| GPS和InSAR同震形变约束的尼泊尔M_W7.9和M_W7.3地震破裂滑动分布 |
| |
| 引用本文： | 谭凯,赵斌,张彩红,杜瑞林,王琪,黄勇,张锐,乔学军.GPS和InSAR同震形变约束的尼泊尔M_W7.9和M_W7.3地震破裂滑动分布[J].地球物理学报,2016,59(6):2080-2093. |
| |
| 作者姓名： | 谭凯 赵斌 张彩红 杜瑞林 王琪 黄勇 张锐 乔学军 |
| |
| 作者单位： | 1. 中国地震局地震研究所, 地震大地测量重点实验室, 武汉 430071;
2. 中国地质大学(武汉)地球物理与空间信息学院, 行星科学研究所, 武汉 430074;
3. 地壳运动监测工程研究中心, 北京 100036 |
| |
| 基金项目： | 地震行业科研专项(201308009),中国地震局地震研究所所长基金(IS201506220),国家自然科学基金(40974012,41274027)联合资助. |
| |
| 摘 要： | 2015年4月25日尼泊尔爆发MW7.9地震,继而引发5月12日MW7.3级余震,GPS、InSAR监测到震源区及周边大范围同震形变.本文以国内外的GPS和InSAR同震形变为约束,考虑喜马拉雅断裂带岩石圈垂向分层和横向差异的影响,反演主喜马拉雅逆冲断裂在这次主震和余震中破裂面形状和滑动分布.结果显示,主震从USGS确定的震中位置向东偏南延伸100km以上,破裂地面迹线与主前缘逆冲断裂迹线基本一致.破裂面倾角约7°~11°,大部分破裂集中在深度8~20km,同余震分布深度一致.主震最大滑动量约6.0~6.6m,位于14km深处.余震破裂集中在震中附近30km范围内,填补了主震东部破裂空区,最大滑动约3.6~4.6 m,位于13km深.深度20km以下基本没有破裂.地壳介质不均匀性对破裂滑动分布的影响较大,介质不均匀模型的观测值不符值比各向同性弹性半空间模型降低10%以上.本文地震破裂模型特征与地震反射剖面、以及根据震间期大地测量数据反演的喜马拉雅深部蠕滑剖面极其相似.跨喜马拉雅断裂剖面的震间形变量与地震破裂滑移量直接相关.以此推算,尼泊尔中部大震原地复发周期在300年以上.
|
| 关 键 词： | 尼泊尔地震 GPS InSAR 同震形变 破裂滑动分布 |
| 收稿时间： | 2016-01-18 |
Rupture models of the Nepal MW7.9 earthquake and MW7.3 aftrershock constrained by GPS and InSAR coseismic deformations |
| |
| TAN Kai,ZHAO Bin,ZHANG Cai-Hong,DU Rui-Lin,WANG Qi,HUANG Yong,ZHANG Rui,QIAO Xue-Jun.Rupture models of the Nepal MW7.9 earthquake and MW7.3 aftrershock constrained by GPS and InSAR coseismic deformations[J].Chinese Journal of Geophysics,2016,59(6):2080-2093. |
| |
| Authors: | TAN Kai ZHAO Bin ZHANG Cai-Hong DU Rui-Lin WANG Qi HUANG Yong ZHANG Rui QIAO Xue-Jun |
| |
| Institution: | 1. Key Laboratory of Earthquake Geodesy, Institute of Seismology, Chinese Earthquake Administration, Wuhan 430071, China;
2. Institute of Geophysics & Geomatics, China University of Geosciences, Wuhan 430074, China;
3. National Earthquake Infrastructure Service, Beijing 100036, China |
| |
| Abstract: | During the Nepal MW7.9 earthquake on 25 April 2015 and the MW7.3 aftershock on 12 May, large surface deformations were recorded by seismographs and geodetic instruments. The inversion of teleseismic waveforms enabled a quick gain of knowledge about spatio-temporal processes of rupture including the onset, rise time, speed of rupture, as well as the location, geometry and seismic moment, providing relatively imprecise solutions of distribution of slip. InSAR and GPS data collected in the near field can be used to constrain fault geometry and slip distribution at a finer resolution. However, previous geodetic models usually provided distributions of main shock slip in either an elastic half space or a spherically symmetric Earth. These solutions varied significantly in slip pattern. It is assumed that radial stratification and lateral heterogeneity of Earth medium might introduce considerable biases onto the geodetic inversions of slip distribution. This paper reanalyzes all available measurements of ground deformation associated with the main shock and its largest aftershock based on a radially layered and laterally heterogeneous Earth model.
