Fractions Skill Research Articles

A new verification approach for nowcasting based on intensity and spatial-temporal feature correction

Forecast verification is very important in the nowcasting operation and technical development of strong convective weather. The current conventional verification method for nowcasting uses a binary classification event verification method, which exists with double punishment, leading to low scoring issues. In order to make up for the shortcomings of conventional verification methods and explore the potential value of forecasting, based on the characteristics and requirements of strong convective weather nowcasting operations, this paper proposes a neighborhood verification method that considers spatial scale, time scale, and intensity error information simultaneously, based on the spatial neighborhood fraction skill score (FSS) verification method. The paper designs an intensity neighborhood correction scheme, a time neighborhood correction scheme, and a comprehensive correction scheme that considers both intensity and time neighborhoods. By introducing spatial neighborhood probability, the strict spatial matching requirement is weakened. Based on the Gaussian membership function in fuzzy logic, the value of forecasting grid below the threshold is explored. Time neighborhood adaptive weighting strategy is adopted to solve the problem of early and delayed forecasts. The evaluation and verification of the optical flow radar extrapolation nowcasting results show that compared to the traditional “point-to-point” verification indicators, the proposed indicators can increase the tolerance of verification, provide more useful information, and are more in line with the needs of practical application.

Ensemble versus deterministic lightning forecast performance at a convective scale over Indian region

The present study quantifies the improvement achieved in lightning forecast skill of the NCMRWF regional ensemble prediction system (NEPS-R) compared to its deterministic counterpart (CNTL). The lightning forecasts over study regions of East and Northeast India (ENEI) and Peninsular India (PI) during the pre-monsoon season and Central-East and Northeast India (CENEI) during the monsoon season have been verified using lightning observations from the Indian Institute of Tropical Meteorology (IITM) Lightning Detection Network (LDN). The persisting systematic negative bias in deterministic and EPS-based forecasts of the ensemble mean (EnsMean) and ensemble maximum (EnsMax) indicate the lack of spread among the members, supported by the low values of ensemble spread over all the study regions. EnsMean has the lowest RMSE, with a decrease in error ranging from 0.8 % to 2.18 % compared to CNTL. Categorical skill scores indicate that the EPS-based forecasts (EnsMean and EnsMax) are more skillful than the deterministic forecast at all thresholds and lead times. Further, Fractions Skill Score (FSS) establishes the superiority of the ensemble forecasts over the deterministic forecasts, where for threshold >1, EnsMean is skillful at comparatively smaller neighborhood sizes (ENEI and PI ∼68 km; CENEI ∼36 km for day-1) than CNTL (ENEI-116 km; PI-196 km; CENEI-68 km). EnsMax at higher thresholds (>5 and >10) is skillful at lesser neighborhood sizes ranging from 116 to 276 km compared to CNTL (>401 km) for day-1. Hence, skillful re-scaled EPS forecasts based on FSS could provide better guidance for the forecasters. The Continuous Ranked Probability Score of EPS forecasts is lower by around 9 % than the Mean Absolute Error of CNTL forecasts, and the ROC of EPS shows better discrimination of events and non-events compared to CNTL. These highlight the merits of using an EPS over a deterministic system for forecasting a field of high spatial variability, like lightning, and thereby, the use of vast computational resources to run a convective scale EPS is justified.

Downscaling, bias correction, and spatial adjustment of extreme tropical cyclone rainfall in ERA5 using deep learning

Hydrological models that are used to analyse flood risk induced by tropical cyclones often input ERA5 reanalysis data. However, ERA5 precipitation has large systematic biases, especially over heavy precipitation events like Tropical Cyclones, compromising its usefulness in such scenarios. Few studies to date have performed bias correction of ERA5 precipitation and none of them for extreme rainfall induced by tropical cyclones. Additionally, most existing works on bias adjustment focus on adjusting pixel-wise metrics of bias, such as the Mean Squared Error (MSE). However, it is equally important to ensure that the rainfall peaks are correctly located within the rainfall maps, especially if these maps are then used as input to hydrological models. In this paper, we describe a novel machine learning model that addresses both gaps, RA-Ucmpd, based on the popular U-Net model. The key novelty of RA-Ucmpd is its loss function, the compound loss, which optimizes both a pixel-wise bias metric (the MSE) and a spatial verification metric (a modified version of the Fractions Skill Score). Our results show how RA-Ucmpd improves ERA5 in almost all metrics by 3-28%—more than the other models we used for comparison which actually worsen the total rainfall bias of ERA5—at the cost of a slightly increased (3%) error on the magnitude of the peak. We analyse the behaviour of RA-Ucmpd by visualizing accumulated maps of four particularly wet tropical cyclones and by dividing our data according to the Saffir-Simpson scale and to whether they made landfall, and we perform an error analysis to understand under what conditions our model performs best.

