Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
31,219
result(s) for
"Precipitation forecasting"
Sort by:
Probabilistic Precipitation Forecast Postprocessing Using Quantile Mapping and Rank-Weighted Best-Member Dressing
Hamill et al. described a multimodel ensemble precipitation postprocessing algorithm that is used operationally by the U.S. National Weather Service (NWS). This article describes further changes that produce improved, reliable, and skillful probabilistic quantitative precipitation forecasts (PQPFs) for single or multimodel prediction systems. For multimodel systems, final probabilities are produced through the linear combination of PQPFs from the constituent models. The new methodology is applied to each prediction system. Prior to adjustment of the forecasts, parametric cumulative distribution functions (CDFs) of model and analyzed climatologies are generated using the previous 60 days’ forecasts and analyses and supplemental locations. The CDFs, which can be stored with minimal disk space, are then used for quantile mapping to correct state-dependent bias for each member. In this stage, the ensemble is also enlarged using a stencil of forecast values from the 5 × 5 surrounding grid points. Different weights and dressing distributions are assigned to the sorted, quantile-mapped members, with generally larger weights for outlying members and broader dressing distributions for members with heavier precipitation. Probability distributions are generated from the weighted sum of the dressing distributions. The NWS Global Ensemble Forecast System (GEFS), the Canadian Meteorological Centre (CMC) global ensemble, and the European Centre for Medium-Range Weather Forecasts (ECMWF) ensemble forecast data are postprocessed for April–June 2016. Single prediction system postprocessed forecasts are generally reliable and skillful. Multimodel PQPFs are roughly as skillful as the ECMWF system alone. Postprocessed guidance was generally more skillful than guidance using the Gamma distribution approach of Scheuerer and Hamill, with coefficients generated from data pooled across the United States.
Journal Article
Toward Probabilistic Post-Fire Debris-Flow Hazard Decision Support
Post-wildfire debris flows (PFDF) threaten life and property in western North America. They are triggered by short-duration, high-intensity rainfall. Following a wildfire, rainfall thresholds are developed that, if exceeded, indicate high likelihood of a PFDF. Existing weather forecast products allow forecasters to identify favorable atmospheric conditions for rainfall intensities that may exceed established thresholds at lead times needed for decision-making (e.g., ≥24 h). However, at these lead times, considerable uncertainty exists regarding rainfall intensity and whether the high-intensity rainfall will intersect the burn area. The approach of messaging on potential hazards given favorable conditions is generally effective in avoiding unanticipated PFDF impacts, but may lead to “messaging fatigue” if favorable triggering conditions are forecast numerous times, yet no PFDF occurs (i.e., false alarm). Forecasters and emergency managers need additional tools that increase their confidence regarding occurrence of short-duration, high-intensity rainfall as well as tools that tie rainfall forecasts to potential PFDF outcomes. We present a concept for probabilistic tools that evaluate PFDF hazards by coupling a high-resolution (1-km), large (100-member) ensemble 24-h precipitation forecast at 5-min resolution with PFDF likelihood and volume models. The observed 15-min maximum rainfall intensities are captured within the ensemble spread, though in highest ∼10% of members. We visualize the model output in several ways to demonstrate most likely and most extreme outcomes and to characterize uncertainty. Our experiment highlights the benefits and limitations of this approach, and provides an initial step toward further developing situational awareness and impact-based decision-support tools for forecasting PFDF hazards.
Journal Article
A Hybrid Analog-Ensemble–Convolutional-Neural-Network Method for Postprocessing Precipitation Forecasts
An ensemble precipitation forecast postprocessing 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 postprocessing 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 postprocessing and neural networks, and is one of only a few studies pertaining to precipitation ensemble postprocessing in BC.
