Three Methods to Determine Profiles of Reflectivity from Volumetric Radar Data to Correct Precipitation Estimates

Abstract

Abstract The vertical variability of radar reflectivity reduces the reliability of precipitation estimation by radar, especially in complex orography. This important source of error can, at least partially, be corrected for, if the vertical profile of radar reflectivity (VPR) is known. This work addresses three ways to determine VPR from volumetric radar data for correcting precipitation estimates. The first way uses a climatological profile. The second method, operational in Switzerland, takes the actual weather conditions into account: a mean profile is estimated directly from volumetric radar data collected close to the radar. The third way determines the identified profile, taking the variability of the VPRs in space into account. This approach yields local estimates of the profile (on areas of about 20 km × 20 km) based on an inverse method. Two cases, a convective event and a stratiform event, are used to illustrate the three ways for determining the VPR, and the resulting improvement, verified with rain gauges. An enlarged dataset of nine cases shows that a correction based on a climatological profile already improves the accuracy of rain estimates by radar significantly: the fractional standard error (FSE) is reduced from the noncorrected 44% to 31%. By correcting with a single, mean profile (averaged over 1 h using real-time data), the FSE is further reduced from 31% to 25%. Last, the use of 70 locally identified profiles leads to best results (FSE = 23%). A higher improvement (lower FSE) is obtained for the stratiform rain event than for the convective case.