Linguistic precursors of sixth-grade geometric and fraction skills in children with and without Developmental Language Disorder

BackgroundPrevious research has shown that foundational linguistic skills (i.e., phonological awareness and grammatical ability) indirectly (through arithmetic skills) predict growth from fifth-to sixth-grade geometric and fraction skills. AimsOur study aimed to investigate the linguistic precursors of sixth-grade geometric and fraction skills in children with and without DLD, while examining potential (cognitive) strengths within the DLD group that may partly compensate for learning geometry and fractions, at both the group and individual level. SampleParticipants were 46 children with DLD and 122 typically developing peers from 9 to 11 years of age. MethodsClassroom and individual measures were administered in both grade 4 and grade 6. ResultsAt the group level, results showed children with DLD to score below their peers on arithmetic, geometric, and fraction skills. Furthermore, indirect effects of phonological awareness and naming speed, via arithmetic skills, on geometric and fraction skills were found to be equally strong for both groups. In addition, similar strengths for both groups were found for nonverbal intelligence, academic vocabulary, and verbal reasoning in directly predicting the scores in geometric and fraction skills. Finally, at the individual level, a strength in verbal reasoning was found to partly compensate the delays in mathematics in children with DLD. ConclusionsThe educational needs of children with and without DLD in mathematics learning might be more quantitative in nature than that they are qualitative. In addition, identifying individual strengths should be integrated into standardized test batteries and treatment approaches.

A holistic investigation of fraction learning: examining the hierarchy of fraction skills, misconceptions, mathematics anxiety and response confidence

ABSTRACT This study aimed to holistically investigate fraction knowledge by considering pupils’ fraction performance on various tasks (i.e. mapping, equivalence, comparison, and arithmetic) and in relation to whole number arithmetic, fraction misconceptions, mathematics anxiety, and confidence in responses. Northern Irish pupils (N = 123; 77 girls; M age = 11.1 years) demonstrated a strong understanding of fraction magnitudes with few pupils showcasing misconceptions across magnitude tasks. In contrast, fraction arithmetic knowledge, which was uniquely predicted by whole number arithmetic and fraction comparison, was still developing with many pupils incorrectly applying whole number strategies to arithmetic problems. Mathematics anxiety was low, but negatively correlated with performance, and confidence reports were more closely aligned with performance on whole number arithmetic compared to fraction tasks. Overall, researchers and educators need to consider not just fraction performance, but also foundational skills, the presence of misconceptions, and pupils’ feelings and beliefs to better understand fraction learning.

The fractions skill score for ensemble forecast verification

AbstractThe fractions skill score (FSS) is a neighbourhood verification method originally designed to verify deterministic forecasts of binary events. Previous studies employed different approaches for computing an ensemble‐based FSS for probabilistic forecast verification. We show that the formulation of an ensemble‐based FSS substantially affects verification results. Comparing four possible approaches, we determine how different ensemble‐based FSS variants depend on ensemble size, neighbourhood size, and forecast event frequency of occurrence. We demonstrate that only one ensemble‐based FSS, which we call the probabilistic FSS (pFSS), is well behaved and reasonably dependent on ensemble size. Furthermore, we derive a relationship to describe how the pFSS behaves with ensemble size. The proposed relationship is similar to a known result for the Brier skill score. Our study uses high‐resolution 1000‐member ensemble precipitation forecasts from a high‐impact weather period. The large ensemble enables us to study the influence of ensemble and neighbourhood size on forecast skill by deriving probabilistic skilful spatial scales.

