- 期刊
結合TMI微波資料及MTSAT紅外線資料反演海上颱風降水
Rainfall Estimation over Ocean by Combining TMI Microwave and MTSAT Infrared Observations
摘要
本研究之目的是結合微波資料可穿透雲的特性以及紅外線資料高時間解析度的優點，建立準即時估算颱風降雨的技術。目前使用1998~2004年Tropical Rainfall Measuring Mission (TRMM)衛星上的TRMM Microwave Imager (TMI)微波資料及日本宮古島附近十一個島嶼測站之地面觀測降雨量，以統計法建立多頻道線性迴歸方程式，估算海上颱風定量降水。本研究首先以TRMM衛星上的Visible Infrared Scanner (VIRS) 11μm高解析度紅外線頻道資料，求出颱風降雨冷雲覆蓋面積率(fractional coverage; FC)；然後，使用紅外線降雨估算遙測技術(GOES Precipitation Index; GPI)，以TMI微波資料所反演的颱風降雨率當作降水真值，並與同時間及地點的紅外線資料建立線性迴歸關係式；最後，採用最佳相關係數所對應的紅外線亮度溫度作爲降雨閾值(threshold)，並使用地球同步衛星紅外線資料(MTSAT-1R)估算海上颱風降雨率。 本文針對2005年瑪莎(MATSA)、卡努(KHANUN)、2007年柯羅莎(KROSA)颱風、聖帕(SEPAT)及2008年薔蜜(JANGMI)，利用微波降雨率求得紅外線降雨閾值在三種不同解析度(1°×1°, 0.5°×0.5°, 0.25°×0.25°)下變化的情形。結果顯示，結合微波及紅外線資料所發展的GPI技術在1°×1°解析度下表現最好，並對判斷颱風降雨強弱持續性變化的趨勢有良好的成果。同時，相對於全球紅外線固定之降雨閾值235K，本研究之動態降雨閾值較能掌握住颱風降雨強度之變化。由此顯示，本研究不僅可結合微波及紅外線資料的優點，亦可有效提升監測颱風強度變化之能力。
並列摘要
This study is to develop a new method for estimating near real-time rain rate of typhoon by combining the advantages of microwave data which can penetrate through clouds and interact with the underlying rain particles and the high temporal resolutions from infrared observations. The TMI (Tropical Rainfall Measuring Mission Microwave Imager) microwave data on board TRMM satellite and a total of 11 rain gauges on small islands spreading in the south of Japan from 1998 to 2004 are used to establish a multi-channel linear regression equation by statistical method for retrieving the rain rate of typhoon. There are three major procedures in this study. Firstly the high resolution Visible and Infrared Scanner (VIRS) data of 11 μm on board TRMM is employed to calculate the fractional coverage (FC) of cold cloud. The GPI (GOES Precipitation Index) technique is then utilized to set up the best relationship among the FCs and rain rates, which are retrieved from TMI data. Finally, the satellite Typhoon rain rate over ocean is estimated by the MTSAT-1R geostationary satellite. The investigations of variations of infrared rainfall threshold are carried out by using a total of five cases (MATSA, KHANUN, 2005; KROSA, SEPAT, 2007; JANGMI, 2008) under three domains of different resolutions (1°×1°, 0.5°×0.5°, 0.25°×0.25°). Results show that for this algorithm, combining microwave and infrared data, the correlation between rain rate and FC for the 1°×1° domain is the best than those for the other domains. From the regularly continuous rain rate products, it is capable of catching the tendency of rainfall intensity. In addition, the flexible infrared threshold calibrated by microwave data is more practically useful than the fixed global threshold of 235 K. Overall, this algorithm not only successfully utilizes the advantage of combining microwave and infrared data, but also has the ability of monitoring the changes of typhoon rainfall intensity