucleus for European Modelling of the Ocean) ocean model with horizontal resolution of 1/4° and 75 vertical levels. The system assimilates salinity and temperature profiles, sea surface temperature (SST), sea surface height (SSH), and sea ice concentration observations on a daily basis. In this study, the FOAM system is updated to assimilate Advanced Microwave Scanning Radiometer 2 (AMSR2) and the Spinning Enhanced Visible and Infrared Imager (SEVIRI) SST data. Model results from one month trials are assessed against observations using verification tools which provide a quantitative description of model performance and error, based on statistical metrics, including mean error, root mean square error (RMSE), correlation coefficient, and Taylor diagrams. A series of hindcast experiments is used to run the FOAM system with AMSR2 and SEVIRI SST data, using a control run for comparison. Results show that all trials perform well on the global ocean and that largest SST mean errors were found in the Southern hemisphere. The geographic distribution of the model error for SST and temperature profiles are discussed using statistical metrics evaluated over sub-regions of the global ocean.

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.

When idols look into the future: fair treatment modulates the affective forecasting error in talent show candidates.

PubMed

Feys, Marjolein; Anseel, Frederik

2015-03-01

People's affective forecasts are often inaccurate because they tend to overestimate how they will feel after an event. As life decisions are often based on affective forecasts, it is crucial to find ways to manage forecasting errors. We examined the impact of a fair treatment on forecasting errors in candidates in a Belgian reality TV talent show. We found that perceptions of fair treatment increased the forecasting error for losers (a negative audition decision) but decreased it for winners (a positive audition decision). For winners, this effect was even more pronounced when candidates were highly invested in their self-view as a future pop idol whereas for losers, the effect was more pronounced when importance was low. The results in this study point to a potential paradox between maximizing happiness and decreasing forecasting errors. A fair treatment increased the forecasting error for losers, but actually made them happier. © 2014 The British Psychological Society.

Optimum employment of satellite indirect soundings as numerical model input

NASA Technical Reports Server (NTRS)

Horn, L. H.; Derber, J. C.; Koehler, T. L.; Schmidt, B. D.

1981-01-01

The characteristics of satellite-derived temperature soundings that would significantly affect their use as input for numerical weather prediction models were examined. Independent evaluations of satellite soundings were emphasized to better define error characteristics. Results of a Nimbus-6 sounding study reveal an underestimation of the strength of synoptic scale troughs and ridges, and associated gradients in isobaric height and temperature fields. The most significant errors occurred near the Earth's surface and the tropopause. Soundings from the TIROS-N and NOAA-6 satellites were also evaluated. Results again showed an underestimation of upper level trough amplitudes leading to weaker thermal gradient depictions in satellite-only fields. These errors show a definite correlation to the synoptic flow patterns. In a satellite-only analysis used to initialize a numerical model forecast, it was found that these synoptically correlated errors were retained in the forecast sequence.

Development, testing, and applications of site-specific tsunami inundation models for real-time forecasting

NASA Astrophysics Data System (ADS)

Tang, L.; Titov, V. V.; Chamberlin, C. D.

2009-12-01

The study describes the development, testing and applications of site-specific tsunami inundation models (forecast models) for use in NOAA's tsunami forecast and warning system. The model development process includes sensitivity studies of tsunami wave characteristics in the nearshore and inundation, for a range of model grid setups, resolutions and parameters. To demonstrate the process, four forecast models in Hawaii, at Hilo, Kahului, Honolulu, and Nawiliwili are described. The models were validated with fourteen historical tsunamis and compared with numerical results from reference inundation models of higher resolution. The accuracy of the modeled maximum wave height is greater than 80% when the observation is greater than 0.5 m; when the observation is below 0.5 m the error is less than 0.3 m. The error of the modeled arrival time of the first peak is within 3% of the travel time. The developed forecast models were further applied to hazard assessment from simulated magnitude 7.5, 8.2, 8.7 and 9.3 tsunamis based on subduction zone earthquakes in the Pacific. The tsunami hazard assessment study indicates that use of a seismic magnitude alone for a tsunami source assessment is inadequate to achieve such accuracy for tsunami amplitude forecasts. The forecast models apply local bathymetric and topographic information, and utilize dynamic boundary conditions from the tsunami source function database, to provide site- and event-specific coastal predictions. Only by combining a Deep-ocean Assessment and Reporting of Tsunami-constrained tsunami magnitude with site-specific high-resolution models can the forecasts completely cover the evolution of earthquake-generated tsunami waves: generation, deep ocean propagation, and coastal inundation. Wavelet analysis of the tsunami waves suggests the coastal tsunami frequency responses at different sites are dominated by the local bathymetry, yet they can be partially related to the locations of the tsunami sources. The study also demonstrates the nonlinearity between offshore and nearshore maximum wave amplitudes.

Impact of Targeted Ocean Observations for Improving Ocean Model Initialization for Coupled Hurricane Forecasting

NASA Astrophysics Data System (ADS)

Halliwell, G. R.; Srinivasan, A.; Kourafalou, V. H.; Yang, H.; Le Henaff, M.; Atlas, R. M.

2012-12-01

The accuracy of hurricane intensity forecasts produced by coupled forecast models is influenced by errors and biases in SST forecasts produced by the ocean model component and the resulting impact on the enthalpy flux from ocean to atmosphere that powers the storm. Errors and biases in fields used to initialize the ocean model seriously degrade SST forecast accuracy. One strategy for improving ocean model initialization is to design a targeted observing program using airplanes and in-situ devices such as floats and drifters so that assimilation of the additional data substantially reduces errors in the ocean analysis system that provides the initial fields. Given the complexity and expense of obtaining these additional observations, observing system design methods such as OSSEs are attractive for designing efficient observing strategies. A new fraternal-twin ocean OSSE system based on the HYbrid Coordinate Ocean Model (HYCOM) is used to assess the impact of targeted ocean profiles observed by hurricane research aircraft, and also by in-situ float and drifter deployments, on reducing errors in initial ocean fields. A 0.04-degree HYCOM simulation of the Gulf of Mexico is evaluated as the nature run by determining that important ocean circulation features such as the Loop Current and synoptic cyclones and anticyclones are realistically simulated. The data-assimilation system is run on a 0.08-degree HYCOM mesh with substantially different model configuration than the nature run, and it uses a new ENsemble Kalman Filter (ENKF) algorithm optimized for the ocean model's hybrid vertical coordinates. The OSSE system is evaluated and calibrated by first running Observing System Experiments (OSEs) to evaluate existing observing systems, specifically quantifying the impact of assimilating more than one satellite altimeter, and also the impact of assimilating targeted ocean profiles taken by the NOAA WP-3D hurricane research aircraft in the Gulf of Mexico during the Deepwater Horizon oil spill. OSSE evaluation and calibration is then performed by repeating these two OSEs with synthetic observations and comparing the resulting observing system impact to determine if it differs from the OSE results. OSSEs are first run to evaluate different airborne sampling strategies with respect to temporal frequency of flights and the horizontal separation of upper-ocean profiles during each flight. They are then run to assess the impact of releasing multiple floats and gliders. Evaluation strategy focuses on error reduction in fields important for hurricane forecasting such as the structure of ocean currents and eddies, upper ocean heat content distribution, and upper-ocean stratification.

Impacts of high resolution model downscaling in coastal regions

NASA Astrophysics Data System (ADS)

Bricheno, Lucy; Wolf, Judith

2013-04-01

With model development and cheaper computational resources ocean forecasts are becoming readily available, high resolution coastal forecasting is now a reality. This can only be achieved, however, by downscaling global or basin-scale products such as the MyOcean reanalyses and forecasts. These model products have resolution ranging from 1/16th - 1/4 degree, which are often insufficient for coastal scales, but can provide initialisation and boundary data. We present applications of downscaling the MyOcean products for use in shelf-seas and the nearshore. We will address the question 'Do coastal predictions improve with higher resolution modelling?' with a few focused examples, while also discussing what is meant by an improved result. Increasing resolution appears to be an obvious route for getting more accurate forecasts in operational coastal models. However, when models resolve finer scales, this may lead to the introduction of high-frequency variability which is not necessarily deterministic. Thus a flow may appear more realistic by generating eddies but the simple statistics like rms error and correlation may become less good because the model variability is not exactly in phase with the observations (Hoffman et al., 1995). By deciding on a specific process to simulate (rather than concentrating on reducing rms error) we can better assess the improvements gained by downscaling. In this work we will select two processes which are dominant in our case-study site: Liverpool Bay. Firstly we consider the magnitude and timing of a peak in tide-surge elevations, by separating out the event into timing (or displacement) and intensity (or amplitude) errors. The model can thus be evaluated on how well it predicts the timing and magnitude of the surge. The second important characteristic of Liverpool Bay is the position of the freshwater front. To evaluate model performance in this case, the location, sharpness, and temperature difference across the front will be considered. We will show that by using intelligent metrics designed with a physical process in mind, we can learn more about model performance than by considering 'bulk' statistics alone. R. M. Hoffman and Z. Liu and J-F. Louic and C. Grassotti (1995) 'Distortion Representation of Forecast Errors' Monthly Weather Review 123: 2758-2770

Ensemble Data Assimilation of Photovoltaic Power Information in the Convection-permitting High-Resolution Model COSMO-DE

NASA Astrophysics Data System (ADS)

Declair, Stefan; Saint-Drenan, Yves-Marie; Potthast, Roland

2017-04-01

Determining the amount of weather dependent renewable energy is a demanding task for transmission system operators (TSOs) and wind and photovoltaic (PV) prediction errors require the use of reserve power, which generate costs and can - in extreme cases - endanger the security of supply. In the project EWeLiNE funded by the German government, the German Weather Service and the Fraunhofer Institute on Wind Energy and Energy System Technology develop innovative weather- and power forecasting models and tools for grid integration of weather dependent renewable energy. The key part in energy prediction process chains is the numerical weather prediction (NWP) system. Irradiation forecasts from NWP systems are however subject to several sources of error. For PV power prediction, weaknesses of the NWP model to correctly forecast i.e. low stratus, absorption of condensed water or aerosol optical depths are the main sources of errors. Inaccurate radiation schemes (i.e. the two-stream parametrization) are also known as a deficit of NWP systems with regard to irradiation forecast. To mitigate errors like these, latest observations can be used in a pre-processing technique called data assimilation (DA). In DA, not only the initial fields are provided, but the model is also synchronized with reality - the observations - and hence forecast errors are reduced. Besides conventional observation networks like radiosondes, synoptic observations or air reports of wind, pressure and humidity, the number of observations measuring meteorological information indirectly by means of remote sensing such as satellite radiances, radar reflectivities or GPS slant delays strongly increases. Numerous PV plants installed in Germany potentially represent a dense meteorological network assessing irradiation through their power measurements. Forecast accuracy may thus be enhanced by extending the observations in the assimilation by this new source of information. PV power plants can provide information on clouds, aerosol optical depth or low stratus in terms of remote sensing: the power output is strongly dependent on perturbations along the slant between sun position and PV panel. Since these data are not limited to the vertical column above or below the detector, it may thus complement satellite data and compensate weaknesses in the radiation scheme. In this contribution, the used DA technique (Local Ensemble Transform Kalman Filter, LETKF) is shortly sketched. Furthermore, the computation of the model power equivalents is described and first results are presented and discussed.

Forecasting Daily Patient Outflow From a Ward Having No Real-Time Clinical Data

PubMed Central

Tran, Truyen; Luo, Wei; Phung, Dinh; Venkatesh, Svetha

2016-01-01

Background: Modeling patient flow is crucial in understanding resource demand and prioritization. We study patient outflow from an open ward in an Australian hospital, where currently bed allocation is carried out by a manager relying on past experiences and looking at demand. Automatic methods that provide a reasonable estimate of total next-day discharges can aid in efficient bed management. The challenges in building such methods lie in dealing with large amounts of discharge noise introduced by the nonlinear nature of hospital procedures, and the nonavailability of real-time clinical information in wards. Objective Our study investigates different models to forecast the total number of next-day discharges from an open ward having no real-time clinical data. Methods We compared 5 popular regression algorithms to model total next-day discharges: (1) autoregressive integrated moving average (ARIMA), (2) the autoregressive moving average with exogenous variables (ARMAX), (3) k-nearest neighbor regression, (4) random forest regression, and (5) support vector regression. Although the autoregressive integrated moving average model relied on past 3-month discharges, nearest neighbor forecasting used median of similar discharges in the past in estimating next-day discharge. In addition, the ARMAX model used the day of the week and number of patients currently in ward as exogenous variables. For the random forest and support vector regression models, we designed a predictor set of 20 patient features and 88 ward-level features. Results Our data consisted of 12,141 patient visits over 1826 days. Forecasting quality was measured using mean forecast error, mean absolute error, symmetric mean absolute percentage error, and root mean square error. When compared with a moving average prediction model, all 5 models demonstrated superior performance with the random forests achieving 22.7% improvement in mean absolute error, for all days in the year 2014. Conclusions In the absence of clinical information, our study recommends using patient-level and ward-level data in predicting next-day discharges. Random forest and support vector regression models are able to use all available features from such data, resulting in superior performance over traditional autoregressive methods. An intelligent estimate of available beds in wards plays a crucial role in relieving access block in emergency departments. PMID:27444059

Performance and quality assessment of the recent updated CMEMS global ocean monitoring and forecasting real-time system

NASA Astrophysics Data System (ADS)

Le Galloudec, Olivier; Lellouche, Jean-Michel; Greiner, Eric; Garric, Gilles; Régnier, Charly; Drévillon, Marie; Drillet, Yann

2017-04-01

Since May 2015, Mercator Ocean opened the Copernicus Marine Environment and Monitoring Service (CMEMS) and is in charge of the global eddy resolving ocean analyses and forecast. In this context, Mercator Ocean currently delivers in real-time daily services (weekly analyses and daily forecast) with a global 1/12° high resolution system. The model component is the NEMO platform driven at the surface by the IFS ECMWF atmospheric analyses and forecasts. Observations are assimilated by means of a reduced-order Kalman filter with a 3D multivariate modal decomposition of the forecast error. It includes an adaptive-error estimate and a localization algorithm. Along track altimeter data, satellite Sea Surface Temperature and in situ temperature and salinity vertical profiles are jointly assimilated to estimate the initial conditions for numerical ocean forecasting. A 3D-Var scheme provides a correction for the slowly-evolving large-scale biases in temperature and salinity. R&D activities have been conducted at Mercator Ocean these last years to improve the real-time 1/12° global system for recent updated CMEMS version in 2016. The ocean/sea-ice model and the assimilation scheme benefited of the following improvements: large-scale and objective correction of atmospheric quantities with satellite data, new Mean Dynamic Topography taking into account the last version of GOCE geoid, new adaptive tuning of some observational errors, new Quality Control on the assimilated temperature and salinity vertical profiles based on dynamic height criteria, assimilation of satellite sea-ice concentration, new freshwater runoff from ice sheets melting, … This presentation will show the impact of some updates separately, with a particular focus on adaptive tuning experiments of satellite Sea Level Anomaly (SLA) and Sea Surface Temperature (SST) observations errors. For the SLA, the a priori prescribed observation error is globally greatly reduced. The median value of the error changed from 5cm to 2.5cm in a few assimilation cycles. For the SST, we chose to maintain the median value of the error to 0.4°C. The spatial distribution of the SST error follows the model physics and atmospheric variability. Either for SLA or SST, we improve the performances of the system using this adaptive tuning. The overall behavior of the system integrating all updates reporting on the products quality improvements will be also discussed, highlighting the level of performance and the reliability of the new system.

The FIM-iHYCOM Model in SubX: Evaluation of Subseasonal Errors and Variability

NASA Astrophysics Data System (ADS)

Green, B.; Sun, S.; Benjamin, S.; Grell, G. A.; Bleck, R.

2017-12-01

NOAA/ESRL/GSD has produced both real-time and retrospective forecasts for the Subseasonal Experiment (SubX) using the FIM-iHYCOM model. FIM-iHYCOM couples the atmospheric Flow-following finite volume Icosahedral Model (FIM) to an icosahedral-grid version of the Hybrid Coordinate Ocean Model (HYCOM). This coupled model is unique in terms of its grid structure: in the horizontal, the icosahedral meshes are perfectly matched for FIM and iHYCOM, eliminating the need for a flux interpolator; in the vertical, both models use adaptive arbitrary Lagrangian-Eulerian hybrid coordinates. For SubX, FIM-iHYCOM initializes four time-lagged ensemble members around each Wednesday, which are integrated forward to provide 32-day forecasts. While it has already been shown that this model has similar predictive skill as NOAA's operational CFSv2 in terms of the RMM index, FIM-iHYCOM is still fairly new and thus its overall performance needs to be thoroughly evaluated. To that end, this study examines model errors as a function of forecast lead week (1-4) - i.e., model drift - for key variables including 2-m temperature, precipitation, and SST. Errors are evaluated against two reanalysis products: CFSR, from which FIM-iHYCOM initial conditions are derived, and the quasi-independent ERA-Interim. The week 4 error magnitudes are similar between FIM-iHYCOM and CFSv2, albeit with different spatial distributions. Also, intraseasonal variability as simulated in these two models will be compared with reanalyses. The impact of hindcast frequency (4 times per week, once per week, or once per day) on the model climatology is also examined to determine the implications for systematic error correction in FIM-iHYCOM.

Seasonality and Trend Forecasting of Tuberculosis Prevalence Data in Eastern Cape, South Africa, Using a Hybrid Model.

PubMed

Azeez, Adeboye; Obaromi, Davies; Odeyemi, Akinwumi; Ndege, James; Muntabayi, Ruffin

2016-07-26

Tuberculosis (TB) is a deadly infectious disease caused by Mycobacteria tuberculosis. Tuberculosis as a chronic and highly infectious disease is prevalent in almost every part of the globe. More than 95% of TB mortality occurs in low/middle income countries. In 2014, approximately 10 million people were diagnosed with active TB and two million died from the disease. In this study, our aim is to compare the predictive powers of the seasonal autoregressive integrated moving average (SARIMA) and neural network auto-regression (SARIMA-NNAR) models of TB incidence and analyse its seasonality in South Africa. TB incidence cases data from January 2010 to December 2015 were extracted from the Eastern Cape Health facility report of the electronic Tuberculosis Register (ERT.Net). A SARIMA model and a combined model of SARIMA model and a neural network auto-regression (SARIMA-NNAR) model were used in analysing and predicting the TB data from 2010 to 2015. Simulation performance parameters of mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE), mean percent error (MPE), mean absolute scaled error (MASE) and mean absolute percentage error (MAPE) were applied to assess the better performance of prediction between the models. Though practically, both models could predict TB incidence, the combined model displayed better performance. For the combined model, the Akaike information criterion (AIC), second-order AIC (AICc) and Bayesian information criterion (BIC) are 288.56, 308.31 and 299.09 respectively, which were lower than the SARIMA model with corresponding values of 329.02, 327.20 and 341.99, respectively. The seasonality trend of TB incidence was forecast to have a slightly increased seasonal TB incidence trend from the SARIMA-NNAR model compared to the single model. The combined model indicated a better TB incidence forecasting with a lower AICc. The model also indicates the need for resolute intervention to reduce infectious disease transmission with co-infection with HIV and other concomitant diseases, and also at festival peak periods.

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.

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.

Aerosol analysis and forecast in the European Centre for Medium-Range Weather Forecasts Integrated Forecast System: 2. Data assimilation

NASA Astrophysics Data System (ADS)

Benedetti, A.; Morcrette, J.-J.; Boucher, O.; Dethof, A.; Engelen, R. J.; Fisher, M.; Flentje, H.; Huneeus, N.; Jones, L.; Kaiser, J. W.; Kinne, S.; Mangold, A.; Razinger, M.; Simmons, A. J.; Suttie, M.

2009-07-01

This study presents the new aerosol assimilation system, developed at the European Centre for Medium-Range Weather Forecasts, for the Global and regional Earth-system Monitoring using Satellite and in-situ data (GEMS) project. The aerosol modeling and analysis system is fully integrated in the operational four-dimensional assimilation apparatus. Its purpose is to produce aerosol forecasts and reanalyses of aerosol fields using optical depth data from satellite sensors. This paper is the second of a series which describes the GEMS aerosol effort. It focuses on the theoretical architecture and practical implementation of the aerosol assimilation system. It also provides a discussion of the background errors and observations errors for the aerosol fields, and presents a subset of results from the 2-year reanalysis which has been run for 2003 and 2004 using data from the Moderate Resolution Imaging Spectroradiometer on the Aqua and Terra satellites. Independent data sets are used to show that despite some compromises that have been made for feasibility reasons in regards to the choice of control variable and error characteristics, the analysis is very skillful in drawing to the observations and in improving the forecasts of aerosol optical depth.

Financial forecasts accuracy in Brazil's social security system.

PubMed

Silva, Carlos Patrick Alves da; Puty, Claudio Alberto Castelo Branco; Silva, Marcelino Silva da; Carvalho, Solon Venâncio de; Francês, Carlos Renato Lisboa

2017-01-01

Long-term social security statistical forecasts produced and disseminated by the Brazilian government aim to provide accurate results that would serve as background information for optimal policy decisions. These forecasts are being used as support for the government's proposed pension reform that plans to radically change the Brazilian Constitution insofar as Social Security is concerned. However, the reliability of official results is uncertain since no systematic evaluation of these forecasts has ever been published by the Brazilian government or anyone else. This paper aims to present a study of the accuracy and methodology of the instruments used by the Brazilian government to carry out long-term actuarial forecasts. We base our research on an empirical and probabilistic analysis of the official models. Our empirical analysis shows that the long-term Social Security forecasts are systematically biased in the short term and have significant errors that render them meaningless in the long run. Moreover, the low level of transparency in the methods impaired the replication of results published by the Brazilian Government and the use of outdated data compromises forecast results. In the theoretical analysis, based on a mathematical modeling approach, we discuss the complexity and limitations of the macroeconomic forecast through the computation of confidence intervals. We demonstrate the problems related to error measurement inherent to any forecasting process. We then extend this exercise to the computation of confidence intervals for Social Security forecasts. This mathematical exercise raises questions about the degree of reliability of the Social Security forecasts.

Financial forecasts accuracy in Brazil’s social security system

PubMed Central

2017-01-01

Long-term social security statistical forecasts produced and disseminated by the Brazilian government aim to provide accurate results that would serve as background information for optimal policy decisions. These forecasts are being used as support for the government’s proposed pension reform that plans to radically change the Brazilian Constitution insofar as Social Security is concerned. However, the reliability of official results is uncertain since no systematic evaluation of these forecasts has ever been published by the Brazilian government or anyone else. This paper aims to present a study of the accuracy and methodology of the instruments used by the Brazilian government to carry out long-term actuarial forecasts. We base our research on an empirical and probabilistic analysis of the official models. Our empirical analysis shows that the long-term Social Security forecasts are systematically biased in the short term and have significant errors that render them meaningless in the long run. Moreover, the low level of transparency in the methods impaired the replication of results published by the Brazilian Government and the use of outdated data compromises forecast results. In the theoretical analysis, based on a mathematical modeling approach, we discuss the complexity and limitations of the macroeconomic forecast through the computation of confidence intervals. We demonstrate the problems related to error measurement inherent to any forecasting process. We then extend this exercise to the computation of confidence intervals for Social Security forecasts. This mathematical exercise raises questions about the degree of reliability of the Social Security forecasts. PMID:28859172

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

Assessing the Predictability of Convection using Ensemble Data Assimilation of Simulated Radar Observations in an LETKF system

NASA Astrophysics Data System (ADS)

Lange, Heiner; Craig, George

2014-05-01

This study uses the Local Ensemble Transform Kalman Filter (LETKF) to perform storm-scale Data Assimilation of simulated Doppler radar observations into the non-hydrostatic, convection-permitting COSMO model. In perfect model experiments (OSSEs), it is investigated how the limited predictability of convective storms affects precipitation forecasts. The study compares a fine analysis scheme with small RMS errors to a coarse scheme that allows for errors in position, shape and occurrence of storms in the ensemble. The coarse scheme uses superobservations, a coarser grid for analysis weights, a larger localization radius and larger observation error that allow a broadening of the Gaussian error statistics. Three hour forecasts of convective systems (with typical lifetimes exceeding 6 hours) from the detailed analyses of the fine scheme are found to be advantageous to those of the coarse scheme during the first 1-2 hours, with respect to the predicted storm positions. After 3 hours in the convective regime used here, the forecast quality of the two schemes appears indiscernible, judging by RMSE and verification methods for rain-fields and objects. It is concluded that, for operational assimilation systems, the analysis scheme might not necessarily need to be detailed to the grid scale of the model. Depending on the forecast lead time, and on the presence of orographic or synoptic forcing that enhance the predictability of storm occurrences, analyses from a coarser scheme might suffice.

Monthly ENSO Forecast Skill and Lagged Ensemble Size

PubMed Central

DelSole, T.; Tippett, M.K.; Pegion, K.

2018-01-01

Abstract The mean square error (MSE) of a lagged ensemble of monthly forecasts of the Niño 3.4 index from the Climate Forecast System (CFSv2) is examined with respect to ensemble size and configuration. Although the real‐time forecast is initialized 4 times per day, it is possible to infer the MSE for arbitrary initialization frequency and for burst ensembles by fitting error covariances to a parametric model and then extrapolating to arbitrary ensemble size and initialization frequency. Applying this method to real‐time forecasts, we find that the MSE consistently reaches a minimum for a lagged ensemble size between one and eight days, when four initializations per day are included. This ensemble size is consistent with the 8–10 day lagged ensemble configuration used operationally. Interestingly, the skill of both ensemble configurations is close to the estimated skill of the infinite ensemble. The skill of the weighted, lagged, and burst ensembles are found to be comparable. Certain unphysical features of the estimated error growth were tracked down to problems with the climatology and data discontinuities. PMID:29937973

Monthly ENSO Forecast Skill and Lagged Ensemble Size

NASA Astrophysics Data System (ADS)

Trenary, L.; DelSole, T.; Tippett, M. K.; Pegion, K.

2018-04-01

The mean square error (MSE) of a lagged ensemble of monthly forecasts of the Niño 3.4 index from the Climate Forecast System (CFSv2) is examined with respect to ensemble size and configuration. Although the real-time forecast is initialized 4 times per day, it is possible to infer the MSE for arbitrary initialization frequency and for burst ensembles by fitting error covariances to a parametric model and then extrapolating to arbitrary ensemble size and initialization frequency. Applying this method to real-time forecasts, we find that the MSE consistently reaches a minimum for a lagged ensemble size between one and eight days, when four initializations per day are included. This ensemble size is consistent with the 8-10 day lagged ensemble configuration used operationally. Interestingly, the skill of both ensemble configurations is close to the estimated skill of the infinite ensemble. The skill of the weighted, lagged, and burst ensembles are found to be comparable. Certain unphysical features of the estimated error growth were tracked down to problems with the climatology and data discontinuities.

Ensemble forecasting of short-term system scale irrigation demands using real-time flow data and numerical weather predictions

NASA Astrophysics Data System (ADS)

Perera, Kushan C.; Western, Andrew W.; Robertson, David E.; George, Biju; Nawarathna, Bandara

2016-06-01

Irrigation demands fluctuate in response to weather variations and a range of irrigation management decisions, which creates challenges for water supply system operators. This paper develops a method for real-time ensemble forecasting of irrigation demand and applies it to irrigation command areas of various sizes for lead times of 1 to 5 days. The ensemble forecasts are based on a deterministic time series model coupled with ensemble representations of the various inputs to that model. Forecast inputs include past flow, precipitation, and potential evapotranspiration. These inputs are variously derived from flow observations from a modernized irrigation delivery system; short-term weather forecasts derived from numerical weather prediction models and observed weather data available from automatic weather stations. The predictive performance for the ensemble spread of irrigation demand was quantified using rank histograms, the mean continuous rank probability score (CRPS), the mean CRPS reliability and the temporal mean of the ensemble root mean squared error (MRMSE). The mean forecast was evaluated using root mean squared error (RMSE), Nash-Sutcliffe model efficiency (NSE) and bias. The NSE values for evaluation periods ranged between 0.96 (1 day lead time, whole study area) and 0.42 (5 days lead time, smallest command area). Rank histograms and comparison of MRMSE, mean CRPS, mean CRPS reliability and RMSE indicated that the ensemble spread is generally a reliable representation of the forecast uncertainty for short lead times but underestimates the uncertainty for long lead times.

Tracking signal test to monitor an intelligent time series forecasting model

NASA Astrophysics Data System (ADS)

Deng, Yan; Jaraiedi, Majid; Iskander, Wafik H.

2004-03-01

Extensive research has been conducted on the subject of Intelligent Time Series forecasting, including many variations on the use of neural networks. However, investigation of model adequacy over time, after the training processes is completed, remains to be fully explored. In this paper we demonstrate a how a smoothed error tracking signals test can be incorporated into a neuro-fuzzy model to monitor the forecasting process and as a statistical measure for keeping the forecasting model up-to-date. The proposed monitoring procedure is effective in the detection of nonrandom changes, due to model inadequacy or lack of unbiasedness in the estimation of model parameters and deviations from the existing patterns. This powerful detection device will result in improved forecast accuracy in the long run. An example data set has been used to demonstrate the application of the proposed method.

Application of Bred Vectors To Data Assimilation

NASA Astrophysics Data System (ADS)

Corazza, M.; Kalnay, E.; Patil, Dj

We introduced a statistic, the BV-dimension, to measure the effective local finite-time dimensionality of the atmosphere. We show that this dimension is often quite low, and suggest that this finding has important implications for data assimilation and the accuracy of weather forecasting (Patil et al, 2001). The original database for this study was the forecasts of the NCEP global ensemble forecasting system. The initial differences between the control forecast and the per- turbed forecasts are called bred vectors. The control and perturbed initial conditions valid at time t=n(t are evolved using the forecast model until time t=(n+1) (t. The differences between the perturbed and the control forecasts are scaled down to their initial amplitude, and constitute the bred vectors valid at (n+1) (t. Their growth rate is typically about 1.5/day. The bred vectors are similar by construction to leading Lya- punov vectors except that they have small but finite amplitude, and they are valid at finite times. The original NCEP ensemble data set has 5 independent bred vectors. We define a local bred vector at each grid point by choosing the 5 by 5 grid points centered at the grid point (a region of about 1100km by 1100km), and using the north-south and east- west velocity components at 500mb pressure level to form a 50 dimensional column vector. Since we have k=5 global bred vectors, we also have k local bred vectors at each grid point. We estimate the effective dimensionality of the subspace spanned by the local bred vectors by performing a singular value decomposition (EOF analysis). The k local bred vector columns form a 50xk matrix M. The singular values s(i) of M measure the extent to which the k column unit vectors making up the matrix M point in the direction of v(i). We define the bred vector dimension as BVDIM={Sum[s(i)]}^2/{Sum[s(i)]^2} For example, if 4 out of the 5 vectors lie along v, and one lies along v, the BV- dimension would be BVDIM[sqrt(4), 1, 0,0,0]=1.8, less than 2 because one direction is more dominant than the other in representing the original data. The results (Patil et al, 2001) show that there are large regions where the bred vectors span a subspace of substantially lower dimension than that of the full space. These low dimensionality regions are dominant in the baroclinic extratropics, typically have a lifetime of 3-7 days, have a well-defined horizontal and vertical structure that spans 1 most of the atmosphere, and tend to move eastward. New results with a large number of ensemble members confirm these results and indicate that the low dimensionality regions are quite robust, and depend only on the verification time (i.e., the underlying flow). Corazza et al (2001) have performed experiments with a data assimilation system based on a quasi-geostrophic model and simulated observations (Morss, 1999, Hamill et al, 2000). A 3D-variational data assimilation scheme for a quasi-geostrophic chan- nel model is used to study the structure of the background error and its relationship to the corresponding bred vectors. The "true" evolution of the model atmosphere is defined by an integration of the model and "rawinsonde observations" are simulated by randomly perturbing the true state at fixed locations. It is found that after 3-5 days the bred vectors develop well organized structures which are very similar for the two different norms considered in this paper (potential vorticity norm and streamfunction norm). The results show that the bred vectors do indeed represent well the characteristics of the data assimilation forecast errors, and that the subspace of bred vectors contains most of the forecast error, except in areas where the forecast errors are small. For example, the angle between the 6hr forecast error and the subspace spanned by 10 bred vectors is less than 10o over 90% of the domain, indicating a pattern correlation of more than 98.5% between the forecast error and its projection onto the bred vector subspace. The presence of low-dimensional regions in the perturbations of the basic flow has important implications for data assimilation. At any given time, there is a difference between the true atmospheric state and the model forecast. Assuming that model er- rors are not the dominant source of errors, in a region of low BV-dimensionality the difference between the true state and the forecast should lie substantially in the low dimensional unstable subspace of the few bred vectors that contribute most strongly to the low BV-dimension. This information should yield a substantial improvement in the forecast: the data assimilation algorithm should correct the model state by moving it closer to the observations along the unstable subspace, since this is where the true state most likely lies. Preliminary experiments have been conducted with the quasi-geostrophic data assim- ilation system testing whether it is possible to add "errors of the day" based on bred vectors to the standard (constant) 3D-Var background error covariance in order to capture these important errors. The results are extremely encouraging, indicating a significant reduction (about 40%) in the analysis errors at a very low computational cost. References: 2 Corazza, M., E. Kalnay, DJ Patil, R. Morss, M Cai, I. Szunyogh, BR Hunt, E Ott and JA Yorke, 2001: Use of the breeding technique to estimate the structure of the analysis "errors of the day". Submitted to Nonlinear Processes in Geophysics. Hamill, T.M., Snyder, C., and Morss, R.E., 2000: A Comparison of Probabilistic Fore- casts from Bred, Singular-Vector and Perturbed Observation Ensembles, Mon. Wea. Rev., 128, 1835­1851. Kalnay, E., and Z. Toth, 1994: Removing growing errors in the analysis cycle. Preprints of the Tenth Conference on Numerical Weather Prediction, Amer. Meteor. Soc., 1994, 212-215. Morss, R. E., 1999: Adaptive observations: Idealized sampling strategies for improv- ing numerical weather prediction. PHD thesis, Massachussetts Institute of technology, 225pp. Patil, D. J. S., B. R. Hunt, E. Kalnay, J. A. Yorke, and E. Ott., 2001: Local Low Dimensionality of Atmospheric Dynamics. Phys. Rev. Lett., 86, 5878. 3

Impact of variational assimilation using multivariate background error covariances on the simulation of monsoon depressions over India

NASA Astrophysics Data System (ADS)

Dhanya, M.; Chandrasekar, A.