A 3D-Var assimilation scheme for vertical velocity with CMA-MESO v5.0

Abstract. Certain vertical motions associated with meso-microscale systems are favorable for convection development and maintenance. Correct initialization of updraft motions is thus significant in convective precipitation forecasts. A three-dimensional variational-based vertical velocity (w) assimilation scheme has been developed within the high-resolution (3 km) CMA-MESO (the Mesoscale Weather Numerical Forecast System of the China Meteorological Administration) model. This scheme utilizes the adiabatic Richardson equation as the observation operator for w, enabling the update of horizontal winds and mass fields of the model's background. The tangent linear and adjoint operators are subsequently developed and undergo an accuracy check. A single-point w observation assimilation experiment reveals that the observational information is effectively spread both horizontally and vertically. Specifically, the assimilation of w contributes to the generation of horizontal wind convergence at lower model levels and divergence at higher model levels, thereby adjusting the locations of convection occurrence. The impact of assimilating w on the forecast is then examined through a series of continuous 10 d runs. Further assimilation of w, in addition to the assimilation of conventional and radial wind data, significantly improves the forecast accuracy of precipitation, resulting in higher FSS (fractions skill score) values and higher ETS (equitable threat score) skills at higher thresholds (5 and 20 mm h−1). However, it should be noted that further assimilation of w can potentially lead to some false precipitation, resulting in slightly lower ETS values at lower thresholds (1 mm h−1) and a neutral impact on BIAS (bias score) skills. An individual case study conducted within the batch experiments reveals that assimilating w has a beneficial impact on the enhancement of vertical motion across different layers of the model, facilitating the transport of moisture from lower to middle–high model levels, thereby leading to an improvement in forecast skills.

There’s a Thin Line Between Numerators and Denominators: Addressing Academic and Behavioral Needs of Students With EBD Using SRSD Fractions

Fractions are an imperative skill for students to master to achieve success in future mathematics concepts and classes. Yet many students, especially those experiencing math difficulties and/or characteristics of emotional and behavioral disorders, continue to exit elementary school without a concrete foundation of fractions skills required to successfully transition into higher level math courses. In this article, we present an overview of SRSD Fractions, a strategy instructional framework to teach adding and subtracting fractions with unlike denominators. We highlight the benefits of using the framework to promote student engagement and skill mastery and identify free-access resources for educators to successfully implement the math intervention in their own classrooms.

КАЧЕСТВО, ПРЕДСКАЗУЕМОСТЬ И ПОЛЕЗНОСТЬ В ЗАДАЧАХ РАДИОЛОКАЦИОННОГО НАУКАСТИНГА ОСАДКОВ

Using the spatial verification Fractions Skill Score (FSS), the quality of forecasts accumulated during the testing of the precipitation nowcasting system of the Hydrometeorological Center of Russia in the warm season (May-September 2017) on the territory of the Central and Northwestern federal districts of Russia is assessed. The precipitation intensity fields based on the products of nine DMRL-C radars serve as input data for the forecast model and control data for quality assessments. FSS values are calculated on full and random samples with generalizations by distribution quartiles. The verification results are interpreted in terms of the usefulness (or utility) of forecasting precipitation categories and the practical predictability of these categories in the radar coverage areas. For median FSS estimates, it is shown that the practical predictability of precipitation intensity categories for the entire system of the radars used reaches the nowcasting limit (2.5 hour) only at the forecast of exceeding the thresholds of 0.5 and 0.75 mm/hour. For the thresholds of 1, 2, 3, and 4 mm/hour, the predictability does not exceed 120, 90, 60, and 30 minutes, respectively. Based on the behavior of a set of FSS curves depending on the spatial scale (in the range of 2-122 km), some systematic features and mutual differences of the radars used have been revealed: thus, according to the ranking results, the highest position in the experiments was occupied by the forecasts in the DMRL-C Voeikovo coverage area.

