| 摘要: | 台灣位處環太平洋地震帶上，由菲律賓海板塊和歐亞板塊碰撞形成的。因此台灣的地質處於不穩定和破碎的狀態，經常發生地震。此外，該島的氣候是屬於亞熱帶季風性氣候，常遭遇颱風和強降雨，增加了邊坡滑動的風險。作為應對山崩等自然災害的預防措施，近年來山崩監測系統顯著增加。然而，當大範圍山崩在發生時，傳統的監測系統因局限性而無法提供有效的監測。為實現監測預防，需要能提供綜合監測系統的技術，例如遙感探測技術。合成孔徑雷達 (SAR)屬於微波成像雷達，而干涉合成孔徑雷達（InSAR）是基於 SAR的新型地面位移監測技術。 InSAR 方法主要於同一地點的多個 SAR 圖像進行配準。許多研究採用 InSAR 來監測滑坡和地表下陷，但InSAR根據不同的情況及研究使用方式不同，並不存在參數最佳值或是範圍以做參考，在滑坡監測中尤其如此。因此，本研究根據需求應用StaMPS方法來生成 PS-InSAR 和 SBAS 結果，透過阿里山GNSS監測地點，檢查PS-InSAR和SBAS方法的參數，並使用 RMSE 方法與GNSS 數據做對比，藉以取得參數建議值。PS-InSAR 和 SBAS 處理的五個主要影響參數分別是amplitude dispersion、unwrap_grid_size、unwrap_gold_alpha、unwrap_gold_n_win 和 unwrap_time_n_win。在PS-InSAR將參數分別設置為0.47 ? 0.48、? 50、 0.8、 <32、<100，在SBAS中將參數分別設置為，?0.6、? 30m、 0.8、 ? 24、?32。 為測試每個參數適用性，本研究進一步透過桃園義盛驗證上述建議參數， 但建議後續可在其他監測地點採用類似的比較方法來驗證。;Monitoring systems have increased significantly in recent years as a preventative step for dealing with natural disasters like landslides. The use of remote sensing for surface displacement measurement has been established for several decades, and this tool has been enhanced by the use of Synthetic Aperture Radar (SAR). A new technique for land movement monitoring was born from the SAR, known as Interferometric SAR (InSAR). The InSAR method involves co-registering multiple SAR images simultaneously and in place, and many researchers have adopted it to monitor landslide and subsidence. However, InSAR is a black box without a guideline for an optimal range or value of its parameters. Based on the requirements, this study carefully examined PS-InSAR and SBAS methods parameters and then to find the optimal values of parameters at Ali-shan, Chiayi by comparing with GNSS data using RMSE approach. Here we applied the STAMPS approach to generate the optimal range of PS and SBAS parameters. Finally, five influential parameters and suggested values of PS and SBAS processing have been found: 0.47 ? amplitude dispersion ? 0.48, unwrap_grid_size ? 50, unwrap_gold_alpha= 0.8, unwrap_gold_n_win =< 32, and unwrap_time_n_win ? 100 for PS, and ? 0.6, ? 30m, 0.8, ? 24, ?32 respectively for SBAS. Moreover, to evaluate the feasibility of each parameter, we tested those parameters in I-Shan location with GNSS as well. At this 2nd study case we found the our optimal value more effective compared to the default value. However, a similar comparison in other places for verification is suggested to propose relevant optimal range of each parameter. |
