Continuous Ranked Probability Skill Score Research Articles

Probabilistic Neural Networks for Ensemble Postprocessing

Abstract Accurate temperature forecasts are critical for various industries and sectors. We propose a probabilistic neural network (PNN), an extension of the distributional regression network (DRN), for 2-m temperature forecasts, consisting of three variants with different inputs and target variables. The first variant, standardized anomaly probabilistic neural network (SAPNN), employs a two-step approach involving standardized anomalies and global PNN modeling to effectively capture underlying features and anomalies. The second variant, PNN with geographical predictors (PNNGE), incorporates raw and static geographical predictors to enhance predictive performance. The third variant, PNN with station one-hot encoding (PNNEN), utilizes raw with station one-hot encoding predictors to represent geographical information effectively. We compare three PNN variants with two benchmarks: 1) standardized anomaly model output statistics (SAMOS) and 2) three DRN variants identical to those applied to PNN. These evaluations utilize ECMWF data from 2019 to 2020 at 6-h intervals up to 72 h over Hebei, China. Results show that SAPNN and PNNGE are better than SAMOS, while PNNEN notably exhibits a significant 14% improvement in the continuous ranked probability skill score (CRPSS). Moreover, the PNN variants exhibit comparable or superior performance to DRN regarding forecast accuracy, CRPSS, and reliability, showcasing a better-calibrated spread–error relationship. This study highlights the value of the proposed PNN variants with a distribution output in capturing nonlinear relationships within different sources of predictors and improving temperature forecast skills. Significance Statement This study aims to improve the accuracy, skills, and reliability of 2-m temperature forecasts, which are crucial in agriculture and energy management. To achieve this, we extend a popular artificial intelligence framework and explore four data sources with two schemes to systematically compare the predictive performances in making temperature forecasts. The findings of this research are vital as they offer novel ways to improve forecast skills. Imagine having a weather app that is significantly more accurate, enabling you to plan your day better. This study is about discovering innovative approaches to enhance forecast skills and reliability, which could benefit various aspects of our daily lives. One of the new methods even exhibits a 14% improvement in forecast skills.

Estudio de teleconexión entre variables hidrológicas y climatológicas en una cuenca de cabecera del Rio Maipo para la aplicación de modelos de pronóstico

This study conducts a teleconnection analysis of the seasonal streamflow during the dry season (winter and summer) at the Olivares River basin, a headwater of the Maipo River basin, with traditional climate indices (Antarctic Oscillation, Niño1+2, and Niño3.4), new indices obtained from sea surface temperature (SST) anomaly spatial fields, and in situ hydrometeorological variables from the previous season to identify potential predictors for implementing seasonal streamflow forecast models in the study area. To illustrate the potential of the predictors identified, we fit multiple linear regression models (MLRM) for seasonal streamflow forecast for 0- and 3-month lead times. The forecasts are validated using the leave-1-year-out cross-validation (LOOCV) approach and performance metrics such as the Pearson correlation coefficient (R), BIAS, Nash-Sutcliffe efficiency (NSE), and continuous rank probability skill score (CRPSS). Results show a good performance of the forecast model for cross-validation with R and NSE values ranging from 0.55 to 0.95 and from 0.28 to 0.88 for 0- and 3-month lead times during the dry season. This early implementation provides good perspectives for implementing probabilistic seasonal streamflow forecasting models, which can provide a powerful output to develop robust water management strategies to tackle water scarcity in the study area.

Adapting subseasonal-to-seasonal (S2S) precipitation forecast at watersheds for hydrologic ensemble streamflow forecasting with a machine learning-based post-processing approach