Part I: Improving Wildfire Occurrence Prediction for CONUS Using Deep Learning and Fire Weather Variables

Abstract The purpose of this research is to build an operational model for predicting wildfire occurrence for the contiguous United States (CONUS) in the 1–10-day range using the U-Net 3+ machine learning model. This paper illustrates the range of model performance resulting from choices made in the modeling process, such as how labels are defined for the model and how input variables are codified for the model. By combining the capabilities of the U-Net 3+ model with a neighborhood loss function, fractions skill score (FSS), we can quantify model success by predictions made both in and around the location of the original fire occurrence label. The model is trained on weather, weather-derived fuel, and topography observational inputs and labels representing fire occurrence. Observational weather, weather-derived fuel, and topography data are sourced from the gridded surface meteorological (gridMET) dataset, a daily, CONUS-wide, high-spatial-resolution dataset of surface meteorological variables. Fire occurrence labels are sourced from the U.S. Department of Agriculture’s Fire Program Analysis Fire-Occurrence Database (FPA-FOD), which contains spatial wildfire occurrence data for CONUS, combining data sourced from the reporting systems of federal, state, and local organizations. By exploring the many aspects of the modeling process with the added context of model performance, this work builds understanding around the use of deep learning to predict fire occurrence in CONUS. Significance Statement Our work seeks to explore the limits to which deep learning can predict wildfire occurrence in CONUS with the ultimate goal of providing decision support to those allocating fire resources during high fire seasons. By exploring with what accuracy and lead time we can provide insights to these persons, we hope to reduce loss of life, reduce damage to property, and improve future event preparedness. We compare two models, one trained on all fires in the continental United States and the other on only large lightning fires. We found that a model trained on all fires produced a higher probability of fire.

Using Machine Learning to Predict Convection-Allowing Ensemble Forecast Skill: Evaluation with the NSSL Warn-on-Forecast System

Abstract Forecasters routinely calibrate their confidence in model forecasts. Ensembles inherently estimate forecast confidence but are often underdispersive, and ensemble spread does not strongly correlate with ensemble-mean error. The misalignment between ensemble spread and skill motivates new methods for “forecasting forecast skill” so that forecasters can better utilize ensemble guidance. We have trained logistic regression and random forest models to predict the skill of composite reflectivity forecasts from the NSSL Warn-on-Forecast System (WoFS), a 3-km ensemble that generates rapidly updating forecast guidance for 0–6-h lead times. The forecast skill predictions are valid at 1-, 2-, or 3-h lead times within localized regions determined by the observed storm locations at analysis time. We use WoFS analysis and forecast output and NSSL Multi-Radar/Multi-Sensor composite reflectivity for 106 cases from the 2017 to 2021 NOAA Hazardous Weather Testbed Spring Forecasting Experiments. We frame the prediction task as a multiclassification problem, where the forecast skill labels are determined by averaging the extended fraction skill scores (eFSSs) for several reflectivity thresholds and verification neighborhoods and then converting to one of three classes based on where the average eFSS ranks within the entire dataset: POOR (bottom 20%), FAIR (middle 60%), or GOOD (top 20%). Initial machine learning (ML) models are trained on 323 predictors; reducing to 10 or 15 predictors in the final models only modestly reduces skill. The final models substantially outperform carefully developed persistence- and spread-based models and are reasonably explainable. The results suggest that ML can be a valuable tool for guiding user confidence in convection-allowing (and larger-scale) ensemble forecasts. Significance Statement Some numerical weather prediction (NWP) forecasts are more likely to verify than others. Forecasters often recognize situations where NWP output should be trusted more or less than usual, but objective methods for “forecasting forecast skill” are notably lacking for thunderstorm-scale models. Better estimates of forecast skill can benefit society through more accurate forecasts of high-impact weather. Machine learning (ML) provides a powerful framework for relating forecast skill to the characteristics of model forecasts and available observations over many previous cases. ML models can leverage these relationships to predict forecast skill for new cases in real time. We demonstrate the effectiveness of this approach to forecasting forecast skill using a cutting-edge thunderstorm prediction system and logistic regression and random forest models. Based on this success, we recommend the adoption of similar ML-based methods for other prediction models.

