Improvement of Precipitation Retrieval for Tropical Cyclones from Passive Microwave Observations

In this paper, we improve a passive microwave precipitation algorithm, the Global Satellite Mapping of Precipitation (GSMaP) algorithm [Aonashi et al., 2009, JMSJ]. GSMaP is available for many airborne microwave radiometers (MWRs) such as Tropical Rainfall Measurement Mission (TRMM) Microwave Imager (TMI). GSMaP comprises two parts: forward calculation and retrieval. In forward calculation, lookup tables (LUTs) or databases showing the relationship between rainfall and MWR brightness temperatures (TBs) are produced using radiative transfer models. Retrieval consists of a rain/no-rain classification and estimation of the rainfall rates with LUTs. The algorithm uses precipitation-related variable models (precipitation profiles, drop-size distribution, etc.). TRMM Precipitation Radar (TRMM/PR) profiles are classified into 10 types (six over land, four over ocean and coast) using the PR precipitation parameters (rain area, stratiform rain-area fraction, precipitation-top level, etc.) and the ratio between the PR precipitation rates and Lightning Imaging Sensor (LIS) flash rates. GSMaP team produces convective and stratiform precipitation profiles for these types by averaging the PR convective and stratiform precipitation profiles over prescribed rainfall bins for each type. In this averaging they use profiles relative to freezing level height (FLH) in order to exclude the influence of atmospheric temperature variations. The LUTs are calculated with the precipitation profile of the dominant precipitation type for every 5 x 5 degree point and 3 month. Tropical cyclones (TCs) are not included in 10 types. In this study, we compared the profiles between TCs and current ocean precipitation types and applied to GSMaP algorithm to improve estimation of rainfall of TCs. We analyzed TCs from Sep. 2001 to Aug. 2010 after TRMM boost. We used JAXA/EORC Tropical Cyclone Database 2A25 produced and supplied by the Earth Observation Research and application Center, Japan Aerospace Exploration Agency (JAXA/EORC). The locations and strengths of TCs were obtained from Joint Typhoon Warning Center and National Hurricane Center data. The profile of TC was taller than current tropical ocean precipitation types (shallow rain, organized system, transition zone). There ware significant differences between convective profiles for weak rain bins. Categorizing data according to the strength of TC and the distance from the storm center, the rain intensity at FLH near the cyclone center of category 3-5 for surface rain 1.0 mm/h was more than 5.0 mm/h. Convective profiles for strong rain bins of weaker TCs were taller than that of stronger TCs. We calculated the LUTs of TCs, leading to distinguished structure of eyes of TC though the eye and eyewall of TC looks blur in the original. We are planning to perform in-depth analysis of images of TCs from MWRs for further improvement of rainfall retrieval.