Statistical Postprocessing of Week-1 and Week-2 Precipitation Forecasts over Taiwan

The recent emergence of the National Digital Forecast Database as the flagship product of the National Weather Service has resulted in an increased demand for forecast guidance products on fine-mesh grids. Unfortunately, fine-grid forecasts with geographically regionalized statistical models are usually plagued by nonmeteorological discontinuities at regional boundaries. This study treats the problem in a regionalized Global Forecast System (GFS)-based model output statistics (MOS) application that produces 6-h probabilistic quantitative precipitation forecasts (PQPFs) on a 4-km grid up to 192 h in advance. The technique involves incorporating areal overlap in the geographical regionalization and weighting multiple PQPFs in region-overlap zones. The degree of overlap ranges from about 20 km along meteorologically significant regional boundaries to about 150 km at quasi-arbitrary boundaries. The forecast-weighting constants for a grid point in an overlap zone vary in direct proportion to the distances to the closest associated regional boundaries. The application of the region-overlap and forecast-weighting techniques resulted in retention of sharp PQPF gradients along meteorologically significant regional boundaries and prevention of artificial discontinuities at quasi-arbitrary boundaries. The eradication of the discontinuities in the forecast patterns was achieved without sacrificing forecast skill. While the regionalization was customized for producing high-spatial-resolution 6-h PQPFs over the contiguous United States with a specialized gridded MOS application, the region-overlap and forecast-weighting techniques may have general applicability. Also, the quality of the 6-h PQPFs was not strongly dependent on customization of the regionalization.

Heavy-rainfall events are common in southern France and frequently result in devastating flash floods. Thus, an appropriate anticipation of future rainfall is required: for early flood warning, at least 12–24 h in advance; for alerting operational services, at least 2–3 days ahead. Precipitation forecasts are generally provided by numerical weather prediction models (NWP), and their associated uncertainty is generally estimated through an ensemble approach. Precipitation forecasts also have to be adapted to hydrological scales. This study describes an alternative approach to commonly used limited-area models. Probabilistic quantitative precipitation forecasts (PQPFs) are provided through an analog sorting technique, which directly links synoptic-scale NWP output to catchment-scale rainfall probability distributions. One issue concerns the latest developments in implementing a daily version of this technique into operational conditions. It is shown that the obtained PQPFs depend on the meteorological forecasts used for selecting analogous days and that the method has to be reoptimized when changing the source of synoptic forecasts, because of the NWP output uncertainties. Second, an evaluation of the PQPFs demonstrates that the analog technique performs well for early warning of heavy-rainfall events and provides useful information as potential input to a hydrological ensemble prediction system. It is shown that the obtained daily rainfall distributions can be unreliable. A statistical correction of the observed bias is proposed as a function of the no-rain frequency values, leading to a significant improvement in PQPF sharpness.

This study investigates the probabilistic quantitative precipitation forecast (PQPF) performance of typhoons that affected Taiwan during 2011–16. In this period, a total of 19 typhoons with a land warning issued by the Central Weather Bureau (CWB) are analyzed. The PQPF is calculated using the ensemble precipitation forecast data from the Taiwan Cooperative Precipitation Ensemble Forecast Experiment (TAPEX), and the verification data, verification thresholds, and typhoon characteristics are obtained from the CWB. The overall PQPF performance of TAPEX has an acceptable reliability and discrimination ability, and the higher probability error is distributed at the mountainous area of Taiwan. The PQPF performance is significantly influenced by typhoon characteristics (e.g., typhoon tracks, sizes, and forward speeds). The PQPFs for westward-moving, large, or slow typhoons have higher reliability and discrimination ability, and lower-probability error than those for northward-moving, small, or fast typhoons, except for similar reliability between fast and slow typhoons. Because northward-moving or small typhoons have larger forecast track error, and their PQPF performance is sensitive to the accuracy of the forecast track, a higher probability error occurs than that for westward-moving or large typhoons. Furthermore, because there is no difference in track error between fast and slow typhoons, the larger track spread for slow typhoons increases the rainfall forecast spread and reduces the probability error. The orientation of Taiwan’s topography and the topographic effect also influence and increase the distribution and value of probability error for northward-moving, small, or fast typhoons. In summary, forecast track characteristics are influenced by typhoon characteristics and further affect the PQPF performance.

