Utilizing Dual-Polarization Instantaneous Precipitation Rate to Predict Flash Flooding

The WSR-88D radar uses a dual polarization quantitative precipitation estimation (QPE) algorithm called Digital (instantaneous) Precipitation Rate (DPR) which can be useful in predicting flash flooding. When re-processed to estimate the duration for which areas meet thresholds of two, three, or four inches per hour respectively, research shows that potential flash flood events may be easier to detect when compared to traditional radar interrogation techniques. This study focused on events from both Buffalo (KBUF) and Cleveland (KCLE) radars during the convective seasons between 2015 and 2017. Radar data were examined to identify cases where DPR exceeded the threshold noted above for 20 minutes or more over the course of an hour. A total of 1710 cases were then compared to observed flash flood events. These cases were categorized further based on a location’s vulnerability to flash floods using a local static Flash Flood Potential Index (FFPI) and stratified by rainfall rate and duration. Results show that a DPR of 2”/hr. for 35 minutes or 3”/hr. for 25 minutes are thresholds which correlate with an increased flash flood risk for areas that have at least a moderate FFPI. Rainfall rates over 4 inches per hour, although impressive from a precipitation rate standpoint, usually do not last long at any one location. These rates show less skill since they do not provide much lead time and may be prone to hail contamination. Precipitation rate thresholds of 2 and 3 inches per hour indicate an increased ability to detect and provide lead time for flash floods. These thresholds have been used operationally and have aided meteorologists with detection and lead time for flash floods. Additional research may support the utility of the DPR product in all NWS offices.