CONSISTENCY CHECK OF DROP SIZE DISTRIBUTION IN RAIN RETRIEVALS WITH A COMBINATION OF TRMM MICROWAVE IMAGER AND PRECIPITATION RADAR OBSERVATIONS Shoichi Shige (1) ∗ , Toshio Iguchi (2) , Shuji Shimizu (3) , Nobuhiro Takahashi (2) , Toshiaki Kozu (3) and Ken’ichi Okamoto (4) (1) Department of Aerospace Engineering, Osaka Prefecture University, Sakai, Japan (2) National Institute of Information and Communications Technology (NICT), Koganei, Tokyo, Japan (3) JAXA/Earth Observation Research Center, Tokyo, Japan (4) Department of Electronic and Control systems Engineering, Shimane University, Matsue, Japan 1. INTRODUCTION The TRMM Microwave Imager (TMI) and Precipi- tation Radar (PR) have been providing distribution of rainfall throughout the Tropics and contributed signif- icantly towards reducing uncertainty in satellite esti- mates of rainfall. Although differences in global av- eraged rainfall between the two sensors have been reducing, regional and seasonal differences still ex- ist (Berg et al., 2002). Possible error sources static model assumptions involved with individual retrieval algorithm. For TMI rain retrieval (Kummerow et al., 1996), the database of cloud/radiative model simu- lations is very important. On the other hand, for PR rain algorithm (Iguchi et al., 2000), the appropriate selection of drop size distribution (DSD) is very im- portant because observed radar reflectivities Z de- pend strongly upon the size of water drops. Here, the consistency in observed and simu- lated brightness temperature is investigated, where the simulated brightness temperature are derived from PR precipitation profiles, similar to Viltard et al. (2000), but for a case over the South China Sea Monsoon Experiment (SCSMEX). We exam- ine whether or not the DSD model assumed by the PR algorithm produces good or poor agreement be- tween observed and simulated brightness tempera- ture. 2. DATA DESCRIPTIONS We use two of TRMM standard data products, referenced as PR rain rate/PR-corrected reflectiv- ity (2A25) and TMI brightness temperature (1B11). Both standard products here are version 5. The re- sults reported in this short paper are based on a ∗ Corresponding author address: Dr. Shoichi Shige, De- partment of Space Engineering, Osaka Prefecture Univer- sity, 1-1 Gakuen-cho, Sakai, Osaka, 599-8531 Japan; e-mail: shige@aero.osakafu-u.ac.jp scene observed on a subset of orbit 2719 on 19 May 1998 from the SCSMEX region. Figure 1: TRMM Z-R Here we introduce a brief summary of Z-R re- lations, or, equivalently, the drop size distribution (DSD) for 2A25. The “globally” averaged Z-R rela- tion used in version 5 of 2A25 are as follows Z = 185R 1.43 (convective), (1) and Z = 300R 1.38 (stratiform). (2) The Z-R relations are obtained from a collection of Z- R relations measured near the oceanic from widely distributed locations around the world (Kozu et al., 1999). As shown in Fig. 1, the same Z translates to R smaller in stratiform compared to convective rain- fall. This is because the presence of a few very large drop (formed from the melting large aggregates) in the stratiform drop spectra increases the radar re- flectivity much more than it increases the rain volume (Rosenfeld and Ulbrich, 2003).