| 摘要: | 發生雙眼牆結構的熱帶氣旋除了本身強度高，雙眼牆成形前後伴隨的特殊風場結構和氣旋強度變化仍時常無法由數值模式補捉。為改善對雙眼牆颱風（颶風）的預報能力，瞭解雙眼牆結構發生的原因和維持機制是必要的，對此學者做了許多數值模式實驗和個案研究，而依實驗結果提出之假說或理論皆需要觀測資料驗證。此外，對觀測資料進行大量、多個案分析（Sitkowski et al. 2011）也可帶給我們對雙眼牆颱風（颶風）的概觀認識。
觀測資料方面，在大西洋從1977年開始，NOAA便對所有生成之熱帶氣旋進行各項動力與熱力參數的飛機觀測；反觀擁有最高雙眼牆發生率（Hawkins and Helveston, 2004）的西北太平洋，卻因缺乏常態飛機觀測，研究時除了少數具計畫支持的個案擁有完整的觀測資料，其餘必須由衛星觀測資料中反演出需要的動力或熱力參數。然而在衛星資料應用上，因擁有較佳透雲性、可獲得較多氣旋資訊的微波觀測搭載於繞極衛星，使覆蓋到颱風完整結構的觀測資料在時間上不連續。所幸Ushio et al.（2009）應用卡曼濾波器整合微波資料和搭載於同步衛星的紅外線觀測資料共同估算降雨量，提供每小時的全球降雨空間分佈產品（Global Satellite Mapping of Precipitation, GSMaP），使我們對本區域雙眼牆結構發生時颱風強度和結構在時間軸上之連續變化的進一步分析變為可行。
本研究即應用GSMaP，收集雙眼牆發生期間位於海上的颱風個案，視剖面上的降雨極值位置作為判斷內、外眼牆位置的依據，並進一步從軸對稱觀點決定各時間點的眼牆位置並解析內、外眼牆各別的降水強度定義眼牆置換（Willoughby et al. 1982）起始，記錄期間歷經的時間、降雨結構和強度變化，和西北太平洋過去的雙眼牆颱風研究相比，本研究的各項分析具有時間連續之優勢。分析中將個案依眼牆置換過程的不同情況分類為置換傾向較強的取代型和置換傾向較弱的非取代型，發現取代型(非取代型)眼牆置換平均擁有較大(小)的外眼牆成型半徑和較短(長)的置換時間，與Didlake and Houze (2011)提出之理論呼應；同時將結果與不同區域、不同觀測平台的雙眼牆研究比較，本研究結果具有合理的平均眼牆置換時間和具關連性的眼牆尺寸，說明本研究提出之分析方法具相當參考價值。
Double eyewall(or concentric eyewall) structure, often appears in strong tropical cyclones. As the strong intensity and the dramatic structure changes they have, to study their physical feature and to improve the accuracy of forecast are very important for coastal area.
Observation data is needed in order to verify the hypothesis raised from model experiments which are created to study double eyewall cyclones. In addition, simply analyzing a great amount of observation data (Sitkowski et al. 2011) can also give us understanding of their structure and intensity changes.
Unlike Atlantic, there is no regular flight observations in West north Pacific. Instead, researchers use satellite-retrieved information as database. However, passive microwave observations are temporal discontinuously in tropical cyclone monitoring.
Fortunately, Ushio et al.（2009）applied Kalman filter to combine infrared and passive microwave data to estimate the hourly precipitation rate, named as Global Satellite Mapping of Precipitation(GSMaP). The development of GSMaP provide possibility to analyze West north Pacific's double eyewall typhoons temporal continuously as Atlantic's.
In this work, hourly GSMaP imageries were divided into 12 sections, and we identified the locations of inner and outer eyewall at each time from the locations of local maximum precipitation in each section, and resolved the precipitation intensity of each eyewall. As the results, the structure and intensity changes of 21 double eyewall cases were recorded, and classified into 2 types: replacement type and non-replacement type. replacement(non-replacement) cases have much(less) significant replacement feature when outer eyewall contracts, and their outer eyewall forms with larger(smaller) radius and replacement takes less(longer) time to be completed, which supports Didlake and Houze (2011)'s theory. The result is compared with previous papers, which have been analyze double eyewall structures through observation data but from different platforms, and the similar eyewall repalcement time and relative eyewall size were found in this study. |
