以吉布斯隨機場整合DInSAR和GNSS資料於地表三維變位之建構

Derivation of Three-Dimensional Surface Displacement Field from the Integration of DInSAR and GNSS Data by Gibbs Random Fields

林筱凡(Hsiao-Fan Lin)

指導教授 ： 韓仁毓

國立臺灣大學/工學院/土木工程學研究所/碩士(2020年)

https://doi.org/10.6342/NTU202000392

摘要

隨著空間資訊科學快速發展，發展兼顧效率以及成本的高精度三維空間資訊獲取技術已成為此領域之關鍵課題。合成孔徑雷達差分干涉技術可快速捕捉大範圍之細微變位訊號，然而其所獲得之視線方向位移並無法完整反映真實三維變位。而全球衛星導航系統可提供接收站穩定且精確的三維坐標時間序列，雖測站分布稀疏，仍能作為高精度且穩定空間約制的依據。本研究利用一期兩組的合成孔徑雷達干涉影像解算多期升軌與降軌視角方向的相對變位量。在各期影像不足以回復三維變位的情況下，引入全球導航衛星系統之三維觀測時間序列。以全球導航衛星系統之觀測資料作為邊界條件，透過保持視角方向變位量梯度場的二維泊松方程，對視角方向變位量進行轉換改正，藉此提升LOS變位量的絕對精度。將各期LOS變位量進行線性擬合以減少因影像品質不佳、地震等突發災害導致變位遽增的情況。最後，在影像稀少而不足解的情況下，透過內插全球導航衛星系統之三維變位觀測，估算區域地表三位變位速度場，並藉由吉布斯隨機場理論，以內插之三維變位速度場和兩軌之視角方向變位量進行區域三維變位回復，透過隨機場模型以避免內插行為過度影響成果解算精度，達到地表完整三維變位建構。
根據研究成果顯示，利用保持視角方向變位量梯度場的改正模型，可有效提升視角方向變位量的絕對精度。吉布斯隨機場模型能夠快速的解算三維變位，且有效抑止因內插行為導致誤差過大的情況。經實驗驗證，本研究以吉布斯隨機場模型所建構之區域三維變位能夠在觀測條件稀少的情況下足夠完整地回復三維變位，且相較於最小二乘法各期三維變位的時間序列更加穩定且一致，能作為另一個低成本、高效率建構區域地表三維變位場的可行方案。

關鍵字

合成孔徑雷達差分干涉 ； 全球導航衛星系統 ； 地表三維變位 ； 吉布斯隨機場 ； 資料融合

並列摘要

Driven by rapid progress in spatial information science, developing an efficient and low-cost 3D spatial data acquisition technique is becoming a key issue. Differential Interferometric Synthetic Aperture Radar （DInSAR） is capable of capturing detailed displacement signals of ground surface of a wide area. However, the line-of-sight （LOS） displacements derived from the DInSAR technique do not provide an accurate representation of the actual 3D deformation field. In this study, multiple sets of SAR images will be used to derive the LOS displacements time series. The 3D observations from the Global Navigation Satellite System （GNSS） technique will be used to provide highly accurate spatial constraints as boundary conditions to adjust LOS displacements by preserving the gradient field of LOS displacements. Finally, probability theory and Gibbs random field will be introduced to integrate all available information and a complete approach for delivering 3D displacement of ground surface will then be established. The experimental results show that despite the insufficient interferograms, the 3D displacement can be resolved by this construction methodology with higher accuracy than least square method. A reliable observation and analysis solution with a high accuracy and good coverage will then become available for the deformation monitoring tasks.

並列關鍵字

Differential Interferometric Synthetic Aperture Radar (DInSAR) ； Global Navigation Satellite System (GNSS) ； 3D Deformation Analysis ； Gibbs Random Field ； Data Fusion

參考文獻

Anzidei, M., Boschi, E., Cannelli, V., Devoti, R., Esposito, A., Galvani, A., Serpelloni, E., Pietrantonio, G., Riguzzi, F., Sepe, V., and Serpelloni, E., 2009. Coseismic deformation of the destructive April 6, 2009 L'Aquila earthquake (central Italy) from GPS data, Geophysical Research Letters, 36:L17307.

Baek, J., Kim, S.W., Park, H.J., Jung, H.S., Kim, K.D., and Kim, J.W., 2008. Analysis of ground subsidence in coal mining area using SAR interferometry, Geosciences Journal, 12(3):277-284.

Bechor, N.B., and Zebker, H.A., 2006. Measuring two‐dimensional movements using a single InSAR pair. Geophysical Research Letters, 33:L16311.

Bekaert, D.P.S., Segall, P., Wright, T.J., and Hooper, A.J., 2016. A network inversion filter combining GNSS and InSAR for tectonic slip modeling, Journal of Geophysical Research: Solid Earth, 121(3):2069-2086.

Calais, E., Han, J.Y., DeMets, C., and Nocquet, J. M., 2006. Deformation of the North American plate interior from a decade of continuous GPS measurements, Journal of Geophysical Research, 111(B6).