2016-02-01

The background error covariance structure influences a variational data assimilation system immensely. The simulation of a weather phenomenon like monsoon depression can hence be influenced by the background correlation information used in the analysis formulation. The Weather Research and Forecasting Model Data assimilation (WRFDA) system includes an option for formulating multivariate background correlations for its three-dimensional variational (3DVar) system (cv6 option). The impact of using such a formulation in the simulation of three monsoon depressions over India is investigated in this study. Analysis and forecast fields generated using this option are compared with those obtained using the default formulation for regional background error correlations (cv5) in WRFDA and with a base run without any assimilation. The model rainfall forecasts are compared with rainfall observations from the Tropical Rainfall Measurement Mission (TRMM) and the other model forecast fields are compared with a high-resolution analysis as well as with European Centre for Medium-Range Weather Forecasts (ECMWF) ERA-Interim reanalysis. The results of the study indicate that inclusion of additional correlation information in background error statistics has a moderate impact on the vertical profiles of relative humidity, moisture convergence, horizontal divergence and the temperature structure at the depression centre at the analysis time of the cv5/cv6 sensitivity experiments. Moderate improvements are seen in two of the three depressions investigated in this study. An improved thermodynamic and moisture structure at the initial time is expected to provide for improved rainfall simulation. The results of the study indicate that the skill scores of accumulated rainfall are somewhat better for the cv6 option as compared to the cv5 option for at least two of the three depression cases studied, especially at the higher threshold levels. Considering the importance of utilising improved flow-dependent correlation structures for efficient data assimilation, the need for more studies on the impact of background error covariances is obvious.

Experiments with a three-dimensional statistical objective analysis scheme using FGGE data

NASA Technical Reports Server (NTRS)

Baker, Wayman E.; Bloom, Stephen C.; Woollen, John S.; Nestler, Mark S.; Brin, Eugenia

1987-01-01

A three-dimensional (3D), multivariate, statistical objective analysis scheme (referred to as optimum interpolation or OI) has been developed for use in numerical weather prediction studies with the FGGE data. Some novel aspects of the present scheme include: (1) a multivariate surface analysis over the oceans, which employs an Ekman balance instead of the usual geostrophic relationship, to model the pressure-wind error cross correlations, and (2) the capability to use an error correlation function which is geographically dependent. A series of 4-day data assimilation experiments are conducted to examine the importance of some of the key features of the OI in terms of their effects on forecast skill, as well as to compare the forecast skill using the OI with that utilizing a successive correction method (SCM) of analysis developed earlier. For the three cases examined, the forecast skill is found to be rather insensitive to varying the error correlation function geographically. However, significant differences are noted between forecasts from a two-dimensional (2D) version of the OI and those from the 3D OI, with the 3D OI forecasts exhibiting better forecast skill. The 3D OI forecasts are also more accurate than those from the SCM initial conditions. The 3D OI with the multivariate oceanic surface analysis was found to produce forecasts which were slightly more accurate, on the average, than a univariate version.

Forecasting space weather over short horizons: Revised and updated estimates

NASA Astrophysics Data System (ADS)

Reikard, Gordon

2018-07-01

Space weather reflects multiple causes. There is a clear influence for the sun on the near-earth environment. Solar activity shows evidence of chaotic properties, implying that prediction may be limited beyond short horizons. At the same time, geomagnetic activity also reflects the rotation of the earth's core, and local currents in the ionosphere. The combination of influences means that geomagnetic indexes behave like multifractals, exhibiting nonlinear variability, with intermittent outliers. This study tests a range of models: regressions, neural networks, and a frequency domain algorithm. Forecasting tests are run for sunspots and irradiance from 1820 onward, for the Aa geomagnetic index from 1868 onward, and the Am index from 1959 onward, over horizons of 1-7 days. For irradiance and sunspots, persistence actually does better over short horizons. None of the other models really dominate. For the geomagnetic indexes, the persistence method does badly, while the neural net also shows large errors. The remaining models all achieve about the same level of accuracy. The errors are in the range of 48% at 1 day, and 54% at all later horizons. Additional tests are run over horizons of 1-4 weeks. At 1 week, the best models reduce the error to about 35%. Over horizons of four weeks, the model errors increase. The findings are somewhat pessimistic. Over short horizons, geomagnetic activity exhibits so much random variation that the forecast errors are extremely high. Over slightly longer horizons, there is some improvement from estimating in the frequency domain, but not a great deal. Including solar activity in the models does not yield any improvement in accuracy.

The effect of domain length and parameter estimation on observation impact in data assimilation for flood inundation forecasting.

NASA Astrophysics Data System (ADS)

Cooper, Elizabeth; Dance, Sarah; Garcia-Pintado, Javier; Nichols, Nancy; Smith, Polly

2017-04-01

Timely and accurate inundation forecasting provides vital information about the behaviour of fluvial flood water, enabling mitigating actions to be taken by residents and emergency services. Data assimilation is a powerful mathematical technique for combining forecasts from hydrodynamic models with observations to produce a more accurate forecast. We discuss the effect of both domain size and channel friction parameter estimation on observation impact in data assimilation for inundation forecasting. Numerical shallow water simulations are carried out in a simple, idealized river channel topography. Data assimilation is performed using an Ensemble Transform Kalman Filter (ETKF) and synthetic observations of water depth in identical twin experiments. We show that reinitialising the numerical inundation model with corrected water levels after an assimilation can cause an initialisation shock if a hydrostatic assumption is made, leading to significant degradation of the forecast for several hours immediately following an assimilation. We demonstrate an effective and novel method for dealing with this. We find that using data assimilation to combine observations of water depth with forecasts from a hydrodynamic model corrects the forecast very effectively at time of the observations. In agreement with other authors we find that the corrected forecast then moves quickly back to the open loop forecast which does not take the observations into account. Our investigations show that the time taken for the forecast to decay back to the open loop case depends on the length of the domain of interest when only water levels are corrected. This is because the assimilation corrects water depths in all parts of the domain, even when observations are only available in one area. Error growth in the forecast step then starts at the upstream part of the domain and propagates downstream. The impact of the observations is therefore longer-lived in a longer domain. We have found that the upstream-downstream pattern of error growth can be due to incorrect friction parameter specification, rather than errors in inflow as shown elsewhere. Our results show that joint state-parameter estimation can recover accurate values for the parameter controlling channel friction processes in the model, even when observations of water level are only available on part of the flood plain. Correcting water levels and the channel friction parameter together leads to a large improvement in the forecast water levels at all simulation times. The impact of the observations is therefore much greater when the channel friction parameter is corrected along with water levels. We find that domain length effects disappear for joint state-parameter estimation.

Comparison of ensemble post-processing approaches, based on empirical and dynamical error modelisation of rainfall-runoff model forecasts

NASA Astrophysics Data System (ADS)

Chardon, J.; Mathevet, T.; Le Lay, M.; Gailhard, J.

2012-04-01

In the context of a national energy company (EDF : Electricité de France), hydro-meteorological forecasts are necessary to ensure safety and security of installations, meet environmental standards and improve water ressources management and decision making. Hydrological ensemble forecasts allow a better representation of meteorological and hydrological forecasts uncertainties and improve human expertise of hydrological forecasts, which is essential to synthesize available informations, coming from different meteorological and hydrological models and human experience. An operational hydrological ensemble forecasting chain has been developed at EDF since 2008 and is being used since 2010 on more than 30 watersheds in France. This ensemble forecasting chain is characterized ensemble pre-processing (rainfall and temperature) and post-processing (streamflow), where a large human expertise is solicited. The aim of this paper is to compare 2 hydrological ensemble post-processing methods developed at EDF in order improve ensemble forecasts reliability (similar to Monatanari &Brath, 2004; Schaefli et al., 2007). The aim of the post-processing methods is to dress hydrological ensemble forecasts with hydrological model uncertainties, based on perfect forecasts. The first method (called empirical approach) is based on a statistical modelisation of empirical error of perfect forecasts, by streamflow sub-samples of quantile class and lead-time. The second method (called dynamical approach) is based on streamflow sub-samples of quantile class and streamflow variation, and lead-time. On a set of 20 watersheds used for operational forecasts, results show that both approaches are necessary to ensure a good post-processing of hydrological ensemble, allowing a good improvement of reliability, skill and sharpness of ensemble forecasts. The comparison of the empirical and dynamical approaches shows the limits of the empirical approach which is not able to take into account hydrological dynamic and processes, i. e. sample heterogeneity. For a same streamflow range corresponds different processes such as rising limbs or recession, where uncertainties are different. The dynamical approach improves reliability, skills and sharpness of forecasts and globally reduces confidence intervals width. When compared in details, the dynamical approach allows a noticeable reduction of confidence intervals during recessions where uncertainty is relatively lower and a slight increase of confidence intervals during rising limbs or snowmelt where uncertainty is greater. The dynamic approach, validated by forecaster's experience that considered the empirical approach not discriminative enough, improved forecaster's confidence and communication of uncertainties. Montanari, A. and Brath, A., (2004). A stochastic approach for assessing the uncertainty of rainfall-runoff simulations. Water Resources Research, 40, W01106, doi:10.1029/2003WR002540. Schaefli, B., Balin Talamba, D. and Musy, A., (2007). Quantifying hydrological modeling errors through a mixture of normal distributions. Journal of Hydrology, 332, 303-315.

Forecasting the onset of an allergic risk to poaceae in Nancy and Strasbourg (France) with different methods.

PubMed

Cassagne, E; Caillaud, P D; Besancenot, J P; Thibaudon, M

2007-10-01

Pollen of Poaceae is among the most allergenic pollen in Europe with pollen of birch. It is therefore useful to elaborate models to help pollen allergy sufferers. The objective of this study was to construct forecast models that could predict the first day characterized by a certain level of allergic risk called here the Starting Date of the Allergic Risk (SDAR). Models result from four forecast methods (three summing and one multiple regression analysis) used in the literature. They were applied on Nancy and Strasbourg from 1988 to 2005 and were tested on 2006. Mean Absolute Error and Actual forecast ability test are the parameters used to choose best models, assess and compare their accuracy. It was found, on the whole, that all the models presented a good forecast accuracy which was equivalent. They were all reliable and were used in order to forecast the SDAR in 2006 with contrasting results in forecasting precision.

Ensemble-sensitivity Analysis Based Observation Targeting for Mesoscale Convection Forecasts and Factors Influencing Observation-Impact Prediction

NASA Astrophysics Data System (ADS)

Hill, A.; Weiss, C.; Ancell, B. C.

2017-12-01

The basic premise of observation targeting is that additional observations, when gathered and assimilated with a numerical weather prediction (NWP) model, will produce a more accurate forecast related to a specific phenomenon. Ensemble-sensitivity analysis (ESA; Ancell and Hakim 2007; Torn and Hakim 2008) is a tool capable of accurately estimating the proper location of targeted observations in areas that have initial model uncertainty and large error growth, as well as predicting the reduction of forecast variance due to the assimilated observation. ESA relates an ensemble of NWP model forecasts, specifically an ensemble of scalar forecast metrics, linearly to earlier model states. A thorough investigation is presented to determine how different factors of the forecast process are impacting our ability to successfully target new observations for mesoscale convection forecasts. Our primary goals for this work are to determine: (1) If targeted observations hold more positive impact over non-targeted (i.e. randomly chosen) observations; (2) If there are lead-time constraints to targeting for convection; (3) How inflation, localization, and the assimilation filter influence impact prediction and realized results; (4) If there exist differences between targeted observations at the surface versus aloft; and (5) how physics errors and nonlinearity may augment observation impacts.Ten cases of dryline-initiated convection between 2011 to 2013 are simulated within a simplified OSSE framework and presented here. Ensemble simulations are produced from a cycling system that utilizes the Weather Research and Forecasting (WRF) model v3.8.1 within the Data Assimilation Research Testbed (DART). A "truth" (nature) simulation is produced by supplying a 3-km WRF run with GFS analyses and integrating the model forward 90 hours, from the beginning of ensemble initialization through the end of the forecast. Target locations for surface and radiosonde observations are computed 6, 12, and 18 hours into the forecast based on a chosen scalar forecast response metric (e.g., maximum reflectivity at convection initiation). A variety of experiments are designed to achieve the aforementioned goals and will be presented, along with their results, detailing the feasibility of targeting for mesoscale convection forecasts.

The NRL relocatable ocean/acoustic ensemble forecast system

NASA Astrophysics Data System (ADS)

Rowley, C.; Martin, P.; Cummings, J.; Jacobs, G.; Coelho, E.; Bishop, C.; Hong, X.; Peggion, G.; Fabre, J.

2009-04-01

A globally relocatable regional ocean nowcast/forecast system has been developed to support rapid implementation of new regional forecast domains. The system is in operational use at the Naval Oceanographic Office for a growing number of regional and coastal implementations. The new system is the basis for an ocean acoustic ensemble forecast and adaptive sampling capability. We present an overview of the forecast system and the ocean ensemble and adaptive sampling methods. The forecast system consists of core ocean data analysis and forecast modules, software for domain configuration, surface and boundary condition forcing processing, and job control, and global databases for ocean climatology, bathymetry, tides, and river locations and transports. The analysis component is the Navy Coupled Ocean Data Assimilation (NCODA) system, a 3D multivariate optimum interpolation system that produces simultaneous analyses of temperature, salinity, geopotential, and vector velocity using remotely-sensed SST, SSH, and sea ice concentration, plus in situ observations of temperature, salinity, and currents from ships, buoys, XBTs, CTDs, profiling floats, and autonomous gliders. The forecast component is the Navy Coastal Ocean Model (NCOM). The system supports one-way nesting and multiple assimilation methods. The ensemble system uses the ensemble transform technique with error variance estimates from the NCODA analysis to represent initial condition error. Perturbed surface forcing or an atmospheric ensemble is used to represent errors in surface forcing. The ensemble transform Kalman filter is used to assess the impact of adaptive observations on future analysis and forecast uncertainty for both ocean and acoustic properties.

Wind Power Forecasting Error Distributions: An International Comparison; Preprint

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

Hodge, B. M.; Lew, D.; Milligan, M.

2012-09-01

Wind power forecasting is expected to be an important enabler for greater penetration of wind power into electricity systems. Because no wind forecasting system is perfect, a thorough understanding of the errors that do occur can be critical to system operation functions, such as the setting of operating reserve levels. This paper provides an international comparison of the distribution of wind power forecasting errors from operational systems, based on real forecast data. The paper concludes with an assessment of similarities and differences between the errors observed in different locations.

Verification of the skill of numerical weather prediction models in forecasting rainfall from U.S. landfalling tropical cyclones

NASA Astrophysics Data System (ADS)

Luitel, Beda; Villarini, Gabriele; Vecchi, Gabriel A.

2018-01-01

The goal of this study is the evaluation of the skill of five state-of-the-art numerical weather prediction (NWP) systems [European Centre for Medium-Range Weather Forecasts (ECMWF), UK Met Office (UKMO), National Centers for Environmental Prediction (NCEP), China Meteorological Administration (CMA), and Canadian Meteorological Center (CMC)] in forecasting rainfall from North Atlantic tropical cyclones (TCs). Analyses focus on 15 North Atlantic TCs that made landfall along the U.S. coast over the 2007-2012 period. As reference data we use gridded rainfall provided by the Climate Prediction Center (CPC). We consider forecast lead-times up to five days. To benchmark the skill of these models, we consider rainfall estimates from one radar-based (Stage IV) and four satellite-based [Tropical Rainfall Measuring Mission - Multi-satellite Precipitation Analysis (TMPA, both real-time and research version); Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN); the CPC MORPHing Technique (CMORPH)] rainfall products. Daily and storm total rainfall fields from each of these remote sensing products are compared to the reference data to obtain information about the range of errors we can expect from "observational data." The skill of the NWP models is quantified: (1) by visual examination of the distribution of the errors in storm total rainfall for the different lead-times, and numerical examination of the first three moments of the error distribution; (2) relative to climatology at the daily scale. Considering these skill metrics, we conclude that the NWP models can provide skillful forecasts of TC rainfall with lead-times up to 48 h, without a consistently best or worst NWP model.

SCADA-based Operator Support System for Power Plant Equipment Fault Forecasting

NASA Astrophysics Data System (ADS)

Mayadevi, N.; Ushakumari, S. S.; Vinodchandra, S. S.

2014-12-01

Power plant equipment must be monitored closely to prevent failures from disrupting plant availability. Online monitoring technology integrated with hybrid forecasting techniques can be used to prevent plant equipment faults. A self learning rule-based expert system is proposed in this paper for fault forecasting in power plants controlled by supervisory control and data acquisition (SCADA) system. Self-learning utilizes associative data mining algorithms on the SCADA history database to form new rules that can dynamically update the knowledge base of the rule-based expert system. In this study, a number of popular associative learning algorithms are considered for rule formation. Data mining results show that the Tertius algorithm is best suited for developing a learning engine for power plants. For real-time monitoring of the plant condition, graphical models are constructed by K-means clustering. To build a time-series forecasting model, a multi layer preceptron (MLP) is used. Once created, the models are updated in the model library to provide an adaptive environment for the proposed system. Graphical user interface (GUI) illustrates the variation of all sensor values affecting a particular alarm/fault, as well as the step-by-step procedure for avoiding critical situations and consequent plant shutdown. The forecasting performance is evaluated by computing the mean absolute error and root mean square error of the predictions.

An Approach to Remove the Systematic Bias from the Storm Surge forecasts in the Venice Lagoon

NASA Astrophysics Data System (ADS)

Canestrelli, A.

2017-12-01

In this work a novel approach is proposed for removing the systematic bias from the storm surge forecast computed by a two-dimensional shallow-water model. The model covers both the Adriatic and Mediterranean seas and provides the forecast at the entrance of the Venice Lagoon. The wind drag coefficient at the water-air interface is treated as a calibration parameter, with a different value for each range of wind velocities and wind directions. This sums up to a total of 16-64 parameters to be calibrated, depending on the chosen resolution. The best set of parameters is determined by means of an optimization procedure, which minimizes the RMS error between measured and modeled water level in Venice for the period 2011-2015. It is shown that a bias is present, for which the peaks of wind velocities provided by the weather forecast are largely underestimated, and that the calibration procedure removes this bias. When the calibrated model is used to reproduce events not included in the calibration dataset, the forecast error is strongly reduced, thus confirming the quality of our procedure. The proposed approach it is not site-specific and could be applied to different situations, such as storm surges caused by intense hurricanes.

[Application of artificial neural networks in forecasting the number of circulatory system diseases death toll].

PubMed

Zhang, Ying; Shao, Yi; Shang, Kezheng; Wang, Shigong; Wang, Jinyan

2014-09-01

Set up the model of forecasting the number of circulatorys death toll based on back-propagation (BP) artificial neural networks discuss the relationship between the circulatory system diseases death toll meteorological factors and ambient air pollution. The data of tem deaths, meteorological factors, and ambient air pollution within the m 2004 to 2009 in Nanjing were collected. On the basis of analyzing the ficient between CSDDT meteorological factors and ambient air pollution, leutral network model of CSDDT was built for 2004 - 2008 based on factors and ambient air pollution within the same time, and the data of 2009 est the predictive power of the model. There was a closely system diseases relationship between meteorological factors, ambient air pollution and the circulatory system diseases death toll. The ANN model structure was 17 -16 -1, 17 input notes, 16 hidden notes and 1 output note. The training precision was 0. 005 and the final error was 0. 004 999 42 after 487 training steps. The results of forecast show that predict accuracy over 78. 62%. This method is easy to be finished with smaller error, and higher ability on circulatory system death toll on independent prediction, which can provide a new method for forecasting medical-meteorological forecast and have the value of further research.

Beyond Rating Curves: Time Series Models for in-Stream Turbidity Prediction

NASA Astrophysics Data System (ADS)

Wang, L.; Mukundan, R.; Zion, M.; Pierson, D. C.

2012-12-01

The New York City Department of Environmental Protection (DEP) manages New York City's water supply, which is comprised of over 20 reservoirs and supplies over 1 billion gallons of water per day to more than 9 million customers. DEP's "West of Hudson" reservoirs located in the Catskill Mountains are unfiltered per a renewable filtration avoidance determination granted by the EPA. While water quality is usually pristine, high volume storm events occasionally cause the reservoirs to become highly turbid. A logical strategy for turbidity control is to temporarily remove the turbid reservoirs from service. While effective in limiting delivery of turbid water and reducing the need for in-reservoir alum flocculation, this strategy runs the risk of negatively impacting water supply reliability. Thus, it is advantageous for DEP to understand how long a particular turbidity event will affect their system. In order to understand the duration, intensity and total load of a turbidity event, predictions of future in-stream turbidity values are important. Traditionally, turbidity predictions have been carried out by applying streamflow observations/forecasts to a flow-turbidity rating curve. However, predictions from rating curves are often inaccurate due to inter- and intra-event variability in flow-turbidity relationships. Predictions can be improved by applying an autoregressive moving average (ARMA) time series model in combination with a traditional rating curve. Since 2003, DEP and the Upstate Freshwater Institute have compiled a relatively consistent set of 15-minute turbidity observations at various locations on Esopus Creek above Ashokan Reservoir. Using daily averages of this data and streamflow observations at nearby USGS gauges, flow-turbidity rating curves were developed via linear regression. Time series analysis revealed that the linear regression residuals may be represented using an ARMA(1,2) process. Based on this information, flow-turbidity regressions with ARMA(1,2) errors were fit to the observations. Preliminary model validation exercises at a 30-day forecast horizon show that the ARMA error models generally improve the predictive skill of the linear regression rating curves. Skill seems to vary based on the ambient hydrologic conditions at the onset of the forecast. For example, ARMA error model forecasts issued before a high flow/turbidity event do not show significant improvements over the rating curve approach. However, ARMA error model forecasts issued during the "falling limb" of the hydrograph are significantly more accurate than rating curves for both single day and accumulated event predictions. In order to assist in reservoir operations decisions associated with turbidity events and general water supply reliability, DEP has initiated design of an Operations Support Tool (OST). OST integrates a reservoir operations model with 2D hydrodynamic water quality models and a database compiling near-real-time data sources and hydrologic forecasts. Currently, OST uses conventional flow-turbidity rating curves and hydrologic forecasts for predictive turbidity inputs. Given the improvements in predictive skill over traditional rating curves, the ARMA error models are currently being evaluated as an addition to DEP's Operations Support Tool.

Analysis/forecast experiments with a flow-dependent correlation function using FGGE data

NASA Technical Reports Server (NTRS)

Baker, W. E.; Bloom, S. C.; Carus, H.; Nestler, M. S.

1986-01-01

The use of a flow-dependent correlation function to improve the accuracy of an optimum interpolation (OI) scheme is examined. The development of the correlation function for the OI analysis scheme used for numerical weather prediction is described. The scheme uses a multivariate surface analysis over the oceans to model the pressure-wind error cross-correlation and it has the ability to use an error correlation function that is flow- and geographically-dependent. A series of four-day data assimilation experiments, conducted from January 5-9, 1979, were used to investigate the effect of the different features of the OI scheme (error correlation) on forecast skill for the barotropic lows and highs. The skill of the OI was compared with that of a successive correlation method (SCM) of analysis. It is observed that the largest difference in the correlation statistics occurred in barotropic and baroclinic lows and highs. The comparison reveals that the OI forecasts were more accurate than the SCM forecasts.

Evaluation of the skill of North-American Multi-Model Ensemble (NMME) Global Climate Models in predicting average and extreme precipitation and temperature over the continental USA

NASA Astrophysics Data System (ADS)

Slater, Louise J.; Villarini, Gabriele; Bradley, Allen A.

2016-08-01

This paper examines the forecasting skill of eight Global Climate Models from the North-American Multi-Model Ensemble project (CCSM3, CCSM4, CanCM3, CanCM4, GFDL2.1, FLORb01, GEOS5, and CFSv2) over seven major regions of the continental United States. The skill of the monthly forecasts is quantified using the mean square error skill score. This score is decomposed to assess the accuracy of the forecast in the absence of biases (potential skill) and in the presence of conditional (slope reliability) and unconditional (standardized mean error) biases. We summarize the forecasting skill of each model according to the initialization month of the forecast and lead time, and test the models' ability to predict extended periods of extreme climate conducive to eight `billion-dollar' historical flood and drought events. Results indicate that the most skillful predictions occur at the shortest lead times and decline rapidly thereafter. Spatially, potential skill varies little, while actual model skill scores exhibit strong spatial and seasonal patterns primarily due to the unconditional biases in the models. The conditional biases vary little by model, lead time, month, or region. Overall, we find that the skill of the ensemble mean is equal to or greater than that of any of the individual models. At the seasonal scale, the drought events are better forecast than the flood events, and are predicted equally well in terms of high temperature and low precipitation. Overall, our findings provide a systematic diagnosis of the strengths and weaknesses of the eight models over a wide range of temporal and spatial scales.

Coupled Data Assimilation in Navy ESPC

NASA Astrophysics Data System (ADS)

Barron, C. N.; Spence, P. L.; Frolov, S.; Rowley, C. D.; Bishop, C. H.; Wei, M.; Ruston, B.; Smedstad, O. M.

2017-12-01

Data assimilation under global coupled Earth System Prediction Capability (ESPC) presents significantly greater challenges than data assimilation in forecast models of a single earth system like the ocean and atmosphere. In forecasts of a single component, data assimilation has broad flexibility in adjusting boundary conditions to reduce forecast errors; coupled ESPC requires consistent simultaneous adjustment of multiple components within the earth system: air, ocean, ice, and others. Data assimilation uses error covariances to express how to consistently adjust model conditions in response to differences between forecasts and observations; in coupled ESPC, these covariances must extend from air to ice to ocean such that changes within one fluid are appropriately balanced with corresponding adjustments in the other components. We show several algorithmic solutions that allow us to resolve these challenges. Specifically, we introduce the interface solver method that augments existing stand-alone systems for ocean and atmosphere by allowing them to be influenced by relevant measurements from the coupled fluid. Plans are outlined for implementing coupled data assimilation within ESPC for the Navy's global coupled model. Preliminary results show the impact of assimilating SST-sensitive radiances in the atmospheric model and first results of hybrid DA in a 1/12 degree model of the global ocean.

Forecasting Inflow and Outflow of Money Currency in East Java Using a Hybrid Exponential Smoothing and Calendar Variation Model

NASA Astrophysics Data System (ADS)

Susanti, Ana; Suhartono; Jati Setyadi, Hario; Taruk, Medi; Haviluddin; Pamilih Widagdo, Putut

2018-03-01

Money currency availability in Bank Indonesia can be examined by inflow and outflow of money currency. The objective of this research is to forecast the inflow and outflow of money currency in each Representative Office (RO) of BI in East Java by using a hybrid exponential smoothing based on state space approach and calendar variation model. Hybrid model is expected to generate more accurate forecast. There are two studies that will be discussed in this research. The first studies about hybrid model using simulation data that contain pattern of trends, seasonal and calendar variation. The second studies about the application of a hybrid model for forecasting the inflow and outflow of money currency in each RO of BI in East Java. The first of results indicate that exponential smoothing model can not capture the pattern calendar variation. It results RMSE values 10 times standard deviation of error. The second of results indicate that hybrid model can capture the pattern of trends, seasonal and calendar variation. It results RMSE values approaching the standard deviation of error. In the applied study, the hybrid model give more accurate forecast for five variables : the inflow of money currency in Surabaya, Malang, Jember and outflow of money currency in Surabaya and Kediri. Otherwise, the time series regression model yields better for three variables : outflow of money currency in Malang, Jember and inflow of money currency in Kediri.

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

Forecasting the Water Demand in Chongqing, China Using a Grey Prediction Model and Recommendations for the Sustainable Development of Urban Water Consumption.

PubMed

Wu, Hua'an; Zeng, Bo; Zhou, Meng

2017-11-15

High accuracy in water demand predictions is an important basis for the rational allocation of city water resources and forms the basis for sustainable urban development. The shortage of water resources in Chongqing, the youngest central municipality in Southwest China, has significantly increased with the population growth and rapid economic development. In this paper, a new grey water-forecasting model (GWFM) was built based on the data characteristics of water consumption. The parameter estimation and error checking methods of the GWFM model were investigated. Then, the GWFM model was employed to simulate the water demands of Chongqing from 2009 to 2015 and forecast it in 2016. The simulation and prediction errors of the GWFM model was checked, and the results show the GWFM model exhibits better simulation and prediction precisions than those of the classical Grey Model with one variable and single order equation GM(1,1) for short and the frequently-used Discrete Grey Model with one variable and single order equation, DGM(1,1) for short. Finally, the water demand in Chongqing from 2017 to 2022 was forecasted, and some corresponding control measures and recommendations were provided based on the prediction results to ensure a viable water supply and promote the sustainable development of the Chongqing economy.

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

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.

Probabilistic forecasts based on radar rainfall uncertainty

NASA Astrophysics Data System (ADS)

Liguori, S.; Rico-Ramirez, M. A.

2012-04-01

The potential advantages resulting from integrating weather radar rainfall estimates in hydro-meteorological forecasting systems is limited by the inherent uncertainty affecting radar rainfall measurements, which is due to various sources of error [1-3]. The improvement of quality control and correction techniques is recognized to play a role for the future improvement of radar-based flow predictions. However, the knowledge of the uncertainty affecting radar rainfall data can also be effectively used to build a hydro-meteorological forecasting system in a probabilistic framework. This work discusses the results of the implementation of a novel probabilistic forecasting system developed to improve ensemble predictions over a small urban area located in the North of England. An ensemble of radar rainfall fields can be determined as the sum of a deterministic component and a perturbation field, the latter being informed by the knowledge of the spatial-temporal characteristics of the radar error assessed with reference to rain-gauges measurements. This approach is similar to the REAL system [4] developed for use in the Southern-Alps. The radar uncertainty estimate can then be propagated with a nowcasting model, used to extrapolate an ensemble of radar rainfall forecasts, which can ultimately drive hydrological ensemble predictions. A radar ensemble generator has been calibrated using radar rainfall data made available from the UK Met Office after applying post-processing and corrections algorithms [5-6]. One hour rainfall accumulations from 235 rain gauges recorded for the year 2007 have provided the reference to determine the radar error. Statistics describing the spatial characteristics of the error (i.e. mean and covariance) have been computed off-line at gauges location, along with the parameters describing the error temporal correlation. A system has then been set up to impose the space-time error properties to stochastic perturbations, generated in real-time at gauges location, and then interpolated back onto the radar domain, in order to obtain probabilistic radar rainfall fields in real time. The deterministic nowcasting model integrated in the STEPS system [7-8] has been used for the purpose of propagating the uncertainty and assessing the benefit of implementing the radar ensemble generator for probabilistic rainfall forecasts and ultimately sewer flow predictions. For this purpose, events representative of different types of precipitation (i.e. stratiform/convective) and significant at the urban catchment scale (i.e. in terms of sewer overflow within the urban drainage system) have been selected. As high spatial/temporal resolution is required to the forecasts for their use in urban areas [9-11], the probabilistic nowcasts have been set up to be produced at 1 km resolution and 5 min intervals. The forecasting chain is completed by a hydrodynamic model of the urban drainage network. The aim of this work is to discuss the implementation of this probabilistic system, which takes into account the radar error to characterize the forecast uncertainty, with consequent potential benefits in the management of urban systems. It will also allow a comparison with previous findings related to the analysis of different approaches to uncertainty estimation and quantification in terms of rainfall [12] and flows at the urban scale [13]. Acknowledgements The authors would like to acknowledge the BADC, the UK Met Office and Dr. Alan Seed from the Australian Bureau of Meteorology for providing the radar data and the nowcasting model. The authors acknowledge the support from the Engineering and Physical Sciences Research Council (EPSRC) via grant EP/I012222/1.

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.

Analysis of Numerical Weather Predictions of Reference Evapotranspiration and Precipitation

NASA Astrophysics Data System (ADS)

Bughici, Theodor; Lazarovitch, Naftali; Fredj, Erick; Tas, Eran

2017-04-01

This study attempts to improve the forecast skill of the evapotranspiration (ET0) and Precipitation for the purpose of crop irrigation management over Israel using the Weather Research and Forecasting (WRF) Model. Optimized crop irrigation, in term of timing and quantities, decreases water and agrochemicals demand. Crop water demands depend on evapotranspiration and precipitation. The common method for computing reference evapotranspiration, for agricultural needs, ET0, is according to the FAO Penman-Monteith equation. The weather variables required for ET0 calculation (air temperature, relative humidity, wind speed and solar irradiance) are estimated by the WRF model. The WRF Model with two-way interacting domains at horizontal resolutions of 27, 9 and 3 km is used in the study. The model prediction was performed in an hourly time resolution and a 3 km spatial resolution, with forecast lead-time of up to four days. The WRF prediction of these variables have been compared against measurements from 29 meteorological stations across Israel for the year 2013. The studied area is small but with strong climatic gradient, diverse topography and variety of synoptic conditions. The forecast skill that was used for forecast validation takes into account the prediction bias, mean absolute error and root mean squared error. The forecast skill of the variables was almost robust to lead time, except for precipitation. The forecast skill was tested across stations with respect to topography and geographic location and for all stations with respect to seasonality and synoptic weather system determined by employing a semi-objective synoptic systems classification to the forecasted days. It was noticeable that forecast skill of some of the variables was deteriorated by seasonality and topography. However, larger impacts in the ET0 skill scores on the forecasted day are achieved by a synoptic based forecast. These results set the basis for increasing the robustness of ET0 to synoptic effects and for more precise crop irrigation over Israel.

Models for short term malaria prediction in Sri Lanka

PubMed Central

Briët, Olivier JT; Vounatsou, Penelope; Gunawardena, Dissanayake M; Galappaththy, Gawrie NL; Amerasinghe, Priyanie H

2008-01-01

Background Malaria in Sri Lanka is unstable and fluctuates in intensity both spatially and temporally. Although the case counts are dwindling at present, given the past history of resurgence of outbreaks despite effective control measures, the control programmes have to stay prepared. The availability of long time series of monitored/diagnosed malaria cases allows for the study of forecasting models, with an aim to developing a forecasting system which could assist in the efficient allocation of resources for malaria control. Methods Exponentially weighted moving average models, autoregressive integrated moving average (ARIMA) models with seasonal components, and seasonal multiplicative autoregressive integrated moving average (SARIMA) models were compared on monthly time series of district malaria cases for their ability to predict the number of malaria cases one to four months ahead. The addition of covariates such as the number of malaria cases in neighbouring districts or rainfall were assessed for their ability to improve prediction of selected (seasonal) ARIMA models. Results The best model for forecasting and the forecasting error varied strongly among the districts. The addition of rainfall as a covariate improved prediction of selected (seasonal) ARIMA models modestly in some districts but worsened prediction in other districts. Improvement by adding rainfall was more frequent at larger forecasting horizons. Conclusion Heterogeneity of patterns of malaria in Sri Lanka requires regionally specific prediction models. Prediction error was large at a minimum of 22% (for one of the districts) for one month ahead predictions. The modest improvement made in short term prediction by adding rainfall as a covariate to these prediction models may not be sufficient to merit investing in a forecasting system for which rainfall data are routinely processed. PMID:18460204

Potential predictability and forecast skill in ensemble climate forecast: the skill-persistence rule

NASA Astrophysics Data System (ADS)

Jin, Y.; Rong, X.; Liu, Z.

