Evaluating California Atmospheric River Event Precipitation Forecasts from the HRRR Ensemble (HRRRE)

Improved forecasts of Atmospheric River (AR) events, which provide nearly half of California's annual precipitation, can reduce impacts to lives, property, and water storage. We evaluate precipitation forecasts of five AR events that occurred in Feb-Mar 2019 using the High Resolution Rapid Refresh Ensemble (HRRRE). Model ensembles such as the HRRRE, which is based on HRRR version 4 and contains nine ensemble members, can provide probabilistic forecasts and insights into forecast uncertainty. We compare quantitative precipitation forecasts (QPF) to widely-used quantitative precipitation estimation (QPE) products such as the Stage IV analysis and the Mesonet gauge network. We quantify forecast performance of the HRRRE at varying lead times via several metrics such as Reliability Diagrams and Rank Histograms, and compare to forecasts from the deterministic HRRR. These results provide understanding of the usefulness of the HRRRE for forecasting AR events along the West Coast, and guide future model development.