|Abstract: ||2010年2月27日於智利馬烏萊發生了MW 8.8的地震，該地震位於南美板塊與納茲卡板塊的聚合帶上，這也是繼1960年MW 9.5智利地震以來於該地區發生的最大地震，因此受到廣泛的注意。對此地震的研究，先前多以運動學觀點分析此地震的震源破裂過程，本研究將藉由震源輻射能量及靜應力降變化，探討此地震的動態破裂特徵。首先，先以表面波的破裂方向性分析建立此地震單一破裂的斷層參數，結果顯示智利地震屬於雙向破裂，其一破裂朝向N19E方向，破裂長度約300 km，另一破裂朝N196E，破裂長度約100 km，破裂總長約400 km，總震源歷時約200 s，是一個並非完全對稱的雙向破裂；另由位於垂直破裂方向的表面波頻譜節點週期，估算智利地震上揚時間為30.6 s，約是整個震源歷時的0.16倍，也符合過去大地震的觀測(0.15-0.20倍)，此時得到往北破裂速度約1.81 km/s，而往南破裂速度約為1.23 km/s，呈現北快南慢的破裂特徵，這些結果也與一些先前的研究相符，只是本研究所評估的破裂速度仍是偏慢。接下來順推模擬位於垂直破裂方向地遠場P波，調查其多重震源特徵，得到智利地震至少由16個子破裂所組成，其總地震矩(M0)為1.021022 Nm (相當MW=8.6)，總震源輻射能量(ES)為2.171017 Nm，ES/M0比值為2.1310-5，符合一般隱沒帶地震的比值(3-510-5)，但並不是所謂海嘯地震的比值(0.7-3.0 10-6)。破裂過程中顯示第四個子破裂有最大的M0=2.661021 Nm，發生在地震一開始後的第35 s，其歷時25 s，但不具有最大的ES及ES/M0，這與靜應力降有關。由各子破裂震源參數評估其靜應力降變化，顯示靜應力降隨ES/M0比值增加而增加，在60-100 s時達到最高的應力降，顯示破裂時斷層面上的強度分布並不均勻。而智利地震平均靜應力降為62 bars，高於板塊交界地震靜應力降(30 bars)，比對由表面波所得偏慢的破裂速度，顯示智利地震有高靜應力降，但有慢破裂速度的特徵存在，此特徵與2011年MW 9.0日本東北地震相同，目前，此特徵亦是一個待解的問題。|
On February 27, 2010, an earthquake with Mw 8.8 occurred in Maule, Chile, where the Nazca plate is subducting eastward beneath the South American plate. This earthquake was called the 2010 Maule (Chile) earthquake or the 2010 MW 8.8 Chile earthquake. It has been the largest earthquake in the Chile region since the 1960 Mw 9.5 Chile earthquake. Hence, the earthquake came to widespread notice by seismologists. Previous studies investigated rupture features of the 2010 MW 8.8 Chile earthquake based on the view in kinematics. In this study, we made an attempt on examining variations in the radiated seismic energy and static stress drop for further understanding the dynamic rupture features. First, the directivity analysis of surface-wave was used to determine the fault parameters based on the single source with uniform rupture process. Results showed that the event had two rupture directions with different rupture lengths and rupture velocities. A rupture toward N19E had a rupture length of ~300 km with a rupture velocity of ~1.81 km/s and the other toward N196E had a rupture length of ~100 km with a rupture velocity of ~1.23 km/s. The source duration for the earthquake was about 200 s. From the Fourier spectral nodes, the rise time was estimated at 30.6 s, ~0.16 times of the source duration, comparable with previous observations for larger earthquakes (0.15-0.20). In terms of the single source, our results exhibited the rupture features of the 2010 Chile earthquake, which is an event with asymmetric bilateral faulting, fast rupture velocity in the northern rupture and slow rupture velocity in the southern one. Such rupture features are in agreement with previous studies, but the rupture velocities are still slower than those. Subsequently, for investigating the complex rupture features, we adopted a forward P-wave modeling method to infer the multiple sources of the 2010 Chile earthquake by using the stations whose azimuth angles are normal to the rupture direction. Results showed that the earthquake consisted of at least 16 sub-events. The total seismic moment (M0) was 1.021022 Nm, corresponding to MW = 8.6. The estimated radiated seismic energy (ES) was 2.171017 Nm and the ES/M0 ratio is 2.1310-5, consistent with the values in the subduction-zone earthquakes (3-510-5), but not meeting those for tsunami earthquakes (0.7-3.010-6). During the faulting, the fourth sub-event with the duration of 25 s was 35 s later after the onset of the earthquake and had the largest M0 = 2.661021 Nm, but there was no the largest ES and ES/M0 which is probably related to the static stress drop. The estimated static stress drop for each sub-event from their source parameters increased with ES/M0, but not with ES. Up to 60-100 s after the onset, the static stress drop reached to the larger values. This implied uneven strength on the fault plane during the earthquake rupture. The average static stress of the 2010 Chile earthquake was estimated to be 62 bars, higher than those of interplate earthquakes (30 bars). As mentioned above, the 2010 Chile earthquake had a slow rupture velocity from the rupture directivity analysis. As a result, an import rupture feature is that there is high static stress drop along with slow rupture velocity for the 2010 Chile earthquake. This was also similar to the 2011 MW 9.0 Tohoku earthquake. Our results revealed a possible anti-correlation between the static stress drop and rupture velocity. Up to now, this is still an open issue.