| 摘要: | 地層下陷為一漸進、持續性的地質災害問題，面對此種災害，精確地、高頻地持續地表變形監測為必要的工作，而合成孔徑雷達干涉(Interferometry Synthetic Aperture Radar, InSAR))技術的應用則為地層下陷提供了一個合適的監測方法。然而由於雷達使用的微波訊號在水氣較高的狀況下可能有遲滯的問題，因此InSAR技術在地表變形的量測上可能會因大氣條件而影響其精確度，有時可達10至14公分。特別是近年的研究顯示了全球增溫使得熱帶地區於近數十年來有水氣增加的情況，使得InSAR技術在這些區域，如印尼等熱帶國家，將面臨其應用上的挑戰。在InSAR的應用上，永久干涉點(persistent scatterer point candidates, PSCs)的決定為影響測量成果的關鍵步驟，而PSCs一般可基於不同時期雷達影像的強度以及相位的同調性(coherence)進行選取。據此，本研究則致力於訂定出一個有效的InSAR應用方法，透過應用現行常見的InSAR技術，包含雷達差分干涉法(Differential Interferometry SAR, DInSAR)、多時序干涉法(Multi-Temporal Interferometry SAR, MTInSAR)以及永久散射體差分干涉法(Persistent Scatterer Interferometry SAR, PSInSAR)，並同時採取不同的PSCs的選取策略，包含進行大氣相位過濾(atmospheric phase screen, APS)、應用強度穩定指標(amplitude stability index, ASI) 以及空間同調性(spatial coherence, SC)等PSCs擷取方法進行試驗。研究區則選定位於熱帶氣候地區的印尼雅加達市以及東爪哇省。使用之雷達資料為Sentinel 1A(降軌)影像，針對2014年11月至2019年間共收集了208組影像對進行分析，並收集6處GPS站之量測資料進行成果檢核。研究成果顯示，應用PSInSAR同時合併ASI以及SC方法選取之PSCs有最好的測量成效(R = 0.74 to 0.99以及RMSE 2.91 to 6.78cm)，此外，在進行InSAR分析時，應用APS分析得之地層下陷速度(6.65 cm/yr)最為接近GPS的觀測數據(10.23 cm/yr)。因此，本研究認為應用InSAR技術進行地表變形監測時，必須考量大氣中水氣含量可能造成的影響，就本研究的成果發現，整合APS分析可得到較為精確的量測結果。;Land subsidence is a slow and gradual geological hazard, thus consistent, precise, and frequent observations are needed to measure it, and Interferometry Synthetic Aperture Radar (InSAR) could be considered as one suitable monitoring technique. InSAR measurement can be affected by the atmosphere because the microwave signal could be delayed (slowed down) once it passes through the troposphere with high water vapor, leading to the error measurement for monitoring surface deformation up to 10 to 14cm. In addition, recent studies have revealed an increasing trend of global water vapor content, especially in a tropical region such as Indonesia, over the past few decades, indicating that the InSAR measurements are getting more challenging to be applied in the tropical region. For InSAR measurement, the persistent scatterer point candidates (PSCs) selection technique is considered as the essential part to ensure the quality of measurement outcomes, and PSCs, in general, can be selected based on amplitude and/or phase signal product. This research aims to determine an effective InSAR measurement technique regarding the tropospheric effect in the tropical climate region over Jakarta Capital City and East Java Province, Indonesia. This study collected 208 SLC images of Sentinel 1A (descending) product from November 2014 to December 2019 and later validated it with 6 available Global Positioning System (GPS) stations projected into the line of sight (LOS). Several techniques, including Differential Interferometry SAR (DInSAR), Multi-Temporal Interferometry SAR (MTInSAR), and Persistent Scatterer Interferometry SAR (PSInSAR) techniques, were compared in this study, and the atmospheric phase screen (APS) was also considered in the PSCs analysis. Both amplitude stability index (ASI) from amplitude signal products and spatial coherence (SC) from phase signal products were tested with different operations. The PSInSAR techniques with the combination of ASI and SC have the closest agreement (R = 0.74 to 0.99) and accuracy (RMSE 2.91 to 6.78cm) compared to other techniques. Besides, InSAR techniques with APS could explain the land subsidence phenomenon with the closest subsidence velocity of 6.65 cm/year compared with GPS observation of 10.23 cm/year. Therefore, it is essential to incorporate the atmospheric phase screen (APS) to ensure the precision of InSAR measurement, especially in the tropical region where the water vapor content is at its peak. |
