Comparisons of precipitation estimates from microwave and infrared satellite imagery and meteorological radar for São Paulo State, Brazil

This study evaluates and compares the capacity to discriminate areas with convective rainfall for two meteorological events in São Paulo (SP) State, using simultaneously the 85.5 GHz Special Sensor Microwave Imager (SSM/I), the GOES-8 satellite and the Bauru/SP meteorological S-band radar data. Specifically, three models are used: (a) the Convective Stratiform Technique (CST) developed by Adler and Negri which uses infrared (IR) images; (b) the Goddard Scattering Algorithm (GSCAT) implemented by Adler et al. and Negri et al., for the SSM/I; and (c) precipitation rates obtained from the Bauru/SP meteorological radar data. Satellite images for two afternoons (24 and 26 January 1997) and the corresponding Bauru/SP radar data were used in this study. The results obtained from the CST technique and the GSCAT algorithm for the central area of SP State were compared with the radar data assumed to be the ground truth. Both the GSCAT and CST algorithms have good skill to discriminate areas with precipitation. The first one shows a better performance in identifying areas with high precipitation rates than the latter one. The results also show that correlation between GSCAT and radar precipitation rates is better than when the CST is used.