2017-12-01

This study investigates the factors that impact the forecast skill for the real world (actual skill) and perfect model (perfect skill) in ensemble climate model forecast with a series of fully coupled general circulation model forecast experiments. It is found that the actual skill of sea surface temperature (SST) in seasonal forecast is substantially higher than the perfect skill on a large part of the tropical oceans, especially the tropical Indian Ocean and the central-eastern Pacific Ocean. The higher actual skill is found to be related to the higher observational SST persistence, suggesting a skill-persistence rule: a higher SST persistence in the real world than in the model could overwhelm the model bias to produce a higher forecast skill for the real world than for the perfect model. The relation between forecast skill and persistence is further examined using a first-order autoregressive model (AR1) analytically for theoretical solutions and numerically for analogue experiments. The AR1 model study shows that the skill-persistence rule is strictly valid in the case of infinite ensemble size, but can be distorted by the sampling error and non-AR1 processes.

The impact of underwater glider observations in the forecast of Hurricane Gonzalo (2014)

NASA Astrophysics Data System (ADS)

Goni, G. J.; Domingues, R. M.; Kim, H. S.; Domingues, R. M.; Halliwell, G. R., Jr.; Bringas, F.; Morell, J. M.; Pomales, L.; Baltes, R.

2017-12-01

The tropical Atlantic basin is one of seven global regions where tropical cyclones (TC) are commonly observed to originate and intensify from June to November. On average, approximately 12 TCs travel through the region every year, frequently affecting coastal, and highly populated areas. In an average year, 2 to 3 of them are categorized as intense hurricanes. Given the appropriate atmospheric conditions, TC intensification has been linked to ocean conditions, such as increased ocean heat content and enhanced salinity stratification near the surface. While errors in hurricane track forecasts have been reduced during the last years, errors in intensity forecasts remain mostly unchanged. Several studies have indicated that the use of in situ observations has the potential to improve the representation of the ocean to correctly initialize coupled hurricane intensity forecast models. However, a sustained in situ ocean observing system in the tropical North Atlantic Ocean and Caribbean Sea dedicated to measuring subsurface thermal and salinity fields in support of TC intensity studies and forecasts has yet to be implemented. Autonomous technologies offer new and cost-effective opportunities to accomplish this objective. We highlight here a partnership effort that utilize underwater gliders to better understand air-sea processes during high wind events, and are particularly geared towards improving hurricane intensity forecasts. Results are presented for Hurricane Gonzalo (2014), where glider observations obtained in the tropical Atlantic: Helped to provide an accurate description of the upper ocean conditions, that included the presence of a low salinity barrier layer; Allowed a detailed analysis of the upper ocean response to hurricane force winds of Gonzalo; Improved the initialization of the ocean in a coupled ocean-atmosphere numerical model; and together with observations from other ocean observing platforms, substantially reduced the error in intensity forecast using the HYCOM-HWRF model. Data collected by this project are transmitted in real-time to the Global Telecommunication System, distributed through the institutional web pages, by the IOOS Glider Data Assembly Center, and by NCEI, and assimilated in real-time numerical weather forecast models.

Implementation of Automatic Clustering Algorithm and Fuzzy Time Series in Motorcycle Sales Forecasting

NASA Astrophysics Data System (ADS)

Rasim; Junaeti, E.; Wirantika, R.

2018-01-01

Accurate forecasting for the sale of a product depends on the forecasting method used. The purpose of this research is to build motorcycle sales forecasting application using Fuzzy Time Series method combined with interval determination using automatic clustering algorithm. Forecasting is done using the sales data of motorcycle sales in the last ten years. Then the error rate of forecasting is measured using Means Percentage Error (MPE) and Means Absolute Percentage Error (MAPE). The results of forecasting in the one-year period obtained in this study are included in good accuracy.

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.

On The Usage Of Fire Smoke Emissions In An Air Quality Forecasting System To Reduce Particular Matter Forecasting Error

NASA Astrophysics Data System (ADS)

Huang, H. C.; Pan, L.; McQueen, J.; Lee, P.; ONeill, S. M.; Ruminski, M.; Shafran, P.; DiMego, G.; Huang, J.; Stajner, I.; Upadhayay, S.; Larkin, N. K.

2016-12-01

Wildfires contribute to air quality problems not only towards primary emissions of particular matters (PM) but also emitted ozone precursor gases that can lead to elevated ozone concentration. Wildfires are unpredictable and can be ignited by natural causes such as lightning or accidently by human negligent behavior such as live cigarette. Although wildfire impacts on the air quality can be studied by collecting fire information after events, it is extremely difficult to predict future occurrence and behavior of wildfires for real-time air quality forecasts. Because of the time constraints of operational air quality forecasting, assumption of future day's fire behavior often have to be made based on observed fire information in the past. The United States (U.S.) NOAA/NWS built the National Air Quality Forecast Capability (NAQFC) based on the U.S. EPA CMAQ to provide air quality forecast guidance (prediction) publicly. State and local forecasters use the forecast guidance to issue air quality alerts in their area. The NAQFC fine particulates (PM2.5) prediction includes emissions from anthropogenic and biogenic sources, as well as natural sources such as dust storms and fires. The fire emission input to the NAQFC is derived from the NOAA NESDIS HMS fire and smoke detection product and the emission module of the US Forest Service BlueSky Smoke Modeling Framework. This study focuses on the error estimation of NAQFC PM2.5 predictions resulting from fire emissions. The comparisons between the NAQFC modeled PM2.5 and the EPA AirNow surface observation show that present operational NAQFC fire emissions assumption can lead to a huge error in PM2.5 prediction as fire emissions are sometimes placed at wrong location and time. This PM2.5 prediction error can be propagated from the fire source in the Northwest U.S. to downstream areas as far as the Southeast U.S. From this study, a new procedure has been identified to minimize the aforementioned error. An additional 24 hours reanalysis-run of NAQFC using same-day observed fire emission are being tested. Preliminary results have shown that this procedure greatly improves the PM2.5 predictions at both nearby and downstream areas from fire sources. The 24 hours reanalysis-run is critical and necessary especially during extreme fire events to provide better PM2.5 predictions.

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.

A Systems Modeling Approach to Forecast Corn Economic Optimum Nitrogen Rate.

PubMed

Puntel, Laila A; Sawyer, John E; Barker, Daniel W; Thorburn, Peter J; Castellano, Michael J; Moore, Kenneth J; VanLoocke, Andrew; Heaton, Emily A; Archontoulis, Sotirios V

2018-01-01

Historically crop models have been used to evaluate crop yield responses to nitrogen (N) rates after harvest when it is too late for the farmers to make in-season adjustments. We hypothesize that the use of a crop model as an in-season forecast tool will improve current N decision-making. To explore this, we used the Agricultural Production Systems sIMulator (APSIM) calibrated with long-term experimental data for central Iowa, USA (16-years in continuous corn and 15-years in soybean-corn rotation) combined with actual weather data up to a specific crop stage and historical weather data thereafter. The objectives were to: (1) evaluate the accuracy and uncertainty of corn yield and economic optimum N rate (EONR) predictions at four forecast times (planting time, 6th and 12th leaf, and silking phenological stages); (2) determine whether the use of analogous historical weather years based on precipitation and temperature patterns as opposed to using a 35-year dataset could improve the accuracy of the forecast; and (3) quantify the value added by the crop model in predicting annual EONR and yields using the site-mean EONR and the yield at the EONR to benchmark predicted values. Results indicated that the mean corn yield predictions at planting time ( R 2 = 0.77) using 35-years of historical weather was close to the observed and predicted yield at maturity ( R 2 = 0.81). Across all forecasting times, the EONR predictions were more accurate in corn-corn than soybean-corn rotation (relative root mean square error, RRMSE, of 25 vs. 45%, respectively). At planting time, the APSIM model predicted the direction of optimum N rates (above, below or at average site-mean EONR) in 62% of the cases examined ( n = 31) with an average error range of ±38 kg N ha -1 (22% of the average N rate). Across all forecast times, prediction error of EONR was about three times higher than yield predictions. The use of the 35-year weather record was better than using selected historical weather years to forecast (RRMSE was on average 3% lower). Overall, the proposed approach of using the crop model as a forecasting tool could improve year-to-year predictability of corn yields and optimum N rates. Further improvements in modeling and set-up protocols are needed toward more accurate forecast, especially for extreme weather years with the most significant economic and environmental cost.

A Systems Modeling Approach to Forecast Corn Economic Optimum Nitrogen Rate

PubMed Central

Puntel, Laila A.; Sawyer, John E.; Barker, Daniel W.; Thorburn, Peter J.; Castellano, Michael J.; Moore, Kenneth J.; VanLoocke, Andrew; Heaton, Emily A.; Archontoulis, Sotirios V.

2018-01-01

Historically crop models have been used to evaluate crop yield responses to nitrogen (N) rates after harvest when it is too late for the farmers to make in-season adjustments. We hypothesize that the use of a crop model as an in-season forecast tool will improve current N decision-making. To explore this, we used the Agricultural Production Systems sIMulator (APSIM) calibrated with long-term experimental data for central Iowa, USA (16-years in continuous corn and 15-years in soybean-corn rotation) combined with actual weather data up to a specific crop stage and historical weather data thereafter. The objectives were to: (1) evaluate the accuracy and uncertainty of corn yield and economic optimum N rate (EONR) predictions at four forecast times (planting time, 6th and 12th leaf, and silking phenological stages); (2) determine whether the use of analogous historical weather years based on precipitation and temperature patterns as opposed to using a 35-year dataset could improve the accuracy of the forecast; and (3) quantify the value added by the crop model in predicting annual EONR and yields using the site-mean EONR and the yield at the EONR to benchmark predicted values. Results indicated that the mean corn yield predictions at planting time (R2 = 0.77) using 35-years of historical weather was close to the observed and predicted yield at maturity (R2 = 0.81). Across all forecasting times, the EONR predictions were more accurate in corn-corn than soybean-corn rotation (relative root mean square error, RRMSE, of 25 vs. 45%, respectively). At planting time, the APSIM model predicted the direction of optimum N rates (above, below or at average site-mean EONR) in 62% of the cases examined (n = 31) with an average error range of ±38 kg N ha−1 (22% of the average N rate). Across all forecast times, prediction error of EONR was about three times higher than yield predictions. The use of the 35-year weather record was better than using selected historical weather years to forecast (RRMSE was on average 3% lower). Overall, the proposed approach of using the crop model as a forecasting tool could improve year-to-year predictability of corn yields and optimum N rates. Further improvements in modeling and set-up protocols are needed toward more accurate forecast, especially for extreme weather years with the most significant economic and environmental cost. PMID:29706974

An ensemble Kalman filter with a high-resolution atmosphere-ocean coupled model for tropical cyclone forecasts

NASA Astrophysics Data System (ADS)

Kunii, M.; Ito, K.; Wada, A.

2015-12-01

An ensemble Kalman filter (EnKF) using a regional mesoscale atmosphere-ocean coupled model was developed to represent the uncertainties of sea surface temperature (SST) in ensemble data assimilation strategies. The system was evaluated through data assimilation cycle experiments over a one-month period from July to August 2014, during which a tropical cyclone as well as severe rainfall events occurred. The results showed that the data assimilation cycle with the coupled model could reproduce SST distributions realistically even without updating SST and salinity during the data assimilation cycle. Therefore, atmospheric variables and radiation applied as a forcing to ocean models can control oceanic variables to some extent in the current data assimilation configuration. However, investigations of the forecast error covariance estimated in EnKF revealed that the correlation between atmospheric and oceanic variables could possibly lead to less flow-dependent error covariance for atmospheric variables owing to the difference in the time scales between atmospheric and oceanic variables. A verification of the analyses showed positive impacts of applying the ocean model to EnKF on precipitation forecasts. The use of EnKF with the coupled model system captured intensity changes of a tropical cyclone better than it did with an uncoupled atmosphere model, even though the impact on the track forecast was negligibly small.

Forecasting influenza in Hong Kong with Google search queries and statistical model fusion.

PubMed

Xu, Qinneng; Gel, Yulia R; Ramirez Ramirez, L Leticia; Nezafati, Kusha; Zhang, Qingpeng; Tsui, Kwok-Leung

2017-01-01

The objective of this study is to investigate predictive utility of online social media and web search queries, particularly, Google search data, to forecast new cases of influenza-like-illness (ILI) in general outpatient clinics (GOPC) in Hong Kong. To mitigate the impact of sensitivity to self-excitement (i.e., fickle media interest) and other artifacts of online social media data, in our approach we fuse multiple offline and online data sources. Four individual models: generalized linear model (GLM), least absolute shrinkage and selection operator (LASSO), autoregressive integrated moving average (ARIMA), and deep learning (DL) with Feedforward Neural Networks (FNN) are employed to forecast ILI-GOPC both one week and two weeks in advance. The covariates include Google search queries, meteorological data, and previously recorded offline ILI. To our knowledge, this is the first study that introduces deep learning methodology into surveillance of infectious diseases and investigates its predictive utility. Furthermore, to exploit the strength from each individual forecasting models, we use statistical model fusion, using Bayesian model averaging (BMA), which allows a systematic integration of multiple forecast scenarios. For each model, an adaptive approach is used to capture the recent relationship between ILI and covariates. DL with FNN appears to deliver the most competitive predictive performance among the four considered individual models. Combing all four models in a comprehensive BMA framework allows to further improve such predictive evaluation metrics as root mean squared error (RMSE) and mean absolute predictive error (MAPE). Nevertheless, DL with FNN remains the preferred method for predicting locations of influenza peaks. The proposed approach can be viewed a feasible alternative to forecast ILI in Hong Kong or other countries where ILI has no constant seasonal trend and influenza data resources are limited. The proposed methodology is easily tractable and computationally efficient.

A comparative study of artificial neural network, adaptive neuro fuzzy inference system and support vector machine for forecasting river flow in the semiarid mountain region

NASA Astrophysics Data System (ADS)

He, Zhibin; Wen, Xiaohu; Liu, Hu; Du, Jun

2014-02-01

Data driven models are very useful for river flow forecasting when the underlying physical relationships are not fully understand, but it is not clear whether these data driven models still have a good performance in the small river basin of semiarid mountain regions where have complicated topography. In this study, the potential of three different data driven methods, artificial neural network (ANN), adaptive neuro fuzzy inference system (ANFIS) and support vector machine (SVM) were used for forecasting river flow in the semiarid mountain region, northwestern China. The models analyzed different combinations of antecedent river flow values and the appropriate input vector has been selected based on the analysis of residuals. The performance of the ANN, ANFIS and SVM models in training and validation sets are compared with the observed data. The model which consists of three antecedent values of flow has been selected as the best fit model for river flow forecasting. To get more accurate evaluation of the results of ANN, ANFIS and SVM models, the four quantitative standard statistical performance evaluation measures, the coefficient of correlation (R), root mean squared error (RMSE), Nash-Sutcliffe efficiency coefficient (NS) and mean absolute relative error (MARE), were employed to evaluate the performances of various models developed. The results indicate that the performance obtained by ANN, ANFIS and SVM in terms of different evaluation criteria during the training and validation period does not vary substantially; the performance of the ANN, ANFIS and SVM models in river flow forecasting was satisfactory. A detailed comparison of the overall performance indicated that the SVM model performed better than ANN and ANFIS in river flow forecasting for the validation data sets. The results also suggest that ANN, ANFIS and SVM method can be successfully applied to establish river flow with complicated topography forecasting models in the semiarid mountain regions.

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.

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.

Boosting Learning Algorithm for Stock Price Forecasting

NASA Astrophysics Data System (ADS)

Wang, Chengzhang; Bai, Xiaoming

2018-03-01

To tackle complexity and uncertainty of stock market behavior, more studies have introduced machine learning algorithms to forecast stock price. ANN (artificial neural network) is one of the most successful and promising applications. We propose a boosting-ANN model in this paper to predict the stock close price. On the basis of boosting theory, multiple weak predicting machines, i.e. ANNs, are assembled to build a stronger predictor, i.e. boosting-ANN model. New error criteria of the weak studying machine and rules of weights updating are adopted in this study. We select technical factors from financial markets as forecasting input variables. Final results demonstrate the boosting-ANN model works better than other ones for stock price forecasting.

Shared investment projects and forecasting errors: setting framework conditions for coordination and sequencing data quality activities.

PubMed

Leitner, Stephan; Brauneis, Alexander; Rausch, Alexandra

2015-01-01

In this paper, we investigate the impact of inaccurate forecasting on the coordination of distributed investment decisions. In particular, by setting up a computational multi-agent model of a stylized firm, we investigate the case of investment opportunities that are mutually carried out by organizational departments. The forecasts of concern pertain to the initial amount of money necessary to launch and operate an investment opportunity, to the expected intertemporal distribution of cash flows, and the departments' efficiency in operating the investment opportunity at hand. We propose a budget allocation mechanism for coordinating such distributed decisions The paper provides guidance on how to set framework conditions, in terms of the number of investment opportunities considered in one round of funding and the number of departments operating one investment opportunity, so that the coordination mechanism is highly robust to forecasting errors. Furthermore, we show that-in some setups-a certain extent of misforecasting is desirable from the firm's point of view as it supports the achievement of the corporate objective of value maximization. We then address the question of how to improve forecasting quality in the best possible way, and provide policy advice on how to sequence activities for improving forecasting quality so that the robustness of the coordination mechanism to errors increases in the best possible way. At the same time, we show that wrong decisions regarding the sequencing can lead to a decrease in robustness. Finally, we conduct a comprehensive sensitivity analysis and prove that-in particular for relatively good forecasters-most of our results are robust to changes in setting the parameters of our multi-agent simulation model.

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.

Forecast of the number of patients with end-stage renal disease in the United States to the year 2010.

PubMed

Xue, J L; Ma, J Z; Louis, T A; Collins, A J

2001-12-01

As the United States end-stage renal disease (ESRD) program enters the new millennium, the continued growth of the ESRD population poses a challenge for policy makers, health care providers, and financial planners. To assist in future planning for the ESRD program, the growth of patient numbers and Medicare costs was forecasted to the year 2010 by modeling of historical data from 1982 through 1997. A stepwise autoregressive method and exponential smoothing models were used. The forecasting models for ESRD patient numbers demonstrated mean errors of -0.03 to 1.03%, relative to the observed values. The model for Medicare payments demonstrated -0.12% mean error. The R(2) values for the forecasting models ranged from 99.09 to 99.98%. On the basis of trends in patient numbers, this forecast projects average annual growth of the ESRD populations of approximately 4.1% for new patients, 6.4% for long-term ESRD patients, 7.1% for dialysis patients, 6.1% for patients with functioning transplants, and 8.2% for patients on waiting lists for transplants, as well as 7.7% for Medicare expenditures. The numbers of patients with ESRD in 2010 are forecasted to be 129,200 +/- 7742 (95% confidence limits) new patients, 651,330 +/- 15,874 long-term ESRD patients, 520,240 +/- 25,609 dialysis patients, 178,806 +/- 4349 patients with functioning transplants, and 95,550 +/- 5478 patients on waiting lists. The forecasted Medicare expenditures are projected to increase to $28.3 +/- 1.7 billion by 2010. These projections are subject to many factors that may alter the actual growth, compared with the historical patterns. They do, however, provide a basis for discussing the future growth of the ESRD program and how the ESRD community can meet the challenges ahead.

Impact of archeomagnetic field model data on modern era geomagnetic forecasts

NASA Astrophysics Data System (ADS)

Tangborn, Andrew; Kuang, Weijia

2018-03-01

A series of geomagnetic data assimilation experiments have been carried out to demonstrate the impact of assimilating archeomagnetic data via the CALS3k.4 geomagnetic field model from the period between 10 and 1590 CE. The assimilation continues with the gufm1 model from 1590 to 1990 and CM4 model from 1990 to 2000 as observations, and comparisons between these models and the geomagnetic forecasts are used to determine an optimal maximum degree for the archeomagnetic observations, and to independently estimate errors for these observations. These are compared with an assimilation experiment that uses the uncertainties provided with CALS3k.4. Optimal 20 year forecasts in 1990 are found when the Gauss coefficients up to degree 3 are assimilated. In addition we demonstrate how a forecast and observation bias correction scheme could be used to reduce bias in modern era forecasts. Initial experiments show that this approach can reduce modern era forecast biases by as much as 50%.

A multiphysical ensemble system of numerical snow modelling

NASA Astrophysics Data System (ADS)

Lafaysse, Matthieu; Cluzet, Bertrand; Dumont, Marie; Lejeune, Yves; Vionnet, Vincent; Morin, Samuel

2017-05-01

Physically based multilayer snowpack models suffer from various modelling errors. To represent these errors, we built the new multiphysical ensemble system ESCROC (Ensemble System Crocus) by implementing new representations of different physical processes in the deterministic coupled multilayer ground/snowpack model SURFEX/ISBA/Crocus. This ensemble was driven and evaluated at Col de Porte (1325 m a.s.l., French alps) over 18 years with a high-quality meteorological and snow data set. A total number of 7776 simulations were evaluated separately, accounting for the uncertainties of evaluation data. The ability of the ensemble to capture the uncertainty associated to modelling errors is assessed for snow depth, snow water equivalent, bulk density, albedo and surface temperature. Different sub-ensembles of the ESCROC system were studied with probabilistic tools to compare their performance. Results show that optimal members of the ESCROC system are able to explain more than half of the total simulation errors. Integrating members with biases exceeding the range corresponding to observational uncertainty is necessary to obtain an optimal dispersion, but this issue can also be a consequence of the fact that meteorological forcing uncertainties were not accounted for. The ESCROC system promises the integration of numerical snow-modelling errors in ensemble forecasting and ensemble assimilation systems in support of avalanche hazard forecasting and other snowpack-modelling applications.

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.

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.

Full Field and Anomaly Initialisation using a low order climate model: a comparison, and proposals for advanced formulations

NASA Astrophysics Data System (ADS)

Weber, Robin; Carrassi, Alberto; Guemas, Virginie; Doblas-Reyes, Francisco; Volpi, Danila

2014-05-01

Full Field (FFI) and Anomaly Initialisation (AI) are two schemes used to initialise seasonal-to-decadal (s2d) prediction. FFI initialises the model on the best estimate of the actual climate state and minimises the initial error. However, due to inevitable model deficiencies, the trajectories drift away from the observations towards the model's own attractor, inducing a bias in the forecast. AI has been devised to tackle the impact of drift through the addition of this bias onto the observations, in the hope of gaining an initial state closer to the model attractor. Its goal is to forecast climate anomalies. The large variety of experimental setups, global coupled models, and observational networks adopted world-wide have led to varying results with regards to the relative performance of AI and FFI. Our research is firstly motivated in a comparison of these two initialisation approaches under varying circumstances of observational errors, observational distributions, and model errors. We also propose and compare two advanced schemes for s2d prediction. Least Square Initialisation (LSI) intends to propagate observational information of partially initialized systems to the whole model domain, based on standard practices in data assimilation and using the covariance of the model anomalies. Exploring the Parameters Uncertainty (EPU) is an online drift correction technique applied during the forecast run after initialisation. It is designed to estimate, and subtract, the bias in the forecast related to parametric error. Experiments are carried out using an idealized coupled dynamics in order to facilitate better control and robust statistical inference. Results show that an improvement of FFI will necessitate refinements in the observations, whereas improvements in AI are subject to model advances. A successful approximation of the model attractor using AI is guaranteed only when the differences between model and nature probability distribution functions (PDFs) are limited to the first order. Significant higher order differences can lead to an initial conditions distribution for AI that is less representative of the model PDF and lead to a degradation of the initalisation skill. Finally, both ad- vanced schemes lead to significantly improved skill scores, encouraging their implementation for models of higher complexity.

Monthly mean forecast experiments with the GISS model

NASA Technical Reports Server (NTRS)

Spar, J.; Atlas, R. M.; Kuo, E.

1976-01-01

The GISS general circulation model was used to compute global monthly mean forecasts for January 1973, 1974, and 1975 from initial conditions on the first day of each month and constant sea surface temperatures. Forecasts were evaluated in terms of global and hemispheric energetics, zonally averaged meridional and vertical profiles, forecast error statistics, and monthly mean synoptic fields. Although it generated a realistic mean meridional structure, the model did not adequately reproduce the observed interannual variations in the large scale monthly mean energetics and zonally averaged circulation. The monthly mean sea level pressure field was not predicted satisfactorily, but annual changes in the Icelandic low were simulated. The impact of temporal sea surface temperature variations on the forecasts was investigated by comparing two parallel forecasts for January 1974, one using climatological ocean temperatures and the other observed daily ocean temperatures. The use of daily updated sea surface temperatures produced no discernible beneficial effect.

Characterizing Macro Scale Patterns Of Uncertainty For Improved Operational Flood Forecasting Over The Conterminous United States

NASA Astrophysics Data System (ADS)

Vergara, H. J.; Kirstetter, P.; Gourley, J. J.; Flamig, Z.; Hong, Y.

2015-12-01

The macro scale patterns of simulated streamflow errors are studied in order to characterize uncertainty in a hydrologic modeling system forced with the Multi-Radar/Multi-Sensor (MRMS; http://mrms.ou.edu) quantitative precipitation estimates for flood forecasting over the Conterminous United States (CONUS). The hydrologic model is centerpiece of the Flooded Locations And Simulated Hydrograph (FLASH; http://flash.ou.edu) real-time system. The hydrologic model is implemented at 1-km/5-min resolution to generate estimates of streamflow. Data from the CONUS-wide stream gauge network of the United States' Geological Survey (USGS) were used as a reference to evaluate the discrepancies with the hydrological model predictions. Streamflow errors were studied at the event scale with particular focus on the peak flow magnitude and timing. A total of 2,680 catchments over CONUS and 75,496 events from a 10-year period are used for the simulation diagnostic analysis. Associations between streamflow errors and geophysical factors were explored and modeled. It is found that hydro-climatic factors and radar coverage could explain significant underestimation of peak flow in regions of complex terrain. Furthermore, the statistical modeling of peak flow errors shows that other geophysical factors such as basin geomorphometry, pedology, and land cover/use could also provide explanatory information. Results from this research demonstrate the utility of uncertainty characterization in providing guidance to improve model adequacy, parameter estimates, and input quality control. Likewise, the characterization of uncertainty enables probabilistic flood forecasting that can be extended to ungauged locations.

Verifying and Postprocesing the Ensemble Spread-Error Relationship

NASA Astrophysics Data System (ADS)

Hopson, Tom; Knievel, Jason; Liu, Yubao; Roux, Gregory; Wu, Wanli

2013-04-01

With the increased utilization of ensemble forecasts in weather and hydrologic applications, there is a need to verify their benefit over less expensive deterministic forecasts. One such potential benefit of ensemble systems is their capacity to forecast their own forecast error through the ensemble spread-error relationship. The paper begins by revisiting the limitations of the Pearson correlation alone in assessing this relationship. Next, we introduce two new metrics to consider in assessing the utility an ensemble's varying dispersion. We argue there are two aspects of an ensemble's dispersion that should be assessed. First, and perhaps more fundamentally: is there enough variability in the ensembles dispersion to justify the maintenance of an expensive ensemble prediction system (EPS), irrespective of whether the EPS is well-calibrated or not? To diagnose this, the factor that controls the theoretical upper limit of the spread-error correlation can be useful. Secondly, does the variable dispersion of an ensemble relate to variable expectation of forecast error? Representing the spread-error correlation in relation to its theoretical limit can provide a simple diagnostic of this attribute. A context for these concepts is provided by assessing two operational ensembles: 30-member Western US temperature forecasts for the U.S. Army Test and Evaluation Command and 51-member Brahmaputra River flow forecasts of the Climate Forecast and Applications Project for Bangladesh. Both of these systems utilize a postprocessing technique based on quantile regression (QR) under a step-wise forward selection framework leading to ensemble forecasts with both good reliability and sharpness. In addition, the methodology utilizes the ensemble's ability to self-diagnose forecast instability to produce calibrated forecasts with informative skill-spread relationships. We will describe both ensemble systems briefly, review the steps used to calibrate the ensemble forecast, and present verification statistics using error-spread metrics, along with figures from operational ensemble forecasts before and after calibration.

Stream-flow forecasting using extreme learning machines: A case study in a semi-arid region in Iraq

NASA Astrophysics Data System (ADS)

Yaseen, Zaher Mundher; Jaafar, Othman; Deo, Ravinesh C.; Kisi, Ozgur; Adamowski, Jan; Quilty, John; El-Shafie, Ahmed

2016-11-01

Monthly stream-flow forecasting can yield important information for hydrological applications including sustainable design of rural and urban water management systems, optimization of water resource allocations, water use, pricing and water quality assessment, and agriculture and irrigation operations. The motivation for exploring and developing expert predictive models is an ongoing endeavor for hydrological applications. In this study, the potential of a relatively new data-driven method, namely the extreme learning machine (ELM) method, was explored for forecasting monthly stream-flow discharge rates in the Tigris River, Iraq. The ELM algorithm is a single-layer feedforward neural network (SLFNs) which randomly selects the input weights, hidden layer biases and analytically determines the output weights of the SLFNs. Based on the partial autocorrelation functions of historical stream-flow data, a set of five input combinations with lagged stream-flow values are employed to establish the best forecasting model. A comparative investigation is conducted to evaluate the performance of the ELM compared to other data-driven models: support vector regression (SVR) and generalized regression neural network (GRNN). The forecasting metrics defined as the correlation coefficient (r), Nash-Sutcliffe efficiency (ENS), Willmott's Index (WI), root-mean-square error (RMSE) and mean absolute error (MAE) computed between the observed and forecasted stream-flow data are employed to assess the ELM model's effectiveness. The results revealed that the ELM model outperformed the SVR and the GRNN models across a number of statistical measures. In quantitative terms, superiority of ELM over SVR and GRNN models was exhibited by ENS = 0.578, 0.378 and 0.144, r = 0.799, 0.761 and 0.468 and WI = 0.853, 0.802 and 0.689, respectively and the ELM model attained lower RMSE value by approximately 21.3% (relative to SVR) and by approximately 44.7% (relative to GRNN). Based on the findings of this study, several recommendations were suggested for further exploration of the ELM model in hydrological forecasting problems.

Ensemble Data Assimilation of Wind and Photovoltaic Power Information in the Convection-permitting High-Resolution Model COSMO-DE

NASA Astrophysics Data System (ADS)

Declair, Stefan; Saint-Drenan, Yves-Marie; Potthast, Roland

2016-04-01

Determining the amount of weather dependent renewable energy is a demanding task for transmission system operators (TSOs) and wind and photovoltaic (PV) prediction errors require the use of reserve power, which generate costs and can - in extreme cases - endanger the security of supply. In the project EWeLiNE funded by the German government, the German Weather Service and the Fraunhofer Institute on Wind Energy and Energy System Technology develop innovative weather- and power forecasting models and tools for grid integration of weather dependent renewable energy. The key part in energy prediction process chains is the numerical weather prediction (NWP) system. Wind speed and irradiation forecast from NWP system are however subject to several sources of error. The quality of the wind power prediction is mainly penalized by forecast error of the NWP model in the planetary boundary layer (PBL), which is characterized by high spatial and temporal fluctuations of the wind speed. For PV power prediction, weaknesses of the NWP model to correctly forecast i.e. low stratus, the absorption of condensed water or aerosol optical depth are the main sources of errors. Inaccurate radiation schemes (i.e. the two-stream parametrization) are also known as a deficit of NWP systems with regard to irradiation forecast. To mitigate errors like these, NWP model data can be corrected by post-processing techniques such as model output statistics and calibration using historical observational data. Additionally, latest observations can be used in a pre-processing technique called data assimilation (DA). In DA, not only the initial fields are provided, but the model is also synchronized with reality - the observations - and hence the model error is reduced in the forecast. Besides conventional observation networks like radiosondes, synoptic observations or air reports of wind, pressure and humidity, the number of observations measuring meteorological information indirectly such as satellite radiances, radar reflectivities or GPS slant delays strongly increases. The numerous wind farm and PV plants installed in Germany potentially represent a dense meteorological network assessing irradiation and wind speed through their power measurements. The accuracy of the NWP data may thus be enhanced by extending the observations in the assimilation by this new source of information. Wind power data can serve as indirect measurements of wind speed at hub height. The impact on the NWP model is potentially interesting since conventional observation network lacks measurements in this part of the PBL. Photovoltaic power plants can provide information on clouds, aerosol optical depth or low stratus in terms of remote sensing: the power output is strongly dependent on perturbations along the slant between sun position and PV panel. Additionally, since the latter kind of data is not limited to the vertical column above or below the detector. It may thus complement satellite data and compensate weaknesses in the radiation scheme. In this contribution, the DA method (Local Ensemble Transform Kalman Filter, LETKF) is shortly sketched. Furthermore, the computation of the model power equivalents is described and first assimilation results are presented and discussed.

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.

Monthly streamflow forecasting with auto-regressive integrated moving average

NASA Astrophysics Data System (ADS)

Nasir, Najah; Samsudin, Ruhaidah; Shabri, Ani

2017-09-01

Forecasting of streamflow is one of the many ways that can contribute to better decision making for water resource management. The auto-regressive integrated moving average (ARIMA) model was selected in this research for monthly streamflow forecasting with enhancement made by pre-processing the data using singular spectrum analysis (SSA). This study also proposed an extension of the SSA technique to include a step where clustering was performed on the eigenvector pairs before reconstruction of the time series. The monthly streamflow data of Sungai Muda at Jeniang, Sungai Muda at Jambatan Syed Omar and Sungai Ketil at Kuala Pegang was gathered from the Department of Irrigation and Drainage Malaysia. A ratio of 9:1 was used to divide the data into training and testing sets. The ARIMA, SSA-ARIMA and Clustered SSA-ARIMA models were all developed in R software. Results from the proposed model are then compared to a conventional auto-regressive integrated moving average model using the root-mean-square error and mean absolute error values. It was found that the proposed model can outperform the conventional model.

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.

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.

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.

Fitting dynamic models to the Geosat sea level observations in the tropical Pacific Ocean. I - A free wave model

NASA Technical Reports Server (NTRS)

Fu, Lee-Lueng; Vazquez, Jorge; Perigaud, Claire

1991-01-01

Free, equatorially trapped sinusoidal wave solutions to a linear model on an equatorial beta plane are used to fit the Geosat altimetric sea level observations in the tropical Pacific Ocean. The Kalman filter technique is used to estimate the wave amplitude and phase from the data. The estimation is performed at each time step by combining the model forecast with the observation in an optimal fashion utilizing the respective error covariances. The model error covariance is determined such that the performance of the model forecast is optimized. It is found that the dominant observed features can be described qualitatively by basin-scale Kelvin waves and the first meridional-mode Rossby waves. Quantitatively, however, only 23 percent of the signal variance can be accounted for by this simple model.

Forecasting the Water Demand in Chongqing, China Using a Grey Prediction Model and Recommendations for the Sustainable Development of Urban Water Consumption

PubMed Central

Wu, Hua’an; Zhou, Meng

2017-01-01

High accuracy in water demand predictions is an important basis for the rational allocation of city water resources and forms the basis for sustainable urban development. The shortage of water resources in Chongqing, the youngest central municipality in Southwest China, has significantly increased with the population growth and rapid economic development. In this paper, a new grey water-forecasting model (GWFM) was built based on the data characteristics of water consumption. The parameter estimation and error checking methods of the GWFM model were investigated. Then, the GWFM model was employed to simulate the water demands of Chongqing from 2009 to 2015 and forecast it in 2016. The simulation and prediction errors of the GWFM model was checked, and the results show the GWFM model exhibits better simulation and prediction precisions than those of the classical Grey Model with one variable and single order equation GM(1,1) for short and the frequently-used Discrete Grey Model with one variable and single order equation, DGM(1,1) for short. Finally, the water demand in Chongqing from 2017 to 2022 was forecasted, and some corresponding control measures and recommendations were provided based on the prediction results to ensure a viable water supply and promote the sustainable development of the Chongqing economy. PMID:29140266

[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.

