Probabilistic Precipitation Forecasts Research Articles

Evaluation of Subseasonal-to-Seasonal (S2S) precipitation forecast performance from NMME-2 over Indonesia

Abstract Global Subseasonal-to-Seasonal (S2S) precipitation forecast was known to bridge the gap between weather and seasonal forecast, can be utilized as supporting information for decision-making related to hydrometeorological disaster mitigation activities. However, uncertainty in global models caused forecast performance can be different across regions and time periods. Therefore, it is important to evaluate forecast performance before utilizing the prediction result in any region. In this study, performance of probabilistic precipitation forecast from Multi-model Ensemble (MME) of three models in The North American Multi-Model Ensemble phase 2 (NMME-2) project was evaluated in Indonesia region based on two evaluation metrics: continuous ranked probability score (CRPS) and reliability diagram. The evaluation was conducted during the boreal summer (May–October) and boreal winter (November–April) periods, as well as during the active period of subseasonal climate variability phenomenon Madden-Julian Oscillation (MJO). Our result shows that S2S precipitation forecast from MME of CFSv2, CanCM4 and GEOS5 models are sufficiently accurate and reliable during the boreal summer period for Central Sumatra, Southern Sumatra, Southern Kalimantan, Java, Southern Sulawesi, and Southern Papua regions, with a range of CRPS values between 4-16 mm/7 days and a ‘perfect’ reliability category. There is no notable distinction in the performance of S2S precipitation forecast between the active and inactive events of the Madden-Julian Oscillation (MJO). The difference in CRPS values between these two periods is only around 0.8-1.2 mm/7 days, and there is no difference in reliability categories across Indonesia as a whole, nor significant spatial pattern differences.

Improvements in the spread–skill relationship of precipitation in a convective‐scale ensemble through blending

AbstractConvective‐scale ensembles are used routinely in operational centres around the world to produce probabilistic precipitation forecasts, but a lack of spread between members is providing forecasts that are frequently overconfident. This deficiency can be corrected by increasing spread, increasing forecast accuracy, or both. A recent development in the Met Office forecasting system is the inclusion of large‐scale blending (LSB) in the convective‐scale data assimilation scheme. This method aims to reduce the synoptic‐scale forecast error in the analysis by reducing the influence of the convective‐scale data assimilation at scales that are too large to be constrained by the limited domain. These scales are instead initialised using output from the global data assimilation scheme, which we expect to reduce the forecast error and thus improve the spread–skill relationship. In this study, we quantify the impact of LSB on the spread–skill relationship of hourly precipitation accumulations by comparing forecast ensembles with and without LSB over a 17‐day summer trial period. This trial found modest but significant improvements to the spread–skill relationship as calculated using metrics based on the Fractions Skill Score. Skill is improved for a lower precipitation centile by an average of 0.56% at the largest scales, but a corresponding degradation of spread limits the overall correction. The spread–skill disparity is reduced the most in the higher centiles due to a more muted spread response, with significant reductions of up to 0.40% obtained at larger scales. Case‐study analysis using a novel extension of the Localised Fractions Skill Score demonstrates how spread–skill improvements transfer to smaller‐scale features, not just the scales that have been blended. There are promising signs that further spread–skill improvements can be made by implementing LSB more fully within the ensemble, and we encourage the Met Office to continue developing this technique.

Probabilistic Forecast of Ecological Drought in Rivers Based on Numerical Weather Forecast from S2S Dataset

Ecological droughts in rivers, as a new type of drought, have been greatly discussed in the past decade. Although various studies have been conducted to identify and evaluate ecological droughts in rivers from different indices, a forecast model for this type of drought is still lacking. In this paper, a numerical weather forecast, a hydrological model, and a generalized Bayesian model are employed to establish a new general framework for the probabilistic forecasting of ecological droughts in rivers, and the Daitou section in China is selected as the study area to examine the performance of the new framework. The results show that the hydrological model can accurately simulate the monthly streamflow with a Nash–Sutcliffe efficiency of 0.91 in the validation period, which means that the model can be used to reconstruct the natural streamflow from the impact of an upstream reservoir. Based on a comparison of ecological drought events from the observed and model-simulated streamflow series, the events from the observed series have a larger deficit volume and a longer duration of ecological droughts after 2014, indicating that human activities may lead to a more severe situation of ecological droughts. Furthermore, due to the uncertainty of precipitation forecasts, a probabilistic precipitation forecast is employed for probabilistic ecological drought forecasting. Compared to the deterministic forecast, the probabilistic ecological drought forecast has better performance, with a Brier score decrease of 0.35 to 0.18 and can provide more information about the risk of ecological droughts. In general, the new probabilistic framework developed in this study can serve as a basis for the development of early-warning systems and countermeasures for ecological droughts.