Accurate and reliable precipitation predictions made by dynamical forecast models could provide crucial information for human socioeconomic activities by enabling hydrologic forecasts at the Subseasonal-to-Seasonal (S2S) timescale. To utilize available S2S precipitation predictions for hydrologic forecasts, post-processing techniques have been applied to adapt the raw S2S precipitation to local watersheds. However, conventional statistical-based post-processing techniques are more focused on correcting the forecast bias, but rather limited in improving the predictive skill of available S2S precipitation forecasts. In this study, we combine the Random Forest classifiers (RF) with the Bias Correction and Spatial Disaggregation (BCSD) to adapt the 10-member ensemble precipitation forecast from the NASA Goddard Earth Observing System model version 5 (GEOS5) at 4 watersheds located in the NCEI South climate region of the United States. The adapted S2S precipitation is further applied for streamflow forecast by forcing a classical lumped hydrologic model. The performance of S2S precipitation as well as the corresponding streamflow predictions are benchmarked with the randomly resampled precipitation and the corresponding Ensemble Streamflow Prediction (ESP) framework-generated streamflow predictions. Evaluation statistics of Kling-Gupta Efficiency (KGE), Continuous Ranked Probability Skill Score (CRPSS), Reliability, Resolution, and Sharpness are employed to evaluate the predictive skill of precipitation and streamflow both deterministically and probabilistically. Our results indicate that dynamical S2S precipitation after forecast adaptation leads to consistently higher deterministic skill over ESP at all forecast lead times and across study watersheds. However, at longer forecast lead times beyond 10–15 days, S2S precipitation with a limited ensemble size does not present higher probabilistic skill than ESP. Our results shows that the joint application of RF and BCSD improves the predictive skill of the raw S2S precipitation at study watersheds in contrast to BCSD. Further, the added predictive skill of S2S precipitation brought by RF propagates into streamflow predictions, predominantly at longer forecast lead times exceeding 10 days. Overall, our results highlight the potential success of future work to apply other data-driven approaches to adapt the raw precipitation to local watersheds for more accurate and reliable streamflow forecasts at the S2S timescale.

Operational low-flow forecasting using LSTMs

This study focuses on exploring the potential of using Long Short-Term Memory networks (LSTMs) for low-flow forecasting for the Rhine River at Lobith on a daily scale with lead times up to 46 days ahead. A novel LSTM-based model architecture is designed to leverage both historical observation and forecasted meteorological data to carry out multi-step discharge time series forecasting. The feature and target selection for this deep learning (DL) model involves evaluating the use of different spatial resolutions for meteorological forcing (basin-averaged or subbasin-averaged), the impact of incorporating past discharge observations, and the use of different target variables (discharge Q or time-differenced discharge dQ). Then, the model is trained using the ERA5 dataset as meteorological forcing, and employed for operational forecast with ECMWF seasonal forecast (SEAS5) data. The forecast results are compared to a benchmark process-based model, wflow_sbm. This study also explores the flexibility of the DL model by fine-tuning the pretrained model with limited SEAS5 dataset. Key findings from feature and target selection include: (1) opting for subbasin-averaged meteorological variables significantly improves model performance compared to a basin-averaged approach. (2) Utilizing dQ as the target variable greatly boosts short-term forecast accuracy compared to using Q, with a mean absolute error (MAE) of 25 m3 s−1 and mean absolute percentage error (MAPE) of 0.02 for the first lead time, ensuring reliability and accuracy at the onset of the forecast horizon. (3) While incorporating historical discharge improves the forecasting of Q, its impact on predicting dQ is less pronounced for short lead times. In the operational forecast with SEAS5, compared to the wflow_sbm model, the DL model exhibits skill in forecasting low flows as evidenced by Continuous Ranked Probability Skill Score (CRPSS) median values of all lead times above zero, and better accuracy in forecasting drought events within short lead times. The wflow_sbm model shows higher accuracy for longer lead times. In the exploration of fine-tuning approach, the fine-tuned model generates marginal short-term enhancements in forecasting low-flow events over a non-fine-tuned model. Overall, this study contributes to advancing the field of low-flow forecasting using deep learning approach.

Can Machine Learning Models Be a Suitable Tool for Predicting Central European Cold Winter Weather on Subseasonal to Seasonal Time Scales?

