Development of Forcing Engine for National Water Center Operational Water Prediction System

In this study, the development of the meteorological forcing engine for the NOAA National Water Model (NWM) is introduced.  The system uses real-time, operational forcing from the NSSL (National Severe Storms Laboratory) Multi-Radar/Multi-sensor System (MRMS) rain gauge corrected radar QPE (quantitative precipitation estimation) product, the NOAA/ESRL High Resolution Rapid Refresh (HRRR) output, and NOAA/ESRL Rapid Refresh (RAP) model results to drive the analysis and short-range forecasting. The NCEP Global Forecast System (GFS) output is used to construct the forcing for the medium range (0-10 day) forecasts, and the NOAA Climate Forecast System (CFS) provides the forcing for the long range (0-30 day) forecasts. Meteorological input data are regridded to a common 1-km CONUS domain and selected variables (temperature, humidity, pressure and shortwave radiation) are topographically downscaled to drive the the community WRF-Hydro model.  The pre-operational modeling system was spun up from Oct. 1, 2011 through May 07, 2016 with NLDAS2 forcing. The meteorological forcing data engine and the WRF-Hydro modeling system were then implemented on NOAA’s WCOSS supercomputing system and then forward executed in real-time to create national hydrologic predictions including streamflow forecasts. Following a presentation of the overall software architecture of the meteorological forcing data engine, the impact of different combinations of the forcing fields and current and experimental downscaling methods are investigated, focusing on temperature, humidity, radiation, wind fields and precipitation.