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.

Wind Power Forecasting Error Distributions over Multiple Timescales: Preprint

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

Hodge, B. M.; Milligan, M.

2011-03-01

In this paper, we examine the shape of the persistence model error distribution for ten different wind plants in the ERCOT system over multiple timescales. Comparisons are made between the experimental distribution shape and that of the normal distribution.

Emergency Department Visit Forecasting and Dynamic Nursing Staff Allocation Using Machine Learning Techniques With Readily Available Open-Source Software.

PubMed

Zlotnik, Alexander; Gallardo-Antolín, Ascensión; Cuchí Alfaro, Miguel; Pérez Pérez, María Carmen; Montero Martínez, Juan Manuel

2015-08-01

Although emergency department visit forecasting can be of use for nurse staff planning, previous research has focused on models that lacked sufficient resolution and realistic error metrics for these predictions to be applied in practice. Using data from a 1100-bed specialized care hospital with 553,000 patients assigned to its healthcare area, forecasts with different prediction horizons, from 2 to 24 weeks ahead, with an 8-hour granularity, using support vector regression, M5P, and stratified average time-series models were generated with an open-source software package. As overstaffing and understaffing errors have different implications, error metrics and potential personnel monetary savings were calculated with a custom validation scheme, which simulated subsequent generation of predictions during a 4-year period. Results were then compared with a generalized estimating equation regression. Support vector regression and M5P models were found to be superior to the stratified average model with a 95% confidence interval. Our findings suggest that medium and severe understaffing situations could be reduced in more than an order of magnitude and average yearly savings of up to €683,500 could be achieved if dynamic nursing staff allocation was performed with support vector regression instead of the static staffing levels currently in use.

Modeling Input Errors to Improve Uncertainty Estimates for Sediment Transport Model Predictions

NASA Astrophysics Data System (ADS)

Jung, J. Y.; Niemann, J. D.; Greimann, B. P.

2016-12-01

Bayesian methods using Markov chain Monte Carlo algorithms have recently been applied to sediment transport models to assess the uncertainty in the model predictions due to the parameter values. Unfortunately, the existing approaches can only attribute overall uncertainty to the parameters. This limitation is critical because no model can produce accurate forecasts if forced with inaccurate input data, even if the model is well founded in physical theory. In this research, an existing Bayesian method is modified to consider the potential errors in input data during the uncertainty evaluation process. The input error is modeled using Gaussian distributions, and the means and standard deviations are treated as uncertain parameters. The proposed approach is tested by coupling it to the Sedimentation and River Hydraulics - One Dimension (SRH-1D) model and simulating a 23-km reach of the Tachia River in Taiwan. The Wu equation in SRH-1D is used for computing the transport capacity for a bed material load of non-cohesive material. Three types of input data are considered uncertain: (1) the input flowrate at the upstream boundary, (2) the water surface elevation at the downstream boundary, and (3) the water surface elevation at a hydraulic structure in the middle of the reach. The benefits of modeling the input errors in the uncertainty analysis are evaluated by comparing the accuracy of the most likely forecast and the coverage of the observed data by the credible intervals to those of the existing method. The results indicate that the internal boundary condition has the largest uncertainty among those considered. Overall, the uncertainty estimates from the new method are notably different from those of the existing method for both the calibration and forecast periods.

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.

Evaluation of precipitation forecasts from 3D-Var and hybrid GSI-based system during Indian summer monsoon 2015

NASA Astrophysics Data System (ADS)

Singh, Sanjeev Kumar; Prasad, V. S.

2018-02-01

This paper presents a systematic investigation of medium-range rainfall forecasts from two versions of the National Centre for Medium Range Weather Forecasting (NCMRWF)-Global Forecast System based on three-dimensional variational (3D-Var) and hybrid analysis system namely, NGFS and HNGFS, respectively, during Indian summer monsoon (June-September) 2015. The NGFS uses gridpoint statistical interpolation (GSI) 3D-Var data assimilation system, whereas HNGFS uses hybrid 3D ensemble-variational scheme. The analysis includes the evaluation of rainfall fields and comparisons of rainfall using statistical score such as mean precipitation, bias, correlation coefficient, root mean square error and forecast improvement factor. In addition to these, categorical scores like Peirce skill score and bias score are also computed to describe particular aspects of forecasts performance. The comparison results of mean precipitation reveal that both the versions of model produced similar large-scale feature of Indian summer monsoon rainfall for day-1 through day-5 forecasts. The inclusion of fully flow-dependent background error covariance significantly improved the wet biases in HNGFS over the Indian Ocean. The forecast improvement factor and Peirce skill score in the HNGFS have also found better than NGFS for day-1 through day-5 forecasts.

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.

Investigating the role of background and observation error correlations in improving a model forecast of forest carbon balance using four dimensional variational data assimilation.

NASA Astrophysics Data System (ADS)

Pinnington, Ewan; Casella, Eric; Dance, Sarah; Lawless, Amos; Morison, James; Nichols, Nancy; Wilkinson, Matthew; Quaife, Tristan

2016-04-01

Forest ecosystems play an important role in sequestering human emitted carbon-dioxide from the atmosphere and therefore greatly reduce the effect of anthropogenic induced climate change. For that reason understanding their response to climate change is of great importance. Efforts to implement variational data assimilation routines with functional ecology models and land surface models have been limited, with sequential and Markov chain Monte Carlo data assimilation methods being prevalent. When data assimilation has been used with models of carbon balance, background "prior" errors and observation errors have largely been treated as independent and uncorrelated. Correlations between background errors have long been known to be a key aspect of data assimilation in numerical weather prediction. More recently, it has been shown that accounting for correlated observation errors in the assimilation algorithm can considerably improve data assimilation results and forecasts. In this paper we implement a 4D-Var scheme with a simple model of forest carbon balance, for joint parameter and state estimation and assimilate daily observations of Net Ecosystem CO2 Exchange (NEE) taken at the Alice Holt forest CO2 flux site in Hampshire, UK. We then investigate the effect of specifying correlations between parameter and state variables in background error statistics and the effect of specifying correlations in time between observation error statistics. The idea of including these correlations in time is new and has not been previously explored in carbon balance model data assimilation. In data assimilation, background and observation error statistics are often described by the background error covariance matrix and the observation error covariance matrix. We outline novel methods for creating correlated versions of these matrices, using a set of previously postulated dynamical constraints to include correlations in the background error statistics and a Gaussian correlation function to include time correlations in the observation error statistics. The methods used in this paper will allow the inclusion of time correlations between many different observation types in the assimilation algorithm, meaning that previously neglected information can be accounted for. In our experiments we compared the results using our new correlated background and observation error covariance matrices and those using diagonal covariance matrices. We found that using the new correlated matrices reduced the root mean square error in the 14 year forecast of daily NEE by 44 % decreasing from 4.22 g C m-2 day-1 to 2.38 g C m-2 day-1.

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.

Assimilation of NUCAPS Retrieved Profiles in GSI for Unique Forecasting Applications

NASA Technical Reports Server (NTRS)

Berndt, Emily Beth; Zavodsky, Bradley; Srikishen, Jayanthi; Blankenship, Clay

2015-01-01

Hyperspectral IR profiles can be assimilated in GSI as a separate observation other than radiosondes with only changes to tables in the fix directory. Assimilation of profiles does produce changes to analysis fields and evidenced by: Innovations larger than +/-2.0 K are present and represent where individual profiles impact the final temperature analysis.The updated temperature analysis is colder behind the cold front and warmer in the warm sector. The updated moisture analysis is modified more in the low levels and tends to be drier than the original model background Analysis of model output shows: Differences relative to 13-km RAP analyses are smaller when profiles are assimilated with NUCAPS errors. CAPE is under-forecasted when assimilating NUCAPS profiles, which could be problematic for severe weather forecasting Refining the assimilation technique to incorporate an error covariance matrix and creating a separate GSI module to assimilate satellite profiles may improve results.

Modeling and predicting historical volatility in exchange rate markets

NASA Astrophysics Data System (ADS)

Lahmiri, Salim

2017-04-01

Volatility modeling and forecasting of currency exchange rate is an important task in several business risk management tasks; including treasury risk management, derivatives pricing, and portfolio risk evaluation. The purpose of this study is to present a simple and effective approach for predicting historical volatility of currency exchange rate. The approach is based on a limited set of technical indicators as inputs to the artificial neural networks (ANN). To show the effectiveness of the proposed approach, it was applied to forecast US/Canada and US/Euro exchange rates volatilities. The forecasting results show that our simple approach outperformed the conventional GARCH and EGARCH with different distribution assumptions, and also the hybrid GARCH and EGARCH with ANN in terms of mean absolute error, mean of squared errors, and Theil's inequality coefficient. Because of the simplicity and effectiveness of the approach, it is promising for US currency volatility prediction tasks.

Ensemble Bayesian forecasting system Part I: Theory and algorithms

NASA Astrophysics Data System (ADS)

Herr, Henry D.; Krzysztofowicz, Roman

2015-05-01

The ensemble Bayesian forecasting system (EBFS), whose theory was published in 2001, is developed for the purpose of quantifying the total uncertainty about a discrete-time, continuous-state, non-stationary stochastic process such as a time series of stages, discharges, or volumes at a river gauge. The EBFS is built of three components: an input ensemble forecaster (IEF), which simulates the uncertainty associated with random inputs; a deterministic hydrologic model (of any complexity), which simulates physical processes within a river basin; and a hydrologic uncertainty processor (HUP), which simulates the hydrologic uncertainty (an aggregate of all uncertainties except input). It works as a Monte Carlo simulator: an ensemble of time series of inputs (e.g., precipitation amounts) generated by the IEF is transformed deterministically through a hydrologic model into an ensemble of time series of outputs, which is next transformed stochastically by the HUP into an ensemble of time series of predictands (e.g., river stages). Previous research indicated that in order to attain an acceptable sampling error, the ensemble size must be on the order of hundreds (for probabilistic river stage forecasts and probabilistic flood forecasts) or even thousands (for probabilistic stage transition forecasts). The computing time needed to run the hydrologic model this many times renders the straightforward simulations operationally infeasible. This motivates the development of the ensemble Bayesian forecasting system with randomization (EBFSR), which takes full advantage of the analytic meta-Gaussian HUP and generates multiple ensemble members after each run of the hydrologic model; this auxiliary randomization reduces the required size of the meteorological input ensemble and makes it operationally feasible to generate a Bayesian ensemble forecast of large size. Such a forecast quantifies the total uncertainty, is well calibrated against the prior (climatic) distribution of predictand, possesses a Bayesian coherence property, constitutes a random sample of the predictand, and has an acceptable sampling error-which makes it suitable for rational decision making under uncertainty.

Forecasting of particulate matter time series using wavelet analysis and wavelet-ARMA/ARIMA model in Taiyuan, China.

PubMed

Zhang, Hong; Zhang, Sheng; Wang, Ping; Qin, Yuzhe; Wang, Huifeng

2017-07-01

Particulate matter with aerodynamic diameter below 10 μm (PM 10 ) forecasting is difficult because of the uncertainties in describing the emission and meteorological fields. This paper proposed a wavelet-ARMA/ARIMA model to forecast the short-term series of the PM 10 concentrations. It was evaluated by experiments using a 10-year data set of daily PM 10 concentrations from 4 stations located in Taiyuan, China. The results indicated the following: (1) PM 10 concentrations of Taiyuan had a decreasing trend during 2005 to 2012 but increased in 2013. PM 10 concentrations had an obvious seasonal fluctuation related to coal-fired heating in winter and early spring. (2) Spatial differences among the four stations showed that the PM 10 concentrations in industrial and heavily trafficked areas were higher than those in residential and suburb areas. (3) Wavelet analysis revealed that the trend variation and the changes of the PM 10 concentration of Taiyuan were complicated. (4) The proposed wavelet-ARIMA model could be efficiently and successfully applied to the PM 10 forecasting field. Compared with the traditional ARMA/ARIMA methods, this wavelet-ARMA/ARIMA method could effectively reduce the forecasting error, improve the prediction accuracy, and realize multiple-time-scale prediction. Wavelet analysis can filter noisy signals and identify the variation trend and the fluctuation of the PM 10 time-series data. Wavelet decomposition and reconstruction reduce the nonstationarity of the PM 10 time-series data, and thus improve the accuracy of the prediction. This paper proposed a wavelet-ARMA/ARIMA model to forecast the PM 10 time series. Compared with the traditional ARMA/ARIMA method, this wavelet-ARMA/ARIMA method could effectively reduce the forecasting error, improve the prediction accuracy, and realize multiple-time-scale prediction. The proposed model could be efficiently and successfully applied to the PM 10 forecasting field.

State-space prediction model for chaotic time series

NASA Astrophysics Data System (ADS)

Alparslan, A. K.; Sayar, M.; Atilgan, A. R.

1998-08-01

A simple method for predicting the continuation of scalar chaotic time series ahead in time is proposed. The false nearest neighbors technique in connection with the time-delayed embedding is employed so as to reconstruct the state space. A local forecasting model based upon the time evolution of the topological neighboring in the reconstructed phase space is suggested. A moving root-mean-square error is utilized in order to monitor the error along the prediction horizon. The model is tested for the convection amplitude of the Lorenz model. The results indicate that for approximately 100 cycles of the training data, the prediction follows the actual continuation very closely about six cycles. The proposed model, like other state-space forecasting models, captures the long-term behavior of the system due to the use of spatial neighbors in the state space.

Incorporating TPC observed parameters and QuikSCAT surface wind observations into hurricane initialization using 4D-VAR approaches

NASA Astrophysics Data System (ADS)

Park, Kyungjeen

This study aims to develop an objective hurricane initialization scheme which incorporates not only forecast model constraints but also observed features such as the initial intensity and size. It is based on the four-dimensional variational (4D-Var) bogus data assimilation (BDA) scheme originally proposed by Zou and Xiao (1999). The 4D-Var BDA consists of two steps: (i) specifying a bogus sea level pressure (SLP) field based on parameters observed by the Tropical Prediction Center (TPC) and (ii) assimilating the bogus SLP field under a forecast model constraint to adjust all model variables. This research focuses on improving the specification of the bogus SLP indicated in the first step. Numerical experiments are carried out for Hurricane Bonnie (1998) and Hurricane Gordon (2000) to test the sensitivity of hurricane track and intensity forecasts to specification of initial vortex. Major results are listed below: (1) A linear regression model is developed for determining the size of initial vortex based on the TPC observed radius of 34kt. (2) A method is proposed to derive a radial profile of SLP from QuikSCAT surface winds. This profile is shown to be more realistic than ideal profiles derived from Fujita's and Holland's formulae. (3) It is found that it takes about 1 h for hurricane prediction model to develop a conceptually correct hurricane structure, featuring a dominant role of hydrostatic balance at the initial time and a dynamic adjustment in less than 30 minutes. (4) Numerical experiments suggest that track prediction is less sensitive to the specification of initial vortex structure than intensity forecast. (5) Hurricane initialization using QuikSCAT-derived initial vortex produced a reasonably good forecast for hurricane landfall, with a position error of 25 km and a 4-h delay at landfalling. (6) Numerical experiments using the linear regression model for the size specification considerably outperforms all the other formulations tested in terms of the intensity prediction for both Hurricanes. For examples, the maximum track error is less than 110 km during the entire three-day forecasts for both hurricanes. The simulated Hurricane Gordon using the linear regression model made a nearly perfect landfall, with no position error and only 1-h error in landfalling time. (7) Diagnosis of model output indicates that the initial vortex specified by the linear regression model produces larger surface fluxes of sensible heat, latent heat and moisture, as well as stronger downward angular momentum transport than all the other schemes do. These enhanced energy supplies offset the energy lost caused by friction and gravity wave propagation, allowing for the model to maintain a strong and realistic hurricane during the entire forward model integration.

Forecasting volcanic air pollution in Hawaii: Tests of time series models

NASA Astrophysics Data System (ADS)

Reikard, Gordon

2012-12-01

Volcanic air pollution, known as vog (volcanic smog) has recently become a major issue in the Hawaiian islands. Vog is caused when volcanic gases react with oxygen and water vapor. It consists of a mixture of gases and aerosols, which include sulfur dioxide and other sulfates. The source of the volcanic gases is the continuing eruption of Mount Kilauea. This paper studies predicting vog using statistical methods. The data sets include time series for SO2 and SO4, over locations spanning the west, south and southeast coasts of Hawaii, and the city of Hilo. The forecasting models include regressions and neural networks, and a frequency domain algorithm. The most typical pattern for the SO2 data is for the frequency domain method to yield the most accurate forecasts over the first few hours, and at the 24 h horizon. The neural net places second. For the SO4 data, the results are less consistent. At two sites, the neural net generally yields the most accurate forecasts, except at the 1 and 24 h horizons, where the frequency domain technique wins narrowly. At one site, the neural net and the frequency domain algorithm yield comparable errors over the first 5 h, after which the neural net dominates. At the remaining site, the frequency domain method is more accurate over the first 4 h, after which the neural net achieves smaller errors. For all the series, the average errors are well within one standard deviation of the actual data at all the horizons. However, the errors also show irregular outliers. In essence, the models capture the central tendency of the data, but are less effective in predicting the extreme events.

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.

Forecasting the portuguese stock market time series by using artificial neural networks

NASA Astrophysics Data System (ADS)

Isfan, Monica; Menezes, Rui; Mendes, Diana A.

2010-04-01

In this paper, we show that neural networks can be used to uncover the non-linearity that exists in the financial data. First, we follow a traditional approach by analysing the deterministic/stochastic characteristics of the Portuguese stock market data and some typical features are studied, like the Hurst exponents, among others. We also simulate a BDS test to investigate nonlinearities and the results are as expected: the financial time series do not exhibit linear dependence. Secondly, we trained four types of neural networks for the stock markets and used the models to make forecasts. The artificial neural networks were obtained using a three-layer feed-forward topology and the back-propagation learning algorithm. The quite large number of parameters that must be selected to develop a neural network forecasting model involves some trial and as a consequence the error is not small enough. In order to improve this we use a nonlinear optimization algorithm to minimize the error. Finally, the output of the 4 models is quite similar, leading to a qualitative forecast that we compare with the results of the application of k-nearest-neighbor for the same time series.

Wind-Farm Forecasting Using the HARMONIE Weather Forecast Model and Bayes Model Averaging for Bias Removal.

NASA Astrophysics Data System (ADS)

O'Brien, Enda; McKinstry, Alastair; Ralph, Adam

2015-04-01

Building on previous work presented at EGU 2013 (http://www.sciencedirect.com/science/article/pii/S1876610213016068 ), more results are available now from a different wind-farm in complex terrain in southwest Ireland. The basic approach is to interpolate wind-speed forecasts from an operational weather forecast model (i.e., HARMONIE in the case of Ireland) to the precise location of each wind-turbine, and then use Bayes Model Averaging (BMA; with statistical information collected from a prior training-period of e.g., 25 days) to remove systematic biases. Bias-corrected wind-speed forecasts (and associated power-generation forecasts) are then provided twice daily (at 5am and 5pm) out to 30 hours, with each forecast validation fed back to BMA for future learning. 30-hr forecasts from the operational Met Éireann HARMONIE model at 2.5km resolution have been validated against turbine SCADA observations since Jan. 2014. An extra high-resolution (0.5km grid-spacing) HARMONIE configuration has been run since Nov. 2014 as an extra member of the forecast "ensemble". A new version of HARMONIE with extra filters designed to stabilize high-resolution configurations has been run since Jan. 2015. Measures of forecast skill and forecast errors will be provided, and the contributions made by the various physical and computational enhancements to HARMONIE will be quantified.

Wind power application research on the fusion of the determination and ensemble prediction

NASA Astrophysics Data System (ADS)

Lan, Shi; Lina, Xu; Yuzhu, Hao

2017-07-01

The fused product of wind speed for the wind farm is designed through the use of wind speed products of ensemble prediction from the European Centre for Medium-Range Weather Forecasts (ECMWF) and professional numerical model products on wind power based on Mesoscale Model5 (MM5) and Beijing Rapid Update Cycle (BJ-RUC), which are suitable for short-term wind power forecasting and electric dispatch. The single-valued forecast is formed by calculating the different ensemble statistics of the Bayesian probabilistic forecasting representing the uncertainty of ECMWF ensemble prediction. Using autoregressive integrated moving average (ARIMA) model to improve the time resolution of the single-valued forecast, and based on the Bayesian model averaging (BMA) and the deterministic numerical model prediction, the optimal wind speed forecasting curve and the confidence interval are provided. The result shows that the fusion forecast has made obvious improvement to the accuracy relative to the existing numerical forecasting products. Compared with the 0-24 h existing deterministic forecast in the validation period, the mean absolute error (MAE) is decreased by 24.3 % and the correlation coefficient (R) is increased by 12.5 %. In comparison with the ECMWF ensemble forecast, the MAE is reduced by 11.7 %, and R is increased 14.5 %. Additionally, MAE did not increase with the prolongation of the forecast ahead.

Verification of Meteorological and Oceanographic Ensemble Forecasts in the U.S. Navy

NASA Astrophysics Data System (ADS)

Klotz, S.; Hansen, J.; Pauley, P.; Sestak, M.; Wittmann, P.; Skupniewicz, C.; Nelson, G.

2013-12-01

The Navy Ensemble Forecast Verification System (NEFVS) has been promoted recently to operational status at the U.S. Navy's Fleet Numerical Meteorology and Oceanography Center (FNMOC). NEFVS processes FNMOC and National Centers for Environmental Prediction (NCEP) meteorological and ocean wave ensemble forecasts, gridded forecast analyses, and innovation (observational) data output by FNMOC's data assimilation system. The NEFVS framework consists of statistical analysis routines, a variety of pre- and post-processing scripts to manage data and plot verification metrics, and a master script to control application workflow. NEFVS computes metrics that include forecast bias, mean-squared error, conditional error, conditional rank probability score, and Brier score. The system also generates reliability and Receiver Operating Characteristic diagrams. In this presentation we describe the operational framework of NEFVS and show examples of verification products computed from ensemble forecasts, meteorological observations, and forecast analyses. The construction and deployment of NEFVS addresses important operational and scientific requirements within Navy Meteorology and Oceanography. These include computational capabilities for assessing the reliability and accuracy of meteorological and ocean wave forecasts in an operational environment, for quantifying effects of changes and potential improvements to the Navy's forecast models, and for comparing the skill of forecasts from different forecast systems. NEFVS also supports the Navy's collaboration with the U.S. Air Force, NCEP, and Environment Canada in the North American Ensemble Forecast System (NAEFS) project and with the Air Force and the National Oceanic and Atmospheric Administration (NOAA) in the National Unified Operational Prediction Capability (NUOPC) program. This program is tasked with eliminating unnecessary duplication within the three agencies, accelerating the transition of new technology, such as multi-model ensemble forecasting, to U.S. Department of Defense use, and creating a superior U.S. global meteorological and oceanographic prediction capability. Forecast verification is an important component of NAEFS and NUOPC. Distribution Statement A: Approved for Public Release; distribution is unlimited

Verification of Meteorological and Oceanographic Ensemble Forecasts in the U.S. Navy

NASA Astrophysics Data System (ADS)

Klotz, S. P.; Hansen, J.; Pauley, P.; Sestak, M.; Wittmann, P.; Skupniewicz, C.; Nelson, G.

2012-12-01

The Navy Ensemble Forecast Verification System (NEFVS) has been promoted recently to operational status at the U.S. Navy's Fleet Numerical Meteorology and Oceanography Center (FNMOC). NEFVS processes FNMOC and National Centers for Environmental Prediction (NCEP) meteorological and ocean wave ensemble forecasts, gridded forecast analyses, and innovation (observational) data output by FNMOC's data assimilation system. The NEFVS framework consists of statistical analysis routines, a variety of pre- and post-processing scripts to manage data and plot verification metrics, and a master script to control application workflow. NEFVS computes metrics that include forecast bias, mean-squared error, conditional error, conditional rank probability score, and Brier score. The system also generates reliability and Receiver Operating Characteristic diagrams. In this presentation we describe the operational framework of NEFVS and show examples of verification products computed from ensemble forecasts, meteorological observations, and forecast analyses. The construction and deployment of NEFVS addresses important operational and scientific requirements within Navy Meteorology and Oceanography (METOC). These include computational capabilities for assessing the reliability and accuracy of meteorological and ocean wave forecasts in an operational environment, for quantifying effects of changes and potential improvements to the Navy's forecast models, and for comparing the skill of forecasts from different forecast systems. NEFVS also supports the Navy's collaboration with the U.S. Air Force, NCEP, and Environment Canada in the North American Ensemble Forecast System (NAEFS) project and with the Air Force and the National Oceanic and Atmospheric Administration (NOAA) in the National Unified Operational Prediction Capability (NUOPC) program. This program is tasked with eliminating unnecessary duplication within the three agencies, accelerating the transition of new technology, such as multi-model ensemble forecasting, to U.S. Department of Defense use, and creating a superior U.S. global meteorological and oceanographic prediction capability. Forecast verification is an important component of NAEFS and NUOPC.

Data Assimilation in the Presence of Forecast Bias: The GEOS Moisture Analysis

NASA Technical Reports Server (NTRS)

Dee, Dick P.; Todling, Ricardo

1999-01-01

We describe the application of the unbiased sequential analysis algorithm developed by Dee and da Silva (1998) to the GEOS DAS moisture analysis. The algorithm estimates the persistent component of model error using rawinsonde observations and adjusts the first-guess moisture field accordingly. Results of two seasonal data assimilation cycles show that moisture analysis bias is almost completely eliminated in all observed regions. The improved analyses cause a sizable reduction in the 6h-forecast bias and a marginal improvement in the error standard deviations.

Assimilation of CryoSat-2 altimetry to a hydrodynamic model of the Brahmaputra river

NASA Astrophysics Data System (ADS)

Schneider, Raphael; Nygaard Godiksen, Peter; Ridler, Marc-Etienne; Madsen, Henrik; Bauer-Gottwein, Peter

2016-04-01

Remote sensing provides valuable data for parameterization and updating of hydrological models, for example water level measurements of inland water bodies from satellite radar altimeters. Satellite altimetry data from repeat-orbit missions such as Envisat, ERS or Jason has been used in many studies, also synthetic wide-swath altimetry data as expected from the SWOT mission. This study is one of the first hydrologic applications of altimetry data from a drifting orbit satellite mission, namely CryoSat-2. CryoSat-2 is equipped with the SIRAL instrument, a new type of radar altimeter similar to SRAL on Sentinel-3. CryoSat-2 SARIn level 2 data is used to improve a 1D hydrodynamic model of the Brahmaputra river basin in South Asia set up in the DHI MIKE 11 software. CryoSat-2 water levels were extracted over river masks derived from Landsat imagery. After discharge calibration, simulated water levels were fitted to the CryoSat-2 data along the Assam valley by adapting cross section shapes and datums. The resulting hydrodynamic model shows accurate spatio-temporal representation of water levels, which is a prerequisite for real-time model updating by assimilation of CryoSat-2 altimetry or multi-mission data in general. For this task, a data assimilation framework has been developed and linked with the MIKE 11 model. It is a flexible framework that can assimilate water level data which are arbitrarily distributed in time and space. Different types of error models, data assimilation methods, etc. can easily be used and tested. Furthermore, it is not only possible to update the water level of the hydrodynamic model, but also the states of the rainfall-runoff models providing the forcing of the hydrodynamic model. The setup has been used to assimilate CryoSat-2 observations over the Assam valley for the years 2010 to 2013. Different data assimilation methods and localizations were tested, together with different model error representations. Furthermore, the impact of different filtering and clustering methods and error descriptions of the CryoSat-2 observations was evaluated. Performance improvement in terms of discharge and water level forecast due to the assimilation of satellite altimetry data was then evaluated. The model forecasts were also compared to climatology and persistence forecasts. Using ensemble based filters, the evaluation was done not only based on performance criteria for the central forecast such as root-mean-square error (RMSE) and Nash-Sutcliffe model efficiency (NSE), but also based on sharpness, reliability and continuous ranked probability score (CRPS) of the ensemble of probabilistic forecasts.

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.

Towards more accurate wind and solar power prediction by improving NWP model physics

NASA Astrophysics Data System (ADS)

Steiner, Andrea; Köhler, Carmen; von Schumann, Jonas; Ritter, Bodo

2014-05-01

The growing importance and successive expansion of renewable energies raise new challenges for decision makers, economists, transmission system operators, scientists and many more. In this interdisciplinary field, the role of Numerical Weather Prediction (NWP) is to reduce the errors and provide an a priori estimate of remaining uncertainties associated with the large share of weather-dependent power sources. For this purpose it is essential to optimize NWP model forecasts with respect to those prognostic variables which are relevant for wind and solar power plants. An improved weather forecast serves as the basis for a sophisticated power forecasts. Consequently, a well-timed energy trading on the stock market, and electrical grid stability can be maintained. The German Weather Service (DWD) currently is involved with two projects concerning research in the field of renewable energy, namely ORKA*) and EWeLiNE**). Whereas the latter is in collaboration with the Fraunhofer Institute (IWES), the project ORKA is led by energy & meteo systems (emsys). Both cooperate with German transmission system operators. The goal of the projects is to improve wind and photovoltaic (PV) power forecasts by combining optimized NWP and enhanced power forecast models. In this context, the German Weather Service aims to improve its model system, including the ensemble forecasting system, by working on data assimilation, model physics and statistical post processing. This presentation is focused on the identification of critical weather situations and the associated errors in the German regional NWP model COSMO-DE. First steps leading to improved physical parameterization schemes within the NWP-model are presented. Wind mast measurements reaching up to 200 m height above ground are used for the estimation of the (NWP) wind forecast error at heights relevant for wind energy plants. One particular problem is the daily cycle in wind speed. The transition from stable stratification during nighttime to well mixed conditions during the day presents a big challenge to NWP models. Fast decrease and successive increase in hub-height wind speed after sunrise, and the formation of nocturnal low level jets will be discussed. For PV, the life cycle of low stratus clouds and fog is crucial. Capturing these processes correctly depends on the accurate simulation of diffusion or vertical momentum transport and the interaction with other atmospheric and soil processes within the numerical weather model. Results from Single Column Model simulations and 3d case studies will be presented. Emphasis is placed on wind forecasts; however, some references to highlights concerning the PV-developments will also be given. *) ORKA: Optimierung von Ensembleprognosen regenerativer Einspeisung für den Kürzestfristbereich am Anwendungsbeispiel der Netzsicherheitsrechnungen **) EWeLiNE: Erstellung innovativer Wetter- und Leistungsprognosemodelle für die Netzintegration wetterabhängiger Energieträger, www.projekt-eweline.de

Applications and Comparisons of Four Time Series Models in Epidemiological Surveillance Data

PubMed Central

Young, Alistair A.; Li, Xiaosong

2014-01-01

Public health surveillance systems provide valuable data for reliable predication of future epidemic events. This paper describes a study that used nine types of infectious disease data collected through a national public health surveillance system in mainland China to evaluate and compare the performances of four time series methods, namely, two decomposition methods (regression and exponential smoothing), autoregressive integrated moving average (ARIMA) and support vector machine (SVM). The data obtained from 2005 to 2011 and in 2012 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 accuracy of the statistical models in forecasting future epidemic disease proved their effectiveness in epidemiological surveillance. Although the comparisons found that no single method is completely superior to the others, the present study indeed highlighted that the SVMs outperforms the ARIMA model and decomposition methods in most cases. PMID:24505382

Variational Assimilation of GOME Total-Column Ozone Satellite Data in a 2D Latitude-Longitude Tracer-Transport Model.

NASA Astrophysics Data System (ADS)

Eskes, H. J.; Piters, A. J. M.; Levelt, P. F.; Allaart, M. A. F.; Kelder, H. M.

1999-10-01

A four-dimensional data-assimilation method is described to derive synoptic ozone fields from total-column ozone satellite measurements. The ozone columns are advected by a 2D tracer-transport model, using ECMWF wind fields at a single pressure level. Special attention is paid to the modeling of the forecast error covariance and quality control. The temporal and spatial dependence of the forecast error is taken into account, resulting in a global error field at any instant in time that provides a local estimate of the accuracy of the assimilated field. The authors discuss the advantages of the 4D-variational (4D-Var) approach over sequential assimilation schemes. One of the attractive features of the 4D-Var technique is its ability to incorporate measurements at later times t > t0 in the analysis at time t0, in a way consistent with the time evolution as described by the model. This significantly improves the offline analyzed ozone fields.

Applications systems verification and transfer project. Volume 1: Operational applications of satellite snow cover observations: Executive summary. [usefulness of satellite snow-cover data for water yield prediction

NASA Technical Reports Server (NTRS)

Rango, A.

1981-01-01

Both LANDSAT and NOAA satellite data were used in improving snowmelt runoff forecasts. When the satellite snow cover data were tested in both empirical seasonal runoff estimation and short term modeling approaches, a definite potential for reducing forecast error was evident. A cost benefit analysis run in conjunction with the snow mapping indicated a $36.5 million annual benefit accruing from a one percent improvement in forecast accuracy using the snow cover data for the western United States. The annual cost of employing the system would be $505,000. The snow mapping has proven that satellite snow cover data can be used to reduce snowmelt runoff forecast error in a cost effective manner once all operational satellite data are available within 72 hours after acquisition. Executive summaries of the individual snow mapping projects are presented.

Uncertainty estimation of long-range ensemble forecasts of snowmelt flood characteristics

NASA Astrophysics Data System (ADS)

Kuchment, L.

2012-04-01

Long-range forecasts of snowmelt flood characteristics with the lead time of 2-3 months have important significance for regulation of flood runoff and mitigation of flood damages at almost all large Russian rivers At the same time, the application of current forecasting techniques based on regression relationships between the runoff volume and the indexes of river basin conditions can lead to serious errors in forecasting resulted in large economic losses caused by wrong flood regulation. The forecast errors can be caused by complicated processes of soil freezing and soil moisture redistribution, too high rate of snow melt, large liquid precipitation before snow melt. or by large difference of meteorological conditions during the lead-time periods from climatologic ones. Analysis of economic losses had shown that the largest damages could, to a significant extent, be avoided if the decision makers had an opportunity to take into account predictive uncertainty and could use more cautious strategies in runoff regulation. Development of methodology of long-range ensemble forecasting of spring/summer floods which is based on distributed physically-based runoff generation models has created, in principle, a new basis for improving hydrological predictions as well as for estimating their uncertainty. This approach is illustrated by forecasting of the spring-summer floods at the Vyatka River and the Seim River basins. The application of the physically - based models of snowmelt runoff generation give a essential improving of statistical estimates of the deterministic forecasts of the flood volume in comparison with the forecasts obtained from the regression relationships. These models had been used also for the probabilistic forecasts assigning meteorological inputs during lead time periods from the available historical daily series, and from the series simulated by using a weather generator and the Monte Carlo procedure. The weather generator consists of the stochastic models of daily temperature and precipitation. The performance of the probabilistic forecasts were estimated by the ranked probability skill scores. The application of Monte Carlo simulations using weather generator has given better results then using the historical meteorological series.

