利用衛星SAR影像資料推測海洋風浪方向波譜

Abstract

遙測技術應用於海洋波浪的測定，尤其是對波浪的頻率和方向而言極具 功效。而藉由波浪之方向頻率能譜，可以瞭解波浪能量在頻率及方向上的 分佈情形，其對海岸結構物的隨機振動及受力或波浪之變形特性的探討等 均極為重要。 本文利用ERS-1（歐洲遙測衛星一號）衛星SAR（合成孔徑 雷達）影像資料，來推測海洋風浪的方向波譜，以輔助傳統測定方法的不 足。合成孔徑雷達為一主動式雷達系統，基於微波的特性，具有可穿透雲 層，不受霧、降雨等天候影響，其主動式遙測的特質更提供不分晝夜，具 有全天候觀測的能力，可以快速地獲得大尺度的地表資訊，對海洋觀測提 供了一個相當方便的觀測方式。 本研究先將衛星影像資料做信號變化主 趨勢消除（detrend）的前置處理，再經由二維快速傅立葉轉換以獲得直 角座標之二維波數能譜，接著透過轉換而得極座標之波數能譜，最後再轉 換為方向頻率能譜。研究中並以波譜分析的觀點來做各種提高解析度的分 析方法，並將分析計算結果與海面實測數據做比對驗證。我們以五筆置於 花蓮港外海的浮球式波高計所量測到的實測數據，應用Extension of Maximum EntropyPrinciple的方法求出波浪的方向頻譜，以便與衛星SAR 影像資料所分析出來的方向頻譜做驗證與比對的工作。 從SAR遙測海洋 波場的分析結果與現場實測方向頻率能譜的比對中可以發現，若實測波浪 的示性波高大於1.6公尺，則在比對中不僅主頻上甚為一致，能譜的形狀 也頗為吻合。而在波向的比對方面存在著約 □0% 的差異，仍有待做進一 步的驗證。從實例的比對中可以証實利用衛星SAR資料推測海洋風浪方向 頻譜的正確與可行性。 Remote sensing is an outstanding technique for ocean wave measurement,especially for the estimation of wave frequency and direction. Directionalwave spectrum is important for understanding the distribution of waveenergy, especially for the application in random vibration of oceanstructure and wave deformation. In this study, the SAR (Synthetic Aperture Radar) data imaged from theESA (European Space Agency) Remote-Sensing Satellite 1 (ERS-1) has beenused to evaluate the directional wave spectrum. This method is developedto resolved the problems encountered by the traditional measuring method.As an active system, the SAR provides its own illumination and isindependent on light resources like sunlight, thus permitting the wholeday operation. Furthermore, microwave is able to penetrate the weatherphenomena such as clouds, fog and precipitation, it thus allows forall-weather imaging. It is especially convenient for ocean observations. In this study, data pre-processing using the concept of detrend wasperformed for the first stage. The secondly, the two-dimensional F.F.T(Fast Fourier Transform) algorithm was applied to create two-dimensionalwave spectrum in Cartesian coordinate system. On the third stage,successive formulas were undertaken to get the wave number spectrum inpolar coordinate system. It was finally converted into directionalfrequency wave spectrum. The most impressing result of this research isthe improvement of the resolution, which has been obtained by implementingthe algorithm of wave spectrum analysis. The estimated results for wave spectrum analysis are compared with theobservations measured by wave-rider. Five data sets surveyed over the oceansurface to the east of Hualien Harbor are used to evaluate the accuracy ofmethod in this study. The verification procedure applies the Extension ofMaximum Entropy Principle to obtain the directional wave spectrum in orderto compare with the results created by analyzing SAR imagery data. The results of comparison suggests that as long as the significant waveheights measured by wave-rider are more than 1.6 meters, the estimated peakfrequency coincides very well with the observed one. Moreover, theirprofiles of directional wave spectrum look quite similar. The error rangeswithin □0% for wave direction, but it need further verification. Throughabove evaluation, this study confirms the feasibility of estimating oceanwave spectrum using SAR imagery data.