Does the Implementation of a Nearby X-band Dual-polarimetric Radar Improve Rainfall Estimates from Distant S-band WSR-88D’s? An In-Depth Analysis

The Next Generation Radar (NEXRAD) system operating within the CONUS is devoid of coverage over sparsely large regions. This poor coverage encompasses densely population regions in addition to hydrologically active domains whereby high resolution precipitation data is necessary for hydrologic simulations for estimating streamflow and, thus, flooding. One example of a densely population area is in Boone County, MO, located in Central Missouri, where the two nearest radars, Kansas City (KEAX) and St. Louis (KLSX) are, approximately, 180- and 160-km from the city, respectively. Studies have previously documented that the quality in radar rainfall estimation significantly decreases beyond ranges of 150-km from the radar, prompting researchers to find a remedy to the problem. One method into the amelioration of these errors at large distance has been proposed to construct nearby, cost-effective, high-resolution X-band radars. However, literature as to whether this solution is valid is lacking. The purpose of the current study is to test whether the installation of a nearby dual-polarimetric X-band radar, MZZU, is significantly more accurate (CI = 0.05) than distant S-band WSR-88D radars, KEAX and KLSX in estimating rain rates at various ranges. Results indicate that, despite the localized X-band radar being better overall, grouping the data into yearly, daily, hourly, and seasonal components produce conflicting results. Furthermore, the effects of implementing dual-polarized radar variables (e.g., differential reflectivity and the specific differential phase shift) do not show a significant (CI = 0.05) advantage in estimating rainfall compared to the Marshall-Palmer Z-R relationships.