Improving High-Resolution Weather Forecasts Using the Weather Research and Forecasting (WRF) Model with an Updated Kain–Fritsch Scheme

EPA Science Inventory

Efforts to improve the prediction accuracy of high-resolution (1–10 km) surface precipitation distribution and variability are of vital importance to local aspects of air pollution, wet deposition, and regional climate. However, precipitation biases and errors can occur at ...

Optimization of Coronal Mass Ejection Ensemble Forecasting Using WSA-ENLIL with Coned Model

DTIC Science & Technology

2013-03-01

previous versions by a large margin. The mean absolute forecast error of the median ensemble results was improved by over 43% over the original Coned...for reference for the six extra CMEs. .............................................................................................54 Figure 19...single-shot runs) with the flare location noted for reference for the six extra CMEs

Probabilistic accounting of uncertainty in forecasts of species distributions under climate change

Treesearch

Seth J. Wenger; Nicholas A. Som; Daniel C. Dauwalter; Daniel J. Isaak; Helen M. Neville; Charles H. Luce; Jason B. Dunham; Michael K. Young; Kurt D. Fausch; Bruce E. Rieman

2013-01-01

Forecasts of species distributions under future climates are inherently uncertain, but there have been few attempts to describe this uncertainty comprehensively in a probabilistic manner. We developed a Monte Carlo approach that accounts for uncertainty within generalized linear regression models (parameter uncertainty and residual error), uncertainty among competing...

Shared Investment Projects and Forecasting Errors: Setting Framework Conditions for Coordination and Sequencing Data Quality Activities

PubMed Central

Leitner, Stephan; Brauneis, Alexander; Rausch, Alexandra

2015-01-01

In this paper, we investigate the impact of inaccurate forecasting on the coordination of distributed investment decisions. In particular, by setting up a computational multi-agent model of a stylized firm, we investigate the case of investment opportunities that are mutually carried out by organizational departments. The forecasts of concern pertain to the initial amount of money necessary to launch and operate an investment opportunity, to the expected intertemporal distribution of cash flows, and the departments’ efficiency in operating the investment opportunity at hand. We propose a budget allocation mechanism for coordinating such distributed decisions The paper provides guidance on how to set framework conditions, in terms of the number of investment opportunities considered in one round of funding and the number of departments operating one investment opportunity, so that the coordination mechanism is highly robust to forecasting errors. Furthermore, we show that—in some setups—a certain extent of misforecasting is desirable from the firm’s point of view as it supports the achievement of the corporate objective of value maximization. We then address the question of how to improve forecasting quality in the best possible way, and provide policy advice on how to sequence activities for improving forecasting quality so that the robustness of the coordination mechanism to errors increases in the best possible way. At the same time, we show that wrong decisions regarding the sequencing can lead to a decrease in robustness. Finally, we conduct a comprehensive sensitivity analysis and prove that—in particular for relatively good forecasters—most of our results are robust to changes in setting the parameters of our multi-agent simulation model. PMID:25803736

A probabilistic approach to the drag-based model

NASA Astrophysics Data System (ADS)

Napoletano, Gianluca; Forte, Roberta; Moro, Dario Del; Pietropaolo, Ermanno; Giovannelli, Luca; Berrilli, Francesco

2018-02-01

The forecast of the time of arrival (ToA) of a coronal mass ejection (CME) to Earth is of critical importance for our high-technology society and for any future manned exploration of the Solar System. As critical as the forecast accuracy is the knowledge of its precision, i.e. the error associated to the estimate. We propose a statistical approach for the computation of the ToA using the drag-based model by introducing the probability distributions, rather than exact values, as input parameters, thus allowing the evaluation of the uncertainty on the forecast. We test this approach using a set of CMEs whose transit times are known, and obtain extremely promising results: the average value of the absolute differences between measure and forecast is 9.1h, and half of these residuals are within the estimated errors. These results suggest that this approach deserves further investigation. We are working to realize a real-time implementation which ingests the outputs of automated CME tracking algorithms as inputs to create a database of events useful for a further validation of the approach.

Potential predictability and forecast skill in ensemble climate forecast: a skill-persistence rule

NASA Astrophysics Data System (ADS)

Jin, Yishuai; Rong, Xinyao; Liu, Zhengyu

2017-12-01

This study investigates the factors relationship between the forecast skills for the real world (actual skill) and perfect model (perfect skill) in ensemble climate model forecast with a series of fully coupled general circulation model forecast experiments. It is found that the actual skill for sea surface temperature (SST) in seasonal forecast is substantially higher than the perfect skill on a large part of the tropical oceans, especially the tropical Indian Ocean and the central-eastern Pacific Ocean. The higher actual skill is found to be related to the higher observational SST persistence, suggesting a skill-persistence rule: a higher SST persistence in the real world than in the model could overwhelm the model bias to produce a higher forecast skill for the real world than for the perfect model. The relation between forecast skill and persistence is further proved using a first-order autoregressive model (AR1) analytically for theoretical solutions and numerically for analogue experiments. The AR1 model study shows that the skill-persistence rule is strictly valid in the case of infinite ensemble size, but could be distorted by sampling errors and non-AR1 processes. This study suggests that the so called "perfect skill" is model dependent and cannot serve as an accurate estimate of the true upper limit of real world prediction skill, unless the model can capture at least the persistence property of the observation.

CBO’s Revenue Forecasting Record

DTIC Science & Technology

2015-11-01

1983 1988 1993 1998 2003 2008 2013 -10 0 10 20 30 CBO Administration CBO’s Mean Forecast Error (1.1%) Forecast Errors for CBO’s and the...Administration’s Two-Year Revenue Projections CONGRESS OF THE UNITED STATES CONGRESSIONAL BUDGET OFFICE CBO CBO’s Revenue Forecasting Record NOVEMBER 2015...

A hybrid wavelet de-noising and Rank-Set Pair Analysis approach for forecasting hydro-meteorological time series.

PubMed

Wang, Dong; Borthwick, Alistair G; He, Handan; Wang, Yuankun; Zhu, Jieyu; Lu, Yuan; Xu, Pengcheng; Zeng, Xiankui; Wu, Jichun; Wang, Lachun; Zou, Xinqing; Liu, Jiufu; Zou, Ying; He, Ruimin

2018-01-01

Accurate, fast forecasting of hydro-meteorological time series is presently a major challenge in drought and flood mitigation. This paper proposes a hybrid approach, wavelet de-noising (WD) and Rank-Set Pair Analysis (RSPA), that takes full advantage of a combination of the two approaches to improve forecasts of hydro-meteorological time series. WD allows decomposition and reconstruction of a time series by the wavelet transform, and hence separation of the noise from the original series. RSPA, a more reliable and efficient version of Set Pair Analysis, is integrated with WD to form the hybrid WD-RSPA approach. Two types of hydro-meteorological data sets with different characteristics and different levels of human influences at some representative stations are used to illustrate the WD-RSPA approach. The approach is also compared to three other generic methods: the conventional Auto Regressive Integrated Moving Average (ARIMA) method, Artificial Neural Networks (ANNs) (BP-error Back Propagation, MLP-Multilayer Perceptron and RBF-Radial Basis Function), and RSPA alone. Nine error metrics are used to evaluate the model performance. Compared to three other generic methods, the results generated by WD-REPA model presented invariably smaller error measures which means the forecasting capability of the WD-REPA model is better than other models. The results show that WD-RSPA is accurate, feasible, and effective. In particular, WD-RSPA is found to be the best among the various generic methods compared in this paper, even when the extreme events are included within a time series. Copyright © 2017 Elsevier Inc. All rights reserved.

An optimized Nash nonlinear grey Bernoulli model based on particle swarm optimization and its application in prediction for the incidence of Hepatitis B in Xinjiang, China.

PubMed

Zhang, Liping; Zheng, Yanling; Wang, Kai; Zhang, Xueliang; Zheng, Yujian

2014-06-01

In this paper, by using a particle swarm optimization algorithm to solve the optimal parameter estimation problem, an improved Nash nonlinear grey Bernoulli model termed PSO-NNGBM(1,1) is proposed. To test the forecasting performance, the optimized model is applied for forecasting the incidence of hepatitis B in Xinjiang, China. Four models, traditional GM(1,1), grey Verhulst model (GVM), original nonlinear grey Bernoulli model (NGBM(1,1)) and Holt-Winters exponential smoothing method, are also established for comparison with the proposed model under the criteria of mean absolute percentage error and root mean square percent error. The prediction results show that the optimized NNGBM(1,1) model is more accurate and performs better than the traditional GM(1,1), GVM, NGBM(1,1) and Holt-Winters exponential smoothing method. Copyright © 2014. Published by Elsevier Ltd.

Sensitivity of physical parameterizations on prediction of tropical cyclone Nargis over the Bay of Bengal using WRF model

NASA Astrophysics Data System (ADS)

Raju, P. V. S.; Potty, Jayaraman; Mohanty, U. C.

2011-09-01

Comprehensive sensitivity analyses on physical parameterization schemes of Weather Research Forecast (WRF-ARW core) model have been carried out for the prediction of track and intensity of tropical cyclones by taking the example of cyclone Nargis, which formed over the Bay of Bengal and hit Myanmar on 02 May 2008, causing widespread damages in terms of human and economic losses. The model performances are also evaluated with different initial conditions of 12 h intervals starting from the cyclogenesis to the near landfall time. The initial and boundary conditions for all the model simulations are drawn from the global operational analysis and forecast products of National Center for Environmental Prediction (NCEP-GFS) available for the public at 1° lon/lat resolution. The results of the sensitivity analyses indicate that a combination of non-local parabolic type exchange coefficient PBL scheme of Yonsei University (YSU), deep and shallow convection scheme with mass flux approach for cumulus parameterization (Kain-Fritsch), and NCEP operational cloud microphysics scheme with diagnostic mixed phase processes (Ferrier), predicts better track and intensity as compared against the Joint Typhoon Warning Center (JTWC) estimates. Further, the final choice of the physical parameterization schemes selected from the above sensitivity experiments is used for model integration with different initial conditions. The results reveal that the cyclone track, intensity and time of landfall are well simulated by the model with an average intensity error of about 8 hPa, maximum wind error of 12 m s-1and track error of 77 km. The simulations also show that the landfall time error and intensity error are decreasing with delayed initial condition, suggesting that the model forecast is more dependable when the cyclone approaches the coast. The distribution and intensity of rainfall are also well simulated by the model and comparable with the TRMM estimates.

Limited Area Predictability: Is There A Limit To The Operational Usefulness of A Lam

NASA Astrophysics Data System (ADS)

Mesinger, F.

The issue of the limited area predictability in the context of the operational experience of the Eta Model, driven by the LBCs of the NCEP global spectral (Avn) model, is examined. The traditional view is that "the contamination at the lateral boundaries ... limits the operational usefulness of the LAM beyond some forecast time range". In the case of the Eta this contamination consists not only of the lower resolution of the Avn LBCs and the much discussed mathematical "lateral boundary error", but also of the use of the LBCs of the previous Avn run, at 0000 and 1200 UTC estimated to amount to about an 8 h loss in accuracy. Looking for the signs of the Eta accuracy in relative terms falling behind that of the Avn we have examined the trend of the Eta vs Avn precipitation scores, the rms fits to raobs of the two models as a function of time, and the errors of these models at extended forecast times in placing the centers of major lows. In none of these efforts, some including forecasts out to 84 h, we were able to notice signs of the Eta accuracy being visibly affected by the inflow of the lateral boundary errors. It is therefore hypothesized that some of the Eta design features compensate for the increasing influence of the Avn LBC errors. Candidate features are discussed, with the eta coordinate being a contender to play a major role. This situation being possible for the pair of models discussed, existence of a general limit for the operational usefulness of a LAM seems questionable.

On the effect of model parameters on forecast objects

NASA Astrophysics Data System (ADS)

Marzban, Caren; Jones, Corinne; Li, Ning; Sandgathe, Scott

2018-04-01

Many physics-based numerical models produce a gridded, spatial field of forecasts, e.g., a temperature map. The field for some quantities generally consists of spatially coherent and disconnected objects. Such objects arise in many problems, including precipitation forecasts in atmospheric models, eddy currents in ocean models, and models of forest fires. Certain features of these objects (e.g., location, size, intensity, and shape) are generally of interest. Here, a methodology is developed for assessing the impact of model parameters on the features of forecast objects. The main ingredients of the methodology include the use of (1) Latin hypercube sampling for varying the values of the model parameters, (2) statistical clustering algorithms for identifying objects, (3) multivariate multiple regression for assessing the impact of multiple model parameters on the distribution (across the forecast domain) of object features, and (4) methods for reducing the number of hypothesis tests and controlling the resulting errors. The final output of the methodology is a series of box plots and confidence intervals that visually display the sensitivities. The methodology is demonstrated on precipitation forecasts from a mesoscale numerical weather prediction model.

Implementation and Research on the Operational Use of the Mesoscale Prediction Model COAMPS in Poland

DTIC Science & Technology

2008-09-30

participated in EGU General Assembly , Vienna Austria 13-18 April 2008, giving a poster presentation. Bogumil Jakubiak, University of Warsaw...participated in EGU General Assembly , Vienna Austria 13-18 April 2008, giving two posters presentation. Mikolaj Sierzega, University of Warwick – participated...model forecast to generate background error statistics. This helps us to identify and understand the uncertainties in high-resolution NWP forecasts

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.

The Weighted-Average Lagged Ensemble.

PubMed

DelSole, T; Trenary, L; Tippett, M K

2017-11-01

A lagged ensemble is an ensemble of forecasts from the same model initialized at different times but verifying at the same time. The skill of a lagged ensemble mean can be improved by assigning weights to different forecasts in such a way as to maximize skill. If the forecasts are bias corrected, then an unbiased weighted lagged ensemble requires the weights to sum to one. Such a scheme is called a weighted-average lagged ensemble. In the limit of uncorrelated errors, the optimal weights are positive and decay monotonically with lead time, so that the least skillful forecasts have the least weight. In more realistic applications, the optimal weights do not always behave this way. This paper presents a series of analytic examples designed to illuminate conditions under which the weights of an optimal weighted-average lagged ensemble become negative or depend nonmonotonically on lead time. It is shown that negative weights are most likely to occur when the errors grow rapidly and are highly correlated across lead time. The weights are most likely to behave nonmonotonically when the mean square error is approximately constant over the range forecasts included in the lagged ensemble. An extreme example of the latter behavior is presented in which the optimal weights vanish everywhere except at the shortest and longest lead times.

Simulation studies of the application of SEASAT data in weather and state of sea forecasting models

NASA Technical Reports Server (NTRS)

Cardone, V. J.; Greenwood, J. A.

1979-01-01

The design and analysis of SEASAT simulation studies in which the error structure of conventional analyses and forecasts is modeled realistically are presented. The development and computer implementation of a global spectral ocean wave model is described. The design of algorithms for the assimilation of theoretical wind data into computers and for the utilization of real wind data and wave height data in a coupled computer system are presented.

A comparative study on GM (1,1) and FRMGM (1,1) model in forecasting FBM KLCI

NASA Astrophysics Data System (ADS)

Ying, Sah Pei; Zakaria, Syerrina; Mutalib, Sharifah Sakinah Syed Abd

2017-11-01

FTSE Bursa Malaysia Kuala Lumpur Composite Index (FBM KLCI) is a group of indexes combined in a standardized way and is used to measure the Malaysia overall market across the time. Although composite index can give ideas about stock market to investors, it is hard to predict accurately because it is volatile and it is necessary to identify a best model to forecast FBM KLCI. The objective of this study is to determine the most accurate forecasting model between GM (1,1) model and Fourier Residual Modification GM (1,1) (FRMGM (1,1)) model to forecast FBM KLCI. In this study, the actual daily closing data of FBM KLCI was collected from January 1, 2016 to March 15, 2016. GM (1,1) model and FRMGM (1,1) model were used to build the grey model and to test forecasting power of both models. Mean Absolute Percentage Error (MAPE) was used as a measure to determine the best model. Forecasted value by FRMGM (1,1) model do not differ much than the actual value compare to GM (1,1) model for in-sample and out-sample data. Results from MAPE also show that FRMGM (1,1) model is lower than GM (1,1) model for in-sample and out-sample data. These results shown that FRMGM (1,1) model is better than GM (1,1) model to forecast FBM KLCI.

Hybrid empirical mode decomposition- ARIMA for forecasting exchange rates

NASA Astrophysics Data System (ADS)

Abadan, Siti Sarah; Shabri, Ani; Ismail, Shuhaida

2015-02-01

This paper studied the forecasting of monthly Malaysian Ringgit (MYR)/ United State Dollar (USD) exchange rates using the hybrid of two methods which are the empirical model decomposition (EMD) and the autoregressive integrated moving average (ARIMA). MYR is pegged to USD during the Asian financial crisis causing the exchange rates are fixed to 3.800 from 2nd of September 1998 until 21st of July 2005. Thus, the chosen data in this paper is the post-July 2005 data, starting from August 2005 to July 2010. The comparative study using root mean square error (RMSE) and mean absolute error (MAE) showed that the EMD-ARIMA outperformed the single-ARIMA and the random walk benchmark model.

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.

Ensemble Streamflow Prediction in Korea: Past and Future 5 Years

NASA Astrophysics Data System (ADS)

Jeong, D.; Kim, Y.; Lee, J.

2005-05-01

The Ensemble Streamflow Prediction (ESP) approach was first introduced in 2000 by the Hydrology Research Group (HRG) at Seoul National University as an alternative probabilistic forecasting technique for improving the 'Water Supply Outlook' That is issued every month by the Ministry of Construction and Transportation in Korea. That study motivated the Korea Water Resources Corporation (KOWACO) to establish their seasonal probabilistic forecasting system for the 5 major river basins using the ESP approach. In cooperation with the HRG, the KOWACO developed monthly optimal multi-reservoir operating systems for the Geum river basin in 2004, which coupled the ESP forecasts with an optimization model using sampling stochastic dynamic programming. The user interfaces for both ESP and SSDP have also been designed for the developed computer systems to become more practical. More projects for developing ESP systems to the other 3 major river basins (i.e. the Nakdong, Han and Seomjin river basins) was also completed by the HRG and KOWACO at the end of December 2004. Therefore, the ESP system has become the most important mid- and long-term streamflow forecast technique in Korea. In addition to the practical aspects, resent research experience on ESP has raised some concerns into ways of improving the accuracy of ESP in Korea. Jeong and Kim (2002) performed an error analysis on its resulting probabilistic forecasts and found that the modeling error is dominant in the dry season, while the meteorological error is dominant in the flood season. To address the first issue, Kim et al. (2004) tested various combinations and/or combining techniques and showed that the ESP probabilistic accuracy could be improved considerably during the dry season when the hydrologic models were combined and/or corrected. In addition, an attempt was also made to improve the ESP accuracy for the flood season using climate forecast information. This ongoing project handles three types of climate forecast information: (1) the Monthly Industrial Meteorology Information Magazine (MIMIM) of the Korea Meteorological Administration (2) the Global Data Assimilation Prediction System (GDAPS), and (3) the US National Centers for Environmental Prediction (NCEP). Each of these forecasts is issued in a unique format: (1) MIMIM is a most-probable-event forecast, (2) GDAPS is a single series of deterministic forecasts, and (3) NCEP is an ensemble of deterministic forecasts. Other minor issues include how long the initial conditions influences the ESP accuracy, and how many ESP scenarios are needed to obtain the best accuracy. This presentation also addresses some future research that is needed for ESP in Korea.

An impact analysis of forecasting methods and forecasting parameters on bullwhip effect

NASA Astrophysics Data System (ADS)

Silitonga, R. Y. H.; Jelly, N.

2018-04-01

Bullwhip effect is an increase of variance of demand fluctuation from downstream to upstream of supply chain. Forecasting methods and forecasting parameters were recognized as some factors that affect bullwhip phenomena. To study these factors, we can develop simulations. There are several ways to simulate bullwhip effect in previous studies, such as mathematical equation modelling, information control modelling, computer program, and many more. In this study a spreadsheet program named Bullwhip Explorer was used to simulate bullwhip effect. Several scenarios were developed to show the change in bullwhip effect ratio because of the difference in forecasting methods and forecasting parameters. Forecasting methods used were mean demand, moving average, exponential smoothing, demand signalling, and minimum expected mean squared error. Forecasting parameters were moving average period, smoothing parameter, signalling factor, and safety stock factor. It showed that decreasing moving average period, increasing smoothing parameter, increasing signalling factor can create bigger bullwhip effect ratio. Meanwhile, safety stock factor had no impact to bullwhip effect.

Data assimilation method based on the constraints of confidence region

NASA Astrophysics Data System (ADS)

Li, Yong; Li, Siming; Sheng, Yao; Wang, Luheng

2018-03-01

The ensemble Kalman filter (EnKF) is a distinguished data assimilation method that is widely used and studied in various fields including methodology and oceanography. However, due to the limited sample size or imprecise dynamics model, it is usually easy for the forecast error variance to be underestimated, which further leads to the phenomenon of filter divergence. Additionally, the assimilation results of the initial stage are poor if the initial condition settings differ greatly from the true initial state. To address these problems, the variance inflation procedure is usually adopted. In this paper, we propose a new method based on the constraints of a confidence region constructed by the observations, called EnCR, to estimate the inflation parameter of the forecast error variance of the EnKF method. In the new method, the state estimate is more robust to both the inaccurate forecast models and initial condition settings. The new method is compared with other adaptive data assimilation methods in the Lorenz-63 and Lorenz-96 models under various model parameter settings. The simulation results show that the new method performs better than the competing methods.

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.; Odstrcil, D.; MacNeice, P. J.; Rastaetter, L.; LaSota, J. A.

2014-12-01

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. Real-time ensemble modeling of CME propagation 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). To estimate the effect of uncertainties in determining CME input parameters on arrival time predictions, a distribution of n (routinely n=48) CME input parameter sets 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, including a probability distribution of CME arrival times (for hits), and geomagnetic storm strength (for Earth-directed hits). We present the results of ensemble simulations for a total of 38 CME events in 2013-2014. For 28 of the ensemble runs containing hits, the observed CME arrival was within the range of ensemble arrival time predictions for 14 runs (half). The average arrival time prediction was computed for each of the 28 ensembles predicting hits and using the actual arrival time, an average absolute error of 10.0 hours (RMSE=11.4 hours) was found for all 28 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 sysem 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. The parameter sensitivity study suggests future directions for the system, such as running ensembles using various magnetogram inputs to the WSA model.

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.

A stochastic post-processing method for solar irradiance forecasts derived from NWPs models

NASA Astrophysics Data System (ADS)

Lara-Fanego, V.; Pozo-Vazquez, D.; Ruiz-Arias, J. A.; Santos-Alamillos, F. J.; Tovar-Pescador, J.

2010-09-01

Solar irradiance forecast is an important area of research for the future of the solar-based renewable energy systems. Numerical Weather Prediction models (NWPs) have proved to be a valuable tool for solar irradiance forecasting with lead time up to a few days. Nevertheless, these models show low skill in forecasting the solar irradiance under cloudy conditions. Additionally, climatic (averaged over seasons) aerosol loading are usually considered in these models, leading to considerable errors for the Direct Normal Irradiance (DNI) forecasts during high aerosols load conditions. In this work we propose a post-processing method for the Global Irradiance (GHI) and DNI forecasts derived from NWPs. Particularly, the methods is based on the use of Autoregressive Moving Average with External Explanatory Variables (ARMAX) stochastic models. These models are applied to the residuals of the NWPs forecasts and uses as external variables the measured cloud fraction and aerosol loading of the day previous to the forecast. The method is evaluated for a set one-moth length three-days-ahead forecast of the GHI and DNI, obtained based on the WRF mesoscale atmospheric model, for several locations in Andalusia (Southern Spain). The Cloud fraction is derived from MSG satellite estimates and the aerosol loading from the MODIS platform estimates. Both sources of information are readily available at the time of the forecast. Results showed a considerable improvement of the forecasting skill of the WRF model using the proposed post-processing method. Particularly, relative improvement (in terms of the RMSE) for the DNI during summer is about 20%. A similar value is obtained for the GHI during the winter.

Exploiting Domain Knowledge to Forecast Heating Oil Consumption

NASA Astrophysics Data System (ADS)

Corliss, George F.; Sakauchi, Tsuginosuke; Vitullo, Steven R.; Brown, Ronald H.

2011-11-01

The GasDay laboratory at Marquette University provides forecasts of energy consumption. One such service is the Heating Oil Forecaster, a service for a heating oil or propane delivery company. Accurate forecasts can help reduce the number of trucks and drivers while providing efficient inventory management by stretching the time between deliveries. Accurate forecasts help retain valuable customers. If a customer runs out of fuel, the delivery service incurs costs for an emergency delivery and often a service call. Further, the customer probably changes providers. The basic modeling is simple: Fit delivery amounts sk to cumulative Heating Degree Days (HDDk = Σmax(0,60 °F—daily average temperature)), with wind adjustment, for each delivery period: sk≈ŝk = β0+β1HDDk. For the first few deliveries, there is not enough data to provide a reliable estimate K = 1/β1 so we use Bayesian techniques with priors constructed from historical data. A fresh model is trained for each customer with each delivery, producing daily consumption forecasts using actual and forecast weather until the next delivery. In practice, a delivery may not fill the oil tank if the delivery truck runs out of oil or the automatic shut-off activates prematurely. Special outlier detection and recovery based on domain knowledge addresses this and other special cases. The error at each delivery is the difference between that delivery and the aggregate of daily forecasts using actual weather since the preceding delivery. Out-of-sample testing yields MAPE = 21.2% and an average error of 6.0% of tank capacity for Company A. The MAPE and an average error as a percentage of tank capacity for Company B are 31.5 % and 6.6 %, respectively. One heating oil delivery company who uses this forecasting service [1] reported instances of a customer running out of oil reduced from about 250 in 50,000 deliveries per year before contracting for our service to about 10 with our service. They delivered slightly more oil with 20 % fewer trucks and drivers, citing 250,000 annual savings in operational costs.

Comparative analysis of operational forecasts versus actual weather conditions in airline flight planning: Summary report

NASA Technical Reports Server (NTRS)

Keitz, J. F.

1982-01-01

The impact of more timely and accurate weather data on airline flight planning with the emphasis on fuel savings is studied. This summary report discusses the results of each of the four major tasks of the study. Task 1 compared airline flight plans based on operational forecasts to plans based on the verifying analyses and found that average fuel savings of 1.2 to 2.5 percent are possible with improved forecasts. Task 2 consisted of similar comparisons but used a model developed for the FAA by SRI International that simulated the impact of ATc diversions on the flight plans. While parts of Task 2 confirm the Task I findings, inconsistency with other data and the known impact of ATC suggests that other Task 2 findings are the result of errors in the model. Task 3 compares segment weather data from operational flight plans with the weather actually observed by the aircraft and finds the average error could result in fuel burn penalties (or savings) of up to 3.6 percent for the average 8747 flight. In Task 4 an in-depth analysis of the weather forecast for the 33 days included in the study finds that significant errors exist on 15 days. Wind speeds in the area of maximum winds are underestimated by 20 to 50 kts., a finding confirmed in the other three tasks.

On an improvement of UV index forecast: UV index diagnosis and forecast for Belsk, Poland, in Spring/Summer 1999

NASA Astrophysics Data System (ADS)

Krzyścin, J. W.; Jaroslawski, J.; Sobolewski, P.

2001-10-01

A forecast of the UV index for the following day is presented. The standard approach to the UV index modelling is applied, i.e., the clear-sky UV index is multiplied by the UV cloud transmission factor. The input to the clear-sky model (tropospheric ultraviolet and visible-TUV model, Madronich, in: M. Tevini (Ed.), Environmental Effects of Ultraviolet Radiation, Lewis Publisher, Boca Raton, /1993, p. 17) consists of the total ozone forecast (by a regression model using the observed and forecasted meteorological variables taken as the initial values of aviation (AVN) global model and their 24-hour forecasts, respectively) and aerosols optical depth (AOD) forecast (assumed persistence). The cloud transmission factor forecast is inferred from the 24-h AVN model run for the total (Sun/+sky) solar irradiance at noon. The model is validated comparing the UV index forecasts with the observed values, which are derived from the daily pattern of the UV erythemal irradiance taken at Belsk (52°N,21°E), Poland, by means of the UV Biometer Solar model 501A for the period May-September 1999. Eighty-one percent and 92% of all forecasts fall into /+/-1 and /+/-2 index unit range, respectively. Underestimation of UV index occurs only in 15%. Thus, the model gives a high security in Sun protection for the public. It is found that in /~35% of all cases a more accurate forecast of AOD is needed to estimate the daily maximum of clear-sky irradiance with the error not exceeding 5%. The assumption of the persistence of the cloud characteristics appears as an alternative to the 24-h forecast of the cloud transmission factor in the case when the AVN prognoses are not available.

Analysis and Synthesis of Load Forecasting Data for Renewable Integration Studies: Preprint

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

Steckler, N.; Florita, A.; Zhang, J.

2013-11-01

As renewable energy constitutes greater portions of the generation fleet, the importance of modeling uncertainty as part of integration studies also increases. In pursuit of optimal system operations, it is important to capture not only the definitive behavior of power plants, but also the risks associated with systemwide interactions. This research examines the dependence of load forecast errors on external predictor variables such as temperature, day type, and time of day. The analysis was utilized to create statistically relevant instances of sequential load forecasts with only a time series of historic, measured load available. The creation of such load forecastsmore » relies on Bayesian techniques for informing and updating the model, thus providing a basis for networked and adaptive load forecast models in future operational applications.« less

Improvement of Storm Forecasts Using Gridded Bayesian Linear Regression for Northeast United States

NASA Astrophysics Data System (ADS)

Yang, J.; Astitha, M.; Schwartz, C. S.

2017-12-01

Bayesian linear regression (BLR) is a post-processing technique in which regression coefficients are derived and used to correct raw forecasts based on pairs of observation-model values. This study presents the development and application of a gridded Bayesian linear regression (GBLR) as a new post-processing technique to improve numerical weather prediction (NWP) of rain and wind storm forecasts over northeast United States. Ten controlled variables produced from ten ensemble members of the National Center for Atmospheric Research (NCAR) real-time prediction system are used for a GBLR model. In the GBLR framework, leave-one-storm-out cross-validation is utilized to study the performances of the post-processing technique in a database composed of 92 storms. To estimate the regression coefficients of the GBLR, optimization procedures that minimize the systematic and random error of predicted atmospheric variables (wind speed, precipitation, etc.) are implemented for the modeled-observed pairs of training storms. The regression coefficients calculated for meteorological stations of the National Weather Service are interpolated back to the model domain. An analysis of forecast improvements based on error reductions during the storms will demonstrate the value of GBLR approach. This presentation will also illustrate how the variances are optimized for the training partition in GBLR and discuss the verification strategy for grid points where no observations are available. The new post-processing technique is successful in improving wind speed and precipitation storm forecasts using past event-based data and has the potential to be implemented in real-time.

Forecasting influenza in Hong Kong with Google search queries and statistical model fusion

PubMed Central

Ramirez Ramirez, L. Leticia; Nezafati, Kusha; Zhang, Qingpeng; Tsui, Kwok-Leung

2017-01-01

Background The objective of this study is to investigate predictive utility of online social media and web search queries, particularly, Google search data, to forecast new cases of influenza-like-illness (ILI) in general outpatient clinics (GOPC) in Hong Kong. To mitigate the impact of sensitivity to self-excitement (i.e., fickle media interest) and other artifacts of online social media data, in our approach we fuse multiple offline and online data sources. Methods Four individual models: generalized linear model (GLM), least absolute shrinkage and selection operator (LASSO), autoregressive integrated moving average (ARIMA), and deep learning (DL) with Feedforward Neural Networks (FNN) are employed to forecast ILI-GOPC both one week and two weeks in advance. The covariates include Google search queries, meteorological data, and previously recorded offline ILI. To our knowledge, this is the first study that introduces deep learning methodology into surveillance of infectious diseases and investigates its predictive utility. Furthermore, to exploit the strength from each individual forecasting models, we use statistical model fusion, using Bayesian model averaging (BMA), which allows a systematic integration of multiple forecast scenarios. For each model, an adaptive approach is used to capture the recent relationship between ILI and covariates. Results DL with FNN appears to deliver the most competitive predictive performance among the four considered individual models. Combing all four models in a comprehensive BMA framework allows to further improve such predictive evaluation metrics as root mean squared error (RMSE) and mean absolute predictive error (MAPE). Nevertheless, DL with FNN remains the preferred method for predicting locations of influenza peaks. Conclusions The proposed approach can be viewed a feasible alternative to forecast ILI in Hong Kong or other countries where ILI has no constant seasonal trend and influenza data resources are limited. The proposed methodology is easily tractable and computationally efficient. PMID:28464015

Interest rate next-day variation prediction based on hybrid feedforward neural network, particle swarm optimization, and multiresolution techniques

NASA Astrophysics Data System (ADS)

Lahmiri, Salim

2016-02-01

Multiresolution analysis techniques including continuous wavelet transform, empirical mode decomposition, and variational mode decomposition are tested in the context of interest rate next-day variation prediction. In particular, multiresolution analysis techniques are used to decompose interest rate actual variation and feedforward neural network for training and prediction. Particle swarm optimization technique is adopted to optimize its initial weights. For comparison purpose, autoregressive moving average model, random walk process and the naive model are used as main reference models. In order to show the feasibility of the presented hybrid models that combine multiresolution analysis techniques and feedforward neural network optimized by particle swarm optimization, we used a set of six illustrative interest rates; including Moody's seasoned Aaa corporate bond yield, Moody's seasoned Baa corporate bond yield, 3-Month, 6-Month and 1-Year treasury bills, and effective federal fund rate. The forecasting results show that all multiresolution-based prediction systems outperform the conventional reference models on the criteria of mean absolute error, mean absolute deviation, and root mean-squared error. Therefore, it is advantageous to adopt hybrid multiresolution techniques and soft computing models to forecast interest rate daily variations as they provide good forecasting performance.

Estimating Inflows to Lake Okeechobee Using Climate Indices: A Machine Learning Modeling Approach

NASA Astrophysics Data System (ADS)

Kalra, A.; Ahmad, S.

2008-12-01

The operation of regional water management systems that include lakes and storage reservoirs for flood control and water supply can be significantly improved by using climate indices. This research is focused on forecasting Lag 1 annual inflow to Lake Okeechobee, located in South Florida, using annual oceanic- atmospheric indices of Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), Atlantic Multidecadal Oscillation (AMO), and El Nino-Southern Oscillations (ENSO). Support Vector Machine (SVM) and Least Square Support Vector Machine (LSSVM), belonging to the class of data driven models, are developed to forecast annual lake inflow using annual oceanic-atmospheric indices data from 1914 to 2003. The models were trained with 80 years of data and tested for 10 years of data. Based on Correlation Coefficient, Root Means Square Error, and Mean Absolute Error model predictions were in good agreement with measured inflow volumes. Sensitivity analysis, performed to evaluate the effect of individual and coupled oscillations, revealed a strong signal for AMO and ENSO indices compared to PDO and NAO indices for one year lead-time inflow forecast. Inflow predictions from the SVM models were better when compared with the predictions obtained from feed forward back propagation Artificial Neural Network (ANN) models.

