以合成孔徑雷達干涉法研究台灣之地殼變形

The case studies of the crustal deformation in Taiwan based on SAR interferometry

黃孟涵(Mong-Han Huang)

指導教授 ： 胡植慶

國立臺灣大學/理學院/地質科學研究所/碩士(2006年)

https://doi.org/10.6342/NTU.2006.00125

摘要

本論文主要探討合成孔徑干涉雷達(InSAR)原理以及在中台灣地區與台南地區地殼變形研究上的應用。此外，本研究亦利用全球衛星定位系統(GPS)觀測資料計算研究區域的應變率(strain rate)以及旋轉率(rotation rate)，接著再利用三維彈性半空間錯移模型(3D dislocation model)模擬斷層面狀態，藉以推估變形之機制與討論。
發生於民國八十八年九月二十一日之集集地震，其地震矩規模達到7.6，對中部台灣造成重大的災情，且造成車籠埔斷層長達100公里的破裂。本論文利用差分合成孔徑雷達干涉(D-InSAR)進行對於集集地震之震前(pre-seismic)、同震(coseismic)、以及震後(post-seismic)分別進行觀測。在震前部分，本研究並沒有觀察到特別的前震位移，但是在同震部份，透過其結果顯示出明顯的在斷層下盤的同震位移，並伴隨著局部地區的地表下沉現象，在震後部份，則觀測到斷層上盤局部區域有超過10公分的地表抬升，顯示著地震發生後車籠埔斷層仍持續著向上抬升，但值遠小於同震(局部超過8公尺)。此外，在中部沿海地區，則有觀測到持續的地盤下降的現象，並主要集中在員林、溪湖、以及二林等地區，且這個現在在集集地震後更趨於穩定。在其他遙測資料方面，本研究利用內政部土測局施測之GPS同震位移資料，探討同震應變及旋轉與地震之關係。其結果指示著最大的壓縮應變方向與車籠埔斷層相互垂直，且整體而言斷層的位移除了逆衝之外，在西側有左移北側有右移的現象，且最大的應變幾集中在斷層的西北角，也造成最大的破壞。在數值模擬方面，我假設斷層為一斷坡-基底滑脫(ramp- décollement)斷層，以GPS觀測結果進行逆推，其結果與原觀測直想當接近，更加強此假設的正確性。
台南台地位於台灣島西南方，屬於推測的變形前緣之最前端，台地本身東西側分別被西傾的後甲里斷層、東傾的台南斷層所通過，因此造成了台南台地開花上拱(pop up)的構造。本論文利用於1996至2000之間拍攝的六張SAR影像進行干涉，以瞭解台南台地的地表變形狀況。研究結果顯示台南台地中央平均每年以大約12.5公厘的速度沿視衛星方向上升，且上升區域與台南台地的形狀相符。而以GPS資料的觀測結果顯示出，台南台地相對於沿海地區平均每年以12±4公厘的速度向西北方移動，此外，五次施測的精密水準測量亦觀察出台南台地平均每年向上抬升大約14公厘。如果以GPS觀測的水平速度，和InSAR觀測的視衛星方向速度加以做座標轉換，則可以得到與水準相同的垂直抬升量，因此可以利用此結果討論台南地區的變形型態、機制，以及可能產生的地震災害。

