2016 年 12 巻 p. 297-301
In this study, we introduce a rain potential map (RPM) that globally estimates rain probabilities every hour. More specifically, we created an RPM by associating the brightness temperature (Tb) of the infrared and water vapor channels observed by five geostationary meteorological satellites (GEO) with rain probabilities observed via rain radar of the Tropical Rainfall Measuring Mission (TRMM). By using our RPM, we improved the accuracy of the Global Satellite Mapping of Precipitation (GSMaP) product, which produces global precipitation data by integrating passive microwave and infrared radiometer data. More specifically, we removed GSMaP rain areas over the ocean in which all microwave sensors were unavailable and rain probabilities according to our RPM were below 14%, which improved the “threat score” of detection in GSMaP from 0.37 to 0.41 over the ocean. Conversely, we added rain areas over land in which all microwave sensors were unavailable and rain probabilities according to our RPM were greater than 37%, which improved the “threat score” of detection from 0.27 to 0.35 over land. Given that a GSMaP “threat score” with microwave observations is approximately 0.44, our improvements here are significant.