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.

Evaluation of NOAA's High Resolution Rapid Refresh (HRRR), 12 km North America Model (NAM12) and 4km North America Model (NAM 4) hub-height wind speed forecasts

NASA Astrophysics Data System (ADS)

Pendergrass, W.; Vogel, C. A.

2013-12-01

As an outcome of discussions between Duke Energy Generation and NOAA/ARL following the 2009 AMS Summer Community Meeting, in Norman Oklahoma, ARL and Duke Energy Generation (Duke) signed a Cooperative Research and Development Agreement (CRADA) which allows NOAA to conduct atmospheric boundary layer (ABL) research using Duke renewable energy sites as research testbeds. One aspect of this research has been the evaluation of forecast hub-height winds from three NOAA atmospheric models. Forecasts of 10m (surface) and 80m (hub-height) wind speeds from (1) NOAA/GSD's High Resolution Rapid Refresh (HRRR) model, (2) NOAA/NCEP's 12 km North America Model (NAM12) and (3) NOAA/NCEP's 4k high resolution North America Model (NAM4) were evaluated against 18 months of surface-layer wind observations collected at the joint NOAA/Duke Energy research station located at Duke Energy's West Texas Ocotillo wind farm over the period April 2011 through October 2012. HRRR, NAM12 and NAM4 10m wind speed forecasts were compared with 10m level wind speed observations measured on the NOAA/ATDD flux-tower. Hub-height (80m) HRRR , NAM12 and NAM4 forecast wind speeds were evaluated against the 80m operational PMM27-28 meteorological tower supporting the Ocotillo wind farm. For each HRRR update, eight forecast hours (hour 01, 02, 03, 05, 07, 10, 12, 15) plus the initialization hour (hour 00), evaluated. For the NAM12 and NAM4 models forecast hours 00-24 from the 06z initialization were evaluated. Performance measures or skill score based on absolute error 50% cumulative probability were calculated for each forecast hour. HRRR forecast hour 01 provided the best skill score with an absolute wind speed error within 0.8 m/s of observed 10m wind speed and 1.25 m/s for hub-height wind speed at the designated 50% cumulative probability. For both NAM4 and NAM12 models, skill scores were diurnal with comparable best scores observed during the day of 0.7 m/s of observed 10m wind speed and 1.1 m/s for hub-height wind speed at the designated 50% cumulative probability level.

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.

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.

Increasing the Reliability of Circulation Model Validation: Quantifying Drifter Slip to See how Currents are Actually Moving

NASA Astrophysics Data System (ADS)

Anderson, T.

2016-02-01

Ocean circulation forecasts can help answer questions regarding larval dispersal, passive movement of injured sea animals, oil spill mitigation, and search and rescue efforts. Circulation forecasts are often validated with GPS-tracked drifter paths, but how accurately do these drifters actually move with ocean currents? Drifters are not only moved by water, but are also forced by wind and waves acting on the exposed buoy and transmitter; this imperfect movement is referred to as drifter slip. The quantification and further understanding of drifter slip will allow scientists to differentiate between drifter imperfections and actual computer model error when comparing trajectory forecasts with actual drifter tracks. This will avoid falsely accrediting all discrepancies between a trajectory forecast and an actual drifter track to computer model error. During multiple deployments of drifters in Nantucket Sound and using observed wind and wave data, we attempt to quantify the slip of drifters developed by the Northeast Fisheries Science Center's (NEFSC) Student Drifters Program. While similar studies have been conducted previously, very few have directly attached current meters to drifters to quantify drifter slip. Furthermore, none have quantified slip of NEFSC drifters relative to the oceanographic-standard "CODE" drifter. The NEFSC drifter archive has over 1000 drifter tracks primarily off the New England coast. With a better understanding of NEFSC drifter slip, modelers can reliably use these tracks for model validation.

Increasing the Reliability of Circulation Model Validation: Quantifying Drifter Slip to See how Currents are Actually Moving

NASA Astrophysics Data System (ADS)

Anderson, T.

2015-12-01

Ocean circulation forecasts can help answer questions regarding larval dispersal, passive movement of injured sea animals, oil spill mitigation, and search and rescue efforts. Circulation forecasts are often validated with GPS-tracked drifter paths, but how accurately do these drifters actually move with ocean currents? Drifters are not only moved by water, but are also forced by wind and waves acting on the exposed buoy and transmitter; this imperfect movement is referred to as drifter slip. The quantification and further understanding of drifter slip will allow scientists to differentiate between drifter imperfections and actual computer model error when comparing trajectory forecasts with actual drifter tracks. This will avoid falsely accrediting all discrepancies between a trajectory forecast and an actual drifter track to computer model error. During multiple deployments of drifters in Nantucket Sound and using observed wind and wave data, we attempt to quantify the slip of drifters developed by the Northeast Fisheries Science Center's (NEFSC) Student Drifters Program. While similar studies have been conducted previously, very few have directly attached current meters to drifters to quantify drifter slip. Furthermore, none have quantified slip of NEFSC drifters relative to the oceanographic-standard "CODE" drifter. The NEFSC drifter archive has over 1000 drifter tracks primarily off the New England coast. With a better understanding of NEFSC drifter slip, modelers can reliably use these tracks for model validation.

Assessing the Impact of Pre-gpm Microwave Precipitation Observations in the Goddard WRF Ensemble Data Assimilation System

NASA Technical Reports Server (NTRS)

Chambon, Philippe; Zhang, Sara Q.; Hou, Arthur Y.; Zupanski, Milija; Cheung, Samson

2013-01-01

The forthcoming Global Precipitation Measurement (GPM) Mission will provide next generation precipitation observations from a constellation of satellites. Since precipitation by nature has large variability and low predictability at cloud-resolving scales, the impact of precipitation data on the skills of mesoscale numerical weather prediction (NWP) is largely affected by the characterization of background and observation errors and the representation of nonlinear cloud/precipitation physics in an NWP data assimilation system. We present a data impact study on the assimilation of precipitation-affected microwave (MW) radiances from a pre-GPM satellite constellation using the Goddard WRF Ensemble Data Assimilation System (Goddard WRF-EDAS). A series of assimilation experiments are carried out in a Weather Research Forecast (WRF) model domain of 9 km resolution in western Europe. Sensitivities to observation error specifications, background error covariance estimated from ensemble forecasts with different ensemble sizes, and MW channel selections are examined through single-observation assimilation experiments. An empirical bias correction for precipitation-affected MW radiances is developed based on the statistics of radiance innovations in rainy areas. The data impact is assessed by full data assimilation cycling experiments for a storm event that occurred in France in September 2010. Results show that the assimilation of MW precipitation observations from a satellite constellation mimicking GPM has a positive impact on the accumulated rain forecasts verified with surface radar rain estimates. The case-study on a convective storm also reveals that the accuracy of ensemble-based background error covariance is limited by sampling errors and model errors such as precipitation displacement and unresolved convective scale instability.

Impact of PBL and convection parameterization schemes for prediction of severe land-falling Bay of Bengal cyclones using WRF-ARW model

NASA Astrophysics Data System (ADS)

Singh, K. S.; Bhaskaran, Prasad K.

2017-12-01

This study evaluates the performance of the Advanced Research Weather Research and Forecasting (WRF-ARW) model for prediction of land-falling Bay of Bengal (BoB) tropical cyclones (TCs). Model integration was performed using two-way interactive double nested domains at 27 and 9 km resolutions. The present study comprises two major components. Firstly, the study explores the impact of five different planetary boundary layer (PBL) and six cumulus convection (CC) schemes on seven land-falling BoB TCs. A total of 85 numerical simulations were studied in detail, and the results signify that the model simulated better both the track and intensity by using a combination of Yonsei University (YSU) PBL and the old simplified Arakawa-Schubert CC scheme. Secondly, the study also investigated the model performance based on the best possible combinations of model physics on the real-time forecasts of four BoB cyclones (Phailin, Helen, Lehar, and Madi) that made landfall during 2013 based on another 15 numerical simulations. The predicted mean track error during 2013 was about 71 km, 114 km, 133 km, 148 km, and 130 km respectively from day-1 to day-5. The Root Mean Square Error (RMSE) for Minimum Central Pressure (MCP) was about 6 hPa and the same noticed for Maximum Surface Wind (MSW) was about 4.5 m s-1 noticed during the entire simulation period. In addition the study also reveals that the predicted track errors during 2013 cyclones improved respectively by 43%, 44%, and 52% from day-1 to day-3 as compared to cyclones simulated during the period 2006-2011. The improvements noticed can be attributed due to relatively better quality data that was specified for the initial mean position error (about 48 km) during 2013. Overall the study signifies that the track and intensity forecast for 2013 cyclones using the specified combinations listed in the first part of this study performed relatively better than the other NWP (Numerical Weather Prediction) models, and thereby finds application in real-time forecast.

Constraining the ensemble Kalman filter for improved streamflow forecasting

NASA Astrophysics Data System (ADS)

Maxwell, Deborah H.; Jackson, Bethanna M.; McGregor, James

2018-05-01

Data assimilation techniques such as the Ensemble Kalman Filter (EnKF) are often applied to hydrological models with minimal state volume/capacity constraints enforced during ensemble generation. Flux constraints are rarely, if ever, applied. Consequently, model states can be adjusted beyond physically reasonable limits, compromising the integrity of model output. In this paper, we investigate the effect of constraining the EnKF on forecast performance. A "free run" in which no assimilation is applied is compared to a completely unconstrained EnKF implementation, a 'typical' hydrological implementation (in which mass constraints are enforced to ensure non-negativity and capacity thresholds of model states are not exceeded), and then to a more tightly constrained implementation where flux as well as mass constraints are imposed to force the rate of water movement to/from ensemble states to be within physically consistent boundaries. A three year period (2008-2010) was selected from the available data record (1976-2010). This was specifically chosen as it had no significant data gaps and represented well the range of flows observed in the longer dataset. Over this period, the standard implementation of the EnKF (no constraints) contained eight hydrological events where (multiple) physically inconsistent state adjustments were made. All were selected for analysis. Mass constraints alone did little to improve forecast performance; in fact, several were significantly degraded compared to the free run. In contrast, the combined use of mass and flux constraints significantly improved forecast performance in six events relative to all other implementations, while the remaining two events showed no significant difference in performance. Placing flux as well as mass constraints on the data assimilation framework encourages physically consistent state estimation and results in more accurate and reliable forward predictions of streamflow for robust decision-making. We also experiment with the observation error, which has a profound effect on filter performance. We note an interesting tension exists between specifying an error which reflects known uncertainties and errors in the measurement versus an error that allows "optimal" filter updating.

A scoping review of malaria forecasting: past work and future directions

PubMed Central

Zinszer, Kate; Verma, Aman D; Charland, Katia; Brewer, Timothy F; Brownstein, John S; Sun, Zhuoyu; Buckeridge, David L

2012-01-01

Objectives There is a growing body of literature on malaria forecasting methods and the objective of our review is to identify and assess methods, including predictors, used to forecast malaria. Design Scoping review. Two independent reviewers searched information sources, assessed studies for inclusion and extracted data from each study. Information sources Search strategies were developed and the following databases were searched: CAB Abstracts, EMBASE, Global Health, MEDLINE, ProQuest Dissertations & Theses and Web of Science. Key journals and websites were also manually searched. Eligibility criteria for included studies We included studies that forecasted incidence, prevalence or epidemics of malaria over time. A description of the forecasting model and an assessment of the forecast accuracy of the model were requirements for inclusion. Studies were restricted to human populations and to autochthonous transmission settings. Results We identified 29 different studies that met our inclusion criteria for this review. The forecasting approaches included statistical modelling, mathematical modelling and machine learning methods. Climate-related predictors were used consistently in forecasting models, with the most common predictors being rainfall, relative humidity, temperature and the normalised difference vegetation index. Model evaluation was typically based on a reserved portion of data and accuracy was measured in a variety of ways including mean-squared error and correlation coefficients. We could not compare the forecast accuracy of models from the different studies as the evaluation measures differed across the studies. Conclusions Applying different forecasting methods to the same data, exploring the predictive ability of non-environmental variables, including transmission reducing interventions and using common forecast accuracy measures will allow malaria researchers to compare and improve models and methods, which should improve the quality of malaria forecasting. PMID:23180505

Impact of using scatterometer and altimeter data on storm surge forecasting

NASA Astrophysics Data System (ADS)

Bajo, Marco; De Biasio, Francesco; Umgiesser, Georg; Vignudelli, Stefano; Zecchetto, Stefano

2017-05-01

Satellite data are rarely used in storm surge models because of the lack of established methodologies. Nevertheless, they can provide useful information on surface wind and sea level, which can potentially improve the forecast. In this paper satellite wind data are used to correct the bias of wind originating from a global atmospheric model, while satellite sea level data are used to improve the initial conditions of the model simulations. In a first step, the capability of global winds (biased and unbiased) to adequately force a storm surge model are assessed against that of a high resolution local wind. Then, the added value of direct assimilation of satellite altimeter data in the storm surge model is tested. Eleven storm surge events, recorded in Venice from 2008 to 2012, are simulated using different configurations of wind forcing and altimeter data assimilation. Focusing on the maximum surge peak, results show that the relative error, averaged over the eleven cases considered, decreases from 13% to 7%, using both the unbiased wind and assimilating the altimeter data, while, if the high resolution local wind is used to force the hydrodynamic model, the altimeter data assimilation reduces the error from 9% to 6%. Yet, the overall capabilities in reproducing the surge in the first day of forecast, measured by the correlation and by the rms error, improve only with the use of the unbiased global wind and not with the use of high resolution local wind and altimeter data assimilation.

Monthly forecasting of agricultural pests in Switzerland

NASA Astrophysics Data System (ADS)

Hirschi, M.; Dubrovsky, M.; Spirig, C.; Samietz, J.; Calanca, P.; Weigel, A. P.; Fischer, A. M.; Rotach, M. W.

2012-04-01

Given the repercussions of pests and diseases on agricultural production, detailed forecasting tools have been developed to simulate the degree of infestation depending on actual weather conditions. The life cycle of pests is most successfully predicted if the micro-climate of the immediate environment (habitat) of the causative organisms can be simulated. Sub-seasonal pest forecasts therefore require weather information for the relevant habitats and the appropriate time scale. The pest forecasting system SOPRA (www.sopra.info) currently in operation in Switzerland relies on such detailed weather information, using hourly weather observations up to the day the forecast is issued, but only a climatology for the forecasting period. Here, we aim at improving the skill of SOPRA forecasts by transforming the weekly information provided by ECMWF monthly forecasts (MOFCs) into hourly weather series as required for the prediction of upcoming life phases of the codling moth, the major insect pest in apple orchards worldwide. Due to the probabilistic nature of operational monthly forecasts and the limited spatial and temporal resolution, their information needs to be post-processed for use in a pest model. In this study, we developed a statistical downscaling approach for MOFCs that includes the following steps: (i) application of a stochastic weather generator to generate a large pool of daily weather series consistent with the climate at a specific location, (ii) a subsequent re-sampling of weather series from this pool to optimally represent the evolution of the weekly MOFC anomalies, and (iii) a final extension to hourly weather series suitable for the pest forecasting model. Results show a clear improvement in the forecast skill of occurrences of upcoming codling moth life phases when incorporating MOFCs as compared to the operational pest forecasting system. This is true both in terms of root mean squared errors and of the continuous rank probability scores of the probabilistic forecasts vs. the mean absolute errors of the deterministic system. Also, the application of the climate conserving recalibration (CCR, Weigel et al. 2009) technique allows for successful correction of the under-confidence in the forecasted occurrences of codling moth life phases. Reference: Weigel, A. P.; Liniger, M. A. & Appenzeller, C. (2009). Seasonal Ensemble Forecasts: Are Recalibrated Single Models Better than Multimodels? Mon. Wea. Rev., 137, 1460-1479.

A Module for Assimilating Hyperspectral Infrared Retrieved Profiles into the Gridpoint Statistical Interpolation System for Unique Forecasting Applications

NASA Technical Reports Server (NTRS)

Berndt, Emily; Zavodsky, Bradley; Srikishen, Jayanthi; Blankenship, Clay

2015-01-01

Hyperspectral infrared sounder radiance data are assimilated into operational modeling systems however the process is computationally expensive and only approximately 1% of available data are assimilated due to data thinning as well as the fact that radiances are restricted to cloud-free fields of view. In contrast, the number of hyperspectral infrared profiles assimilated is much higher since the retrieved profiles can be assimilated in some partly cloudy scenes due to profile coupling other data, such as microwave or neural networks, as first guesses to the retrieval process. As the operational data assimilation community attempts to assimilate cloud-affected radiances, it is possible that the use of retrieved profiles might offer an alternative methodology that is less complex and more computationally efficient to solve this problem. The NASA Short-term Prediction Research and Transition (SPoRT) Center has assimilated hyperspectral infrared retrieved profiles into Weather Research and Forecasting Model (WRF) simulations using the Gridpoint Statistical Interpolation (GSI) System. Early research at SPoRT demonstrated improved initial conditions when assimilating Atmospheric Infrared Sounder (AIRS) thermodynamic profiles into WRF (using WRF-Var and assigning more appropriate error weighting to the profiles) to improve regional analysis and heavy precipitation forecasts. Successful early work has led to more recent research utilizing WRF and GSI for applications including the assimilation of AIRS profiles to improve WRF forecasts of atmospheric rivers and assimilation of AIRS, Cross-track Infrared and Microwave Sounding Suite (CrIMSS), and Infrared Atmospheric Sounding Interferometer (IASI) profiles to improve model representation of tropopause folds and associated non-convective wind events. Although more hyperspectral infrared retrieved profiles can be assimilated into model forecasts, one disadvantage is the retrieved profiles have traditionally been assigned the same error values as the rawinsonde observations when assimilated with GSI. Typically, satellitederived profile errors are larger and more difficult to quantify than traditional rawinsonde observations (especially in the boundary layer), so it is important to appropriately assign observation errors within GSI to eliminate potential spurious innovations and analysis increments that can sometimes arise when using retrieved profiles. The goal of this study is to describe modifications to the GSI source code to more appropriately assimilate hyperspectral infrared retrieved profiles and outline preliminary results that show the differences between a model simulation that assimilated the profiles as rawinsonde observations and one that assimilated the profiles in a module with the appropriate error values.

Application of RBFN network and GM (1, 1) for groundwater level simulation

NASA Astrophysics Data System (ADS)

Li, Zijun; Yang, Qingchun; Wang, Luchen; Martín, Jordi Delgado

2017-10-01

Groundwater is a prominent resource of drinking and domestic water in the world. In this context, a feasible water resources management plan necessitates acceptable predictions of groundwater table depth fluctuations, which can help ensure the sustainable use of a watershed's aquifers for urban and rural water supply. Due to the difficulties of identifying non-linear model structure and estimating the associated parameters, in this study radial basis function neural network (RBFNN) and GM (1, 1) models are used for the prediction of monthly groundwater level fluctuations in the city of Longyan, Fujian Province (South China). The monthly groundwater level data monitored from January 2003 to December 2011 are used in both models. The error criteria are estimated using the coefficient of determination ( R 2), mean absolute error (E) and root mean squared error (RMSE). The results show that both the models can forecast the groundwater level with fairly high accuracy, but the RBFN network model can be a promising tool to simulate and forecast groundwater level since it has a relatively smaller RMSE and MAE.

Integrated Wind Power Planning Tool

NASA Astrophysics Data System (ADS)

Rosgaard, Martin; Giebel, Gregor; Skov Nielsen, Torben; Hahmann, Andrea; Sørensen, Poul; Madsen, Henrik

2013-04-01

This poster presents the current state of the public service obligation (PSO) funded project PSO 10464, with the title "Integrated Wind Power Planning Tool". The goal is to integrate a mesoscale numerical weather prediction (NWP) model with purely statistical tools in order to assess wind power fluctuations, with focus on long term power system planning for future wind farms as well as short term forecasting for existing wind farms. Currently, wind power fluctuation models are either purely statistical or integrated with NWP models of limited resolution. Using the state-of-the-art mesoscale NWP model Weather Research & Forecasting model (WRF) the forecast error is sought quantified in dependence of the time scale involved. This task constitutes a preparative study for later implementation of features accounting for NWP forecast errors in the DTU Wind Energy maintained Corwind code - a long term wind power planning tool. Within the framework of PSO 10464 research related to operational short term wind power prediction will be carried out, including a comparison of forecast quality at different mesoscale NWP model resolutions and development of a statistical wind power prediction tool taking input from WRF. The short term prediction part of the project is carried out in collaboration with ENFOR A/S; a Danish company that specialises in forecasting and optimisation for the energy sector. The integrated prediction model will allow for the description of the expected variability in wind power production in the coming hours to days, accounting for its spatio-temporal dependencies, and depending on the prevailing weather conditions defined by the WRF output. The output from the integrated short term prediction tool constitutes scenario forecasts for the coming period, which can then be fed into any type of system model or decision making problem to be solved. The high resolution of the WRF results loaded into the integrated prediction model will ensure a high accuracy data basis is available for use in the decision making process of the Danish transmission system operator. The need for high accuracy predictions will only increase over the next decade as Denmark approaches the goal of 50% wind power based electricity in 2025 from the current 20%.

On-line estimation of error covariance parameters for atmospheric data assimilation

NASA Technical Reports Server (NTRS)

Dee, Dick P.

1995-01-01

A simple scheme is presented for on-line estimation of covariance parameters in statistical data assimilation systems. The scheme is based on a maximum-likelihood approach in which estimates are produced on the basis of a single batch of simultaneous observations. Simple-sample covariance estimation is reasonable as long as the number of available observations exceeds the number of tunable parameters by two or three orders of magnitude. Not much is known at present about model error associated with actual forecast systems. Our scheme can be used to estimate some important statistical model error parameters such as regionally averaged variances or characteristic correlation length scales. The advantage of the single-sample approach is that it does not rely on any assumptions about the temporal behavior of the covariance parameters: time-dependent parameter estimates can be continuously adjusted on the basis of current observations. This is of practical importance since it is likely to be the case that both model error and observation error strongly depend on the actual state of the atmosphere. The single-sample estimation scheme can be incorporated into any four-dimensional statistical data assimilation system that involves explicit calculation of forecast error covariances, including optimal interpolation (OI) and the simplified Kalman filter (SKF). The computational cost of the scheme is high but not prohibitive; on-line estimation of one or two covariance parameters in each analysis box of an operational bozed-OI system is currently feasible. A number of numerical experiments performed with an adaptive SKF and an adaptive version of OI, using a linear two-dimensional shallow-water model and artificially generated model error are described. The performance of the nonadaptive versions of these methods turns out to depend rather strongly on correct specification of model error parameters. These parameters are estimated under a variety of conditions, including uniformly distributed model error and time-dependent model error statistics.

Using Ensemble Short-Term Initialized Coupled NASA GEOS5 Climate Model Integrations to Study Convective Bias Growth

NASA Technical Reports Server (NTRS)

Cohen, Charlie; Robertson, Franklin; Molod, Andrea

2014-01-01

The representation of convective processes, particularly deep convection in the tropics, remains a persistent problem in climate models. In fact structural biases in the distribution of tropical rainfall in the CMIP5 models is hardly different than that of the CMIP3 versions. Given that regional climate change at higher latitudes is sensitive to the configuration of tropical forcing, this persistent bias is a major issue for the credibility of climate change projections. In this study we use model output from integrations of the NASA Global Earth Observing System Five (GEOS5) climate modeling system to study the evolution of biases in the location and intensity of convective processes. We take advantage of a series of hindcast experiments done in support of the US North American Multi-Model Ensemble (NMME) initiative. For these experiments a nine-month forecast using a coupled model configuration is made approximately every five days over the past 30 years. Each forecast is started with an updated analysis of the ocean, atmosphere and land states. For a given calendar month we have approximately 180 forecasts with daily means of various quantities. These forecasts can be averaged to essentially remove "weather scales" and highlight systematic errors as they evolve. Our primary question is to ask how the spatial structure of daily mean precipitation over the tropics evolves from the initial state and what physical processes are involved. Errors in parameterized convection, various water and energy fluxes and the divergent circulation are found to set up on fast time scales (order five days) compared to errors in the ocean, although SST changes can be non-negligible over that time. For the month of June the difference between forecast day five versus day zero precipitation looks quite similar to the difference between the June precipitation climatology and that from the Global Precipitation Climatology Project (GPCP). We focus much of our analysis on the influence of SST gradients, associated PBL baroclinicity enabled by turbulent mixing, the ensuing PBL moisture convergence, and how changes in these processes relate to convective precipitation bias growth over this short period.

Probabilistic In Situ Stress Estimation and Forecasting using Sequential Data Assimilation

NASA Astrophysics Data System (ADS)

Fichtner, A.; van Dinther, Y.; Kuensch, H. R.

2017-12-01

Our physical understanding and forecasting ability of earthquakes, and other solid Earth dynamic processes, is significantly hampered by limited indications on the evolving state of stress and strength on faults. Integrating observations and physics-based numerical modeling to quantitatively estimate this evolution of a fault's state is crucial. However, systematic attempts are limited and tenuous, especially in light of the scarcity and uncertainty of natural data and the difficulty of modelling the physics governing earthquakes. We adopt the statistical framework of sequential data assimilation - extensively developed for weather forecasting - to efficiently integrate observations and prior knowledge in a forward model, while acknowledging errors in both. To prove this concept we perform a perfect model test in a simplified subduction zone setup, where we assimilate synthetic noised data on velocities and stresses from a single location. Using an Ensemble Kalman Filter, these data and their errors are assimilated to update 150 ensemble members from a Partial Differential Equation-driven seismic cycle model. Probabilistic estimates of fault stress and dynamic strength evolution capture the truth exceptionally well. This is possible, because the sampled error covariance matrix contains prior information from the physics that relates velocities, stresses and pressure at the surface to those at the fault. During the analysis step, stress and strength distributions are thus reconstructed such that fault coupling can be updated to either inhibit or trigger events. In the subsequent forecast step the physical equations are solved to propagate the updated states forward in time and thus provide probabilistic information on the occurrence of the next event. At subsequent assimilation steps, the system's forecasting ability turns out to be significantly better than that of a periodic recurrence model (requiring an alarm 17% vs. 68% of the time). This thus provides distinct added value with respect to using observations or numerical models separately. Although several challenges for applications to a natural setting remain, these first results indicate the large potential of data assimilation techniques for probabilistic seismic hazard assessment and other challenges in dynamic solid earth systems.

Using Ensemble Short-Term Initialized Coupled NASA GEOS5 Climate Model Integrations to Study Convective Bias Growth

NASA Astrophysics Data System (ADS)

Robertson, F. R.; Cohen, C.

2014-12-01

The representation of convective processes, particularly deep convection in the tropics, remains a persistent problem in climate models. In fact structural biases in the distribution of tropical rainfall in the CMIP5 models is hardly different than that of the CMIP3 versions. Given that regional climate change at higher latitudes is sensitive to the configuration of tropical forcing, this persistent bias is a major issue for the credibility of climate change projections. In this study we use model output from integrations of the NASA Global Earth Observing System Five (GEOS5) climate modeling system to study the evolution of biases in the location and intensity of convective processes. We take advantage of a series of hindcast experiments done in support of the US North American Multi-Model Ensemble (NMME) initiative. For these experiments a nine-month forecast using a coupled model configuration is made approximately every five days over the past 30 years. Each forecast is started with an updated analysis of the ocean, atmosphere and land states. For a given calendar month we have approximately 180 forecasts with daily means of various quantities. These forecasts can be averaged to essentially remove "weather scales" and highlight systematic errors as they evolve. Our primary question is to ask how the spatial structure of daily mean precipitation over the tropics evolves from the initial state and what physical processes are involved. Errors in parameterized convection, various water and energy fluxes and the divergent circulation are found to set up on fast time scales (order five days) compared to errors in the ocean, although SST changes can be non-negligible over that time. For the month of June the difference between forecast day five versus day zero precipitation looks quite similar to the difference between the June precipitation climatology and that from the Global Precipitation Climatology Project (GPCP). We focus much of our analysis on the influence of SST gradients, associated PBL baroclinicity enabled by turbulent mixing, the ensuing PBL moisture convergence, and how changes in these processes relate to convective precipitation bias growth over this short period.

Increased Accuracy in Statistical Seasonal Hurricane Forecasting

NASA Astrophysics Data System (ADS)

Nateghi, R.; Quiring, S. M.; Guikema, S. D.

2012-12-01

Hurricanes are among the costliest and most destructive natural hazards in the U.S. Accurate hurricane forecasts are crucial to optimal preparedness and mitigation decisions in the U.S. where 50 percent of the population lives within 50 miles of the coast. We developed a flexible statistical approach to forecast annual number of hurricanes in the Atlantic region during the hurricane season. Our model is based on the method of Random Forest and captures the complex relationship between hurricane activity and climatic conditions through careful variable selection, model testing and validation. We used the National Hurricane Center's Best Track hurricane data from 1949-2011 and sixty-one candidate climate descriptors to develop our model. The model includes information prior to the hurricane season, i.e., from the last three months of the previous year (Oct. through Dec.) and the first five months of the current year (January through May). Our forecast errors are substantially lower than other leading forecasts such as that of the National Oceanic and Atmospheric Administration (NOAA).

Forecasting Temporal Dynamics of Cutaneous Leishmaniasis in Northeast Brazil

PubMed Central

Lewnard, Joseph A.; Jirmanus, Lara; Júnior, Nivison Nery; Machado, Paulo R.; Glesby, Marshall J.; Ko, Albert I.; Carvalho, Edgar M.; Schriefer, Albert; Weinberger, Daniel M.

2014-01-01

Introduction Cutaneous leishmaniasis (CL) is a vector-borne disease of increasing importance in northeastern Brazil. It is known that sandflies, which spread the causative parasites, have weather-dependent population dynamics. Routinely-gathered weather data may be useful for anticipating disease risk and planning interventions. Methodology/Principal Findings We fit time series models using meteorological covariates to predict CL cases in a rural region of Bahía, Brazil from 1994 to 2004. We used the models to forecast CL cases for the period 2005 to 2008. Models accounting for meteorological predictors reduced mean squared error in one, two, and three month-ahead forecasts by up to 16% relative to forecasts from a null model accounting only for temporal autocorrelation. Significance These outcomes suggest CL risk in northeastern Brazil might be partially dependent on weather. Responses to forecasted CL epidemics may include bolstering clinical capacity and disease surveillance in at-risk areas. Ecological mechanisms by which weather influences CL risk merit future research attention as public health intervention targets. PMID:25356734

Forecasting temporal dynamics of cutaneous leishmaniasis in Northeast Brazil.

PubMed

Lewnard, Joseph A; Jirmanus, Lara; Júnior, Nivison Nery; Machado, Paulo R; Glesby, Marshall J; Ko, Albert I; Carvalho, Edgar M; Schriefer, Albert; Weinberger, Daniel M

2014-10-01

Cutaneous leishmaniasis (CL) is a vector-borne disease of increasing importance in northeastern Brazil. It is known that sandflies, which spread the causative parasites, have weather-dependent population dynamics. Routinely-gathered weather data may be useful for anticipating disease risk and planning interventions. We fit time series models using meteorological covariates to predict CL cases in a rural region of Bahía, Brazil from 1994 to 2004. We used the models to forecast CL cases for the period 2005 to 2008. Models accounting for meteorological predictors reduced mean squared error in one, two, and three month-ahead forecasts by up to 16% relative to forecasts from a null model accounting only for temporal autocorrelation. These outcomes suggest CL risk in northeastern Brazil might be partially dependent on weather. Responses to forecasted CL epidemics may include bolstering clinical capacity and disease surveillance in at-risk areas. Ecological mechanisms by which weather influences CL risk merit future research attention as public health intervention targets.

Steam-load-forecasting technique for central-heating plants. Final report

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

Lin, M.C.; Carnahan, J.V.

Because boilers generally are most efficient at full loads, the Army could achieve significant savings by running fewer boilers at high loads rather than more boilers at low loads. A reliable load prediction technique could help ensure that only those boilers required to meet demand are on line. This report presents the results of an investigation into the feasibility of forecasting heat plant steam loads from historical patterns and weather information. Using steam flow data collected at Fort Benjamin Harrison, IN, a Box-Jenkins transfer function model with an acceptably small prediction error was initially identified. Initial investigation of forecast modelmore » development appeared successful. Dynamic regression methods using actual ambient temperatures yielded the best results. Box-Jenkins univariate models' results appeared slightly less accurate. Since temperature information was not needed for model building and forecasting, however, it is recommended that Box-Jenkins models be considered prime candidates for load forecasting due to their simpler mathematics.« less

Soil moisture assimilation using a modified ensemble transform Kalman filter with water balance constraint

NASA Astrophysics Data System (ADS)

Wu, Guocan; Zheng, Xiaogu; Dan, Bo

2016-04-01

The shallow soil moisture observations are assimilated into Common Land Model (CoLM) to estimate the soil moisture in different layers. The forecast error is inflated to improve the analysis state accuracy and the water balance constraint is adopted to reduce the water budget residual in the assimilation procedure. The experiment results illustrate that the adaptive forecast error inflation can reduce the analysis error, while the proper inflation layer can be selected based on the -2log-likelihood function of the innovation statistic. The water balance constraint can result in reducing water budget residual substantially, at a low cost of assimilation accuracy loss. The assimilation scheme can be potentially applied to assimilate the remote sensing data.

Post-processing of a low-flow forecasting system in the Thur basin (Switzerland)

NASA Astrophysics Data System (ADS)

Bogner, Konrad; Joerg-Hess, Stefanie; Bernhard, Luzi; Zappa, Massimiliano

2015-04-01

Low-flows and droughts are natural hazards with potentially severe impacts and economic loss or damage in a number of environmental and socio-economic sectors. As droughts develop slowly there is time to prepare and pre-empt some of these impacts. Real-time information and forecasting of a drought situation can therefore be an effective component of drought management. Although Switzerland has traditionally been more concerned with problems related to floods, in recent years some unprecedented low-flow situations have been experienced. Driven by the climate change debate a drought information platform has been developed to guide water resources management during situations where water resources drop below critical low-flow levels characterised by the indices duration (time between onset and offset), severity (cumulative water deficit) and magnitude (severity/duration). However to gain maximum benefit from such an information system it is essential to remove the bias from the meteorological forecast, to derive optimal estimates of the initial conditions, and to post-process the stream-flow forecasts. Quantile mapping methods for pre-processing the meteorological forecasts and improved data assimilation methods of snow measurements, which accounts for much of the seasonal stream-flow predictability for the majority of the basins in Switzerland, have been tested previously. The objective of this study is the testing of post-processing methods in order to remove bias and dispersion errors and to derive the predictive uncertainty of a calibrated low-flow forecast system. Therefore various stream-flow error correction methods with different degrees of complexity have been applied and combined with the Hydrological Uncertainty Processor (HUP) in order to minimise the differences between the observations and model predictions and to derive posterior probabilities. The complexity of the analysed error correction methods ranges from simple AR(1) models to methods including wavelet transformations and support vector machines. These methods have been combined with forecasts driven by Numerical Weather Prediction (NWP) systems with different temporal and spatial resolutions, lead-times and different numbers of ensembles covering short to medium to extended range forecasts (COSMO-LEPS, 10-15 days, monthly and seasonal ENS) as well as climatological forecasts. Additionally the suitability of various skill scores and efficiency measures regarding low-flow predictions will be tested. Amongst others the novel 2afc (2 alternatives forced choices) score and the quantile skill score and its decompositions will be applied to evaluate the probabilistic forecasts and the effects of post-processing. First results of the performance of the low-flow predictions of the hydrological model PREVAH initialised with different NWP's will be shown.

