Assessment of Distributed Hydrological Model Products for Flash Flood Operations

National Weather Service (NWS) forecasters at the River Forecast Centers (RFCs) and Weather Forecast Offices (WFOs) provide timely and reliable severe weather and hydrological forecast services. NWS flash flood operations may be improved through applying advances in distributed hydrological modeling (DHM). Our project extends on-going DHM research and development efforts within the NOAA Hydrometeorological Testbed (HMT) and NWS National Water Center (NWC) to conduct a US Weather Research Program-sponsored assessment of the DHM for flash flood forecast and warning operations.

The HMT has worked with the NWC to implement the Research Distributed Hydrological Model (RDHM) to support research on evaluation of various precipitation monitoring, data assimilation, and forecasting techniques. We applied the RDHM to the Russian-Napa Rivers (RR-N) basin to support assessment of model accuracy; that work has characterized the uncertainty of flood peak predictions and water budget. Further, we implemented the DHM in the CHPS-FEWS computing environment, which allowed a streamlined, near-real time data ingest and simulation capability. The CHPS-FEWS instance has been configured to ingest multiple Quantitative Precipitation Estimations (QPE) and Quantitative Precipitation Forecasts (QPF) forcings, including radar-rainfall products generated by the Multi-Radar Multi-Sensor product.

Export of the CHPS-FEWS time series to a web service interface allowed visualizations of input datasets and DHM outputs including precipitation, surface runoff, threshold frequency level, and soil moisture grids as well as animations of these. Downstream applications use the FEWS output to identify at-risk road crossings and other flood impact features. The system has been operated in real-time for two years to support an R2O assessment of the DHM approach for WFO flash flood operations.

For this presentation, we will show the prototype and illustrate flood threat information at local scales and at locations not currently served by RFC operations. We will describe case study storms and summarize feedback from forecasters. A detailed assessment of users’ opinions of the DHM approach and the various information products will be summarized. We will also spend time discussing the equally important intent to examine WFO flash flood operations concept-of-operations. Lessons learned from this effort are expected to inform future regional assessment activities of the NWC's National Water Model.