The potential of Bayesian model averaging (BMA) and heteroscedastic censored logistic regression (HCLR) to postprocess precipitation ensembles is investigated. For this, outputs from the National Oceanic and Atmospheric Administration's (NOAA's) National Centers for Environmental Prediction (NCEP) 11-member Global Ensemble Forecast System Reforecast, version 2 (GEFSRv2), dataset are used. As part of the experimental setting, 24-h precipitation accumulations and forecast lead times of 24 to 120 h are used, over the mid-Atlantic region (MAR) of the United States. In contrast with previous postprocessing studies, a wider range of forecasting conditions is considered here when evaluating BMA and HCLR. Additionally, BMA and HCLR have not yet been compared against each other under a common and consistent experimental setting. To compare and verify the postprocessors, different metrics are used (e.g., skills scores and reliability diagrams) conditioned upon the forecast lead time, precipitation threshold, and season. Overall, HCLR tends to slightly outperform BMA but the differences among the postprocessors are not as significant. In the future, an alternative approach could be to combine HCLR with BMA to take advantage of their relative strengths.
All Science Journal Classification (ASJC) codes
- Atmospheric Science