Skilful rainfall forecasts from artificial neural networks with long duration series and single-month optimization

NASA Astrophysics Data System (ADS)

Abbot, John; Marohasy, Jennifer

2017-11-01

General circulation models, which forecast by first modelling actual conditions in the atmosphere and ocean, are used extensively for monthly rainfall forecasting. We show how more skilful monthly and seasonal rainfall forecasts can be achieved through the mining of historical climate data using artificial neural networks (ANNs). This technique is demonstrated for two agricultural regions of Australia: the wheat belt of Western Australia and the sugar growing region of coastal Queensland. The most skilful monthly rainfall forecasts measured in terms of Ideal Point Error (IPE), and a score relative to climatology, are consistently achieved through the use of ANNs optimized for each month individually, and also by choosing to input longer historical series of climate indices. Using the longer series restricts the number of climate indices that can be used.

Quantifying and Reducing Uncertainty in Correlated Multi-Area Short-Term Load Forecasting

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

Sun, Yannan; Hou, Zhangshuan; Meng, Da

2016-07-17

In this study, we represent and reduce the uncertainties in short-term electric load forecasting by integrating time series analysis tools including ARIMA modeling, sequential Gaussian simulation, and principal component analysis. The approaches are mainly focusing on maintaining the inter-dependency between multiple geographically related areas. These approaches are applied onto cross-correlated load time series as well as their forecast errors. Multiple short-term prediction realizations are then generated from the reduced uncertainty ranges, which are useful for power system risk analyses.

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.

Ensemble averaging and stacking of ARIMA and GSTAR model for rainfall forecasting

NASA Astrophysics Data System (ADS)

Anggraeni, D.; Kurnia, I. F.; Hadi, A. F.

2018-04-01

Unpredictable rainfall changes can affect human activities, such as in agriculture, aviation, shipping which depend on weather forecasts. Therefore, we need forecasting tools with high accuracy in predicting the rainfall in the future. This research focus on local forcasting of the rainfall at Jember in 2005 until 2016, from 77 rainfall stations. The rainfall here was not only related to the occurrence of the previous of its stations, but also related to others, it’s called the spatial effect. The aim of this research is to apply the GSTAR model, to determine whether there are some correlations of spatial effect between one to another stations. The GSTAR model is an expansion of the space-time model that combines the time-related effects, the locations (stations) in a time series effects, and also the location it self. The GSTAR model will also be compared to the ARIMA model that completely ignores the independent variables. The forcested value of the ARIMA and of the GSTAR models then being combined using the ensemble forecasting technique. The averaging and stacking method of ensemble forecasting method here provide us the best model with higher acuracy model that has the smaller RMSE (Root Mean Square Error) value. Finally, with the best model we can offer a better local rainfall forecasting in Jember for the future.

Evaluating the spatio-temporal performance of sky imager based solar irradiance analysis and forecasts

NASA Astrophysics Data System (ADS)

Schmidt, T.; Kalisch, J.; Lorenz, E.; Heinemann, D.

2015-10-01

Clouds are the dominant source of variability in surface solar radiation and uncertainty in its prediction. However, the increasing share of solar energy in the world-wide electric power supply increases the need for accurate solar radiation forecasts. In this work, we present results of a shortest-term global horizontal irradiance (GHI) forecast experiment based on hemispheric sky images. A two month dataset with images from one sky imager and high resolutive GHI measurements from 99 pyranometers distributed over 10 km by 12 km is used for validation. We developed a multi-step model and processed GHI forecasts up to 25 min with an update interval of 15 s. A cloud type classification is used to separate the time series in different cloud scenarios. Overall, the sky imager based forecasts do not outperform the reference persistence forecasts. Nevertheless, we find that analysis and forecast performance depend strongly on the predominant cloud conditions. Especially convective type clouds lead to high temporal and spatial GHI variability. For cumulus cloud conditions, the analysis error is found to be lower than that introduced by a single pyranometer if it is used representatively for the whole area in distances from the camera larger than 1-2 km. Moreover, forecast skill is much higher for these conditions compared to overcast or clear sky situations causing low GHI variability which is easier to predict by persistence. In order to generalize the cloud-induced forecast error, we identify a variability threshold indicating conditions with positive forecast skill.

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

Decision Aids for Multiple-Decision Disease Management as Affected by Weather Input Errors

USDA-ARS?s Scientific Manuscript database

Many disease management decision support systems (DSS) rely, exclusively or in part, on weather inputs to calculate an indicator for disease hazard. Error in the weather inputs, typically due to forecasting, interpolation or estimation from off-site sources, may affect model calculations and manage...

Improving Weather Forecasts Through Reduced Precision Data Assimilation

NASA Astrophysics Data System (ADS)

Hatfield, Samuel; Düben, Peter; Palmer, Tim

2017-04-01

We present a new approach for improving the efficiency of data assimilation, by trading numerical precision for computational speed. Future supercomputers will allow a greater choice of precision, so that models can use a level of precision that is commensurate with the model uncertainty. Previous studies have already indicated that the quality of climate and weather forecasts is not significantly degraded when using a precision less than double precision [1,2], but so far these studies have not considered data assimilation. Data assimilation is inherently uncertain due to the use of relatively long assimilation windows, noisy observations and imperfect models. Thus, the larger rounding errors incurred from reducing precision may be within the tolerance of the system. Lower precision arithmetic is cheaper, and so by reducing precision in ensemble data assimilation, we can redistribute computational resources towards, for example, a larger ensemble size. Because larger ensembles provide a better estimate of the underlying distribution and are less reliant on covariance inflation and localisation, lowering precision could actually allow us to improve the accuracy of weather forecasts. We will present results on how lowering numerical precision affects the performance of an ensemble data assimilation system, consisting of the Lorenz '96 toy atmospheric model and the ensemble square root filter. We run the system at half precision (using an emulation tool), and compare the results with simulations at single and double precision. We estimate that half precision assimilation with a larger ensemble can reduce assimilation error by 30%, with respect to double precision assimilation with a smaller ensemble, for no extra computational cost. This results in around half a day extra of skillful weather forecasts, if the error-doubling characteristics of the Lorenz '96 model are mapped to those of the real atmosphere. Additionally, we investigate the sensitivity of these results to observational error and assimilation window length. Half precision hardware will become available very shortly, with the introduction of Nvidia's Pascal GPU architecture and the Intel Knights Mill coprocessor. We hope that the results presented here will encourage the uptake of this hardware. References [1] Peter D. Düben and T. N. Palmer, 2014: Benchmark Tests for Numerical Weather Forecasts on Inexact Hardware, Mon. Weather Rev., 142, 3809-3829 [2] Peter D. Düben, Hugh McNamara and T. N. Palmer, 2014: The use of imprecise processing to improve accuracy in weather & climate prediction, J. Comput. Phys., 271, 2-18

Nonlinear problems in data-assimilation : Can synchronization help?

NASA Astrophysics Data System (ADS)

Tribbia, J. J.; Duane, G. S.

2009-12-01

Over the past several years, operational weather centers have initiated ensemble prediction and assimilation techniques to estimate the error covariance of forecasts in the short and the medium range. The ensemble techniques used are based on linear methods. The theory This technique s been shown to be a useful indicator of skill in the linear range where forecast errors are small relative to climatological variance. While this advance has been impressive, there are still ad hoc aspects of its use in practice, like the need for covariance inflation which are troubling. Furthermore, to be of utility in the nonlinear range an ensemble assimilation and prediction method must be capable of giving probabilistic information for the situation where a probability density forecast becomes multi-modal. A prototypical, simplest example of such a situation is the planetary-wave regime transition where the pdf is bimodal. Our recent research show how the inconsistencies and extensions of linear methodology can be consistently treated using the paradigm of synchronization which views the problems of assimilation and forecasting as that of optimizing the forecast model state with respect to the future evolution of the atmosphere.

Seasonal simulations using a coupled ocean-atmosphere model with data assimilation

NASA Astrophysics Data System (ADS)

Larow, Timothy Edward

1997-10-01

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 coupled model is used for seasonal predictions of the boreal summers of 1987 and 1988. The atmosphere model is a modified version of the Florida State University global spectral model, resolution triangular truncation 42 waves. The ocean general circulation model consists of a slightly modified version developed by Latif (1987). Coupling is synchronous with exchange of information every two model hours. Using daily analysis from ECMWF and observed monthly mean SSTs from NCEP, two - one year, time dependent, Newtonian relaxation were conducted using the coupled model prior to the seasonal forecasts. Relaxation was selectively applied to the atmospheric vorticity, divergence, temperature, and dew point depression equations, and to the ocean's surface temperature equation. The ocean's initial conditions are from a six year ocean-only simulation which used observed wind stresses and a relaxation towards observed SSTs for forcings. Coupled initialization was conducted from 1 June 1986 to 1 June 1987 for the 1987 boreal forecast and from 1 June 1987 to 1 June 1988 for the 1988 boreal forecast. Examination of annual means of net heat flux, freshwater flux and wind stress obtained by from the initialization show close agreement with Oberhuber (1988) climatology and the Florida State University pseudo wind stress analysis. Sensitivity of the initialization/assimilation scheme was tested by conducting two - ten member ensemble integrations. Each member was integrated for 90 days (June-August) of the respective year. Initial conditions for the ensembles consisted of the same ocean state as used by the initialize forecasts, while the atmospheric initial conditions were from ECMWF analysis centered on 1 June of the respective year. Root mean square error and anomaly correlations between observed and forecasted SSTs in the Nino 3 and Nino 4 regions show greater skill between the initialized forecasts than the ensemble forecasts. It is hypothesized that differences in the specific humidity within the planetary boundary layer are responsible for the large SST errors noted with the ensembles.

Evaluations of Extended-Range tropical Cyclone Forecasts in the Western North Pacific by using the Ensemble Reforecasts: Preliminary Results

NASA Astrophysics Data System (ADS)

Tsai, Hsiao-Chung; Chen, Pang-Cheng; Elsberry, Russell L.

2017-04-01

The objective of this study is to evaluate the predictability of the extended-range forecasts of tropical cyclone (TC) in the western North Pacific using reforecasts from National Centers for Environmental Prediction (NCEP) Global Ensemble Forecast System (GEFS) during 1996-2015, and from the Climate Forecast System (CFS) during 1999-2010. Tsai and Elsberry have demonstrated that an opportunity exists to support hydrological operations by using the extended-range TC formation and track forecasts in the western North Pacific from the ECMWF 32-day ensemble. To demonstrate this potential for the decision-making processes regarding water resource management and hydrological operation in Taiwan reservoir watershed areas, special attention is given to the skill of the NCEP GEFS and CFS models in predicting the TCs affecting the Taiwan area. The first objective of this study is to analyze the skill of NCEP GEFS and CFS TC forecasts and quantify the forecast uncertainties via verifications of categorical binary forecasts and probabilistic forecasts. The second objective is to investigate the relationships among the large-scale environmental factors [e.g., El Niño Southern Oscillation (ENSO), Madden-Julian Oscillation (MJO), etc.] and the model forecast errors by using the reforecasts. Preliminary results are indicating that the skill of the TC activity forecasts based on the raw forecasts can be further improved if the model biases are minimized by utilizing these reforecasts.

Municipal water consumption forecast accuracy

NASA Astrophysics Data System (ADS)

Fullerton, Thomas M.; Molina, Angel L.

2010-06-01

Municipal water consumption planning is an active area of research because of infrastructure construction and maintenance costs, supply constraints, and water quality assurance. In spite of that, relatively few water forecast accuracy assessments have been completed to date, although some internal documentation may exist as part of the proprietary "grey literature." This study utilizes a data set of previously published municipal consumption forecasts to partially fill that gap in the empirical water economics literature. Previously published municipal water econometric forecasts for three public utilities are examined for predictive accuracy against two random walk benchmarks commonly used in regional analyses. Descriptive metrics used to quantify forecast accuracy include root-mean-square error and Theil inequality statistics. Formal statistical assessments are completed using four-pronged error differential regression F tests. Similar to studies for other metropolitan econometric forecasts in areas with similar demographic and labor market characteristics, model predictive performances for the municipal water aggregates in this effort are mixed for each of the municipalities included in the sample. Given the competitiveness of the benchmarks, analysts should employ care when utilizing econometric forecasts of municipal water consumption for planning purposes, comparing them to recent historical observations and trends to insure reliability. Comparative results using data from other markets, including regions facing differing labor and demographic conditions, would also be helpful.

Comparison of Flow-Dependent and Static Error Correlation Models in the DAO Ozone Data Assimilation System

NASA Technical Reports Server (NTRS)

Wargan, K.; Stajner, I.; Pawson, S.

2003-01-01

In a data assimilation system the forecast error covariance matrix governs the way in which the data information is spread throughout the model grid. Implementation of a correct method of assigning covariances is expected to have an impact on the analysis results. The simplest models assume that correlations are constant in time and isotropic or nearly isotropic. In such models the analysis depends on the dynamics only through assumed error standard deviations. In applications to atmospheric tracer data assimilation this may lead to inaccuracies, especially in regions with strong wind shears or high gradient of potential vorticity, as well as in areas where no data are available. In order to overcome this problem we have developed a flow-dependent covariance model that is based on short term evolution of error correlations. The presentation compares performance of a static and a flow-dependent model applied to a global three- dimensional ozone data assimilation system developed at NASA s Data Assimilation Office. We will present some results of validation against WMO balloon-borne sondes and the Polar Ozone and Aerosol Measurement (POAM) III instrument. Experiments show that allowing forecast error correlations to evolve with the flow results in positive impact on assimilated ozone within the regions where data were not assimilated, particularly at high latitudes in both hemispheres and in the troposphere. We will also discuss statistical characteristics of both models; in particular we will argue that including evolution of error correlations leads to stronger internal consistency of a data assimilation ,

Spatio-temporal behaviour of medium-range ensemble forecasts

NASA Astrophysics Data System (ADS)

Kipling, Zak; Primo, Cristina; Charlton-Perez, Andrew

2010-05-01

Using the recently-developed mean-variance of logarithms (MVL) diagram, together with the TIGGE archive of medium-range ensemble forecasts from nine different centres, we present an analysis of the spatio-temporal dynamics of their perturbations, and show how the differences between models and perturbation techniques can explain the shape of their characteristic MVL curves. We also consider the use of the MVL diagram to compare the growth of perturbations within the ensemble with the growth of the forecast error, showing that there is a much closer correspondence for some models than others. We conclude by looking at how the MVL technique might assist in selecting models for inclusion in a multi-model ensemble, and suggest an experiment to test its potential in this context.

What might we learn from climate forecasts?

PubMed Central

Smith, Leonard A.

2002-01-01

Most climate models are large dynamical systems involving a million (or more) variables on big computers. Given that they are nonlinear and not perfect, what can we expect to learn from them about the earth's climate? How can we determine which aspects of their output might be useful and which are noise? And how should we distribute resources between making them “better,” estimating variables of true social and economic interest, and quantifying how good they are at the moment? Just as “chaos” prevents accurate weather forecasts, so model error precludes accurate forecasts of the distributions that define climate, yielding uncertainty of the second kind. Can we estimate the uncertainty in our uncertainty estimates? These questions are discussed. Ultimately, all uncertainty is quantified within a given modeling paradigm; our forecasts need never reflect the uncertainty in a physical system. PMID:11875200

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.

Forecasting experiments of a dynamical-statistical model of the sea surface temperature anomaly field based on the improved self-memorization principle

NASA Astrophysics Data System (ADS)

Hong, Mei; Chen, Xi; Zhang, Ren; Wang, Dong; Shen, Shuanghe; Singh, Vijay P.

2018-04-01

With the objective of tackling the problem of inaccurate long-term El Niño-Southern Oscillation (ENSO) forecasts, this paper develops a new dynamical-statistical forecast model of the sea surface temperature anomaly (SSTA) field. To avoid single initial prediction values, a self-memorization principle is introduced to improve the dynamical reconstruction model, thus making the model more appropriate for describing such chaotic systems as ENSO events. The improved dynamical-statistical model of the SSTA field is used to predict SSTA in the equatorial eastern Pacific and during El Niño and La Niña events. The long-term step-by-step forecast results and cross-validated retroactive hindcast results of time series T1 and T2 are found to be satisfactory, with a Pearson correlation coefficient of approximately 0.80 and a mean absolute percentage error (MAPE) of less than 15 %. The corresponding forecast SSTA field is accurate in that not only is the forecast shape similar to the actual field but also the contour lines are essentially the same. This model can also be used to forecast the ENSO index. The temporal correlation coefficient is 0.8062, and the MAPE value of 19.55 % is small. The difference between forecast results in spring and those in autumn is not high, indicating that the improved model can overcome the spring predictability barrier to some extent. Compared with six mature models published previously, the present model has an advantage in prediction precision and length, and is a novel exploration of the ENSO forecast method.

Ensemble streamflow assimilation with the National Water Model.

NASA Astrophysics Data System (ADS)

Rafieeinasab, A.; McCreight, J. L.; Noh, S.; Seo, D. J.; Gochis, D.

2017-12-01

Through case studies of flooding across the US, we compare the performance of the National Water Model (NWM) data assimilation (DA) scheme to that of a newly implemented ensemble Kalman filter approach. The NOAA National Water Model (NWM) is an operational implementation of the community WRF-Hydro modeling system. As of August 2016, the NWM forecasts of distributed hydrologic states and fluxes (including soil moisture, snowpack, ET, and ponded water) over the contiguous United States have been publicly disseminated by the National Center for Environmental Prediction (NCEP) . It also provides streamflow forecasts at more than 2.7 million river reaches up to 30 days in advance. The NWM employs a nudging scheme to assimilate more than 6,000 USGS streamflow observations and provide initial conditions for its forecasts. A problem with nudging is how the forecasts relax quickly to open-loop bias in the forecast. This has been partially addressed by an experimental bias correction approach which was found to have issues with phase errors during flooding events. In this work, we present an ensemble streamflow data assimilation approach combining new channel-only capabilities of the NWM and HydroDART (a coupling of the offline WRF-Hydro model and NCAR's Data Assimilation Research Testbed; DART). Our approach focuses on the single model state of discharge and incorporates error distributions on channel-influxes (overland and groundwater) in the assimilation via an ensemble Kalman filter (EnKF). In order to avoid filter degeneracy associated with a limited number of ensemble at large scale, DART's covariance inflation (Anderson, 2009) and localization capabilities are implemented and evaluated. The current NWM data assimilation scheme is compared to preliminary results from the EnKF application for several flooding case studies across the US.

On using scatterometer and altimeter data to improve storm surge forecasting in the Adriatic Sea

NASA Astrophysics Data System (ADS)

Bajo, Marco; Umgiesser, Georg; De Biasio, Francesco; Vignudelli, Stefano; Zecchetto, Stefano

2017-04-01

Satellite data are seldom used in storm surge forecasting. Among the most important issues related to the storm surge forecasting are the quality of the model wind forcing and the initial condition of the sea surface elevation. In this work, focused on storm surge forecasting in the Adriatic Sea, satellite scatterometer wind data are used to correct the wind speed and direction biases of the ECMWF global atmospheric model by tuning the spatial fields, as an alternative to data assimilation. The capability of such an unbiased wind is tested against that of a high resolution wind, produced by a regional non-hydrostatic model. On the other hand, altimeter Total Water Level Envelope (TWLE) data, which provide the sea level elevation, are used to improve the accuracy of the initial state of the model simulations. This is done by assimilating into a storm surge model the TWLE obtained by the altimeter observations along ground tracks, after subtraction of the tidal components. In order to test the methodology, eleven storm surge events recorded in Venice, from 2008 to 2012, have been simulated using different configurations of forcing wind and altimeter data assimilation. Results show that the relative error on the estimation of the maximum surge peak, averaged over the cases considered, decreases from 13% to 7% using both the unbiased wind and the altimeter data assimilation, while forcing the hydrodynamic model with the high resolution wind (no tuning), the altimeter data assimilation reduces the error from 9% to 6%.

Forecasting the Incidence and Prevalence of Patients with End-Stage Renal Disease in Malaysia up to the Year 2040

PubMed Central

Adnan, Tassha Hilda; Hashim, Nadiah Hanis; Mohan, Kirubashni; Kim Liong, Ang; Ahmad, Ghazali; Bak Leong, Goh; Bavanandan, Sunita; Haniff, Jamaiyah

2017-01-01

Background. The incidence of patients with end-stage renal disease (ESRD) requiring dialysis has been growing rapidly in Malaysia from 18 per million population (pmp) in 1993 to 231 pmp in 2013. Objective. To forecast the incidence and prevalence of ESRD patients who will require dialysis treatment in Malaysia until 2040. Methodology. Univariate forecasting models using the number of new and current dialysis patients, by the Malaysian Dialysis and Transplant Registry from 1993 to 2013 were used. Four forecasting models were evaluated, and the model with the smallest error was selected for the prediction. Result. ARIMA (0, 2, 1) modeling with the lowest error was selected to predict both the incidence (RMSE = 135.50, MAPE = 2.85, and MAE = 87.71) and the prevalence (RMSE = 158.79, MAPE = 1.29, and MAE = 117.21) of dialysis patients. The estimated incidences of new dialysis patients in 2020 and 2040 are 10,208 and 19,418 cases, respectively, while the estimated prevalence is 51,269 and 106,249 cases. Conclusion. The growth of ESRD patients on dialysis in Malaysia can be expected to continue at an alarming rate. Effective steps to address and curb further increase in new patients requiring dialysis are urgently needed, in order to mitigate the expected financial and health catastrophes associated with the projected increase of such patients. PMID:28348890

Evaluation of Thompson-type trend and monthly weather data models for corn yields in Iowa, Illinois, and Indiana

NASA Technical Reports Server (NTRS)

French, V. (Principal Investigator)

1982-01-01

An evaluation was made of Thompson-Type models which use trend terms (as a surrogate for technology), meteorological variables based on monthly average temperature, and total precipitation to forecast and estimate corn yields in Iowa, Illinois, and Indiana. Pooled and unpooled Thompson-type models were compared. Neither was found to be consistently superior to the other. Yield reliability indicators show that the models are of limited use for large area yield estimation. The models are objective and consistent with scientific knowledge. Timely yield forecasts and estimates can be made during the growing season by using normals or long range weather forecasts. The models are not costly to operate and are easy to use and understand. The model standard errors of prediction do not provide a useful current measure of modeled yield reliability.

[Real-time irrigation forecast of cotton mulched with plastic film under drip irrigation based on meteorological date].

PubMed

Shen, Xiao-jun; Sun, Jing-sheng; Li, Ming-si; Zhang, Ji-yang; Wang, Jing-lei; Li, Dong-wei

2015-02-01

It is important to improve the real-time irrigation forecasting precision by predicting real-time water consumption of cotton mulched with plastic film under drip irrigation based on meteorological data and cotton growth status. The model parameters for calculating ET0 based on Hargreaves formula were determined using historical meteorological data from 1953 to 2008 in Shihezi reclamation area. According to the field experimental data of growing season in 2009-2010, the model of computing crop coefficient Kc was established based on accumulated temperature. On the basis of crop water requirement (ET0) and Kc, a real-time irrigation forecast model was finally constructed, and it was verified by the field experimental data in 2011. The results showed that the forecast model had high forecasting precision, and the average absolute values of relative error between the predicted value and measured value were about 3.7%, 2.4% and 1.6% during seedling, squaring and blossom-boll forming stages, respectively. The forecast model could be used to modify the predicted values in time according to the real-time meteorological data and to guide the water management in local film-mulched cotton field under drip irrigation.

Spatial Pattern Classification for More Accurate Forecasting of Variable Energy Resources

NASA Astrophysics Data System (ADS)

Novakovskaia, E.; Hayes, C.; Collier, C.

2014-12-01

The accuracy of solar and wind forecasts is becoming increasingly essential as grid operators continue to integrate additional renewable generation onto the electric grid. Forecast errors affect rate payers, grid operators, wind and solar plant maintenance crews and energy traders through increases in prices, project down time or lost revenue. While extensive and beneficial efforts were undertaken in recent years to improve physical weather models for a broad spectrum of applications these improvements have generally not been sufficient to meet the accuracy demands of system planners. For renewables, these models are often used in conjunction with additional statistical models utilizing both meteorological observations and the power generation data. Forecast accuracy can be dependent on specific weather regimes for a given location. To account for these dependencies it is important that parameterizations used in statistical models change as the regime changes. An automated tool, based on an artificial neural network model, has been developed to identify different weather regimes as they impact power output forecast accuracy at wind or solar farms. In this study, improvements in forecast accuracy were analyzed for varying time horizons for wind farms and utility-scale PV plants located in different geographical regions.

Improving emissions inventories in North America through systematic analysis of model performance during ICARTT and MILAGRO

NASA Astrophysics Data System (ADS)

Mena, Marcelo Andres

During 2004 and 2006 the University of Iowa provided air quality forecast support for flight planning of the ICARTT and MILAGRO field campaigns. A method for improvement of model performance in comparison to observations is showed. The method allows identifying sources of model error from boundary conditions and emissions inventories. Simultaneous analysis of horizontal interpolation of model error and error covariance showed that error in ozone modeling is highly correlated to the error of its precursors, and that there is geographical correlation also. During ICARTT ozone modeling error was improved by updating from the National Emissions Inventory from 1999 and 2001, and furthermore by updating large point source emissions from continuous monitoring data. Further improvements were achieved by reducing area emissions of NOx y 60% for states in the Southeast United States. Ozone error was highly correlated to NOy error during this campaign. Also ozone production in the United States was most sensitive to NOx emissions. During MILAGRO model performance in terms of correlation coefficients was higher, but model error in ozone modeling was high due overestimation of NOx and VOC emissions in Mexico City during forecasting. Large model improvements were shown by decreasing NOx emissions in Mexico City by 50% and VOC by 60%. Recurring ozone error is spatially correlated to CO and NOy error. Sensitivity studies show that Mexico City aerosol can reduce regional photolysis rates by 40% and ozone formation by 5-10%. Mexico City emissions can enhance NOy and O3 concentrations over the Gulf of Mexico in up to 10-20%. Mexico City emissions can convert regional ozone production regimes from VOC to NOx limited. A method of interpolation of observations along flight tracks is shown, which can be used to infer on the direction of outflow plumes. The use of ratios such as O3/NOy and NOx/NOy can be used to provide information on chemical characteristics of the plume, such as age, and ozone production regime. Interpolated MTBE observations can be used as a tracer of urban mobile source emissions. Finally procedures for estimating and gridding emissions inventories in Brazil and Mexico are presented.

Assessing the Impact of Observations on the Prediction of Effective Atmospheric Angular Momentum from NAVGEM

NASA Astrophysics Data System (ADS)

Baker, N. L.; Langland, R.

2016-12-01

Variations in Earth rotation are measured by comparing a time based on Earth's variable rotation rate about its axis to a time standard based on an internationally coordinated ensemble of atomic clocks that provide a uniform time scale. The variability of Earth's rotation is partly due to the changes in angular momentum that occur in the atmosphere and ocean as weather patterns and ocean features develop, propagate, and dissipate. The NAVGEM Effective Atmospheric Angular Momentum Functions (EAAMF) and their predictions are computed following Barnes et al. (1983), and provided to the U.S. Naval Observatory daily. These along with similar data from the NOAA GFS model are used to calculate and predict the Earth orientation parameters (Stamatakos et al., 2016). The Navy's high-resolution global weather prediction system consists of the Navy Global Environmental Model (NAVGEM; Hogan et al., 2014) and a hybrid four-dimensional variational data assimilation system (4DVar) (Kuhl et al., 2013). An important component of NAVGEM is the Forecast Sensitivity Observation Impact (FSOI). FSOI is a mathematical method to quantify the contribution of individual observations or sets of observations to the reduction in the 24-hr forecast error (Langland and Baker, 2004). The FSOI allows for dynamic monitoring of the relative quality and value of the observations assimilated by NAVGEM, and the relative ability of the data assimilation system to effectively use the observation information to generate an improved forecast. For this study, along with the FSOI based on the global moist energy error norm, we computed the FSOI using an error norm based on the Effective Angular Momentum Functions. This modification allowed us to assess which observations were most beneficial in reducing the 24-hr forecast error for the atmospheric angular momentum.

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.

An application of seasonal ARIMA models on group commodities to forecast Philippine merchandise exports performance

NASA Astrophysics Data System (ADS)

Natividad, Gina May R.; Cawiding, Olive R.; Addawe, Rizavel C.

2017-11-01

The increase in the merchandise exports of the country offers information about the Philippines' trading role within the global economy. Merchandise exports statistics are used to monitor the country's overall production that is consumed overseas. This paper investigates the comparison between two models obtained by a) clustering the commodity groups into two based on its proportional contribution to the total exports, and b) treating only the total exports. Different seasonal autoregressive integrated moving average (SARIMA) models were then developed for the clustered commodities and for the total exports based on the monthly merchandise exports of the Philippines from 2011 to 2016. The data set used in this study was retrieved from the Philippine Statistics Authority (PSA) which is the central statistical authority in the country responsible for primary data collection. A test for significance of the difference between means at 0.05 level of significance was then performed on the forecasts produced. The result indicates that there is a significant difference between the mean of the forecasts of the two models. Moreover, upon a comparison of the root mean square error (RMSE) and mean absolute error (MAE) of the models, it was found that the models used for the clustered groups outperform the model for the total exports.

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.

Forecasting Error Calculation with Mean Absolute Deviation and Mean Absolute Percentage Error

NASA Astrophysics Data System (ADS)

Khair, Ummul; Fahmi, Hasanul; Hakim, Sarudin Al; Rahim, Robbi

2017-12-01

Prediction using a forecasting method is one of the most important things for an organization, the selection of appropriate forecasting methods is also important but the percentage error of a method is more important in order for decision makers to adopt the right culture, the use of the Mean Absolute Deviation and Mean Absolute Percentage Error to calculate the percentage of mistakes in the least square method resulted in a percentage of 9.77% and it was decided that the least square method be worked for time series and trend data.

Addressing model error through atmospheric stochastic physical parametrizations: impact on the coupled ECMWF seasonal forecasting system

PubMed Central

Weisheimer, Antje; Corti, Susanna; Palmer, Tim; Vitart, Frederic

2014-01-01

The finite resolution of general circulation models of the coupled atmosphere–ocean system and the effects of sub-grid-scale variability present a major source of uncertainty in model simulations on all time scales. The European Centre for Medium-Range Weather Forecasts has been at the forefront of developing new approaches to account for these uncertainties. In particular, the stochastically perturbed physical tendency scheme and the stochastically perturbed backscatter algorithm for the atmosphere are now used routinely for global numerical weather prediction. The European Centre also performs long-range predictions of the coupled atmosphere–ocean climate system in operational forecast mode, and the latest seasonal forecasting system—System 4—has the stochastically perturbed tendency and backscatter schemes implemented in a similar way to that for the medium-range weather forecasts. Here, we present results of the impact of these schemes in System 4 by contrasting the operational performance on seasonal time scales during the retrospective forecast period 1981–2010 with comparable simulations that do not account for the representation of model uncertainty. We find that the stochastic tendency perturbation schemes helped to reduce excessively strong convective activity especially over the Maritime Continent and the tropical Western Pacific, leading to reduced biases of the outgoing longwave radiation (OLR), cloud cover, precipitation and near-surface winds. Positive impact was also found for the statistics of the Madden–Julian oscillation (MJO), showing an increase in the frequencies and amplitudes of MJO events. Further, the errors of El Niño southern oscillation forecasts become smaller, whereas increases in ensemble spread lead to a better calibrated system if the stochastic tendency is activated. The backscatter scheme has overall neutral impact. Finally, evidence for noise-activated regime transitions has been found in a cluster analysis of mid-latitude circulation regimes over the Pacific–North America region. PMID:24842026

Addressing model error through atmospheric stochastic physical parametrizations: impact on the coupled ECMWF seasonal forecasting system.

PubMed

Weisheimer, Antje; Corti, Susanna; Palmer, Tim; Vitart, Frederic

2014-06-28

The finite resolution of general circulation models of the coupled atmosphere-ocean system and the effects of sub-grid-scale variability present a major source of uncertainty in model simulations on all time scales. The European Centre for Medium-Range Weather Forecasts has been at the forefront of developing new approaches to account for these uncertainties. In particular, the stochastically perturbed physical tendency scheme and the stochastically perturbed backscatter algorithm for the atmosphere are now used routinely for global numerical weather prediction. The European Centre also performs long-range predictions of the coupled atmosphere-ocean climate system in operational forecast mode, and the latest seasonal forecasting system--System 4--has the stochastically perturbed tendency and backscatter schemes implemented in a similar way to that for the medium-range weather forecasts. Here, we present results of the impact of these schemes in System 4 by contrasting the operational performance on seasonal time scales during the retrospective forecast period 1981-2010 with comparable simulations that do not account for the representation of model uncertainty. We find that the stochastic tendency perturbation schemes helped to reduce excessively strong convective activity especially over the Maritime Continent and the tropical Western Pacific, leading to reduced biases of the outgoing longwave radiation (OLR), cloud cover, precipitation and near-surface winds. Positive impact was also found for the statistics of the Madden-Julian oscillation (MJO), showing an increase in the frequencies and amplitudes of MJO events. Further, the errors of El Niño southern oscillation forecasts become smaller, whereas increases in ensemble spread lead to a better calibrated system if the stochastic tendency is activated. The backscatter scheme has overall neutral impact. Finally, evidence for noise-activated regime transitions has been found in a cluster analysis of mid-latitude circulation regimes over the Pacific-North America region.

Improving Energy Use Forecast for Campus Micro-grids using Indirect Indicators

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

Aman, Saima; Simmhan, Yogesh; Prasanna, Viktor K.

2011-12-11

The rising global demand for energy is best addressed by adopting and promoting sustainable methods of power consumption. We employ an informatics approach towards forecasting the energy consumption patterns in a university campus micro-grid which can be used for energy use planning and conservation. We use novel indirect indicators of energy that are commonly available to train regression tree models that can predict campus and building energy use for coarse (daily) and fine (15-min) time intervals, utilizing 3 years of sensor data collected at 15min intervals from 170 smart power meters. We analyze the impact of individual features used inmore » the models to identify the ones best suited for the application. Our models show a high degree of accuracy with CV-RMSE errors ranging from 7.45% to 19.32%, and a reduction in error from baseline models by up to 53%.« less

Short-term forecasting of emergency inpatient flow.