Abstract Skillful weather prediction on subseasonal to seasonal time scales is crucial for many socioeconomic ventures. But forecasting, especially extremes, on these time scales is very challenging because the information from initial conditions is gradually lost. Therefore, data-driven methods are discussed as an alternative to numerical weather prediction models. Here, quantile regression forests (QRFs) and random forest classifiers (RFCs) are used for probabilistic forecasting of central European mean wintertime 2-m temperatures and cold wave days at lead times of 14, 21, and 28 days. ERA5 reanalysis meteorological predictors are used as input data for the machine learning models. For the winters of 2000/01–2019/20, the predictions are compared with a climatological ensemble obtained from E-OBS observational data. The evaluation is performed as full distribution predictions for continuous values using the continuous ranked probability skill score and as binary categorical forecasts using the Brier skill score. We find skill at lead times up to 28 days in the 20-winter mean and for individual winters. Case studies show that all used machine learning models are able to learn patterns in the data beyond climatology. A more detailed analysis using Shapley additive explanations suggests that both random forest (RF)-based models are able to learn physically known relationships in the data. This underlines that RF-based data-driven models can be a suitable tool for forecasting central European mean wintertime 2-m temperatures and the occurrence of cold wave days. Significance Statement Because of the chaotic nature of weather, it is very complicated to make predictions with traditional numerical methods 2–4 weeks in advance. Therefore, we use alternative, interpretable methods that “learn” to find statistically relevant patterns in meteorological data that can be used for forecasting central European mean surface wintertime temperatures and cold wave days. These methods are part of the so-called machine learning methods that do not rely on the traditional numerical equations anymore. We test our methods for 20 winters between 2000/01 and 2019/20 against a static weather prediction consisting of the past 30 winters. For single winters and in a mean over the 20 predicted winters, we find improved predictions up to 4 weeks in advance.

Less is more? A novel method for identifying and evaluating non-informative tracers in sediment source mixing models

PurposeAccelerated soil erosion poses a global hazard to soil health. Understanding soil and sediment behaviour through sediment fingerprinting enables the monitoring and identification of areas with high sediment delivery. Land-use specific sediment source apportionment is increasingly determined using the Bayesian mixing model MixSIAR with compound-specific stable isotopes (CSSI). Here, we investigate CSSIs of fatty acid (FA) tracer selection with a novel method to identify and investigate the effect of non-informative tracers on model performance.MethodsTo evaluate CSSI tracer selection, mathematical mixtures were generated using source soils (n = 28) from the Rhine catchment upstream of Basel (Switzerland). Using the continuous ranked probability (CRP) skill score, MixSIAR’s performance was evaluated for 11 combinations of FAs and 15 combinations of FAs with δ15N as a mixing line offset tracer. A novel scaling and discrimination analysis (SDA) was also developed to identify tracers with non-unique mixing spaces.ResultsFA only tracer combinations overestimated pasture contributions while underestimating arable contributions. When compared to models with only FA tracers, utilizing δ15N to offset the mixing line resulted in a 28% improvement in the CRP skill score. δ15N + δ13C FA26 was the optimal tracer set resulting in a 62% model improvement relative to δ15N + all δ13C FAs. The novel SDA method demonstrated how δ13C FA tracers have a non-unique mixing space and thus behave as non-informative tracers. Importantly, the inclusion of non-informative tracers decreased model performance.ConclusionsThese results indicate that MixSIAR did not handle non-informative CSSI tracers effectively. Accordingly, it may be advantageous to remove non-informative tracers, and where feasible, all combinations and permutations of tracers should be assessed to optimize tracer selection. Application of these tracer selection steps can help improve and advance the performance of sediment fingerprinting models and ultimately aid in improving erosion mitigation and management strategies.

