MODIS影像波段萃取應用於乾旱監測之研究

Study of MODIS Imagery Band Extraction Applied to Drought Monitoring

鍾玉龍(Yuh-Lurng Chung) ； 陳朝圳(Chaur-Tzuhn Chen) ； 奚珍妮(Chen-Ni His) ； 劉士銘(Shih-Ming Liu)

《台灣林業科學》 20卷3期 (2005/09) Pp. 239-247

https://doi.org/10.7075/TJFS.200509.0239

摘要

台灣地區乾旱期間，農業用水常須轉移，迫使農田休耕或以轉作方式來調節用水。因此乾旱管理對台灣地區水資源之運用調配產品日漸重要。本研究利用美國NASA近年研發的MODIS影像，應用主軸轉換法來挑選MODIS影像的36波段中可適用於監測乾旱之波段，期望能為監測乾旱之乾旱警示系統提供一些基本資訊。研究結果顯示，萃取草生地樣點利用250m解析力之MODIS影像第l、2波段所推導的NDVI值，於乾旱與非乾旱季節間差異並不顯著，500 m（第3~7波段）及l ,000 m（第20~25及第27~36波段）解析力的PC1第一主軸影像對乾旱及非乾旱影像具有極顯著性之差異，因此將來可應用主軸轉換分析的方法，應用MODIS影像於即時之乾旱監測。

關鍵字

主軸轉換法 ； 預警系統 ； 水資源 ； 常態化植生指標

並列摘要

During drought periods in general, irrigation water must be transferred to civilian use, and farmland forced is compelled to remain fallow. Therefore water resource arrangements have gradually become important in drought management in Taiwan. MODIS images procured from the US National Aeronautics and Space Administration (NASA) were used in this study, and a principal component transformation (PCT) was applied to extract useful information from the MODIS 36-band imagery for a drought warning system. The results revealed that the normalized difference vegetation index (NOW) derived from MODIS imagery bands 1 and 2 which were extracted from a grass sample plot between the dry and wet seasons did not significantly differ, and this means that just using the NOW for drought monitoring seems improper. Both the 500-m (bands 3~7) and 1000-m (bands 20~25 and 27~36) PC1 images in the dry and wet seasons had extremely significant differences. Therefore applying the PCT method for band extraction from MODIS images may be a proper way to instantaneously monitor thoughts in the future.

並列關鍵字

principal component transformation PCT ； warning system ； water resources ； normalized difference vegetation index NOVI