Impacts of Moisture Advection Scheme on Precipitation in the Steep Topography Region between the Tibetan Plateau and the Sichuan Basin

Abstract Water vapor transport is a crucial process in modeling and can contribute to errors in precipitation forecasts. To investigate the sensitivity of precipitation to the moisture advection scheme, this study introduced the two-step shape-preserving advection scheme (TSPAS), which has been proven to improve precipitation simulation over steep topography at lower resolutions, into the Southwest Center Weather Research and Forecast (WRF)-based Intelligent Numerical Grid Forecast System (SWC-WINGS) at a convection-permitting resolution. According to experiments conducted throughout the summer of 2021, the precipitation over the eastern slope of the Tibetan Plateau (TP) is highly sensitive to the moisture advection scheme. TSPAS successfully improved precipitation over the eastern slope of the TP, especially for torrential rainfall. The fractions skill score (FSS) is improved by 0.075 (27.78%) for daily precipitation with a threshold of 100 mm. Compared with the experiment with the original WRF advection scheme, the TSPAS reduced the overestimation of precipitation in the topographic region and excessive water vapor transport in a low-level atmosphere. To understand the precipitation improvement contributed by the advection scheme, additional experiments were conducted for a particular precipitation process from two approaches: switching advection schemes during the rainfall evolution and updating the variables related to moisture advection individually. Results demonstrate that the precipitation improvement is mainly contributed by the moisture advection scheme before the precipitation. Among the different variables, the combination of wind and water vapor was the most influential factor causing the precipitation improvement under the TSPAS.

Spatial solar forecast verification with the neighborhood method and automatic threshold segmentation

Numerical weather prediction (NWP) has hitherto been the default tool for providing day-ahead forecasting services to the solar energy industry. Rapid advancements in solar forecasting using NWP call for more appropriate forecast verification procedures. Current solar forecast verification is almost always carried out through ground-based radiometric data collected at point locations. Consequently, spatial features embedded in the gridded NWP forecasts cannot be verified. This study presents the spatial verification of solar irradiance forecasts using the neighborhood method, with the main goal of emphasizing the importance of such verification procedures. By applying spatial smoothing one establishes a way to directly compare the observed and forecast fields, and concurrently, mitigate verification errors that may arise from small-scale spatial displacements. Within this framework, two variants of the neighborhood-based verification, namely, the fraction-field method and the upscaling method, are examined with respect to two reanalysis products, namely, ERA5 and MERRA-2. The results suggest that, in comparison to the upscaling method, the fraction-field method can better quantify forecast performance by providing fractions skill scores. On top of the traditional neighborhood approach, which involves the subjective selection of threshold for dichotomization, an automatic threshold segmentation method based on the three-component skew-normal mixture model is proposed to resolve the issue, which can also lead to substantial time savings in data processing. Given the spatio-temporal attributes and benefits of visualization, spatial verification is anticipated to serve as a complementary practice to the current mainstream point-location forecast verification.

A general comprehensive evaluation method for cross-scale precipitation forecasts

Abstract. With the development of refined numerical forecasts, problems such as score distortion due to the division of precipitation thresholds in both traditional and improved scoring methods for precipitation forecasts and the increasing subjective risk arising from the scale setting of the neighborhood spatial verification method have become increasingly prominent. To address these issues, a general comprehensive evaluation method (GCEM) is developed for cross-scale precipitation forecasts by directly analyzing the proximity of precipitation forecasts and observations in this study. In addition to the core indicator of the precipitation accuracy score (PAS), the GCEM system also includes score indices for insufficient precipitation forecasts, excessive precipitation forecasts, precipitation forecast biases, and clear/rainy forecasts. The PAS does not distinguish the magnitude of precipitation and does not delimit the area of influence; it constitutes a fair scoring formula with objective performance and can be suitable for evaluating rainfall events such as general and extreme precipitation. The PAS can be used to calculate the accuracy of numerical models or quantitative precipitation forecasts, enabling the quantitative evaluation of the comprehensive capability of various refined precipitation forecasting products. Based on the GCEM, comparative experiments between the PAS and threat score (TS) are conducted for two typical precipitation weather processes. The results show that relative to the TS, the PAS better aligns with subjective expectations, indicating that the PAS is more reasonable than the TS. In the case of an extreme-precipitation event in Henan, China, two high-resolution models were evaluated using the PAS, TS, and fraction skill score (FSS), verifying the evaluation ability of PAS scoring for predicting extreme-precipitation events. In addition, other indices of the GCEM are utilized to analyze the range and extent of both insufficient and excessive forecasts of precipitation, as well as the precipitation forecasting ability for different weather processes. These indices not only provide overall scores similar to those of the TS for individual cases but also support two-dimensional score distribution plots which can comprehensively reflect the performance and characteristics of precipitation forecasts. Both theoretical and practical applications demonstrate that the GCEM exhibits distinct advantages and potential promotion and application value compared to the various mainstream precipitation forecast verification methods.

