同震持續式地下水位變化機制之探討

Abstract

本研究基於台灣地區三口高頻監測井之水位記錄，對於觀測期間內發生之地震，分析水位變化與地震規模、距離、分布等之相關性，並與體積應變互相驗證，探討同震水位變化機制。2004年至2009年觀測期間內，花蓮、壯圍井二號及赤山三號井分別記錄到280、97及47次地震引發的水位振盪現象，其中18、25及5次同時出現同震持續式水位變化。花蓮井只對附近地震反應出持續式水位變化，壯圍二號井能反應遠在1935公里外的汶川地震，而赤山三號則反應台灣東部及南部的地震。同震地下水位變化的發生不僅與震源距離及地震規模相關，也可能受到地質及構造的影響。大幅度的同震水位振盪未必導致持續式水位變化；有些同震水位振盪較小，卻出現持續式水位變化。由於三口井之持續式地下水位變化幅度與水位振盪幅度的相關係數分別為0.51、0.53及0.48，顯示震波振盪難以解釋同震持續式水位變化。土壤液化可以解釋同震水位上升現象，然而花蓮、壯圍井二號及赤山三號井持續式水位變化中僅有17%、 16%及0%出現上升。震動導致地層透水性增加則難以解釋花蓮井及壯圍二號井既觀測到同震水位上升，也出現同震水位下降，更難以解釋赤山三號井同震持續式水位變化速率的差異。至於同震持續式水位變化方向對映體積應變計算的結果，相符的程度為83%、60%及80%。由於在不同地質狀況下，含水層的特性會影響應力重新調整的狀況，可能導致實際上觀測到的同震水位變化可能與簡化條件下錯位模型計算出的體積應變不盡相符。因此，震波振盪難以解釋三口監測井的同震持續式水位變化，同震持續式水位變化可能是靜態體積應變所造成，但計算體積應變的模型仍需改進。

Base on high sampling rate records of three monitoring wells in Taiwan, we analyze the relation between coseismic water-level change and earthquake magnitude, hypocentral distance and distribution, and compare the calculated volumetric strain change with observed coseismic water-level change, in order to discuss possible mechanisms to sustained water-level change. Between 2004 and 2009, the Hualien, Zhuagnwei-2 and Chishan-3 wells recorded 280, 97, 47 oscillatory water-level changes, and 18, 25, 5 sustained water-level changes, respectively. Those earthquakes, which caused sustained water-level change, were located near the Hualien well. While the Zhuangwei-2 well recorded earthquakes as far as the 2008 Wenchuan earthquake 1935 km away from the well. And the Chishan-3 well recorded earthquakes located in the southern and eastern Taiwan. Geology and structure as well as hypocentral distance and earthquake magnitude affect coseismic water-level changes. There are earthquakes that induced large water level oscillations with no sustained water-level change, but some earthquakes induced small water level oscillations with induced sustained water-level change. To the Hualien, Zhuangwei-2 and Chishan-3 wells, the square correlation coefficient between sustained water-level change and oscillation range are 0.51, 0.53 and 0.48, respectively. Therefore, seismic shaking may not account for sustained water-level change. Liquefaction can account for coseismic rises. However, only 17% of sustained water-level changes at the Hualien well, 16% at the Zhuangwei-2 well, and 0% at the Chishan-3 well showed coseismic rises. Enhanced permeability may not account for the coseismic changes in the three wells, because it can’t apply to the Hualien and Zhuangwei-2 well that recorded coseismic rises and falls, and the different rates in sustained water-level changes at the Chishan-3 well. Static strain change can account for 83% of coseismic changes at the Hualien well, 60% at the Zhuangwei-2 well, and 80% at the Chishan-3 well. The inconsistency between calculated strains and observations could be caused by different physical properties of aquifers and the complexity of stress redistribution. Therefore, coseismic sustained water-level changes at the three wells may due to static strain changes, but a simple dislocation model may be insufficient to predict pore pressure change at a specific site.