Rainfall estimation over the Taiwan Island from TRMM/TMI data

Abstract In this study, TMI (TRMM Microwave Imager) data on board TRMM (Tropical Rainfall Measuring Mission) satellite were used to create a new algorithm for satellite microwave rainfall retrievals over the land of Taiwan. The concept of scattering index method was accepted in this study. The rain gauge data of Automatic Rainfall and Meteorological Telemetry System (ARMTS) measurements of Taiwan were used as ground truth to evaluate the satellite rainfall retrievals. There are two processes during the procedure of satellite rainfall retrievals in this study. The first step is to distinguish the raining pixel from the non-raining one and the second step is to estimate the quantitative precipitation for the raining pixels. The GOES-9(Geostationary Operational Environmental Satellite)infrared data were used to recognize the clear, cloud-free, weather system type. The clear TMI data were used to establish the equation of Scattering Index over Land (SIL) in Taiwan area. Finally, the SIL of raining pixels and the corresponding rain gauge data were used to establish the satellite rainfall retrievals algorithm. Two typhoon cases data were used to validate the usability of the satellite rainfall retrievals algorithm developed by this study. Results show that there exists a highly correlation between the estimated values and rain gauge measurements. For cases of MINDULLE and AERE typhoon, their root mean square errors are only 5.14 and 3.48 mm/hr, respectively, with the means of 14.14 and 6.10 mm/hr, respectively. The validation shows that the new algorithm for satellite rainfall retrievals is suitable for Taiwan's land. It was also shown that the rainfall retrievals are underestimated at the upslope side of mountains and overestimated at the lee side. For this reason, one should have take into account the effects caused by terrain and wind field for the improvement of satellite rainfall retrievals in the future.