應用ALOS PALSAR雷達干涉分析屏東地區地表變形時間序列

Abstract

屏東平原是台灣目前持續有地層下陷之區域，其中最為嚴重的下陷區域是林邊鄉和佳冬鄉沿岸區域。屏東平原地層下陷的原因主要是超抽地下水所導致。本研究使用兩種不同的InSAR的技術進行屏東地區地層下陷的監測，並結合水文、地層下陷監測井與部分地質資料進行綜合分析。研究中使用了由2006年至2011年18幅ALOS衛星影像獲取屏東地區地表變形，使用TCPInSAR與PSInSAR技術監測並將其成果與水準測量成果進行比較，PSInSAR成果與水準成果差異之均方跟誤差為0.98公分/每年，TCPInSAR成果與水準成果差異之均方跟誤差為1.01公分/每年，顯示InSAR之技術可成功應用於屏東地區。屏東之最大下陷量約在-3公分/每年，在時間序列的比較中發現在乾季時地表變形多是屬於下陷，在濕季時則有地層回脹趨勢。利用地層下陷監測井的總壓縮量與PSInSAR結果相比對，結果差異為毫米等級，PSInSAR所測得之變形量略大於地層下陷監測井測得之壓縮量。

Pingtung Plain is the one of the counties in Taiwan suffering severe subsidence caused by groundwater withdrawal. The most serious area is located at Linpien and Chiatung. In this study, we use two InSAR methods, including Temporarily Coherent Point InSAR (TCPInSAR) and Persistent Scatterer InSAR (StaMPS), with 18 ALOS PALSAR acquisitions from 2007 to 2011 to determine land deformation. We combine the result with hydrological and land subsidence monitoring well data to analyze surface deformations in the Pingtung Plain. Results of this study show that the RMSE differences between the subsidence rates from leveling and InSAR are 0.98cm/yr and 1.01 cm/yr for PSInSAR and TCPInSAR, respectively, demonstrating the effectiveness of the two methods. The largest subsidence rate in Pingtung plain is about -3 cm/yr. The time series analysis confirms that the displacement rates of Pingtung vary with seasons. In the dry season, surface displacement rate are the largest due to settlement and the surface redounds during wet seasons. The subsidence values from monitoring well data and from PSInSAR agree to mm. The subsidence values from PSInSAR are slightly larger than that from the land subsidence monitoring well data.