Catchment‐scale skill assessment of seasonal precipitation forecasts across South Korea

AbstractClimate change is expected to make droughts more frequent and severe all over the world. South Korea is no exception and is already suffering from extreme droughts such as one that prolonged from 2013 to 2015 and caused nation‐wide damage. To mitigate drought damages, better management of existing water infrastructure is essential. A promising opportunity to improve operational decisions is to make use of seasonal weather forecasts that are provided by general circulation models and can be downscaled to the catchment scale. This study hence assesses the skill of seasonal forecasts over 20 catchments in South Korea, where the largest reservoirs are located. Datasets from four weather forecasting centres (ECMWF, UK Met Office, Météo France and DWD) have been evaluated over the period 2011–2020, and their skill quantified using the Continuous Ranked Probability Skill Score (CRPSS). We analyse how skill varies across the seasons and years, and if it can be linked to catchments characteristics. In doing so, we develop a methodology and a Python package to implement it, which is freely available for future applications to other regions. As for the study case, our results showed that among the four forecasting centres, ECMWF's forecasts were the most skilful in South Korea. In particular, seasonal forecasts outperform the climatology for 2 months of lead time and are more skilful during the wet season and in dry years. Linear bias correction is found to be useful to correct systematic seasonal biases, whereas we found no significant correlation between the catchment characteristics and forecast skill. We also investigated the possibility of anticipating dry years from seasonal forecasts and/or ENSO indices but found no significant link.

Daily ensemble river discharge reforecasts and real-time forecasts from the operational Global Flood Awareness System

Abstract. Operational global-scale hydrological forecasting systems are used to help manage hydrological extremes such as floods and droughts. The vast amounts of raw data that underpin forecast systems and the ability to generate information on forecast skill have, until now, not been publicly available. As part of the Global Flood Awareness System (GloFAS; https://www.globalfloods.eu/, last access: 3 December 2022) service evolution, in this paper daily ensemble river discharge reforecasts and real-time forecast datasets are made free and openly available through the Copernicus Climate Change Service (C3S) Climate Data Store (CDS). They include real-time forecast data starting on 1 January 2020 updated operationally every day and a 20-year set of reforecasts and associated metadata. This paper describes the model components and configuration used to generate the real-time river discharge forecasts and the reforecasts. An evaluation of ensemble forecast skill using the continuous ranked probability skill score (CRPSS) was also undertaken for river points around the globe. Results show that GloFAS is skilful in over 93 % of catchments in the short (1 to 3 d) and medium range (5 to 15 d) against a persistence benchmark forecast and skilful in over 80 % of catchments out to the extended range (16 to 30 d) against a climatological benchmark forecast. However, the strength of skill varies considerably by location with GloFAS found to have no or negative skill at longer lead times in broad hydroclimatic regions in tropical Africa, western coast of South America, and catchments dominated by snow and ice in high northern latitudes. Forecast skill is summarised as a new headline skill score available as a new layer on the GloFAS forecast Web Map Viewer to aid user interpretation and understanding of forecast quality.

Calibration of ECMWF SEAS5 based streamflow forecast in Seasonal hydrological forecasting for Citarum river basin, West Java, Indonesia

Study region:Citarum river basin, West Java — Indonesia. Study focus:We look at the skill of Empirical Quantile Mapping corrected ECMWF SEAS5 (SEAS5 EQM bias-corrected) based streamflow forecasts in the Citarum river basin. We focus on July to October because these are agriculturally important months in Java. We use a high-resolution hydrologic model (wflow_sbm) with data for the period 1989–2009. New hydrological insights for the region:Water users and agricultural practitioners commonly need monthly to seasonal hydrological forecasts. The forecasts should be sufficiently skillful and provide information that is relevant to the decisions makers in order to have practical value to them. We assess if skilful SEAS5 EQM bias-corrected based seasonal forecasts are available with the purpose to support rice production. In this streamflow forecast calibration, we look at different aggregation days and different lead times. For the verification, we use the Continuous Ranked Probability Skill Score (CRPSS), Brier Skill Score (BSS), and Mean Average Error (MAE). We also look at the correlation, the Root Mean Square Error (RMSE), and the Receiver operating characteristic Skill (ROCS). The LT1 and LT2 forecast show higher skills than longer lead times. Meanwhile, streamflow based on the aggregated forecast at 30 to 60 days aggregation days is more skillful than larger aggregations. In Indonesia, this study is a study that initiates using a hydrological model with inputs from a seasonal rainfall forecast.

