熱帶氣旋所伴隨降雨強度觀測估計的檢驗與比較

Abstract

西北太平洋為颱風活動旺盛的地區。而以模式模擬颱風也是目前研究氣候變遷的重要課題。為了驗證模式是否準確，常會利用區域覆蓋率高的再分析資料或是衛星觀測資料，例如ERA-interim、áphrodite、TRMM 等作為標準。本研究即是著重於TRMM 3B42這筆多重衛星合成的資料相對於地面測站網格化資料，在颱風侵台期間雨量資料的表現為何。

在TRMM 3B42 3-hrly v6, v7, GSMaP-MVK和測站觀測資料相比的部分，選用了1998-2010(TRMM), 2000-2010(GSMaP-MVK)年颱風侵台其間作為比較之用。為了公平，首先需要解決兩筆資料在時間解析度、空間解析度、空間覆蓋率的不同。經過remap調整時間解析度、調整資料範圍以及將兩筆資料的時間解析度統一後，例用測站觀測資料的平均雨量強度排名並分組後，得到TRMM 3B42 3hrly v6, v7系統性地比測站觀測資料弱，且v6除了在小於2.5mm/hr 的降雨，皆比v7強；而GSMaP-MVK也弱於測站觀測資料。

為了進一步知道TRMM 3B42 3-hrly v6資料中，衛星本身的PR (Precipitation Radar) , TMI (TRMM Microwave Imager) 兩個sensor對於資料本身的貢獻為何，在研究後半段引入了TRMM 3G68來分別觀察PR, TMI, 對於西北太平洋1998-2010年海上颱風降雨的表現差異，並發現TMI相對於PR, 對於颱風中心附近的降雨表現相較於PR，更為接近真實觀測值。

Western North Pacific is an area with numberous typhoon activities. Using model to simulate typhoon is one of the important issues to research climate change. In order to verify the accuracy of models, reanalysis data or satellite observation data with high regional coverage is needed as a standard, for example: ERA-interim, áphrodite, and TRMM. This study highlights TRMM 3B42, a satellite-merge data, comparing to station observation data during typhoon affecting Taiwan.

First choosing 1998-2010 for TRMM 3B42 (both v6 and v7), and 2000-2010 for GSMaP-MVK as comparison purposes. To achieve the goal of fair comparison, adjusting these three datasets to the same both time and space resolution, also space coverage is the first priority. After ranking with station observation data’s rainfall average and separate rainfall data to several groups, both TRMM 3B42 3hrly v6 and v7 are systematically weaker than station observation data, and so does GSMaP-MVK.

To further understand TRMM’s contribution to TRMM 3B42 3-hrly v6, this study also uses TRMM 3G68 to separate PR (Precipitation Radar), TMI (TRMM Microwave Imager), the two variables for observing their behavior during 1998-2010 when typhoons were on Western North Pacific overseas. In contrast to PR, TMI performs more closely to reality around the eye of typhoons.