關鍵字

合成孔徑雷達干涉 ； 地殼變形 ； 應變率計算

並列摘要

I apply the D-InSAR technique to monitor the active crustal deformation both in Tainan and Central Taiwan by using ERS SAR images during 1996-2000 (Tainan) and 1999-2000 (Central Taiwan). The Mw 7.6 Chi-Chi earthquake occurred on 21th September, 1999 in central Taiwan and a ~100 km surface rupture along the Chelungpu fault was observed. I apply InSAR technique to measure the coseismic and postseismic deformation in the footwall and hanging wall of the Chelungpu fault. The 2-pass method is used to obtain interferograms from DEM data and 5 ERS-1/2 SAR images passing central Taiwan between February 1999 and August 2000. The result of D-InSAR (Differential-InSAR) reveals the significant coseismic deformation on the footwall area of the Chelungpu fault along the line of sight (LOS) of ERS satellites. The coseismic slant range displacement (SRD) on the footwall of the Chelungpu fault demonstrates a difference of 32.6 cm from the coastal area to the west side of the Chelungpu fault. This observation shows the maximum coseismic uplift is close to the west side of the fault. The postseismic interferogram shows no significant postseismic slip on the footwall of the Chelungpu fault, however a SRD of ~20 cm is observed on Shinshou tableland located on the hanging wall. Besides, a SRD of ~10 cm is also observed along the E-W reverbanks of the Choshui river. A significant land subsidence area was with total subsidence ~10 to 20 mm during a period of one or two months detected near the Changhua County both in coseismic and postseismic InSAR pairs. In addition, we recalculate the coseismic deformation revealed by GPS observations on the footwall of Chelungpu fault into the SRD in order to compare with the deformation pattern of D-InSAR. I also try to exploit 3-D dislocation models to estimate movement of the fault plane modified from a ramp-décollement model suggested by Johnson et al. (2004). Thus the coseismic deformation inferred from this dislocation model can be transferred into the coordinate of the SRD. The results of the simulated intergerograms show the same trend as those from D-InSAR. However, low coherence of SAR interferograms in the southern part of the study area in Chunghua County limits the accurate comparison with the predicted models.
The Tainan tableland is located in-between a blind fault in the west and the Houchiali Fault in the east, thus the Tainan tableland can be interpreted as a pop-up structure in a fold-thrust belt at active tectonic margin. Interferometric processing of six SAR images reveals the average SRD as ~12.5 mm/yr, and it increases from west edge of Tainan tableland and decreases across the Houchiali fault. The campaign-mode GPS data set from 1999 to 2003 indicate an average horizontal movement of 12±4 mm/yr in the direction of N44°W for the Tainan tableland with respect to western coastline. Furthermore 5 times precise leveling surveys across Tainan tableland over a period of 2 years show an uplift rate of ~14 mm/yr for the benchmarks on the tableland. The horizontal strain rates calculated by GPS horizontal data showed the extensional on the Tainan tableland with the rate of ~6 μstrain/yr and compressional regime on the coast line and the eastern Tawan lowland with the rate of 5 μstrain/yr. By combining with the horizontal velocity of GPS data and the SRD of D-InSAR we transfer the SRD into vertical deformation and discuss the deformation pattern and seismic hazards of Tainan area.

並列關鍵字

SAR interferometry ； crustal deformation ； strain rate estimation

參考文獻

Abdelfattah, R., J. M. Nicolas, Topographic SAR interferometry formulation for high-precision DEM generation, IEEE Trans. Geosci. Rem. Sens., 40, 2415-2426, 2002.

Amelung, F., C. Oppenheimer, P. Segall, and H. Zebker, Ground deformation near Gada’ale Vocano‚ Afar, observed by radar interferometry, Geophys. Res. Lett., 27, 3093-3096, 2000.

Bos, A. G., W. Spakma, and M. C. J. Nyst, Surface deformation and tectonic setting of Taiwan inferred from a GPS velocity field, J. Geophys. Res., 108(B10), 2458, doi: 10.1029/2002JB002336, 2003.

Bürgmann, R., P. A. Rosen, E. J. Fielding, Synthetic aperture radar interferometry to measure Earth’s surface topography and its deformation, Annu. Rev. Earth Planet. Sci., 28, 169-209, 2000.

Chang, C.-P., T. Y. Chang, J. Angelier, H. Kao, J.-C. Lee, and S.-B. Yu, Strain and stress field in Taiwan oblique convergent system: constraints from GPS observation and tectonic data, Earth Plan. Sci. Lett., 214, 115-127, 2003.