Developing an Improved Flood Prediction System
Flood prediction has traditionally relied on weather forecasts, past events, and local rainfall-runoff models, where available. But what if a weather event is unprecedented, land conditions change due to urbanization or wildfire, or local data is scarce? Because flooding is dynamic, highly localized, and influenced by numerous factors, there's a clear need for consistent, continuous predictions across all regions—urban and rural, mountain and coastal, flood-prone and newly vulnerable areas.
WRF-Hydro® System
Hydrometeorological research has led to the development of an advanced hydrologic prediction system called WRF-Hydro®. WRF-Hydro® was originally designed as a framework to facilitate easier coupling between the Weather Research and Forecasting (WRF) numerical weather prediction model and elements of terrestrial hydrological models. WRF-Hydro® is both a stand-alone hydrological model as well as an architecture for coupling hydrological and atmospheric models. It can model atmospheric, land surface, and hydrological processes on different spatial units and at different time scales, enabling the translation from global dynamics to local impacts within a common framework. WRF-Hydro® has been coupled with coastal models to assess compound flood risk and integrated with data assimilation tools to incorporate remote sensing, streamflow data, and forecasts of reservoir and glacial lake releases.
Highlights:
- Capable of capturing river flooding as well as inundation in low-lying areas located away from a river or stream.
- Suitable for generating dynamic maps of potential flood inundation.
- Can be coupled to storm-surge models for coastal-zone compound flooding.
- Built-in data assimilation capabilities.
- Scalable from desktop to cloud to high-performance computing platforms.
A custom configuration of WRF-Hydro® was adopted by the U.S. National Weather Service in 2016 as the operational NOAA National Water Model (NWM), which continuously forecasts hydrologic risk across the contiguous United States, Hawaii, Puerto Rico, and South-Central Alaska. In the contiguous U.S. alone, the NWM expanded hydrologic forecasting coverage from about 3,800 locations to over 2.7 million, providing critical guidance to local forecasters and emergency responders, particularly in traditionally "hydro blind" areas. The NWM aided forecasting and response operations in Texas during 2017 Hurricane Harvey, in the Midwest during the 2019 floods, in Puerto Rico during 2022 Hurricane Fiona, and in North Carolina and Tennessee during 2024 Hurricane Helene, among others.
WRF-Hydro® also provides operational flood forecasts for Israel and the United Arab Emirates, as well as community research in flood prediction and risk assessments in many locations around the globe.