Comparison of probabilistic forecasts of extreme precipitation for a global and convection‐permitting ensemble and hybrid statistical–dynamical method based on equatorial wave information

AbstractRecent work has demonstrated that skilful hybrid statistical–dynamical forecasts of heavy rainfall events in Southeast Asia can be made by combining model forecasts of the phases and amplitudes of Kelvin, Rossby, and westward‐moving Rossby gravity waves with climatological rainfall statistics conditioned on these waves. This study explores the sensitivity of this hybrid forecast to its parameter choices and compares its skill in forecasting extreme rainfall events in the Philippines, Malaysia, Indonesia, and Vietnam to that of the Met Office Global and Regional Ensemble Prediction System (MOGREPS). The hybrid forecast is found to outperform both the global and convection‐permitting ensemble in some regions when forecasting the most extreme events; however, for less extreme events, the ensemble is found more skilful. A weighted blend of the MOGREPS forecasts and the hybrid forecast was found to have the highest skill of all for almost all definitions of extreme event and in most regions. To quantify the influence of errors in the predicted wave state on the skill of the hybrid forecast, the skill of a hypothetical best‐case forecast was also calculated using reanalysis data to specify the wave amplitudes and phases. This best‐case forecast indicates that errors in the forecasts of all wave types reduce the skill of hybrid forecast; however, the reduction in skill is largest for Kelvin waves. The skill in convection‐permitting models is greater than for global models in the regions where Kelvin waves dominate, but the added value of limited‐area high‐resolution forecasts is hampered by the poor representation of Kelvin waves in the parent global model.

The Roles of Chaos Seeding and Multiple Perturbations in Convection-Permitting Ensemble Forecasting over Southern China

Abstract The roles of chaos seeding and multiple perturbations, including model perturbations and topographic perturbations, in convection-permitting ensemble forecasting, are assessed. Six comparison experiments were conducted for 14 heavy rainfall events over southern China. Chaos seeding was run as a benchmark experiment to compare their effects to the intended perturbations. The results first reveal the chaos seeding phenomenon. That is, the tiny and local perturbations of the skin soil moisture propagate into the whole analysis domain within an hour and expand to every prognostic variable, and the perturbations derived from chaos seeding develop when moist convection is active. Second, the chaos seeding has a statistically significant difference from our intended perturbations for the ensemble spread magnitudes of precipitation and the spread–skill relationships and probabilistic forecast skills of dynamical variables. Additionally, for the probabilistic forecasts of precipitation, initial and lateral boundary perturbations and model perturbations can yield statistically larger FSS and AROC scores than chaos seeding; topographic perturbations can only improve FSS and AROC scores a little. The different performances may be related to the different degrees of the real dynamical influence of our intended perturbations. Finally, model perturbations can increase the ensemble spreads of precipitation, and improve FSS and AROC scores of precipitation and the consistency of mid- and low-level dynamical variables. However, the effects of topographic perturbations are small.

A Bayesian spatio‐temporal model for short‐term forecasting of precipitation fields

