摘要 : In this study, rainfall data are prepared at a 0.01 degrees scale using 16-yr spaceborne radar data over the area of 36.13 degrees S-36.13 degrees N as provided by the Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar... 展开 In this study, rainfall data are prepared at a 0.01 degrees scale using 16-yr spaceborne radar data over the area of 36.13 degrees S-36.13 degrees N as provided by the Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar (PR). A spatial resolution that is finer than the field of view is obtained by assuming rainfall uniformity within an instantaneous footprint centered on the PR footprint geolocation. These ultra-high-resolution data reveal local rainfall concentrations over slope areas. A new estimate of the maximum rainfall at Cherrapunji, India, was observed on the valley side, approximately 5 km east of the gauge station, and is approximately 50% higher than the value indicated by the 0.1 degrees-scale data. A case study of Yakushima Island, Japan, indicates that several percent of the sampling error arising from the spatial mismatch may be contained in conventional 0.05 degrees-scale datasets generated without footprint areal information. The differences attributable to the enhancement in the resolution are significant in complex terrain such as the Himalayas. The differences in rainfall averaged for the 0.1 degrees and 0.01 degrees scales exceed 10 mm day(-1) over specific slope areas. In the case of New Guinea, the mean rainfall on a mountain ridge can be 30 times smaller than that on an adjacent slope at a distance of 0.25 degrees; this is not well represented by other high-resolution datasets based on gauges and infrared radiometers. The substantial nonuniformity of rainfall climatology highlights the need for a better understanding of kilometer-scale geographic constraints on rainfall and retrieval approaches. 收起