Precipitation detection by the TOPEX/Poseidon dual frequency radar altimeter, TOPEX microwave radiometer, Special Sensor Microwave/Imager and climatological shipboard reports

The athors evaluate the ability of a dual-frequency radar (C and Ku band) altimeter to detect rain events. A TOPEX/Poseidon (T/P) altimeter rain flag for the year 1994 is compared to collocated rain rate (RR) from the Defense Meteorological Satellite Program's Special Sensor Microwave Imager (DMSP SSM/I), as processed to the TOPEX/Poseidon passive radiometer's (TMR) liquid-water content, and to a 34-year climatology of shipboard present-weather reports compiled by G. W. Petty (1995). The altimeter-SSM/I analysis is couched in terms of the tradeoff between the probability of a false positive and the probability of a failure to detect rain. The authors show that the ability of the SSM/I and TMR datasets to detect precipitation are closer to each other than to either the altimeter or the shipboard climatology, and this difference is accentuated at latitudes poleward of 45/spl deg/. They argue that the different footprint sizes explain only part of this discrepancy. They propose that the difference at high latitudes is caused by the altimeter data's sensitivity to snow. In order to detect precipitation (as opposed to detecting bad altimetric values or out-of-range altimetric corrections), a TMR-only flag with liquid-water content of 600 /spl mu/m recovers too few rain events, 400 /spl mu/m is close to climatological moderate-to heavy intensity rains, and 200 /spl mu/m is close to rain of any intensity. For the same purpose, a combined altimeter and TMR flag, with a TMR threshold of 100 /spl mu/m and with the Ku radar cross section 1.5 standard deviations below an average Ku/C curve, gives the best match for climatological precipitation of any intensity class.