A Bayesian Hierarchical Model Combination Framework for Real‐Time Daily Ensemble Streamflow Forecasting Across a Rainfed River Basin

AbstractThe frequent occurrence of floods during the rainy season is one of the threats in rainfed river basins, especially in river basins of India. This study implemented a Bayesian hierarchical model combination (BHMC) framework to generate skillful and reliable real‐time daily ensemble streamflow forecast and peak flow and demonstrates its utility in the Narmada River basin in Central India for the peak monsoon season (July–August). The framework incorporates information from multiple sources (e.g., deterministic hydrological forecast, meteorological forecast, and observed data) as predictors. The forecasts were validated with a leave‐1‐year‐out cross‐validation using accuracy metrics such as BIAS and Pearson correlation coefficient (R) and probabilistic metrics such as continuous ranked probability skill score, probability integral transform (PIT) plots, and the average width of the 95% confidence intervals (AWCI) plots. The results show that the BHMC framework can increase the forecast skill by 40% and reduce absolute bias by at least 28% compared to the raw deterministic forecast from a physical model, the Variable Infiltration Capacity model. In addition, PIT and AWCI show that the framework can provide sharp and reliable streamflow forecast ensembles for short lead times (1–3‐day lead time) and provide useful skills beyond up to 5‐day lead time. These will be of immense help in emergency and disaster preparedness.

Long-Term Streamflow Prediction Using Hybrid SVR-ANN Based on Bayesian Model Averaging

The Volga River, as the primary supplier of the Caspian Sea, plays a huge role in its ecosystem sustainability. In this study, we analyze its runoff predictability for different monthly forecast horizons. Additionally, the meteorological and hydrological variables affecting runoff in each month are identified. A wide range of the potential variables was first collected and the Boruta variable preprocessing method was employed to select the important ones. Then the hybrid models were created by combining the selected variables and the data-driven models [i.e., support vector regression (SVR), artificial neural network (ANN), and multiple linear regression (MLR)]. To postprocess the predicted data, the Bayesian model averaging (BMA) method was employed using the combination of the Boruta–artificial neural network (B-ANN) and the Boruta–support vector regression (B-SVR) models. Finally, the Kling-Gupta efficiency (KGE) and the continuous ranked probability skill score (CRPSS) probabilistic evaluation criteria were applied to evaluate the hybrid models. The results showed that the streamflow of the previous steps is the most crucial variable in predicting the streamflow of all next horizons, while its significance decreases as the forecast time horizon increases. Moreover, the temperature variables have the unlike effect on the streamflow prediction and the minimum temperature for winter and spring and the maximum and the average temperatures for summer and autumn are the most effective ones. Correspondingly, BMA-B-ANN-SVR presents the best performance among the hybrid models (for example, the median of 0.8 and 0.91 for CRPSS and KGE in the first horizon, respectively); the reliability of its predicted runoff for different forecast horizons is much better than other hybrid models (B-ANN and B-SVR).

How Skillful Are the European Subseasonal Predictions of Wind Speed and Surface Temperature?

Abstract Subseasonal forecasts of 100-m wind speed and surface temperature, if skillful, can be beneficial to the energy sector as they can be used to plan asset availability and maintenance, assess risks of extreme events, and optimally trade power on the markets. In this study, we evaluate the skill of the European Centre for Medium-Range Weather Forecasts’ subseasonal predictions of 100-m wind speed and 2-m temperature. To the authors’ knowledge, this assessment is the first for the 100-m wind speed, which is an essential variable of practical importance to the energy sector. The assessment is carried out on both forecasts and reforecasts over European domain gridpoint wise and also by considering several spatially averaged domains, using several metrics to assess different attributes of forecast quality. We propose a novel way of synthesizing the continuous ranked probability skill score. The results show that the skill of the forecasts and reforecasts depends on the choice of the climate variable, the period of the year, and the geographical domain. Indeed, the predictions of temperature are better than those of wind speed, with enhanced skill found for both variables in the winter relative to other seasons. The results also indicate significant differences between the skill of forecasts and reforecasts, arising mainly due to the differing ensemble sizes. Overall, depending on the choice of the geographical domain and the forecast attribute, the results show skillful predictions beyond 2 weeks, and in certain cases, up to 6 weeks for both variables, thereby encouraging their implementation in operational decision-making.