AbstractWith extreme weather events becoming more common, the risk posed by surface water flooding is ever increasing. In this work we propose a model, and associated Bayesian inference scheme, for generating short‐term, probabilistic forecasts of localised precipitation on a spatial grid. Our generative hierarchical dynamic model is formulated in discrete space and time with a lattice‐Markov spatio‐temporal auto‐regressive structure, inspired by continuous models of advection and diffusion. Observations from both weather radar and ground based rain gauges provide information from which we can learn the precipitation field through a latent process in addition to unknown model parameters. Working in the Bayesian paradigm provides a coherent framework for capturing uncertainty, both in the underlying model parameters and in our forecasts. Further, appealing to simulation based sampling using MCMC yields a straightforward solution to handling zeros, treated as censored observations, via data augmentation. Both the underlying state and the observations are of moderately large dimension ( and respectively) and this renders standard inference approaches computationally infeasible. Our solution is to embed the ensemble Kalman smoother within a Gibbs sampling scheme to facilitate approximate Bayesian inference in reasonable time. Both the methodology and the effectiveness of our posterior sampling scheme are demonstrated via simulation studies and by a case study of real data from the Urban Observatory project based in Newcastle upon Tyne, UK.

Improving National Blend of Models Probabilistic Precipitation Forecasts Using Long Time Series of Reforecasts and Precipitation Reanalyses. Part I: Methods

Abstract This article describes proposed revised methods for the statistical postprocessing of precipitation amount intended for the NOAA’s National Blend of Models using the Global Ensemble Forecast System version 12 data (GEFSv12). The procedure updates the previously established procedure of quantile mapping, weighting of sorted members, and dressing of the ensemble. The revised method leverages the long reforecast training dataset that has become available to improve quantile mapping of GEFSv12 data by eliminating the use of supplemental locations, that is, training data from other grid points. It establishes improved definitions of cumulative distributions through a spline-fitting approach. It provides updated algorithms for the weighting of sorted members based on closest-member histogram statistics, and it establishes an objective method for the dressing of the quantile-mapped, weighted ensemble. Verification statistics and case studies are provided in the accompanying article (Part II).

Improving National Blend of Models Probabilistic Precipitation Forecasts Using Long Time Series of Reforecasts and Precipitation Reanalyses. Part II: Results

Abstract This second part of the series presents results from verifying a precipitation forecast calibration method discussed in the first part, based on quantile mapping (QM), weighting of sorted members, and dressing of the ensemble. NOAA’s Global Ensemble Forecast System, version 12 (GEFSv12), reforecasts were used in this study. The method was validated with preoperational GEFSv12 forecasts from December 2017 to November 2019. The method is proposed as an enhancement for GEFSv12 precipitation postprocessing in NOAA’s National Blend of Models. The first part described adaptations to the methodology to leverage the ∼20-yr GEFSv12 reforecast data. As shown here in this part, when compared with probabilistic quantitative precipitation forecasts from the raw ensemble, the adapted method produced downscaled, high-resolution forecasts that were significantly more reliable and skillful than raw ensemble-derived probabilities, especially at shorter lead times (i.e., <5 days) and for forecasts of events from light precipitation to >10 mm (6 h)−1. Cool-season events in the western United States were especially improved when the QM algorithm was applied, providing a statistical downscaling with realistic smaller-scale detail related to terrain features. The method provided less value added for forecasts of longer lead times and for the heaviest precipitation.

Performance of a Convective-Scale Ensemble Prediction System on 2017 Warm-Season Afternoon Thunderstorms over Taiwan

Abstract A 16-member convective-scale ensemble prediction system (CEPS) developed at the Central Weather Bureau (CWB) of Taiwan is evaluated for probability forecasts of convective precipitation. To address the issues of limited predictability of convective systems, the CEPS provides short-range forecasts using initial conditions from a rapid-updated ensemble data assimilation system. This study aims to identify the behavior of the CEPS forecasts, especially the impact of different ensemble configurations and forecast lead times. Warm-season afternoon thunderstorms (ATs) from 30 June to 4 July 2017 are selected. Since ATs usually occur between 1300 and 2000 LST, this study compares deterministic and probabilistic quantitative precipitation forecasts (QPFs) launched at 0500, 0800, and 1100 LST. This study demonstrates that initial and boundary perturbations (IBP) are crucial to ensure good spread–skill consistency over the 18-h forecasts. On top of IBP, additional model perturbations have insignificant impacts on upper-air and precipitation forecasts. The deterministic QPFs launched at 1100 LST outperform those launched at 0500 and 0800 LST, likely because the most-recent data assimilation analyses enhance the practical predictability. However, it cannot improve the probabilistic QPFs launched at 1100 LST due to inadequate ensemble spreads resulting from limited error growth time. This study points out the importance of sufficient initial condition uncertainty on short-range probabilistic forecasts to exploit the benefits of rapid-update data assimilation analyses. Significance Statement This study aims to understand the behavior of convective-scale short-range probabilistic forecasts in Taiwan and the surrounding area. Taiwan is influenced by diverse weather systems, including typhoons, mei-yu fronts, and local thunderstorms. During the past decade, there has been promising improvement in predicting mesoscale weather systems (e.g., typhoons and mei-yu fronts). However, it is still challenging to provide timely and accurate forecasts for rapid-evolving high-impact convection. This study provides a reference for the designation of convective-scale ensemble prediction systems; in particular, those with a goal to provide short-range probabilistic forecasts. While the findings cannot be extrapolated to all ensemble prediction systems, this study demonstrates that initial and boundary perturbations are the most important factors, while the model perturbation has an insignificant effect. This study suggests that in-depth studies are required to improve the convective-scale initial condition accuracy and uncertainty to provide reliable probabilistic forecasts within short lead times.

