本研究係使用TRMM(Tropical Rainfall Measuring Mission)衛星上的微波資料(TMI, TRMM Microwave Imager)及台灣自動雨量觀測站(ARMTS, Automatic Rainfall and Meteorological telemetry System)降雨資料，並利用散射法以求得在台灣陸上的衛星估算降雨量。此外，以地面降雨觀測資料為真值進行本研究(LRCT, Land Rainfall Retrieval by Chen and Tsai)反演成果的驗證，以進一步了解衛星估算降雨量在陸上降雨反演上的特徵。 對於陸地降雨反演的研究，首先，利用GOES-9紅外線衛星雲圖，辨識台灣地區無雲晴空的天氣形態，以作為大氣無散射條件的個案，並建立台灣地區之散射指數(SIL, Scattering Index over Land)方程。其次，針對無雨狀況的散射指數SIL，進行統計分析以定義台灣地區散射指數的降雨閾值(threshold)為8K。最後，建立台灣陸地散射指數SIL與地面降雨量的關係，以求得台灣陸地降雨的方程式。 在估算陸地降雨量的驗證方面，以敏督利及艾莉颱風侵台期間為個案。其定量降水反演的驗證結果，估算的平均降雨量分別為14.14、6.10mm/hr，其均方根誤差分別為5.14、348mm/hr。此外，在本研究中也發現在地形的迎風面上衛星估算的雨量值有明顯的低估而背風面的卻有高估的情形。因此，對於陸地降雨反演的研究必須考慮到地形及風場的影響，方能提升降雨估算的準確度。

In this study, TMI (TRMM Microwave Imager) on board the TRMM (Tropical Rainfall Measuring Mission) satellite data and rainfall rate of Automatic Rainfall and Meteorological Telemetry System (ARMTS) measurements of Taiwan were used to estimate quantitative precipitation over the Taiwan land by scattering indices method. And ARMTS rainfall rate are used as a ground truth to validate the accuracy of rainfall retrievals derived from LRCT (Land Rainfall Retrieval by Chen and Tsai), to understand precipitation characteristics from LRCT over the Taiwan land. The study of rainfall retrievals over land, the first to develop Scattering Index over Land(SIL) equation for Taiwan land under non-scattering atmosphere conditions, that identified weather as cloud-free on Taiwan region with GOES-9 IR images. Second To determine rainfall threshold to distinguish from rainfall area is 8 K, and derived from an analysis of SIL statistics under non-rainfall conditions. Final found the regressive relationship between SIL and ground-gauge rain rate for get an equation to retrieved rainfall over Taiwan land. Using two typhoon cases to validate the accuracy of estimated rainfall over land during the MINDULLE and AERE typhoon hit Taiwan. Respectively, the estimate value of mean rain rate is 14.14 and 6.10 mm/hr, the root-mean-squared error is 5.14 and 3.48 mm/hr. In addition, the result indicates that LRCT rainfall retrievals are underestimated at facing the wind side and overestimated at the lee side. For this reason, it have to consideration parameter of terrain and wind field to influence the rainfall retrievals over land, and that can increase accuracy of estimated rainfall.