We used GPS-derived co-seismic displacements at a total of 33 sites in Nepal and south Tibet, together with two swaths of the ALOS-2 InSAR interferograms for constraining fault geometry and slip distribution. Firstly, using a genetic algorithm, we solved for rupture parameters (centroid location, depth, length, width, strike, dip, rake and slip magnitude) for a uniform slip model and estimated their statistic confidence intervals. Then, we divided the best-fitting fault plane further into numerous sub-fault patches and estimated slip values of these patches adopting a nonnegative least squares algorithm, which revealed in detail the heterogeneous features of slip on this optimal rupture plane. In the inversions of slip distribution, we minimized misfits to surface displacements while maintaining smoothness of slip across neighboring patches based on a trade-off curve of data misfits vs. slip roughness.
The modeled surface displacements associated with the main shock are all directed to the epicenter, consistent with a simple thrusting mechanism. The co-seismic horizontal offset was 1.89 m at the KKN4 GPS station in Nepal where the maximum observed displacement was documented. In south Tibet, 100~400 km away from the epicenter, the modeled co-seismic horizontal displacements varied from several millimeters to approximately half a meter, consistent with the largest horizontal offset of 54 cm recorded by the J041 station. The modeled offsets of the largest aftershock match the observed ones up to 2 cm at GPS sites as far as approximately 200 km away from the epicenter and InSAR line-of-sight (LOS) range offsets of up to 1 m are fitted within their formal errors.
Under the same constraint conditions, the slip amount inverted in the elastic half space (the homogeneous model) is usually larger than that for a layered crust (the inhomogeneous model), although they appear to be little changes in the distribution of slip. Furthermore, the homogeneous model yields an overall misfit to the data better that the inhomogeneous model does. The former misfits are estimated only 67% for GPS, and 84% for InSAR of the latter ones. Compared with the homogeneous model, the inhomogeneous model reduces, in particular, the postfit residuals by more than 10% of the GPS displacements on the hanging wall north of Kathmandu. These GPS sites including those in Tibet are sensitive to the effect of laterally inhomogeneous media on slip behavior.
The modeled fault geometry shows a reverse fault striking 285° from the north, dipping 7°~11° to the north, and extending east-southeasterly over 100 kilometers from the USGS epicenter. The geometry of the Main Himalayan Thrust inferred from the coseismic deformation is similar to that as illustrated by a seismic reflection profile obtained in the INDEPTH project and is in agreement with what is adopted for a creeping segment of the Main Himalayan Thrust beneath the Himalaya in fitting GPS velocities. The modeling of coseismic and interseismic deformation shows that the India plate plunges into the Tibetan Plateau at a sub-horizontal dip angle.
Our model of slip distribution of the main shock reveals a main patch with local peak slip about 6.0~6.6 m at 14 km depth. The main shock geodetic moment is estimated to be 7.65×1020 N·m, corresponding to the moment magnitude MW7.9, in accordance with the GCMT solution. Most of slip induced by the largest aftershock is localized within a circular region of 30 kilometers in radius around its epicenter, filling in the gap (area of no any slip) left by the main shock. This aftershock is characterized by a main patch with local peak slip of about 3.6~4.6 m at 13 km depth and its geodetic moment is estimated to be 1.12×1020 N·m, corresponding to MW7.3. In the 2015 event, most of slip is confined to the brittle crust at depths of 8 to 20 km, and slip below 20 km depth is barely observed inconsistent with the depth distribution of aftershocks. We analyzed the stress loading triggered by the mainshock according to our slip model. The calculation of Coulomb stress distribution indicates that the majority of the aftershocks is restricted to the areas that have experienced an enhanced stress loading of 0.2~0.8 MPa due to the main shock. Our model suggests that the 2015 earthquake in Nepal has released stresses accumulated in interseismic period due to the underthrusting of the India plate, and hypothesizes that the recurrence of such a large earthquake is more than 300 years in central Nepal where the slip rate of the seismogenic thrust fault is about 20 mm·a-1. |
| |
| Keywords: | Nepal earthquake GPS InSAR Coseismic deformation Rupture slip distribution |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《地球物理学报》浏览原始摘要信息 |
| 点击此处可从《地球物理学报》下载免费的PDF全文 |
|