Global Precipitation Nowcasting of Integrated Multi-satellitE Retrievals for GPM: A U-Net Convolutional LSTM Architecture

Abstract This paper presents a deep supervised learning architecture for 30-min global precipitation nowcasts with a 4-h lead time. The architecture follows a U-Net structure with convolutional long short-term memory (ConvLSTM) cells empowered by ConvLSTM-based skip connections to reduce information loss due to the pooling operation. The training uses data from the Integrated Multi-satellitE Retrievals for GPM (IMERG) and a few key drivers of precipitation from the Global Forecast System (GFS). The impacts of different training loss functions, including the mean-squared error (regression) and the focal loss (classification), on the quality of precipitation nowcasts are studied. The results indicate that the regression network performs well in capturing light precipitation (<1.6 mm h−1), while the classification network can outperform the regression counterpart for nowcasting of high-intensity precipitation (>8 mm h−1), in terms of the critical success index (CSI). It is uncovered that including the forecast variables can improve precipitation nowcasting, especially at longer lead times in both networks. Taking IMERG as a relative reference, a multiscale analysis, in terms of fractions skill score (FSS), shows that the nowcasting machine remains skillful for precipitation rate above 1 mm h−1 at the resolution of 10 km compared to 50 km for GFS. For precipitation rates greater than 4 mm h−1, only the classification network remains FSS skillful on scales greater than 50 km within a 2-h lead time. Significance Statement This study presents a deep neural network architecture for global precipitation nowcasting with a 4-h lead time, using sequences of past satellite precipitation data and simulations from a numerical weather prediction model. The results show that the nowcasting machine can improve short-term predictions of high-intensity global precipitation. The research outcomes will enable us to expand our understanding of how modern artificial intelligence can improve the predictability of extreme weather and benefit flood early warning systems for saving lives and properties.

The added value of Himawari-8 satellite radiance data assimilation for very heavy rainfall prediction in the Indonesian Maritime Continent (Case study: East Kalimantan)

Abstract This study evaluates the impact of assimilating Advanced Himawari Imager (AHI) radiance data on the 8-10 water vapor channel using the three-dimensional variational (3DVAR) technique for the very heavy rainfall events in East Kalimantan on June 2-4, 2019. We utilized the Weather Research and Forecasting (WRF) and WRF Data Assimilation (WRFDA) system to run assimilation and prediction model simulations. To compare the assimilated model with AHI data and without assimilation, we devised two experimental schemes: 3DVAR and NODA. We assessed the enhancement in model prediction by comparing simulation results based on convection evolution, vertically integrated moisture flux convergence (VIMFC) increment, and rain prediction skill metrics, including Threat Score (TS), Equitable Threat Score (ETS), and Fractions Skill Score (FSS). Our findings suggest that the incorporation of AHI data results in a substantial improvement of the WRF model’s prediction of the very heavy rainfall events in East Kalimantan.

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.

A global observing‐system simulation experiment for the EPS–Sterna microwave constellation

AbstractA constellation of microwave sounders named the EUMETSAT Polar System–Sterna (EPS–Sterna) is under study at the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), with the aim of complementing the backbone orbits of the global observing system in low Earth orbit. The satellites of this constellation would be similar to the Arctic Weather Satellite (AWS), which is being developed by the European Space Agency (ESA). The microwave sounder on board AWS is equipped with temperature sounding channels around the 50‐GHz oxygen absorption band, water‐vapour sounding channels around the 183‐ and 325‐GHz absorption bands, and also window channels at 89 and 165 GHz. An observing‐system simulation experiment (OSSE) has been conducted at the Centre National de Recherches Météorologiques (CNRM) to evaluate the impact of this constellation on numerical weather prediction (NWP) at the global scale with the Action de Recherche Petite Echelle Grande Echelle (ARPEGE) model. Two periods ranging from August–October 2021 and December 2021–February 2022 have been chosen to compute the nature run and to run 4D‐Var data assimilation experiments. As validation of the OSSE framework, the impact of a Metop‐B denial experiment in the OSSE is compared with the impact of a Metop‐B denial with real observations. This comparison shows that the Metop‐B denial impacts are very similar in the OSSE and with real observations, with the OSSE slightly overestimating the impact. Then, the impacts of various scenarios for the EPS–Sterna constellation are assessed by computing forecast errors, fractions skill scores, and moist global energy norms, and comparing these with the results of a baseline experiment without the EPS–Sterna constellation. Significant and positive improvements of the forecasts are found up to 96 h, for every variable tested, with an impact increasing with the number of satellites.