Bias correction of ensemble precipitation forecasts in the improvement of summer streamflow prediction skill

Hydrologic uncertainty processor for probabilistic river stage forecasting: precipitation-dependent model

To study the impact of training sample heterogeneity on the performance of Bayesian model averaging (BMA), two BMA experiments were performed on probabilistic quantitative precipitation forecasts (PQPFs) in the northern China region in July and August of 2010 generated from an 11-member short-range ensemble forecasting system. One experiment, as in many conventional BMA studies, used an overall training sample that consisted of all available cases in the training period, while the second experiment used stratified sampling BMA by first dividing all available training cases into subsamples according to their ensemble spread, and then performing BMA on each subsample. The results showed that ensemble spread is a good criterion to divide ensemble precipitation cases into subsamples, and that the subsamples have different statistical properties. Pooling the subsamples together forms a heterogeneous overall sample. Conventional BMA is incapable of interpreting heterogeneous samples, and produces unreliable PQPF. It underestimates the forecast probability at high-threshold PQPF and local rainfall maxima in BMA percentile forecasts. BMA with stratified sampling according to ensemble spread overcomes the problem reasonably well, producing sharper predictive probability density functions and BMA percentile forecasts, and more reliable PQPF than the conventional BMA approach. The continuous ranked probability scores, Brier skill scores, and reliability diagrams of the two BMA experiments were examined for all available forecast days, along with a logistic regression experiment. Stratified sampling BMA outperformed the raw ensemble and conventional BMA in all verifications, and also showed better skill than logistic regression in low-threshold forecasts.

Predicting the location and intensity of precipitation still remains a main issue in numerical weather prediction (NWP). Resolution is a very important component of precipitation forecasts in NWP. Compared with a lower resolution model, a higher resolution model can predict small scale (i.e., storm scale) precipitation and depict convection structures more precisely. In addition, an ensemble technique can be used to improve the precipitation forecast because it can estimate uncertainties associated with forecasts. Therefore, NWP using both a higher resolution model and ensemble technique is expected to represent inherent uncertainties of convective scale motion better and lead to improved forecasts. In this study, the limited area ensemble prediction system for the convective-scale (i.e., high resolution) operational Unified Model (UM) in Korea Meteorological Administration (KMA) was developed and evaluated for the ensemble forecasts during August 2012. The model domain covers the limited area over the Korean Peninsula. The high resolution limited area ensemble prediction system developed showed good skill in predicting precipitation, wind, and temperature at the surface as well as meteorological variables at 500 and 850 hPa. To investigate which combination of horizontal resolution and ensemble member is most skillful, the system was run with three different horizontal resolutions (1.5, 2, and 3 km) and ensemble members (8, 12, and 16), and the forecasts from the experiments were evaluated. To assess the quantitative precipitation forecast (QPF) skill of the system, the precipitation forecasts for two heavy rainfall cases during the study period were analyzed using the Fractions Skill Score (FSS) and Probability Matching (PM) method. The PM method was effective in representing the intensity of precipitation and the FSS was effective in verifying the precipitation forecast for the high resolution limited area ensemble prediction system in KMA.

A hydrometeorologic forecast system (HFS) has been developed that takes advantage of new high-resolution rainfall datasets from the WSR-88D radar system, the Oklahoma Mesonet, and Oklahoma Local Analysis and Prediction System (OLAPS). New schemes to analyze precipitation and to adjust radar rainfall rates have been proposed to improve the quantitative precipitation forecast (QPF) for hydrologic purposes. Adjusted WSR-88D rainfall rates were advected by the 700-mb wind field from OLAPS to produce an extrapolation QPF. Several experiments were conducted to evaluate the effect of the rainfall adjustment and wind field upon the extrapolation QPF. In addition, mesoscale model‐produced QPFs were generated using The Pennsylvania State University‐ National Center for Atmospheric Research Mesoscale Model. Control and rainfall assimilation experiments were performed using both Kuo and Kain‐Fritsch cumulus parameterization schemes for three rainfall events from April 1994. All model runs were integrated forward 12 h and then verified against the analyzed precipitation field. Both the extrapolation and model-produced QPFs were used to produce hydrologic forecasts for the Dry Creek watershed in north-central Oklahoma. Results indicate that extrapolation QPFs degrade exponentially with time and become inferior to the QPF from a mesoscale model after 2 h. When the extrapolated rainfall estimates were input into a hydrologic model, an underestimate of the peak flow occurred since the time evolution of precipitating systems is not handled by extrapolation. Due to the lag time between the peak in precipitation and the peak in streamflow, the greatest impact upon the accuracy of hydrologic forecasts resulted from improvements in the analyzed precipitation field. On the other hand, mesoscale forecast simulations revealed that the assimilation of analyzed rainfall had a limited impact upon the evolution of model-produced precipitation forecasts out to 4 h. However, model-produced QPFs improved after 8 h into the integration. While the Kuo scheme produced less dispersion error, the Kain‐ Fritsch scheme created less amplitude error. The assimilation of analyzed rainfall through the convergence factor of the Kuo scheme had a greater impact upon the performance of the mesoscale model than did the Kain‐Fritsch rainfall assimilation through the adjustment of its precipitation efficiency factor. Therefore, a new generation HFS has been developed to take advantage of new technology and new scientific methods in an attempt to mitigate the age-old issue of devastating floods that occur without warning. Each component has been tested and evaluated. The results of testing and evaluating each component of the proposed HFS are presented in this paper.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]ARCEAU

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.

A feed-forward neural network is configured to calibrate the bias of a high-resolution probabilistic quantitative precipitation forecast (PQPF) produced by a 12-km version of the NCEP Regional Spectral Model (RSM) ensemble forecast system. Twice-daily forecasts during the 2002–2003 cool season (1 November–31 March, inclusive) are run over four U.S. Geological Survey (USGS) hydrologic unit regions of the southwest United States. Calibration is performed via a cross-validation procedure, where four months are used for training and the excluded month is used for testing. The PQPFs before and after the calibration over a hydrological unit region are evaluated by comparing the joint probability distribution of forecasts and observations. Verification is performed on the 4-km stage IV grid, which is used as “truth.” The calibration procedure improves the Brier score (BrS), conditional bias (reliability) and forecast skill, such as the Brier skill score (BrSS) and the ranked probability skill score (RPSS), relative to the sample frequency for all geographic regions and most precipitation thresholds. However, the procedure degrades the resolution of the PQPFs by systematically producing more forecasts with low nonzero forecast probabilities that drive the forecast distribution closer to the climatology of the training sample. The problem of degrading the resolution is most severe over the Colorado River basin and the Great Basin for relatively high precipitation thresholds where the sample of observed events is relatively small.

Quantitative hydrologic forecasting usually requires knowledge of the spatial and temporal distribution of precipitation. First, it is important to accurately measure the precipitation falling over a particular watershed of interest. Second, especially for small watersheds and/or for longer forecast lead times, forecasts of precipitation are critical to the achievement of the greatest possible hydrologic forecast accuracy and longest possible lead time. This paper describes the current hydrologic forecasting program of the U.S. National Weather Service (NWS) and highlights the relevance of Quantitative Precipitation Forecasting (QPF) products to real-time hydrologic forecasting. Specific requirements for QPF products in support of hydrologic forecasting applications are defined and current operational QPF procedures are reviewed to determine to what extent they meet these requirements. It is concluded that no known QPF procedures capable of fulfilling all desired requirements are currently available operationally, although much valuable QPF information is available to meet parts of these requirements. Some recent advances in mesoscale QPF research are examined and these techniques are treated in two categories: those uncoupled dynamically from and those dynamically coupled to hydrologic forecasting procedures. Finally, a summary of possible future directions toward achieving improved use of QPF information in hydrologic forecasting applications is presented.

