SSM/I衛星資料在估算侵台颱風降雨之初步研究

The Preliminary Study of the Distribution of Typhoon Rainfall for Taiwan Areas using SSM/I Data

劉崇治(Chung-Chih Liu) ； 林裕豐(Yu-Feng Lin) ； 戴世忠(Shi-Zhong Dai) ； 趙俊傑(Jun-Jie Chao) ； 劉振榮(Gin-Rong Liu)

《大氣科學》 35卷1期 (2007/03) Pp. 51-67

摘要

本文利用1988至2004年SSM/I (Special Sensor Microwave/Imager)衛星資料，透過回歸統計建立颱風侵襲台灣地區期間，定量降水估算之客觀方法。研究選取36個侵台颱風個案，以個案衛星資料反演之可降水量及颱風滯留時間作為預報要素，透過樣本分佈的統計方法將上述要素分別轉換成無因次之預報因子a與b，由a與b合成之預報指數I(下标 x) (I(下标 x)=a×b)與台灣地區總累積降水量之間的相關係數達0.76，顯示利用預報指數I(下标 x)來估算颱風侵台期間台灣地區的總累積降水量有相當程度的可靠性。另研究中分別針對颱風大小、強度及侵台路徑進行分類，並計算預報指數I(下标 x)與各類別中台灣地區實際總降雨之相關係數。研究結果發現，以30kt暴風半徑做為分類，將有助於提升預報指數與台灣地區總累積降水量之間的相關性，其中暴風半徑介於101km至200km的中型颱風相關性高達0.96，小型颱風（暴風半徑為100km及以下）及大型颱風（暴風半徑介於201km至300km）的相關係數亦有超過0.8以上的表現，顯示透過颱風暴風半徑的分類，可以進一步提升估算台灣地區總降雨量的可信度。

關鍵字

SSM/I ； 可降水量 ； 預報指數

並列摘要

An algorithm to determine the distribution of rainfall of typhoons for affecting Taiwan was proposed by using SSM/I microwave and rain gauge data. There were totally thirty six typhoon cases in this study. At first, the distribution of the rainfall for each possible moving direction of typhoon (totally six different moving directions) was determined in advance by using rain gauge data. The precipitable water of typhoon and the affecting time were considered to play a role to determine the total accumulate rainfall of Taiwan. Therefore, once the precipitable water of typhoon and affecting time were determined, the prediction of rainfall can be determined. A non-dimensional factor of prediction (I(subscript x)) with meaning of total precipitable water and affecting time of typhoon was used to forecast the total accumulate rainfall of Taiwan. The results show that the coefficient of correlation between the non-dimensional factor of prediction and total accumulate rainfall of Taiwan was 0.76. Moreover, It will contribute to improve the accuracy of estimation of total accumulate rainfall of Taiwan by distinguishing the radius of 30kt wind of typhoon. The coefficient of correlation for medium-sized typhoon (101 km to 200 km) was up to 0.96. The small-sized typhoon (smaller than 100 km), and large-sized typhoon (201 km to 300 km ) also exceed more than 0.8. That shows that it can improve and estimate the credibility of total rainfall of Taiwan further through the classification of the radius of storm wind of typhoon.

並列關鍵字

SSM/I ； Precipitable water ； A non-dimensional factor of prediction