Journal Article
Deep Learning Forecast Uncertainty for Precipitation over the Western United States
Reliably quantifying uncertainty in precipitation forecasts remains a critical challenge. This work examines the application of a deep learning (DL) architecture, Unet, for postprocessing deterministic numerical weather predictions of precipitation to improve their skills and for deriving forecast uncertainty. Daily accumulated 0–4-day precipitation forecasts are generated from a 34-yr reforecast based on the West Weather Research and Forecasting (West-WRF) mesoscale model, developed by the Center for Western Weather and Water Extremes. The Unet learns the distributional parameters associated with a censored, shifted gamma distribution. In addition, the DL framework is tested against state-of-the-art benchmark methods, including an analog ensemble, nonhomogeneous regression, and mixed-type meta-Gaussian distribution. These methods are evaluated over four years of data and the western United States. The Unet outperforms the benchmark methods at all lead times as measured by continuous ranked probability and Brier skill scores. The Unet also produces a reliable estimation of forecast uncertainty, as measured by binned spread–skill relationship diagrams. Additionally, the Unet has the best performance for extreme events (i.e., the 95th and 99th percentiles of the distribution) and for these cases, its performance improves as more training data are available.
Journal Article
Different Initial Condition Perturbation Methods for Convection-Permitting Ensemble Forecasting over South China during the Rainy Season
In this study, downscaling, ensemble data assimilation, time lagging, and their combination were used to generate initial condition (IC) perturbations for 12-h convection-permitting ensemble forecasting for heavy-rainfall events over South China during the rainy season in 2013–20. These events were classified as weak- and strong-forcing cases based on synoptic-scale forcing during the presummer rainy season and as landfalling tropical cyclone (TC) cases. This study investigated the impacts of various IC perturbation methods on multiscale characteristics of perturbations and the forecast performance for both nonprecipitation and precipitation variables. These perturbation methods represented different source IC uncertainties and thus differed in multiscale characteristics of perturbations in vertical structures, horizontal distributions, and time evolution. The combination of various IC perturbation methods evidently increased perturbations or spreads of precipitation in both magnitude and location and thus improved the forecast-error estimation. Such an improvement was most and least evident for TC cases during the early and late forecasts, respectively, and was more evident for strong- than weak-forcing cases beyond 6 h. The combination of various IC perturbation methods generally improved both the ensemble-mean and probabilistic forecasts with case-dependent improvements. For heavy rainfall forecasting, 1–6-h improvements were most prominent for TC cases in terms of discrimination and accuracy, while 7–12-h improvements were least prominent for weak-forcing cases in terms of reliability and accuracy. In particular, the improvements in predicting weak-forcing cases increased with spatial errors. In contrast, for strong-forcing cases, the improvements were least and most prominent before and beyond 6 h, respectively.
Journal Article
Conditional Ensemble Model Output Statistics for Postprocessing of Ensemble Precipitation Forecasting
Forecasts produced by EPSs provide the potential state of the future atmosphere and quantify uncertainty. However, the raw ensemble forecasts from a single EPS are typically characterized by underdispersive predictions, especially for precipitation that follows a right-skewed gamma distribution. In this study, censored and shifted gamma distribution ensemble model output statistics (CSG-EMOS) is performed as one of the state-of-the-art methods for probabilistic precipitation postprocessing across China. Ensemble forecasts from multiple EPSs, including the European Centre for Medium-Range Weather Forecasts, the National Centers for Environmental Prediction, and the Met Office, are collected as raw ensembles. A conditional CSG EMOS (Cond-CSG-EMOS) model is further proposed to calibrate the ensemble forecasts for heavy-precipitation events, where the standard CSG-EMOS is insufficient. The precipitation samples from the training period are divided into two categories, light- and heavy-precipitation events, according to a given precipitation threshold and prior ensemble forecast. Then individual models are, respectively, optimized for adequate parameter estimation. The results demonstrate that the Cond-CSG-EMOS is superior to the raw EPSs and the standard CSG-EMOS, especially for the calibration of heavy-precipitation events. The spatial distribution of forecast skills shows that the Cond-CSG-EMOS outperforms the others over most of the study region, particularly in North and Central China. A sensitivity testing on the precipitation threshold shows that a higher threshold leads to better outcomes for the regions that have more heavy-precipitation events, i.e., South China. Our results indicate that the proposed Cond-CSG-EMOS model is a promising approach for the statistical postprocessing of ensemble precipitation forecasts.
Journal Article
Improving National Blend of Models Probabilistic Precipitation Forecasts Using Long Time Series of Reforecasts and Precipitation Reanalyses. Part I: Methods
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).
Journal Article
Improving National Blend of Models Probabilistic Precipitation Forecasts Using Long Time Series of Reforecasts and Precipitation Reanalyses. Part II: Results
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.
Journal Article