Evaluation of Probabilistic Forecasts of Binary Events with the Neighborhood Brier Divergence Skill Score

Abstract A procedure for evaluating the quality of probabilistic forecasts of binary events has been developed. This is based on a two-step procedure: pooling of forecasts on the one hand and observations on the other hand, on all the points of a neighborhood in order to obtain frequencies at the neighborhood length scale and then to calculate the Brier divergence for these neighborhood frequencies. This score allows the comparison of a probabilistic forecast and observations at the neighborhood length scale, and therefore, the rewarding of event forecasts shifted from the location of the observed event by a distance smaller than the neighborhood size. A new decomposition of this score generalizes that of the Brier score and allows the separation of the generalized resolution, reliability, and uncertainty terms. The neighborhood Brier divergence skill score (BDnSS) measures the performance of the probabilistic forecast against the sample climatology. BDnSS and its decomposition have been used for idealized and real cases in order to show the utility of neighborhoods when comparing at different scales the performances of ensemble forecasts between themselves or with deterministic forecasts or of deterministic forecasts between themselves. Significance Statement A pooling of forecasts on the one hand and observations on the other hand, on all the points of a neighborhood, is performed in order to obtain frequencies at the neighborhood scale. The Brier divergence is then calculated for these neighborhood frequencies to compare a probabilistic forecast and observations at the neighborhood scale. A new decomposition of this score generalizes that of the Brier score and allows the separation of the generalized resolution, reliability, and uncertainty terms. This uncertainty term is used to define the neighborhood Brier divergence skill score which is an alternative to the popular fractions skill score, with a more appropriate denominator.

Challenges in Simulating Prevailing Fog Types Over Urban Region of Delhi

AbstractAccurately predicting fog is challenging due to interplay of myriad processes in its formation and high spatiotemporal variability. This study compares the performance of the Weather Research and Forecasting model with control (CNTL‐WRF) and assimilated fine‐grid (HRLDAS‐WRF) soil fields in the Ingo‐Gangetic Plain (IGP) over a 2‐years winter period (2019–2020 and 2020–2021). Results show HRLDAS‐WRF enhances accuracy in representing surface fog's heterogeneity and lifecycle across the IGP, demonstrating a spatial skill improvement of approximately 18% with a Fraction Skill Score of 0.44, compared to CNTL‐WRF's (0.36). Employing fog classification algorithm identifies 25 dense fog episodes (Vis < 500 m) over Delhi's urban boundary layer, including 14 radiation (RAD), 5 cloud‐base lowering (CBL), 3 advection + radiation (ADV + RAD), and 3 evaporation (EVA) episodes. CNTL‐WRF predicts 20 episodes but misses five due to a dry bias in the initial moisture conditions. However, HRLDAS‐WRF demonstrates limited vertical fog growth in various occurrences, highlighting the crucial role of fine‐gridded soil states for enhanced land‐surface feedback. Detailed analysis shows a 40% reduction in mean onset error for RAD fog occurrences in HRLDAS‐WRF when compared to CNTL‐WRF. In CBL fog episodes, both models exhibit significant radiative cooling and inversion before fog onset, leading to inaccurate predictions as RAD fog. Similarly, forecasting the abrupt development of ADV + RAD fog episodes is challenging as models struggle to replicate moisture intrusion over radiatively cooled surfaces in windy conditions. Predicting EVA fog, forms within an hour after sunrise, remains difficult due to the current model parameterization that rapidly dissipates fog soon after sunrise.