Assessment of stochastic weather forecast of precipitation near European cities, based on analogs of circulation

Abstract. In this study, we assess the skill of a stochastic weather generator (SWG) to forecast precipitation in several cities in western Europe. The SWG is based on a random sampling of analogs of the geopotential height at 500 hPa (Z500). The SWG is evaluated for two reanalyses (NCEP and ERA5). We simulate 100-member ensemble forecasts on a daily time increment. We evaluate the performance of SWG with forecast skill scores and we compare it to ECMWF forecasts. Results show significant positive skill score (continuous rank probability skill score and correlation) compared with persistence and climatology forecasts for lead times of 5 and 10 d for different areas in Europe. We find that the low predictability episodes of our model are related to specific weather regimes, depending on the European region. Comparing the SWG forecasts to ECMWF forecasts, we find that the SWG shows a good performance for 5 d. This performance varies from one region to another. This paper is a proof of concept for a stochastic regional ensemble precipitation forecast. Its parameters (e.g., region for analogs) must be tuned for each region in order to optimize its performance.

A hybrid analog-ensemble, convolutional-neural-network method for post-processing precipitation forecasts

Abstract An ensemble precipitation forecast post-processing method is proposed by hybridizing the Analog Ensemble (AnEn), Minimum Divergence Schaake Shuffle (MDSS), and Convolutional Neural Network (CNN) methods. This AnEn-CNN hybrid takes the ensemble mean of Global Ensemble Forecast System (GEFS) 3-hourly precipitation forecasts as input and produces bias-corrected, probabilistically calibrated, and physically realistic gridded precipitation forecast sequences out to 7-days. The AnEn-CNN hybrid post-processing is trained on the European Centre for Medium-Range Weather Forecasts Reanalysis version 5 (ERA5), and verified against station observations across British Columbia (BC), Canada, from 2017 to 2019. The AnEn-CNN hybrid produces more skillful forecasts than a quantile-mapped GEFS baseline and other conventional AnEn methods, with a roughly 10% increase in Continuous Ranked Probability Skill Score. Further, it outperforms other AnEn methods by 0-60% in terms of Brier Skill Score (BSS) for heavy precipitation periods across disparate hydrological regions. Longer forecast lead times exhibit larger performance gains. Verification against 7-day accumulated precipitation totals for heavy precipitation periods also demonstrates that precipitation sequences are realistically reconstructed. Case studies further show that the AnEn-CNN hybrid scheme produces more realistic spatial precipitation patterns and precipitation intensity spectra. This work pioneers the combination of conventional statistical post-processing and neural networks, and is one of only a few studies pertaining to precipitation ensemble post-processing in BC.

Probabilistic Predictions from Deterministic Atmospheric River Forecasts with Deep Learning

Abstract Deep-learning (DL) postprocessing methods are examined to obtain reliable and accurate probabilistic forecasts from single-member numerical weather predictions of integrated vapor transport (IVT). Using a 34-yr reforecast, based on the Center for Western Weather and Water Extremes West-WRF mesoscale model of North American West Coast IVT, the dynamically/statistically derived 0–120-h probabilistic forecasts for IVT under atmospheric river (AR) conditions are tested. These predictions are compared with the Global Ensemble Forecast System (GEFS) dynamic model and the GEFS calibrated with a neural network. In addition, the DL methods are tested against an established, but more rigid, statistical–dynamical ensemble method (the analog ensemble). The findings show, using continuous ranked probability skill score and Brier skill score as verification metrics, that the DL methods compete with or outperform the calibrated GEFS system at lead times from 0 to 48 h and again from 72 to 120 h for AR vapor transport events. In addition, the DL methods generate reliable and skillful probabilistic forecasts. The implications of varying the length of the training dataset are examined, and the results show that the DL methods learn relatively quickly and ∼10 years of hindcast data are required to compete with the GEFS ensemble.

