Evaluation of GPROF-SSMI/S rainfall estimates over land during the Brazilian CHUVA-VALE campaign

One of the major goals of the CHUVA project (Cloud processes of tHe main precipitation systems in Brazil: A contribUtion to cloud resolVing modeling and to the GPM [GlobAl Precipitation Measurement]) is to collect information about the cloud processes of the main precipitating systems over Brazil in order to evaluate and improve the quality of satellite-based precipitation estimates. Thus, this paper evaluates the performance of the Goddard Profiling Algorithm (GPROF) version 2004 for the Special Sensor Microwave Imager/Sounder (SSMIS) sensor, carried onboard the Defense Meteorological Satellite Program (DMSP) F16, F17 and F18 satellites, for instantaneous rain rates over land by comparing with other remote sensing based estimates such as X-band dual polarization rainfall retrievals from the CHUVA Project (named here X-band CHUVA radar) and TRMM (Tropical Rainfall Measurement Mission) Precipitation Radar - version 7 (PR-V7) precipitation estimates (named here the 2A25_V7 product) for the CHUVA-VALE campaign. The analyses were performed on an area within a radius of 60 km from the X-band CHUVA radar, located at Sao Jose dos Campos, SP - Brazil (centered at 23.2 degrees S and 45.95 degrees W) and another delimited by 41 degrees W-51 degrees W and 18 degrees S-28 degrees S (10 degrees x 10 degrees box). GPROF showed an overestimate of light/moderate rain rate intensities, while underestimating the rainfall rates above 10 mm h(-1), considering the X-band CHUVA radar as reference. The same behavior was observed when compared with the 2A25_V7 database. Through a PDF analysis, GPROF was found to overestimate the frequencies of moderate rain rates (between 2 and 10 mm h(-1), above 15%), and underestimate the frequencies of light and high rain rates (<2 mm h(-1) and >10 mm h(-1), respectively) when compared to both the X-band CHUVA radar and 2A25_V7 reference databases. The results for the studied region suggest that GPROF has a relatively good agreement (spatial distribution and accumulated rainfall), especially for convective rain cases, due the significant presence of ice scattering. However, the intensity of light/moderate rains is overestimated. (C) 2014 Elsevier B.V. All rights reserved.