This paper outlines a methodology to produce probabilistic quantitative precipitation forecasts by means of a dedicated uncertainty processor for weather model output. The uncertainty processor is developed as a component of a Bayesian forecasting system for river flow prediction. In this context the quantitative precipitation forecast is envisaged as a mixed binary–continuous predictand. The processor is applied to the quantitative precipitation forecasts and to precipitation observations covering a 5-yr period, whereby the forecasted and observed relative air humidity are used as ancillary meteorological indicators. The application of the processor to the selected dataset highlights a significantly larger skill of the quantitative precipitation forecast in predicting event occurrence rather than event depth and provides an objective quantification of the forecast uncertainty. The methodology applied here remains restricted to small basins, in which spatial variability of precipitation can be considered negligible. The need for processing the uncertainty induced by spatial variability of rainfall is briefly addressed.

Conventional statistical postprocessing techniques offer limited ability to improve the skills of probabilistic guidance for heavy precipitation. This paper introduces two artificial neural network (ANN)-based, geographically aware, and computationally efficient postprocessing schemes, namely, the ANN-multiclass (ANN-Mclass) and the ANN–censored, shifted gamma distribution (ANN-CSGD). Both schemes are implemented to postprocess Global Ensemble Forecast System (GEFS) forecasts to produce probabilistic quantitative precipitation forecasts (PQPFs) over the contiguous United States (CONUS) using a short (60 days), rolling training window. The performances of these schemes are assessed through a set of hindcast experiments, wherein postprocessed 24-h PQPFs from the two ANN schemes were compared against those produced using the benchmark quantile mapping algorithm for lead times ranging from 1 to 8 days. Outcomes of the hindcast experiments show that ANN schemes overall outperform the benchmark as well as the raw forecast over the CONUS in predicting probability of precipitation over a range of thresholds. The relative performance varies among geographic regions, with the two ANN schemes broadly improving upon quantile mapping over the central, south, and southeast, and slightly underperforming along the Pacific coast where skills of raw forecasts are the highest. Between the two schemes, the hybrid ANN-CSGD outperforms at higher rainfall thresholds (i.e., >50 mm day−1), though the outperformance comes at a slight expense of sharpness and spatial specificity. Collectively, these results confirm the ability of the ANN algorithms to produce skillful PQPFs with a limited training window and point to the prowess of the hybrid scheme for calibrating PQPFs for rare-to-extreme rainfall events.

This paper summarizes the results of an experiment in which National Weather Service forecasters formulated probabilistic quantitative precipitation forecasts (QPFs) during a 17-month period in 1981–82. These forecasts expressed the likelihood that certain threshold amounts of precipitation would be equaled or exceeded in 12-hour periods at four locations in Texas. The forecasters had no previous experience in quantifying the uncertainty in such forecasts, but they did receive feedback regarding their collective performance at the end of the first year of the experiment. In the evaluation of the experimental results, particular attention is focused on three issues: 1) the reliability and skill of the subjective QPFs; 2) the effects of feedback and experience on the quality of these forecasts; and 3) the relative performance of the subjective probabilistic QPFs and objective probabilistic QPFs produced by the model output statistics system. The subjective probabilistic QPFs possess positive skill, although they exhibit considerable overforecasting for larger precipitation amounts. Moreover, the feedback provided to the forecasters evidently contributed to modest increases in the reliability and skill of their forecasts. In this regard, the quality of the subjective and objective QPFs is generally comparable in the first year of the experiment. However, after the receipt of the feedback, the skill of the subjective forecasts exceeded the skill of the objective forecasts. These results are considered to be encouraging regarding the ability of forecasters to formulate reliable and skillful probabilistic QPFS, but more extensive experiments should be undertaken to investigate this and related issues in greater detail.