Assimilation of Ground and Satellite Snow Observations in a Distributed Hydrologic Model for Water Supply Forecasting

AbstractRunoff from the Upper Colorado River Basin (UCRB) is an important water resource in the western United States. The majority of annual runoff is derived from spring snowmelt, and therefore April–July water supply volume (WSV) forecasts produced by the Colorado Basin River Forecast Center are critical to basin water management. The primary objective of this study was to evaluate the impact of snow data assimilation (DA) in a distributed hydrologic model on WSV forecasting error and skill in headwater catchments of the UCRB. To do this, a framework was built to use the Localized Ensemble transform Kalman filter to update modeled snow water equivalent (SWE) states in the Hydrology Laboratory‐Research Distributed Hydrologic Model with SNOTEL SWE observations and spatially and temporally complete MODIS Snow Covered‐Area and Grain size data. The DA approach was assessed by evaluating ensemble streamflow prediction forecasts over a 20‐year period for 23 catchments in the UCRB. Overall, the DA approach improved water supply forecast skill in 80% of pilot basins with an average improvement in the median continuous rank probability skill score of 4%. A research‐to‐operations transition was facilitated by automating the implementation of the DA approach. This work demonstrates the capacity for gridded and point snow products to be used to objectively update model states in an operational forecasting setting.

Evaluation of subseasonal to seasonal forecasts over India for renewable energy applications

Abstract. This study evaluates subseasonal to seasonal scale (S2S) forecasts of meteorological variables relevant for the renewable energy (RE) sector of India from six ocean-atmosphere coupled models: ECMWF SEAS5, DWD GCFS 2.0, Météo-France's System 6, NCEP CFSv2, UKMO GloSea5 GC2-LI, and CMCC SPS3. The variables include 10 m wind speed, incoming solar radiation, 2 m temperature, and 2 m relative humidity because they are critical for estimating the supply and demand of renewable energy. The study is conducted over seven homogenous regions of India for 1994–2016. The target months are April and May when the electricity demand is the highest and June–September when the renewable resources outstrip the demand. The evaluation is done by comparing the forecasts at 1, 2, 3, 4, and 5-months lead-times with the ERA5 reanalysis spatially averaged over each region. The fair continuous ranked probability skill score (FCRPSS) is used to quantitatively assess the forecast skill. Results show that incoming surface solar radiation predictions are the best, while 2 m relative humidity is the worst. Overall SEAS5 is the best performing model for all variables, for all target months in all regions at all lead times while GCFS 2.0 performs the worst. Predictability is higher over the southern regions of the country compared to the north and north-eastern parts. Overall, the quality of the raw S2S forecasts from numerical models over India are not good. These forecasts require calibration for further skill improvement before being deployed for applications in the RE sector.

Calibration of ECMWF Seasonal Ensemble Precipitation Reforecasts in Java (Indonesia) Using Bias-Corrected Precipitation and Climate Indices

AbstractThe seasonal precipitation forecast is one of the essential inputs for economic and agricultural activities and has significant impact on decision-making. Large-scale modes of climate variability have strong relationships with seasonal rainfall in Java and are natural candidates for use as potential predictors in a statistical postprocessing application. We explore whether using climate indices as additional predictors in the statistical postprocessing of ECMWF Seasonal Forecast System 5 (SEAS5) precipitation can improve skill. We use parametric statistical postprocessing by applying a logistic distribution-based ensemble model output statistics (EMOS) technique. We add a variety of potential predictors in the analysis, namely SEAS5 raw and empirical quantile mapping (EQM) bias-corrected precipitation, Niño-3.4 index, dipole mode index (DMI), Madden–Julian oscillation (MJO) indices, sea surface temperature (SST) around Java, and several other predictors. We analyze the period of 1981–2010, focusing on July, August, September, and October. We use the continuous ranked probability skill score (CRPSS) and Brier skill score (BSS) in a comparative verification of raw, EQM, and EMOS seasonal precipitation forecasts. We have found that it is essential to use EQM-corrected precipitation as a predictor instead of raw precipitation in the latter. Besides, Niño-3.4 and DMI forecasts are not needed as extra predictors to improve monthly precipitation forecasts for the first lead month, except for September. However, for somewhat longer lead months, in September and October when there is more skill than climatology, the model that includes only Niño-3.4 and DMI forecasts as potential predictors performs about the same compared to the model that uses only EQM-corrected precipitation as a predictor.

Conditioning ensemble streamflow prediction with the North Atlantic Oscillation improves skill at longer lead times

Abstract. Skilful hydrological forecasts can benefit decision-making in water resources management and other water-related sectors that require long-term planning. In Ireland, no such service exists to deliver forecasts at the catchment scale. In order to understand the potential for hydrological forecasting in Ireland, we benchmark the skill of ensemble streamflow prediction (ESP) for a diverse sample of 46 catchments using the GR4J (Génie Rural à 4 paramètres Journalier) hydrological model. Skill is evaluated within a 52-year hindcast study design over lead times of 1 d to 12 months for each of the 12 initialisation months, January to December. Our results show that ESP is skilful against a probabilistic climatology benchmark in the majority of catchments up to several months ahead. However, the level of skill was strongly dependent on lead time, initialisation month, and individual catchment location and storage properties. Mean ESP skill was found to decay rapidly as a function of lead time, with a continuous ranked probability skill score (CRPSS) of 0.8 (1 d), 0.32 (2-week), 0.18 (1-month), 0.05 (3-month), and 0.01 (12-month). Forecasts were generally more skilful when initialised in summer than other seasons. A strong correlation (ρ=0.94) was observed between forecast skill and catchment storage capacity (baseflow index), with the most skilful regions, the Midlands and the East, being those where slowly responding, high-storage catchments are located. Forecast reliability and discrimination were also assessed with respect to low- and high-flow events. In addition to our benchmarking experiment, we conditioned ESP with the winter North Atlantic Oscillation (NAO) using adjusted hindcasts from the Met Office's Global Seasonal Forecasting System version 5. We found gains in winter forecast skill (CRPSS) of 7 %–18 % were possible over lead times of 1 to 3 months and that improved reliability and discrimination make NAO-conditioned ESP particularly effective at forecasting dry winters, a critical season for water resources management. We conclude that ESP is skilful in a number of different contexts and thus should be operationalised in Ireland given its potential benefits for water managers and other stakeholders.

Hydrological ensemble forecasting using a multi-model framework

Ensemble streamflow predictions (ESP) from a single hydrological model tend to under-sample the variability needed to provide a good representation of streamflow observations. ESPs typically end up being biased and under-dispersed, making them imprecise and less valuable in an operational context. In this study, a proposed methodology based on a multi-hydrological model approach is applied to address these issues. The forecast methodology includes eight lumped hydrological models whose states were updated through data assimilation using an Ensemble Kalman Filter (EnKF) to provide initial states to the models, allowing them to better sample uncertainty in the hydrometeorological observations and reduce initial condition errors. ESP forecasts are then driven by numerical weather prediction ensemble forecasts of the European Centre for Medium-Range Weather Forecasts (ECMWF) for a nine-day lead-time on five snowmelt-dominated catchments in Quebec, Canada. These eight assimilated individual ESPs are then post-processed using a quantile mapping bias correction method, after which they are combined in a larger ensemble (known as a Grand Ensemble Streamflow Prediction (GESP)) using a multi-model approach. Results indicate that the proposed methodology allowed reducing both forecast biases and under-dispersion problems for all catchments and four periods corresponding to the seasonality of the streamflow. On their own, the data assimilation (DA) and post-processing manipulations allowed shifting the hydrological ensembles to better encompass the observed streamflow. The multi-model approach further improved the forecasts by increasing their reliability. Talagrand diagrams, supported by the Kolmogorov-Smirnov test for uniformity, as well as the Average Bin Distance to Uniformity (ABDU) and the Continuous Ranked Probability Skill Score (CRPSS), were chosen as performance metrics to evaluate the robustness of the methodology. These findings suggest that it is possible to improve hydrological short-term forecast reliability through a multi-model approach in an operational forecasting context.