An object-based comparison of reflectivity patterns in tropical cyclone rainbands detected by different radar types

Keywords: hurricane, radar, landfall, shape analysis
Abstract Type: Paper Abstract

Abstract

The different scanning angles and atmospheric cross-sections through which energy passes can produce different reflectivity values as rainfall regions are detected by radars. The Dual-frequency Precipitation Radar (DPR) on board the Global Precipitation Mission’s satellite scans vertically with 5 km horizontal resolution in a swath spanning 250 km. The Weather Surveillance Radar 1988-Doppler (WSR) units scan more horizontally over 360 degrees at varying tilts, with the highest resolution data nearest the station. This study uses an object-based approach to examine how differently these instruments detect the rainbands in tropical cyclones (TCs). We create a mosaic to detect the entire TC by retaining the highest reflectivity value detected by each WSR. We then extract values along a 3 km altitude from this and the DPR dataset and map them to a 5 km grid retaining pixels that contain values from both datasets. We delineate objects by retaining pixels equal to or greater than 20 dBZ up to 40 dBZ and calculate spatial metrics to facilitate comparisons. Among the cases examined, Hurricane Laura’s compact shape prior to landfall was detected the most similarly between the two types of radars. Two WSRs stopped working during a second DPR overpass of Laura, resulting in DPR objects being larger and more compact at the same reflectivity value.