NOAA Hydrometeorology Testbed

Improving flood prediction models in mountainous regions

October 27, 2014

California’s Russian River watershed has a history of some of the most severe flood damages in California. Because of its steep topography, which can enhance the streamflow response during heavy precipitation events, and limited availability of streamflow information, accurately forecasting flash floods can be challenging. ESRL's Physical Sciences Division (PSD) is collaborating with the National Weather Service (NWS) Weather Forecast Office (WFO) in Monterey, CA to evaluate distributed hydrologic modeling for improved prediction of flooding events in the Russian River watershed.

"Heavy rain falling in mountainous terrain can quickly change small streams into dangerous torrents, and it can be difficult for forecasters to predict the risk of flash flooding," says Rob Cifelli, a meteorologist at PSD.

The Research Distributed Hydrologic Model (RDHM) incorporates detailed information, such as precipitation, terrain, soils and vegetation, which is needed to accurately simulate the complex water balance dynamics of tributary watersheds, especially in mountainous regions. In contrast to traditional methods, an advantage of the distributed approach is that channel flow estimates are available at any location in the watershed stream network. Use and evaluation of the RDHM for hydro forecasting in the Russian River watershed will help inform flash flood warnings at WFOs nationwide.

Russian River flooding in Guerneville, California (Photo credit: USACE)

NWS hydrological services are provided primarily by the River Forecast Centers (RFCs) and WFOs. The RFCs currently apply the NWS River Forecast System for selected forecast points on major rivers. The RFCs are moving to the new Community Hydrologic Prediction System-Flood Early Warning System (CHPS-FEWS) computing platform, which has enhanced capabilities.

"PSD has been working on improving precipitation and streamflow forecasts in mountainous regions like the Russian River watershed," says Cifelli.

To support WFO hydrologic service responsibilities in the Russian River watershed, PSD established a prototype real-time modeling environment in conjunction with the NWS Office of Hydrologic Development (OHD) and Riverside Technologies. The RDHM has recently been integrated into the CHPS-FEWS system to supplement the RFC forecasts for small tributaries.

"The CHPS-FEWS system provides a framework to better understand how uncertainty in a precipitation forecast translates into uncertainty in streamflow response," says Cifelli. "At the same time RDHM can be evaluated as a tool for flash flood forecasting."

Contact: Rob Cifelli