PubMed

Abraham, Gad; Byrnes, Graham B; Bain, Christopher A

2009-05-01

Hospital managers have to manage resources effectively, while maintaining a high quality of care. For hospitals where admissions from the emergency department to the wards represent a large proportion of admissions, the ability to forecast these admissions and the resultant ward occupancy is especially useful for resource planning purposes. Since emergency admissions often compete with planned elective admissions, modeling emergency demand may result in improved elective planning as well. We compare several models for forecasting daily emergency inpatient admissions and occupancy. The models are applied to three years of daily data. By measuring their mean square error in a cross-validation framework, we find that emergency admissions are largely random, and hence, unpredictable, whereas emergency occupancy can be forecasted using a model combining regression and autoregressive integrated moving average (ARIMA) model, or a seasonal ARIMA model, for up to one week ahead. Faced with variable admissions and occupancy, hospitals must prepare a reserve capacity of beds and staff. Our approach allows estimation of the required reserve capacity.

Improving Global Reanalyses and Short Range Forecast Using TRMM and SSM/I-Derived Precipitation and Moisture Observations

NASA Technical Reports Server (NTRS)

Hou, Arthur Y.; Zhang, Sara Q.; deSilva, Arlindo M.

2000-01-01

Global reanalyses currently contain significant errors in the primary fields of the hydrological cycle such as precipitation, evaporation, moisture, and the related cloud fields, especially in the tropics. The Data Assimilation Office (DAO) at the NASA Goddard Space Flight Center has been exploring the use of tropical rainfall and total precipitable water (TPW) observations from the TRMM Microwave Imager (TMI) and the Special Sensor Microwave/ Imager (SSM/I) instruments to improve short-range forecast and reanalyses. We describe a "1+1"D procedure for assimilating 6-hr averaged rainfall and TPW in the Goddard Earth Observing System (GEOS) Data Assimilation System (DAS). The algorithm is based on a 6-hr time integration of a column version of the GEOS DAS, hence the "1+1"D designation. The scheme minimizes the least-square differences between the observed TPW and rain rates and those produced by the column model over the 6-hr analysis window. This 1+lD scheme, in its generalization to four dimensions, is related to the standard 4D variational assimilation but uses analysis increments instead of the initial condition as the control variable. Results show that assimilating the TMI and SSM/I rainfall and TPW observations improves not only the precipitation and moisture fields but also key climate parameters such as clouds, the radiation, the upper-tropospheric moisture, and the large-scale circulation in the tropics. In particular, assimilating these data reduce the state-dependent systematic errors in the assimilated products. The improved analysis also provides better initial conditions for short-range forecasts, but the improvements in forecast are less than improvements in the time-averaged assimilation fields, indicating that using these data types is effective in correcting biases and other errors of the forecast model in data assimilation.

Improving Global Reanalyses and Short-Range Forecast Using TRMM and SSM/I-Derived Precipitation and Moisture Observations

NASA Technical Reports Server (NTRS)

Hou, Arthur Y.; Zhang, Sara Q.; daSilva, Arlindo M.

1999-01-01

Global reanalyses currently contain significant errors in the primary fields of the hydrological cycle such as precipitation, evaporation, moisture, and the related cloud fields, especially in the tropics. The Data Assimilation Office (DAO) at the NASA Goddard Space Flight Center has been exploring the use of tropical rainfall and total precipitable water (TPW) observations from the TRMM Microwave Imager (TMI) and the Special Sensor Microwave/ Imager (SSM/I) instruments to improve short-range forecast and reanalyses. We describe a 1+1D procedure for assimilating 6-hr averaged rainfall and TPW in the Goddard Earth Observing System (GEOS) Data Assimilation System (DAS). The algorithm is based on a 6-hr time integration of a column version of the GEOS DAS, hence the 1+1D designation. The scheme minimizes the least-square differences between the observed TPW and rain rates and those produced by the column model over the 6-hr analysis window. This 1+1D scheme, in its generalization to four dimensions, is related to the standard 4D variational assimilation but uses analysis increments instead of the initial condition as the control variable. Results show that assimilating the TMI and SSW rainfall and TPW observations improves not only the precipitation and moisture fields but also key climate parameters such as clouds, the radiation, the upper-tropospheric moisture, and the large-scale circulation in the tropics. In particular, assimilating these data reduce the state-dependent systematic errors in the assimilated products. The improved analysis also provides better initial conditions for short-range forecasts, but the improvements in forecast are less than improvements in the time-averaged assimilation fields, indicating that using these data types is effective in correcting biases and other errors of the forecast model in data assimilation.

Real-time flood forecasting by employing artificial neural network based model with zoning matching approach

NASA Astrophysics Data System (ADS)

Sulaiman, M.; El-Shafie, A.; Karim, O.; Basri, H.

2011-10-01

Flood forecasting models are a necessity, as they help in planning for flood events, and thus help prevent loss of lives and minimize damage. At present, artificial neural networks (ANN) have been successfully applied in river flow and water level forecasting studies. ANN requires historical data to develop a forecasting model. However, long-term historical water level data, such as hourly data, poses two crucial problems in data training. First is that the high volume of data slows the computation process. Second is that data training reaches its optimal performance within a few cycles of data training, due to there being a high volume of normal water level data in the data training, while the forecasting performance for high water level events is still poor. In this study, the zoning matching approach (ZMA) is used in ANN to accurately monitor flood events in real time by focusing the development of the forecasting model on high water level zones. ZMA is a trial and error approach, where several training datasets using high water level data are tested to find the best training dataset for forecasting high water level events. The advantage of ZMA is that relevant knowledge of water level patterns in historical records is used. Importantly, the forecasting model developed based on ZMA successfully achieves high accuracy forecasting results at 1 to 3 h ahead and satisfactory performance results at 6 h. Seven performance measures are adopted in this study to describe the accuracy and reliability of the forecasting model developed.

Replacement Beef Cow Valuation under Data Availability Constraints

PubMed Central

Hagerman, Amy D.; Thompson, Jada M.; Ham, Charlotte; Johnson, Kamina K.

2017-01-01

Economists are often tasked with estimating the benefits or costs associated with livestock production losses; however, lack of available data or absence of consistent reporting can reduce the accuracy of these valuations. This work looks at three potential estimation techniques for determining the value for replacement beef cows with varying types of market data to proxy constrained data availability and discusses the potential margin of error for each technique. Oklahoma bred replacement cows are valued using hedonic pricing based on Oklahoma bred cow data—a best case scenario—vector error correction modeling (VECM) based on national cow sales data and cost of production (COP) based on just a representative enterprise budget and very limited sales data. Each method was then used to perform a within-sample forecast of 2016 January to December, and forecasts are compared with the 2016 monthly observed market prices in Oklahoma using the mean absolute percent error (MAPE). Hedonic pricing methods tend to overvalue for within-sample forecasting but performed best, as measured by MAPE for high quality cows. The VECM tended to undervalue cows but performed best for younger animals. COP performed well, compared with the more data intensive methods. Examining each method individually across eight representative replacement beef female types, the VECM forecast resulted in a MAPE under 10% for 33% of forecasted months, followed by hedonic pricing at 24% of the forecasted months and COP at 14% of the forecasted months for average quality beef females. For high quality females, the hedonic pricing method worked best producing a MAPE under 10% in 36% of the forecasted months followed by the COP method at 21% of months and the VECM at 14% of the forecasted months. These results suggested that livestock valuation method selection was not one-size-fits-all and may need to vary based not only on the data available but also on the characteristics (e.g., quality or age) of the livestock being valued. PMID:29164141

Using a Software Tool in Forecasting: a Case Study of Sales Forecasting Taking into Account Data Uncertainty

NASA Astrophysics Data System (ADS)

Fabianová, Jana; Kačmáry, Peter; Molnár, Vieroslav; Michalik, Peter

2016-10-01

Forecasting is one of the logistics activities and a sales forecast is the starting point for the elaboration of business plans. Forecast accuracy affects the business outcomes and ultimately may significantly affect the economic stability of the company. The accuracy of the prediction depends on the suitability of the use of forecasting methods, experience, quality of input data, time period and other factors. The input data are usually not deterministic but they are often of random nature. They are affected by uncertainties of the market environment, and many other factors. Taking into account the input data uncertainty, the forecast error can by reduced. This article deals with the use of the software tool for incorporating data uncertainty into forecasting. Proposals are presented of a forecasting approach and simulation of the impact of uncertain input parameters to the target forecasted value by this case study model. The statistical analysis and risk analysis of the forecast results is carried out including sensitivity analysis and variables impact analysis.

Debiasing affective forecasting errors with targeted, but not representative, experience narratives.

PubMed

Shaffer, Victoria A; Focella, Elizabeth S; Scherer, Laura D; Zikmund-Fisher, Brian J

2016-10-01

To determine whether representative experience narratives (describing a range of possible experiences) or targeted experience narratives (targeting the direction of forecasting bias) can reduce affective forecasting errors, or errors in predictions of experiences. In Study 1, participants (N=366) were surveyed about their experiences with 10 common medical events. Those who had never experienced the event provided ratings of predicted discomfort and those who had experienced the event provided ratings of actual discomfort. Participants making predictions were randomly assigned to either the representative experience narrative condition or the control condition in which they made predictions without reading narratives. In Study 2, participants (N=196) were again surveyed about their experiences with these 10 medical events, but participants making predictions were randomly assigned to either the targeted experience narrative condition or the control condition. Affective forecasting errors were observed in both studies. These forecasting errors were reduced with the use of targeted experience narratives (Study 2) but not representative experience narratives (Study 1). Targeted, but not representative, narratives improved the accuracy of predicted discomfort. Public collections of patient experiences should favor stories that target affective forecasting biases over stories representing the range of possible experiences. Copyright © 2016 Elsevier Ireland Ltd. All rights reserved.

EMC: Air Quality Forecast Home page

Science.gov Websites

archive NAM Verification Meteorology Error Time Series EMC NAM Spatial Maps Real Time Mesoscale Analysis Precipitation verification NAQFC VERIFICATION CMAQ Ozone & PM Error Time Series AOD Error Time Series HYSPLIT Smoke forecasts vs GASP satellite Dust and Smoke Error Time Series HYSPLIT WCOSS Upgrade (July

Ocean regional circulation model sensitizes to resolution of the lateral boundary conditions

NASA Astrophysics Data System (ADS)

Pham, Van Sy; Hwang, Jin Hwan

2017-04-01

Dynamical downscaling with nested regional oceanographic models is an effective approach for forecasting operationally coastal weather and projecting long term climate on the ocean. Nesting procedures deliver the unwanted in dynamic downscaling due to the differences of numerical grid sizes and updating steps. Therefore, such unavoidable errors restrict the application of the Ocean Regional Circulation Model (ORCMs) in both short-term forecasts and long-term projections. The current work identifies the effects of errors induced by computational limitations during nesting procedures on the downscaled results of the ORCMs. The errors are quantitatively evaluated for each error source and its characteristics by the Big-Brother Experiments (BBE). The BBE separates identified errors from each other and quantitatively assess the amount of uncertainties employing the same model to simulate for both nesting and nested model. Here, we focus on discussing errors resulting from two main matters associated with nesting procedures. They should be the spatial grids' differences and the temporal updating steps. After the diverse cases from separately running of the BBE, a Taylor diagram was adopted to analyze the results and suggest an optimization intern of grid size and updating period and domain sizes. Key words: lateral boundary condition, error, ocean regional circulation model, Big-Brother Experiment. Acknowledgement: This research was supported by grants from the Korean Ministry of Oceans and Fisheries entitled "Development of integrated estuarine management system" and a National Research Foundation of Korea (NRF) Grant (No. 2015R1A5A 7037372) funded by MSIP of Korea. The authors thank the Integrated Research Institute of Construction and Environmental Engineering of Seoul National University for administrative support.

Optimal Power Flow for Distribution Systems under Uncertain Forecasts: Preprint

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

Dall'Anese, Emiliano; Baker, Kyri; Summers, Tyler

2016-12-01

The paper focuses on distribution systems featuring renewable energy sources and energy storage devices, and develops an optimal power flow (OPF) approach to optimize the system operation in spite of forecasting errors. The proposed method builds on a chance-constrained multi-period AC OPF formulation, where probabilistic constraints are utilized to enforce voltage regulation with a prescribed probability. To enable a computationally affordable solution approach, a convex reformulation of the OPF task is obtained by resorting to i) pertinent linear approximations of the power flow equations, and ii) convex approximations of the chance constraints. Particularly, the approximate chance constraints provide conservative boundsmore » that hold for arbitrary distributions of the forecasting errors. An adaptive optimization strategy is then obtained by embedding the proposed OPF task into a model predictive control framework.« less

Evaluating Snow Data Assimilation Framework for Streamflow Forecasting Applications Using Hindcast Verification

NASA Astrophysics Data System (ADS)

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

2012-12-01

Snow water equivalent (SWE) estimation is a key factor in producing reliable streamflow simulations and forecasts in snow dominated areas. However, measuring or predicting SWE has significant uncertainty. Sequential data assimilation, which updates states using both observed and modeled data based on error estimation, has been shown to reduce streamflow simulation errors but has had limited testing for forecasting applications. In the current study, a snow data assimilation framework integrated with the National Weather System River Forecasting System (NWSRFS) is evaluated for use in ensemble streamflow prediction (ESP). Seasonal water supply ESP hindcasts are generated for the North Fork of the American River Basin (NFARB) in northern California. Parameter sets from the California Nevada River Forecast Center (CNRFC), the Differential Evolution Adaptive Metropolis (DREAM) algorithm and the Multistep Automated Calibration Scheme (MACS) are tested both with and without sequential data assimilation. The traditional ESP method considers uncertainty in future climate conditions using historical temperature and precipitation time series to generate future streamflow scenarios conditioned on the current basin state. We include data uncertainty analysis in the forecasting framework through the DREAM-based parameter set which is part of a recently developed Integrated Uncertainty and Ensemble-based data Assimilation framework (ICEA). Extensive verification of all tested approaches is undertaken using traditional forecast verification measures, including root mean square error (RMSE), Nash-Sutcliffe efficiency coefficient (NSE), volumetric bias, joint distribution, rank probability score (RPS), and discrimination and reliability plots. In comparison to the RFC parameters, the DREAM and MACS sets show significant improvement in volumetric bias in flow. Use of assimilation improves hindcasts of higher flows but does not significantly improve performance in the mid flow and low flow categories.

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).

Alternatives to accuracy and bias metrics based on percentage errors for radiation belt modeling applications

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

Morley, Steven Karl

This report reviews existing literature describing forecast accuracy metrics, concentrating on those based on relative errors and percentage errors. We then review how the most common of these metrics, the mean absolute percentage error (MAPE), has been applied in recent radiation belt modeling literature. Finally, we describe metrics based on the ratios of predicted to observed values (the accuracy ratio) that address the drawbacks inherent in using MAPE. Specifically, we define and recommend the median log accuracy ratio as a measure of bias and the median symmetric accuracy as a measure of accuracy.

Non-parametric data-based approach for the quantification and communication of uncertainties in river flood forecasts

NASA Astrophysics Data System (ADS)

Van Steenbergen, N.; Willems, P.

2012-04-01

Reliable flood forecasts are the most important non-structural measures to reduce the impact of floods. However flood forecasting systems are subject to uncertainty originating from the input data, model structure and model parameters of the different hydraulic and hydrological submodels. To quantify this uncertainty a non-parametric data-based approach has been developed. This approach analyses the historical forecast residuals (differences between the predictions and the observations at river gauging stations) without using a predefined statistical error distribution. Because the residuals are correlated with the value of the forecasted water level and the lead time, the residuals are split up into discrete classes of simulated water levels and lead times. For each class, percentile values are calculated of the model residuals and stored in a 'three dimensional error' matrix. By 3D interpolation in this error matrix, the uncertainty in new forecasted water levels can be quantified. In addition to the quantification of the uncertainty, the communication of this uncertainty is equally important. The communication has to be done in a consistent way, reducing the chance of misinterpretation. Also, the communication needs to be adapted to the audience; the majority of the larger public is not interested in in-depth information on the uncertainty on the predicted water levels, but only is interested in information on the likelihood of exceedance of certain alarm levels. Water managers need more information, e.g. time dependent uncertainty information, because they rely on this information to undertake the appropriate flood mitigation action. There are various ways in presenting uncertainty information (numerical, linguistic, graphical, time (in)dependent, etc.) each with their advantages and disadvantages for a specific audience. A useful method to communicate uncertainty of flood forecasts is by probabilistic flood mapping. These maps give a representation of the probability of flooding of a certain area, based on the uncertainty assessment of the flood forecasts. By using this type of maps, water managers can focus their attention on the areas with the highest flood probability. Also the larger public can consult these maps for information on the probability of flooding for their specific location, such that they can take pro-active measures to reduce the personal damage. The method of quantifying the uncertainty was implemented in the operational flood forecasting system for the navigable rivers in the Flanders region of Belgium. The method has shown clear benefits during the floods of the last two years.

Sustainable Odds

NASA Astrophysics Data System (ADS)

Smith, L. A.

2016-12-01

While probability forecasting has many philosophical and mathematical attractions, it is something of a dishonest nonsense if acting on such forecasts is expected to lead to rapid ruin. Model-based probabilities, when interpreted as actionable, are shown to lead to the rapid ruin of a cooperative entity offering odds interpreting the probability forecasts at face value. Arguably, these odds would not be considered "fair", but inasmuch as some definitions of "fair odds" include this case, this presentation will focus on "sustainable odds": Odds which are not expected to lead to the rapid ruin of the cooperative under the assumption that those placing bets have no information beyond that available to the forecast system. It is argued that sustainable odds will not correspond to probabilities outside the Perfect Model Scenario, that the "implied probabilities" determined from sustainable odds will always sum to more than one, and that the excess of this sum over one reflects the skill of the forecast system, being a quantitative measure of structural model error.

Flood Forecast Accuracy and Decision Support System Approach: the Venice Case

NASA Astrophysics Data System (ADS)

Canestrelli, A.; Di Donato, M.

2016-02-01

In the recent years numerical models for weather predictions have experienced continuous advances in technology. As a result, all the disciplines making use of weather forecasts have made significant steps forward. In the case of the Safeguard of Venice, a large effort has been put in order to improve the forecast of tidal levels. In this context, the Istituzione Centro Previsioni e Segnalazioni Maree (ICPSM) of the Venice Municipality has developed and tested many different forecast models, both of the statistical and deterministic type, and has shown to produce very accurate forecasts. For Venice, the maximum admissible forecast error should be (ideally) of the order of ten centimeters at 24 hours. The entity of the forecast error clearly affects the decisional process, which mainly consists of alerting the population, activating the movable barriers installed at the three tidal inlets and contacting the port authority. This process becomes more challenging whenever the weather predictions, and therefore the water level forecasts, suddenly change. These new forecasts have to be quickly transformed into operational tasks. Therefore, it is of the utter importance to set up scheduled alerts and emergency plans by means of easy-to-follow procedures. On this direction, Technital has set up a Decision Support System based on expert procedures that minimizes the human mistakes and, as a consequence, reduces the risk of flooding of the historical center. Moreover, the Decision Support System can communicate predefined alerts to all the interested subjects. The System uses the water levels forecasts produced by the ICPSM by taking into account the accuracy at different leading times. The Decision Support System has been successfully tested with 8 years of data, 6 of them in real time. Venice experience shows that the Decision Support System is an essential tool which assesses the risks associated with a particular event, provides clear operational procedures and minimizes the impact of natural floods on human lives, private properties and historical monuments.

A revised radiation package of G-packed McICA and two-stream approximation: Performance evaluation in a global weather forecasting model

NASA Astrophysics Data System (ADS)

Baek, Sunghye

2017-07-01

For more efficient and accurate computation of radiative flux, improvements have been achieved in two aspects, integration of the radiative transfer equation over space and angle. First, the treatment of the Monte Carlo-independent column approximation (MCICA) is modified focusing on efficiency using a reduced number of random samples ("G-packed") within a reconstructed and unified radiation package. The original McICA takes 20% of CPU time of radiation in the Global/Regional Integrated Model systems (GRIMs). The CPU time consumption of McICA is reduced by 70% without compromising accuracy. Second, parameterizations of shortwave two-stream approximations are revised to reduce errors with respect to the 16-stream discrete ordinate method. Delta-scaled two-stream approximation (TSA) is almost unanimously used in Global Circulation Model (GCM) but contains systematic errors which overestimate forward peak scattering as solar elevation decreases. These errors are alleviated by adjusting the parameterizations of each scattering element—aerosol, liquid, ice and snow cloud particles. Parameterizations are determined with 20,129 atmospheric columns of the GRIMs data and tested with 13,422 independent data columns. The result shows that the root-mean-square error (RMSE) over the all atmospheric layers is decreased by 39% on average without significant increase in computational time. Revised TSA developed and validated with a separate one-dimensional model is mounted on GRIMs for mid-term numerical weather forecasting. Monthly averaged global forecast skill scores are unchanged with revised TSA but the temperature at lower levels of the atmosphere (pressure ≥ 700 hPa) is slightly increased (< 0.5 K) with corrected atmospheric absorption.

Ultra-Short-Term Wind Power Prediction Using a Hybrid Model

NASA Astrophysics Data System (ADS)

Mohammed, E.; Wang, S.; Yu, J.

2017-05-01

This paper aims to develop and apply a hybrid model of two data analytical methods, multiple linear regressions and least square (MLR&LS), for ultra-short-term wind power prediction (WPP), for example taking, Northeast China electricity demand. The data was obtained from the historical records of wind power from an offshore region, and from a wind farm of the wind power plant in the areas. The WPP achieved in two stages: first, the ratios of wind power were forecasted using the proposed hybrid method, and then the transformation of these ratios of wind power to obtain forecasted values. The hybrid model combines the persistence methods, MLR and LS. The proposed method included two prediction types, multi-point prediction and single-point prediction. WPP is tested by applying different models such as autoregressive moving average (ARMA), autoregressive integrated moving average (ARIMA) and artificial neural network (ANN). By comparing results of the above models, the validity of the proposed hybrid model is confirmed in terms of error and correlation coefficient. Comparison of results confirmed that the proposed method works effectively. Additional, forecasting errors were also computed and compared, to improve understanding of how to depict highly variable WPP and the correlations between actual and predicted wind power.

An operational wave forecasting system for the east coast of India

NASA Astrophysics Data System (ADS)

Sandhya, K. G.; Murty, P. L. N.; Deshmukh, Aditya N.; Balakrishnan Nair, T. M.; Shenoi, S. S. C.

2018-03-01

Demand for operational ocean state forecasting is increasing, owing to the ever-increasing marine activities in the context of blue economy. In the present study, an operational wave forecasting system for the east coast of India is proposed using unstructured Simulating WAves Nearshore model (UNSWAN). This modelling system uses very high resolution mesh near the Indian east coast and coarse resolution offshore, and thus avoids the necessity of nesting with a global wave model. The model is forced with European Centre for Medium-Range Weather Forecasts (ECMWF) winds and simulates wave parameters and wave spectra for the next 3 days. The spatial pictures of satellite data overlaid on simulated wave height show that the model is capable of simulating the significant wave heights and their gradients realistically. Spectral validation has been done using the available data to prove the reliability of the model. To further evaluate the model performance, the wave forecast for the entire year 2014 is evaluated against buoy measurements over the region at 4 waverider buoy locations. Seasonal analysis of significant wave height (Hs) at the four locations showed that the correlation between the modelled and observed was the highest (in the range 0.78-0.96) during the post-monsoon season. The variability of Hs was also the highest during this season at all locations. The error statistics showed clear seasonal and geographical location dependence. The root mean square error at Visakhapatnam was the same (0.25) for all seasons, but it was the smallest for pre-monsoon season (0.12 m and 0.17 m) for Puducherry and Gopalpur. The wind sea component showed higher variability compared to the corresponding swell component in all locations and for all seasons. The variability was picked by the model to a reasonable level in most of the cases. The results of statistical analysis show that the modelling system is suitable for use in the operational scenario.

Analysis of forecasting and inventory control of raw material supplies in PT INDAC INT’L

NASA Astrophysics Data System (ADS)

Lesmana, E.; Subartini, B.; Riaman; Jabar, D. A.

2018-03-01

This study discusses the data forecasting sales of carbon electrodes at PT. INDAC INT L uses winters and double moving average methods, while for predicting the amount of inventory and cost required in ordering raw material of carbon electrode next period using Economic Order Quantity (EOQ) model. The result of error analysis shows that winters method for next period gives result of MAE, MSE, and MAPE, the winters method is a better forecasting method for forecasting sales of carbon electrode products. So that PT. INDAC INT L is advised to provide products that will be sold following the sales amount by the winters method.

Improving the Forecast Accuracy of an Ocean Observation and Prediction System by Adaptive Control of the Sensor Network

NASA Astrophysics Data System (ADS)

Talukder, A.; Panangadan, A. V.; Blumberg, A. F.; Herrington, T.; Georgas, N.

2008-12-01

The New York Harbor Observation and Prediction System (NYHOPS) is a real-time, estuarine and coastal ocean observing and modeling system for the New York Harbor and surrounding waters. Real-time measurements from in-situ mobile and stationary sensors in the NYHOPS networks are assimilated into marine forecasts in order to reduce the discrepancy with ground truth. The forecasts are obtained from the ECOMSED hydrodynamic model, a shallow water derivative of the Princeton Ocean Model. Currently, all sensors in the NYHOPS system are operated in a fixed mode with uniform sampling rates. This technology infusion effort demonstrates the use of Model Predictive Control (MPC) to autonomously adapt the operation of both mobile and stationary sensors in response to changing events that are -automatically detected from the ECOMSED forecasts. The controller focuses sensing resources on those regions that are expected to be impacted by the detected events. The MPC approach involves formulating the problem of calculating the optimal sensor parameters as a constrained multi-objective optimization problem. We have developed an objective function that takes into account the spatiotemporal relationship of the in-situ sensor locations and the locations of events detected by the model. Experiments in simulation were carried out using data collected during a freshwater flooding event. The location of the resulting freshwater plume was calculated from the corresponding model forecasts and was used by the MPC controller to derive control parameters for the sensing assets. The operational parameters that are controlled include the sampling rates of stationary sensors, paths of unmanned underwater vehicles (UUVs), and data transfer routes between sensors and the central modeling computer. The simulation experiments show that MPC-based sensor control reduces the RMS error in the forecast by a factor of 380% as compared to uniform sampling. The paths of multiple UUVs were simultaneously calculated such that measurements from on-board sensors would lead to maximal reduction in the forecast error after data assimilation. The MPC controller also reduces the consumption of system resources such as energy expended in sampling and wireless communication. The MPC-based control approach can be generalized to accept data from remote sensing satellites. This will enable in-situ sensors to be regulated using forecasts generated by assimilating local high resolution in-situ measurements with wide-area observations from remote sensing satellites.

Microgrid optimal scheduling considering impact of high penetration wind generation

NASA Astrophysics Data System (ADS)

Alanazi, Abdulaziz

The objective of this thesis is to study the impact of high penetration wind energy in economic and reliable operation of microgrids. Wind power is variable, i.e., constantly changing, and nondispatchable, i.e., cannot be controlled by the microgrid controller. Thus an accurate forecasting of wind power is an essential task in order to study its impacts in microgrid operation. Two commonly used forecasting methods including Autoregressive Integrated Moving Average (ARIMA) and Artificial Neural Network (ANN) have been used in this thesis to improve the wind power forecasting. The forecasting error is calculated using a Mean Absolute Percentage Error (MAPE) and is improved using the ANN. The wind forecast is further used in the microgrid optimal scheduling problem. The microgrid optimal scheduling is performed by developing a viable model for security-constrained unit commitment (SCUC) based on mixed-integer linear programing (MILP) method. The proposed SCUC is solved for various wind penetration levels and the relationship between the total cost and the wind power penetration is found. In order to reduce microgrid power transfer fluctuations, an additional constraint is proposed and added to the SCUC formulation. The new constraint would control the time-based fluctuations. The impact of the constraint on microgrid SCUC results is tested and validated with numerical analysis. Finally, the applicability of proposed models is demonstrated through numerical simulations.

Performance and Quality Assessment of the Forthcoming Copernicus Marine Service Global Ocean Monitoring and Forecasting Real-Time System

NASA Astrophysics Data System (ADS)

Lellouche, J. M.; Le Galloudec, O.; Greiner, E.; Garric, G.; Regnier, C.; Drillet, Y.

2016-02-01

Mercator Ocean currently delivers in real-time daily services (weekly analyses and daily forecast) with a global 1/12° high resolution system. The model component is the NEMO platform driven at the surface by the IFS ECMWF atmospheric analyses and forecasts. Observations are assimilated by means of a reduced-order Kalman filter with a 3D multivariate modal decomposition of the forecast error. It includes an adaptive-error estimate and a localization algorithm. Along track altimeter data, satellite Sea Surface Temperature and in situ temperature and salinity vertical profiles are jointly assimilated to estimate the initial conditions for numerical ocean forecasting. A 3D-Var scheme provides a correction for the slowly-evolving large-scale biases in temperature and salinity.Since May 2015, Mercator Ocean opened the Copernicus Marine Service (CMS) and is in charge of the global ocean analyses and forecast, at eddy resolving resolution. In this context, R&D activities have been conducted at Mercator Ocean these last years in order to improve the real-time 1/12° global system for the next CMS version in 2016. The ocean/sea-ice model and the assimilation scheme benefit among others from the following improvements: large-scale and objective correction of atmospheric quantities with satellite data, new Mean Dynamic Topography taking into account the last version of GOCE geoid, new adaptive tuning of some observational errors, new Quality Control on the assimilated temperature and salinity vertical profiles based on dynamic height criteria, assimilation of satellite sea-ice concentration, new freshwater runoff from ice sheets melting …This presentation doesn't focus on the impact of each update, but rather on the overall behavior of the system integrating all updates. This assessment reports on the products quality improvements, highlighting the level of performance and the reliability of the new system.

An Open-Book Modular Watershed Modeling Framework for Rapid Prototyping of GPM- based Flood Forecasting in International River Basins

NASA Astrophysics Data System (ADS)

Katiyar, N.; Hossain, F.

2006-05-01

Floods have always been disastrous for human life. It accounts for about 15 % of the total death related to natural disasters. There are around 263 transboundary river basins listed by UNESCO, wherein at least 30 countries have more than 95% of their territory locked in one or more such transboundary basins. For flood forecasting in the lower riparian nations of these International River Basins (IRBs), real-time rainfall data from upstream nations is naturally the most critical factor governing the forecasting effectiveness. However, many upstream nations fail to provide data to the lower riparian nations due to a lack of in-situ rainfall measurement infrastructure or a lack of a treaty for real-time sharing of rainfall data. A potential solution is therefore to use satellites that inherently measure rainfall across political boundaries. NASA's proposed Global Precipitation Measurement (GPM) mission appears very promising in providing this vital rainfall information under the data- limited scenario that will continue to prevail in most IRBs. However, satellite rainfall is associated with uncertainty and hence, proper characterization of the satellite rainfall error propagation in hydrologic models for flood forecasting is a critical priority that should be resolved in the coming years in anticipation of GPM. In this study, we assess an open book modular watershed modeling approach for estimating the expected error in flood forecasting related to GPM rainfall data. Our motivation stems from the critical challenge in identifying the specific IRBs that would benefit from a pre-programmed satellite-based forecasting system in anticipation of GPM. As the number of flood-prone IRBs is large, conventional data-intensive implementation of existing physically-based distributed hydrologic models on case-by-case IRBs is considered time-consuming for completing such a global assessment. A more parsimonious approach is justified at the expense of a tolerable loss of detail and accuracy. Through assessment of our proposed modular modeling framework, we present our initial understanding in resolving the fundamental question - Can a parsimonious open-book watershed modeling framework be a physically consistent proxy for rapid and global identification of IRBs in greater need of a GPM-based flood forecasting system?

Calibration of limited-area ensemble precipitation forecasts for hydrological predictions

NASA Astrophysics Data System (ADS)

Diomede, Tommaso; Marsigli, Chiara; Montani, Andrea; Nerozzi, Fabrizio; Paccagnella, Tiziana

2015-04-01

The main objective of this study is to investigate the impact of calibration for limited-area ensemble precipitation forecasts, to be used for driving discharge predictions up to 5 days in advance. A reforecast dataset, which spans 30 years, based on the Consortium for Small Scale Modeling Limited-Area Ensemble Prediction System (COSMO-LEPS) was used for testing the calibration strategy. Three calibration techniques were applied: quantile-to-quantile mapping, linear regression, and analogs. The performance of these methodologies was evaluated in terms of statistical scores for the precipitation forecasts operationally provided by COSMO-LEPS in the years 2003-2007 over Germany, Switzerland, and the Emilia-Romagna region (northern Italy). The analog-based method seemed to be preferred because of its capability of correct position errors and spread deficiencies. A suitable spatial domain for the analog search can help to handle model spatial errors as systematic errors. However, the performance of the analog-based method may degrade in cases where a limited training dataset is available. A sensitivity test on the length of the training dataset over which to perform the analog search has been performed. The quantile-to-quantile mapping and linear regression methods were less effective, mainly because the forecast-analysis relation was not so strong for the available training dataset. A comparison between the calibration based on the deterministic reforecast and the calibration based on the full operational ensemble used as training dataset has been considered, with the aim to evaluate whether reforecasts are really worthy for calibration, given that their computational cost is remarkable. The verification of the calibration process was then performed by coupling ensemble precipitation forecasts with a distributed rainfall-runoff model. This test was carried out for a medium-sized catchment located in Emilia-Romagna, showing a beneficial impact of the analog-based method on the reduction of missed events for discharge predictions.

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.

Ensemble superparameterization versus stochastic parameterization: A comparison of model uncertainty representation in tropical weather prediction

NASA Astrophysics Data System (ADS)

Subramanian, Aneesh C.; Palmer, Tim N.

2017-06-01

Stochastic schemes to represent model uncertainty in the European Centre for Medium-Range Weather Forecasts (ECMWF) ensemble prediction system has helped improve its probabilistic forecast skill over the past decade by both improving its reliability and reducing the ensemble mean error. The largest uncertainties in the model arise from the model physics parameterizations. In the tropics, the parameterization of moist convection presents a major challenge for the accurate prediction of weather and climate. Superparameterization is a promising alternative strategy for including the effects of moist convection through explicit turbulent fluxes calculated from a cloud-resolving model (CRM) embedded within a global climate model (GCM). In this paper, we compare the impact of initial random perturbations in embedded CRMs, within the ECMWF ensemble prediction system, with stochastically perturbed physical tendency (SPPT) scheme as a way to represent model uncertainty in medium-range tropical weather forecasts. We especially focus on forecasts of tropical convection and dynamics during MJO events in October-November 2011. These are well-studied events for MJO dynamics as they were also heavily observed during the DYNAMO field campaign. We show that a multiscale ensemble modeling approach helps improve forecasts of certain aspects of tropical convection during the MJO events, while it also tends to deteriorate certain large-scale dynamic fields with respect to stochastically perturbed physical tendencies approach that is used operationally at ECMWF.