39HYDRO The Integration of Probabilistic Precipitation Forecasts from the National Blend of Models into NOAA's National Water Model: A Case Study of Hurricane Beryl

The accuracy of forecasting extreme hydrologic events, including flash floods and long-duration floods, is significantly influenced by the skill of quantitative precipitation forecasts (QPFs). NOAA's operational National Water Model (NWM), which generates medium-range (10-day ahead) hydrologic forecasts for over 3 million stream channels across the OCONUS, currently lacks probabilistic forecasting and relies solely on deterministic QPF from the Global Forecast System (GFS). In this work, the operational National Blend of Models (NBM) probabilistic QPF products has been explored and incorporated into the NOAA’s Meteorological Forcing Engine (MFE) and NWM. This study seeks to quantify the impact of probabilistic information from NBM Quantile Mapping and Dressing (QMD) on ensemble streamflow forecasting and compare it with the currently operational NWM GFS-based deterministic streamflow forecasts. As a case study, MR ensemble forecasts of hourly streamflow will be generated and evaluated over a period from late June to early July 2024, capturing Hurricane Beryl’s landfall in the South Central region of the United States.