Adaptive Blending of Probabilistic Precipitation Forecasts with Emphasis on Calibration and Temporal Forecast Consistency

Abstract A wealth of forecasting models is available for operational weather forecasting. Their strengths often depend on the lead time considered, which generates the need for a seamless combination of different forecast methods. The combined and continuous products are made in order to retain or even enhance the forecast quality of the individual forecasts and to extend the lead time to potentially hazardous weather events. In this study, we further improve an artificial neural network–based combination model that was recently proposed in a previous paper. This model combines two initial precipitation ensemble forecasts and produces exceedance probabilities for a set of thresholds for hourly precipitation amounts. Both initial forecasts perform differently well for different lead times, whereas the combined forecast is calibrated and outperforms both initial forecasts with respect to various validation scores and for all considered lead times (from +1 to +6 h). Moreover, the robustness of the combination model is tested by applying it to a new dataset and by evaluating the spatial and temporal consistency of its forecasts. The changes proposed further improve the forecast quality and make it more useful for practical applications. Temporal consistency of the combined product is evaluated using a flip-flop index. It is shown that the combination provides a higher persistence with decreasing lead times compared to both input systems.

A New Hybrid Framework for Error Correction and Uncertainty Analysis of Precipitation Forecasts with Combined Postprocessors

With the rapid development of meteorological models, numerical weather prediction is increasingly used in flood forecasting and reservoir regulation, but its forecasting ability is limited by the large amount of uncertainty from meteorological systems. In this paper, a new, hybrid framework is developed to improve numerical precipitation forecasting by combining the multimodel ensemble and probabilistic postprocessing methods. The results show that the multimodel ensemble method used in this paper is an efficient way to reduce prediction errors, especially missing alarm errors. In a comparison of the probabilistic postprocessors based the generalized Bayesian model (GBM) and bivariate probabilistic model (BPM), the GBM shows better performance from the aspects of indicators and is more suitable for real-time applications. Meanwhile, the assessment of probabilistic results shows that the skill of probabilistic precipitation forecasts is related to the quality of their inputs. According to these results, a new hybrid framework is proposed by taking the results from multimodel ensemble as the input of probabilistic postprocessor. Compared to using the raw numerical in GBM, the hybrid framework improves the accuracy, sharpness, reliability, and resolution ability from different lead times by 2–13%, 1–22%, and 0–12% respectively, especially when the lead time is less than 4 d, the improvement can reach 9–13%, 10–22%, and 5–12% respectively. In conclusion, the hybrid two-step framework can provide a more skillful precipitation forecast, which can be useful for flood forecasting and reservoir regulation.

