Experimental Test of the Effects of Z–R Law Variations on Comparison of WSR-88D Rainfall Amounts with Surface Rain Gauge and Disdrometer Data

Abstract

Reflectivity factors and rainfall rates found from Level II WSR-88D data for the National Weather Service (NWS) radar in Greer, South Carolina (KGSP), are compared with similar parameters found from disdrometer data collected at the Clemson Atmospheric Research Laboratory. These comparisons are used to determine experimentally the sensitivity of rainfall amounts found from the WSR-88D data to variations in the parameters A and b of the Z?R law (Z = ARb) that is used in analysis of the data. Analyses of data for nine storms in upstate South Carolina are described. These nine cases encompass a variety of rainfall types including stratiform rain, airmass thunderstorms, and strong cold front convective activity. It is found, after correction of the radar reflectivity factors for obvious calibration offset, that the rainfall depths found by radar are in good agreement with those found from the disdrometer when the NWS default values of A and b (A = 300, b = 1.4) are used. If the values of A and b found from an empirical Z?R analysis of the disdrometer data are used, then, as expected, the agreement is even better. It is important to recognize that this good agreement was obtained only after adjustment of the radar-measured reflectivity factors for calibration offset. A similar analysis is also described for one other storm but with the addition of rain gauge data from locations in upstate South Carolina, which are remote from the Clemson laboratory and where there are no collocated disdrometers. The radar data are used to determine rainfall amounts for each of the rain gauge locations as well as for the Clemson laboratory. It is found that the agreement between radar-measured rainfall amounts and those found from the rain gauges is very good, even when the default values of A and b are used. Again, this agreement is obtained only after correction of the radar data for the calibration offset found from comparison of the radar reflectivity factors determined from the WSR-88D data and disdrometer data at the Clemson laboratory. When the values of A and b determined from the disdrometer data are used, the agreement is further improved. It is concluded that the commonly observed large differences between KGSP WSR-88D-measured rainfall amounts and surface rain gauge data are not due primarily to variations in Z?R law parameters but are the result of hardware calibration offsets. These offsets could probably be eliminated by performing an accurate calibration of the WSR-88D with special emphasis on antenna gain. The results are also of importance to the ground validation field campaigns, which are a part of the Tropical Rainfall Measuring Mission operated by the National Aeronautics and Space Administration. Similar large differences between WSR-88D and surface measurements have been found in these field campaigns.