利用海嘯模擬討論大地震之震源

Abstract

有別於地震波記錄，海嘯波模擬可以提供另一種獨立的觀測於地震震源的討論。本研究中針對三個大地震事件（分別為2009年薩摩亞地震、2010年智利大地震、2011年日本東北地震），藉由不同的斷層面的破裂模式，模擬海嘯波波型與實際記錄的海嘯波形比對，驗證和推斷真實的破裂面位態，討論大地震震源機制。本研究所使用的模擬程式為COMCOT，而實際海嘯觀測記錄來自佈放於大洋中的深海壓力計記錄以及測高衛星資料。根據美國地質調查所（USGS）所解算的震源機制解，2010年智利大地震的海嘯模擬結果與實際觀測波形相符合，證明地震資料解可信，其描述此地震的斷層面上有兩個主要滑移量分布且大多位於沿海地區；另外，本研究也嘗試使用測高衛星資料討論沿海地區的海嘯共振信號，其最大震幅在地震發生後3小時達1.5米。在 2011年日本東北地震案例中，海嘯模擬的結果顯示最大滑動量的區域非常接近海溝，本研究除了考慮不同斷層面的破裂模式，也同時採用靜態的同震位移分佈和動態破裂隨時間演變的模型，藉以討論地震破裂過程對海嘯波初始條件的影響。另外，海嘯也能夠有助於判別實際斷層面破裂的位態。關於2009年薩摩亞地震震源中逆衝斷層的討論，首先藉由海嘯模擬的結果，本研究確認不僅發生在外部隆起的正斷層引發海嘯，隱沒處的逆斷層也是產生海嘯的重要來源；此外，本研究利用海嘯波模擬區分出此系列地震中的正斷層，其位態為朝向東北傾斜的斷層面。本研究也檢驗GPS站觀測之同震位移資料和計算的同震錯動量相互比較，其結果顯示在2009年薩摩亞地震中，附近島嶼上的GPS的資料無法幫助分辨真正的斷層面位態，這不確定性可以用來幫助思考，為何前人使用不同的資料（地震、GPS等），對這個地震的卻有不同的區域應力的解讀。

Parallel to the seismological observations, recordings of tsunami waves can provide additional clues for the seismic source. In this study, with the aid of tsunami simulations using the tsunami modeling package – COMCOT, we study three big earthquakes and their focal characteristics: the 2009 Samoa Earthquake, the 2010 Chile Earthquake, and the 2011 Tohoku-Oki Earthquake. By testing varied rupture models of fault planes in simulation, we can verify and infer the orientation and geometry of the real fault planes. The observed records are from the ocean bottom pressure recorders and the altimetry satellites. The simulated results based on the USGS solution for the slip distribution of the 2010 Chile Earthquake fit well with the observations, supporting the seismic source model that has two asperities and mostly located in the coastal region. In addition, we confirm that the altimetry satellite can successfully detect the shelf resonance signal of the tsunami, whose height can be up to 1.5 meters even 3 hours after the earthquake. In the case of the 2011 Tohoku-Oki Earthquake, the simulated tsunami waveforms show that the maximum slip located much closer to the trench. Except for using different static slip distributions, we also adopt the rupture model involving with time evolution to discuss the effect on the initial sea surface height of tsunami from the progress of rupture. Also, tsunami waves can help to distinguish the actual fault plane. In the 2009 Samoa Earthquake, firstly, by the means of tsunami simulation, it shows that not only the normal fault, but also the thrust should be taken as to provoke the tsunami waves. Furthermore, we can discriminate the real orientation of the normal faulting as the one dipping to the north-east. Besides, we calculate the coseismic dislocation at the near-by GPS stations as well, and show that the modeling of coseismic dislocation cannot help to tell apart the real fault plane. This experience is taken to examine the former researches who explained the stress transfer in the 2009 Samoa Earthquake by means of GPS dislocation analysis.