Deep-learning-based post-processing for probabilistic precipitation forecasting

Ensemble prediction systems (EPSs) serve as a popular technique to provide probabilistic precipitation prediction in short- and medium-range forecasting. However, numerical models still suffer from imperfect configurations associated with data assimilation and physical parameterization, which can lead to systemic bias. Even state-of-the-art models often fail to provide high-quality precipitation forecasting, especially for extreme events. In this study, two deep-learning-based models—a shallow neural network (NN) and a deep NN with convolutional layers (CNN)—were used as alternative post-processing approaches to further improve the probabilistic forecasting of precipitation over China with 1–7 lead days. A popular conventional method—the censored and shifted gamma distribution-based ensemble model output statistics (CSG EMOS)—was used as the baseline. Re-forecasts run using a frozen EPS—Global Ensemble Forecast System version 12—were collected as the raw ensembles spanning from 2000 to 2019. The re-forecast data were generated once per day and consisted of one control run and four perturbed members. We used the calendar year 2018 as the validation period and 2019 as the testing period, and the remaining 18 years of data were used for training. According to the results, in terms of the continuous ranked probability score (CRPS) and the Brier score, the CNN model significantly outperforms the shallow NN model, as well as the CSG EMOS approach and the raw ensemble, especially for heavy or extreme precipitation events (those exceeding 50 mm/day). A remarkable degradation was seen when reducing the size of training samples from 18 years of data to two years. The spatial distribution of the CRPS shows that the stations in central China were better calibrated than those in other regions. With a lead time of 1 day, the CNN model was found to be superior to the other models (in terms of the CRPS) at 74.5% of the study stations. These results indicate that deep NNs can serve as a promising approach to the statistical post-processing of probabilistic precipitation forecasting.

The Development of the NCEP Global Ensemble Forecast System Version 12

Abstract The Global Ensemble Forecast System (GEFS) is upgraded to version 12, in which the legacy Global Spectral Model (GSM) is replaced by a model with a new dynamical core—the Finite Volume Cubed-Sphere Dynamical Core (FV3). Extensive tests were performed to determine the optimal model and ensemble configuration. The new GEFS has cubed-sphere grids with a horizontal resolution of about 25 km and an increased ensemble size from 20 to 30. It extends the forecast length from 16 to 35 days to support subseasonal forecasts. The stochastic total tendency perturbation (STTP) scheme is replaced by two model uncertainty schemes: the stochastically perturbed physics tendencies (SPPT) scheme and stochastic kinetic energy backscatter (SKEB) scheme. Forecast verification is performed on a period of more than two years of retrospective runs. The results show that the upgraded GEFS outperforms the operational-at-the-time version by all measures included in the GEFS verification package. The new system has a better ensemble error–spread relationship, significantly improved skills in large-scale environment forecasts, precipitation probability forecasts over CONUS, tropical cyclone track and intensity forecasts, and significantly reduced 2-m temperature biases over North America. GEFSv12 was implemented on 23 September 2020.

Validation of HWRF-based Probabilistic TC Wind and Precipitation forecasts

AbstractTropical cyclones are associated with a variety of significant social hazards, including wind, rain, and storm surge. Despite this, most of the model validation effort has been directed toward track and intensity forecasts. In contrast, few studies have investigated the skill of state-of-the-art, high-resolution ensemble prediction systems in predicting associated TC hazards, which is crucial since TC position and intensity do not always correlate with the TC-related hazards, and can result in impacts far from the actual TC center. Furthermore, dynamic models can provide flow-dependent uncertainty estimates, which in turn can provide more specific guidance to forecasters than statistical uncertainty estimates based on past errors. This study validates probabilistic forecasts of wind speed and precipitation hazards derived from the HWRF ensemble prediction system and compares its skill to forecasts by the stochastically-based operational Monte Carlo Model (NHC), the IFS (ECMWF), and the GEFS (NOAA) in use 2017-2019. Wind and Precipitation forecasts are validated against NHC best track wind radii information and the National Stage IV QPE Product. The HWRF 34 kn wind forecasts have comparable skill to the global models up to 60 h lead time before HWRF skill decreases, possibly due to detrimental impacts of large track errors. In contrast, HWRF has comparable quality to its competitors for higher thresholds of 50 kn and 64 kn throughout 120 h lead time. In terms of precipitation hazards, HWRF performs similar or better than global models, but depicts higher, although not perfect, reliability, especially for events over 5 in120h−1. Post-processing, like Quantile Mapping, improves forecast skill for all models significantly and can alleviate reliability issues of the global models.

