應用衛星影像及測高資料於灘線變遷及波候特性之研究

Abstract

由本研究預計以三年時間進行。第一年將透過色彩融合技術提升多頻譜衛星影像的 解析度，再透過多頻譜影像資料比較常態化植生指數標準差(NDVI)與常態化水指數標準 差(NDWI)在灘線位置辨識上的成效，最後探討監督式分類法所辨識出的海灘線位置與往 昔邊界擷取方法的差異；第二年利用衛星測高資料針對不同季節以及空間區塊的波浪特 性作分析與比較，並以群集分析法在空間上將台灣周邊海域分為不同的區域，最後再將 每個分區進行分年、分季及分月的波浪統計特性分析，希望能透過本技術了解台灣周邊 海域的波浪特性；第三年以目前高解析度的光學衛星影像配合二維頻譜分析來擷取影像 中的海面波浪資訊，採用人造數據與衛星影像資料探討本方法中可能會產生的誤差，了 解本法的適用性及其使用限制。

This study will be carried out for three years to investigate the shoreline evolution and wave climate using satellite images and altimetry data. In the first-year study, a method for accurately extracting shorelines of satellite images is investigated using fused HIS (Hue, Saturation, Intensity). The shoreline of a satellite image is detected by supervised classification on fused HIS with a newly-defined NDWI, which is like the traditional NDVI, using maximum likelihood method. The accuracy of detected shoreline of a chosen satellite image is evaluated by comparing that with the measured shoreline. Altimetry data of five satellites, those are Topex, Trasson-1, ERS-2, GFO-1 and Envisat-1, from 1992 to 2010 were collected to investigate the seasonal and monthly characteristics of wave climate around Taiwan through cluster analysis and distribution fitting. For engineering application, the relation between the measured offshore wave data by data buoy or pressure transducer and wave climate of altimetry data will be established in empirical implicit expressions. In the third year, two-dimensional wave spectrum will be developed to describe the wave pattern of satellite images with high resolution. The accuracy of proposed method will be examined by the error between the artificial data and satellite altimetry data (or information) to assess the validity and limitation of the proposed method in actual application.