Non-systematic errors of monthly oceanic rain rate from microwave radiometry

Abstract

Global monthly rainfall maps derived from microwave sensors are now routinely produced from data collected by the Defense Meteorological Satellite Program (DMSP) Special Sensor Microwave Imager (SSM/I) and Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI). Since the DMSP (F-13 and F-14) and TMI satellites have different orbits and swath structure, they have different sampling errors. In this paper the non-systematic errors, which are dominated by sampling errors, of monthly oceanic rainfall over 5/spl deg/ latitude by 5/spl deg/ longitude boxes are calculated separately for SSM/Is and TMI for three years (1998 to 2000). The non-systematic errors are 21.4%, 21.9% and 19.1% for SSM/I of F-13 and F-14, and TMI, respectively. The nonsystematic error for TMI is less than 1 mm day/sup -1/ for rain rates up to 6 mm day/sup -1/, and is about 15-20% at higher rain rates. This is smaller than that for either F-13 or F-14 SSM/I alone at the low rain rates, but is comparable at the high rain rates (>6 mm day/sup -1/). A combination of F-13 and F-14 SSM/I rain rates reduces the nonsystematic error to slightly lower (15.2%) than that of the TMI alone. It is argued that a refined space-borne sampling strategy, such as that proposed by the Global Precipitation Mission (GPM), is needed to reduce the sampling errors to better than 10% at high rain rates.