Assimilation of Real-Time Streamflow Observations for the National Water Model Using Ensemble Kalman Filter

It is widely recognized that real-time assimilation of streamflow observations is one of the most effective means of improving operational streamflow forecasting. Currently, the National Water Model (NWM) employs a nudging scheme to assimilate more than 6,000 USGS streamflow observations in real time. A problem with nudging is how the forecasts relax quickly to open-loop bias in the forecast. In this work, we present an ensemble streamflow data assimilation (DA) approach intended for operational implementation in the NWM. The approach utilizes the new channel-only capabilities of the NWM and HydroDART, a coupled system of the offline WRF-Hydro model and NCAR’s Data Assimilation Research Testbed, DART. Our approach involves updating the single model state of channel discharge with overland and groundwater influxes as uncertain input in assimilation of streamflow data via ensemble Kalman filter. In this presentation, we describe the streamflow DA module under development and present preliminary comparisons with the current NWM DA scheme for multiple flooding events.