Forecasting excessive rainfall with random forests and a deterministic convection-allowing model

AbstractApproximately seven years of daily initializations from the convection-allowing National Severe Storms Laboratory Weather Research and Forecasting model are used as inputs to train random forest (RF) machine learning models to probabilistically predict instances of excessive rainfall. Unlike other hazards, excessive rainfall does not have an accepted definition, so multiple definitions of excessive rainfall and flash flooding – including flash flood reports and 24-hr average recurrence intervals (ARIs) – are used to explore RF configuration forecast sensitivities. RF forecasts are analogous to operational Weather Prediction Center (WPC) day-1 Excessive Rainfall Outlooks (EROs) and their resolution, reliability, and skill are strongly influenced by rainfall definitions and how inputs are assembled for training. Models trained with 1-y ARI exceedances defined by the Stage-IV (ST4) precipitation analysis perform poorly in the northern Great Plains and southwest U.S., in part due to a high bias in the number of training events in these regions. Increasing the ARI threshold to 2 years or removing ST4 data from training, optimizing forecast skill geographically, and spatially averaging meteorological inputs for training generally results in improved CONUS-wide RF forecast skill. Both EROs and RF forecasts have seasonal skill – poor forecasts in the late fall and winter and skillful forecasts in the summer and early fall. However, the EROs are consistently and significantly better than their RF counterparts, regardless of RF configuration, particularly in the summer months. The results suggest careful consideration should be made when developing ML-based probabilistic precipitation forecasts with convection-allowing model inputs, and further development is necessary to consider these forecast products for operational implementation.

Using stratified Bayesian model averaging in probabilistic forecasts of precipitation over the middle and lower Yangtze River region

Using stratified Bayesian model averaging in probabilistic forecasts of precipitation over the middle and lower Yangtze River region

Improved multi‐model ensemble forecasts of Iran's precipitation and temperature using a hybrid dynamical‐statistical approach during fall and winter seasons

AbstractSkillful seasonal climate forecasts can support decision making in water resources management and agricultural planning. In arid and semi‐arid regions, tailoring reliable forecasts has the potential to improve water management by using key hydroclimate variables months in advance. This article analyses and compares the performance of two common approaches (empirical and hybrid dynamical‐statistical) in seasonal climate forecasting over a drought‐prone area located in Southwest Asia including Iran. Empirical models are framed as a baseline skill that hybrid models need to outperform. Both approaches provide probabilistic forecasts of precipitation and temperature using canonical correlation analysis to provide forecasts at 0.25° resolution. Empirical models are developed based on the large‐scale observed atmosphere–ocean patterns for forecasting using antecedent climate anomalies as predictors, while the hybrid approach makes use of model output statistics to correct systematic errors in dynamical climate model forecast outputs. Eight state‐of‐the‐art dynamical models from the North American Multi‐Model Ensemble project are analysed. Individual models with the highest goodness index are weighted to develop seven different hybrid dynamical‐statistical Multi‐model Ensembles. In this study, (October–December) and (January–February) are considered as target seasons which are the most important periods within the water year for water resource allocation to the agriculture sector. The results show that the hybrid approach has improved performance compared to the raw general circulation models and purely empirical models, and that the performance of the hybrid models is season‐dependent. Seasonal forecasts of precipitation (temperature) have a higher skill in OND (JFM). In addition, in most cases, Multi‐model Ensemble (MME) is more skillful than the empirical models and outperforms individual dynamical models. However, the best individual model might be as skillful as the MME given the target season and region of interest.

Case Dependence of Multiscale Interactions between Multisource Perturbations for Convection-Permitting Ensemble Forecasting during SCMREX

AbstractThis study examines the case dependence of the multiscale characteristics of initial condition (IC) and model physics (MO) perturbations and their interactions in a convection-permitting ensemble prediction system (CPEPS), focusing on the 12-h forecasts of precipitation perturbation energy. The case dependence of forecast performances of various ensemble configurations is also examined to gain guidance for CPEPS design. Heavy-rainfall cases over Southern China during the Southern China Monsoon Rainfall Experiment (SCMREX) in May 2014 were discriminated between the strongly and weakly forced events in terms of synoptic-scale forcing, each of which included 10 cases. In the cases with weaker forcing, MO perturbations showed larger influences while the enhancements of convective activities relative to the control member due to IC perturbations were less evident, leading to smaller dispersion reduction due to adding MO perturbations to IC perturbations. Such dispersion reduction was more sensitive to IC and MO perturbation methods in the weakly and strongly forced cases, respectively. The dispersion reduction improved the probabilistic forecasts of precipitation, with more evident improvements in the cases with weaker forcing. To improve the benefits of dispersion reduction in forecasts, it is instructive to elaborately consider the case dependence of dispersion reduction, especially the various sensitivities of dispersion reduction to different-source perturbation methods in various cases, in CPEPS design.

