Development of a Gridded Flash Flood Climatology for Use by the Weather Prediction Center

Despite recent improvements in high resolution Convection Allowing Models (CAMs), flooding events remain one of the deadliest weather-related hazards in the United States. While considerable focus has been on refining the performance of the CAMs, it is also critical to have well calibrated probabilistic forecasts that can be referenced against a climatological background probability. Numerous flash flood climatologies have been explored regionally, but a CONUS wide flash flood climatology has not been developed due to the difficulty and inconsistencies with reporting flash flooding. This talk, which is part of the Forecasting a Continuum of Environmental Threats (FACETs) project, will focus on the development of a CONUS flash flood climatology using multiple data sources.

The basis of the flash flood climatology is a 11-year hydrological simulation from the Flooded Locations and Simulated Hydrographs (FLASH) project developed at the Hydrometeroology and Remote Sensing (HyDROS) laboratory at Oklahoma University (OU). FLASH is forced by Multi-Radar Multi-Sensor (MRMS) precipitation estimates and produces high resolution hydrologic forecasts nationally. Since the FLASH model is subject to error, additional observations are assimilated from the United States Geological Survey (USGS) streamflow gauges and National Weather Service (NWS) Local Storm Reports (LSRs) throughout the 11-year climatology.

This talk will highlight details of the flooding climatology, including the average probability and number of floods within 40 km of a point. The regionality and seasonality of flooding will be explored, with emphasis on urban versus rural areas. This work will be further fine-tuned with the goal of redefining WPCs operational flooding forecast probabilities from the Excessive Rainfall Outlooks (EROs) so that it is based on the flooding climatology.