87th AMS Annual Meeting

Saturday, 13 January 2007
GIS-based Radar Rainfall Verification
Braxton Lee Edwards, University of Oklahoma, Norman, Oklahoma; and O. Wilhelmi, D. N. Yates, and C. Worster
Hydrological models and flash flood warning systems are largely dependent on accurate precipitation inputs. In the Colorado Front Range estimation of rainfall has been problematic due to the varying intensity and spatial distribution of the precipitation fields. The goal of this project was to conduct a Geographic Information Systems-based spatial analysis and verification of the radar-derived precipitation. Rain gage measurements were used for correcting radar rainfall estimates over a 24-hour period for convective and stratiform precipitation events over the Denver Urban Drainage and Flood Control District. Two methods were tested for correcting radar-derived precipitation: 1) mean difference between recorded gage values and the radar measurements and 2) mean difference of the inverse distance weighted (IDW) interpolated gage values and the radar measurements. An overall comparison of radar and gage measurements for the two rain events showed that the radar produced more spatially accurate precipitation estimates during a convective event. The IDW interpolated precipitation method was found more appropriate for regional scale verification. The methodology developed in this study provides a framework for spatial rainfall verification, which can aid in automated correction of radar rainfall estimates. This can assist flood control and emergency managers in mitigating and responding to flash flood events.

Supplementary URL: