Detecting Azimuth-Dependent Biases in Radar Precipitation Estimates

Across the United States, regions experiencing drought face many challenges, including agricultural woes and water use restrictions; however, no single preferred methodology for monitoring drought conditions currently exists in the United States. In Texas, drought conditions are monitored using high-resolution, multi-sensor precipitation estimates (MPEs) that originate from radar precipitation data and are combined with long-term climate station data to produce objective assessments of drought intensity. However, MPEs suffer from systematic biases due to the nature of radar technology. The purpose of this research is to improve the accuracy of MPEs used for monitoring drought by detecting such biases. One prominent type of systematic bias occurs when a radar beam encounters an obstruction, such as a tall building or mountain, and appears as an azimuthal discontinuity in the estimated precipitation data. Additional discontinuities occur in boundary regions where spatial coverage suddenly shifts from one radar to another. The Canny edge detection technique was selected from various methodologies and will be used to locate regions where azimuth-dependent discontinuities and radar boundary errors exist. Future research is necessary to determine the appropriate corrections needed to minimize the detected biases and provide an optimal MPE dataset for drought monitoring.