The regional model‐based Mesoscale Ensemble Prediction System, MEPS, at the Japan Meteorological Agency

AbstractThe regional model‐based Mesoscale Ensemble Prediction System (MEPS) has been operational since June 2019 at the Japan Meteorological Agency (JMA). The primary objective of the newly operational MEPS is to provide uncertainty information for JMA's operational regional model, Mesoscale Model (MSM), which provides information to support disaster prevention and aviation safety. This article describes MEPS in detail and discusses issues to be addressed in the future. For effective evaluation of uncertainties in MSM, the forecast model in MEPS is configured in the same way as that in MSM, except for the initial and lateral boundary conditions. Initial perturbations for all 20 ensemble runs are generated by a linear combination of singular vectors (SVs) with three different spatial and temporal resolutions, with the aim of capturing multi‐scale uncertainties in the initial conditions simultaneously. The SVs from a global model are also used as lateral boundary perturbations to ensure consistency between the initial and boundary conditions of each ensemble member. The verification results showed that MEPS achieved the expected performance of an ensemble prediction system: the ensemble mean outperformed the control forecast with a good spread–skill relationship; moreover, the skill scores of probabilistic precipitation forecasts were evaluated as valid for rainfall of up to 30 mm·(3 hr)−1. In an additional experiment conducted without using the two smaller‐scale initial perturbations, the skill was substantially reduced compared with that of the original MEPS, especially for larger precipitation thresholds. Therefore, the smaller‐scale perturbations were essential to capture uncertainties associated with local heavy rainfall events.

Assessing the skill of NCMRWF global ensemble prediction system in predicting Indian summer monsoon during 2018

The quality of probabilistic precipitation and zonal wind forecasts from National Centre for Medium Range Weather Forecasting (NCMRWF) Global Ensemble Prediction System (NEPS-G) is investigated for Indian summer monsoon period between June–September 2018. Integrated Multi-satellite Retrievals for Global Precipitation Measurement (IMERG GPM) are used for verification of precipitation forecasts. The predictive skill of different categories of rainfall is examined with respect to daily climatology based on Tropical Rainfall Measuring Mission (TRMM) observations and reanalysis data from the Indian Monsoon Data Assimilation and Analysis (IMDAA). ERA Interim and IMDAA reanalysis daily climatologies are used to compute skill for the zonal wind forecasts at 850 hPa (u850). The model has a systematic tendency to over-predict the low level westerlies associated with the monsoon circulation. RMSE over Gangetic plains near Himalayan foothills is more in day-3 as compared to subsequent forecast lead times due to its overestimation of the easterly zonal wind flow. Spread in u850 is comparable to RMSE in day-1 forecast. The ensemble forecasting system is slightly under-dispersive for longer forecast lead times, since the rate of growth of forecast uncertainty is larger than that could be predicted by the ensemble system. Forecasts are sharper for lower thresholds of rainfall and exhibit more reliability and better discrimination of events over shorter lead times. Similar to reliability, the rank distribution depends on forecast lead time as well as ensemble spread. The positive Brier skill score and Continuous Ranked Probability Skill Score values above 0.4 for probabilistic wind as well as precipitation forecasts of light to moderate category, consistently show high predictive skill till day-7, with reference to the long-term climatology. NEPS-G could predict an extreme rainfall event with high probabilities of precipitation exceeding thresholds classified by India Meteorological department, which are in good correspondence with that of rainfall observed by GPM IMERG. A monsoon index based on large-scale features of monsoon circulation could be predicted by the EPS with high probabilistic skill during the peak monsoon.