InSAR和GPS大地测量形变数据反演地震断层滑动的影响特征研究

发布时间：2018-01-03 10:24

本文关键词：InSAR和GPS大地测量形变数据反演地震断层滑动的影响特征研究　出处：《西南交通大学》2017年硕士论文　论文类型：学位论文

【摘要】：地震是断层运动的一种直接表征,强震具有破坏力大、致灾性广等特点,这使得对于探索地震发震机理及防震减灾成为地球科学的重要研究目标之一。InSAR和GPS卫星大地测量技术正成为一种高效的地壳形变监测手段,可为地震研究提供丰富的形变观测数据。尤其是大地测量与地震断层位错理论的结合与成功应用,使得以大地测量形变为约束的震源模型反演成为科学认识地震机理的一种重要手段。然而,利用大地测量形变反演地震断层滑动仍存在精度匹配与可靠性问题,如基于InSAR单一维的LOS方向形变数据对反演断层滑动分布的精度,断层近场InSAR干涉失相关现象对反演滑动分布的影响,基于不同维度和不同台站分布密度的GPS形变数据反演断层滑动分布的精度,以及以大地测量形变数据为约束反演断层滑动分布对深部滑动的敏感性。因此,本文以典型的走滑断层、正断层、逆冲断层为具体实验对象,对上述问题开展系统性的实验论证与定量分析。为验证InSAR单一维的LOS方向形变观测精度对反演断层滑动分布的可靠性,本文反演了不同噪声水平LOS维形变为约束的滑动分布,结果显示,在噪声强度达到60mm时,滑动反演残差为可接受残差的30%至90%,且北、东、竖直三方向形变场解释度均大于95%,说明在常规InSAR观测精度条件下,对于近Mw7.0级及以上强震,使用单一 LOS维形变数据可反演获得可靠的断层滑动分布。为分析断层近场InSAR干涉失相关对反演断层滑动分布的影响,本文以LOS维绝对形变量级为梯度构建不同程度干涉失相关形变场反演滑动分布。结果表明,当大于20%的InSAR LOS方向最大绝对形变参与反演时,反演滑动残差小于1/10绝对平均滑动量,可获得可靠的滑动分布。为评估不同维度GPS形变数据反演断层滑动的精度,本文开展了使用GPS三维形变、北东二维形变、以及单独使用一维形变来反演滑动分布的五组实验。结果表明,使用北东二维形变反演精度略高于GPS三维形变反演精度,均高于单独使用一维形变反演精度。以GPS水平观测精度3mm、竖直观测精度10mm为参考,使用数量充足的任一维GPS形变数据均可反演获得可靠的断层滑动分布。为分析GPS台站分布密度对反演断层滑动分布的影响,论文反演了不同GPS格网密度条件下的滑动分布。结果表明,考虑拉普拉斯平滑约束前提下,当GPS台站均匀分布且数量不小于1/7的离散子断层数量时,可获得可靠的断层滑动分布。以日本2011年地震为例,反演了此次地震不同GPS台站分布密度条件下的断层滑动分布,验证了实例结果与模拟结果具有较高的一致性。为分析InSAR与GPS大地测量形变数据反演断层滑动分布对深部滑动的敏感性,本文反演了深部滑动中心沿深度方向变化的滑动分布。结果显示,常规InSAR观测精度(30mm随机噪声)形变数据对深部滑动反演的探测性相对较低,而高精度(3mm随机噪声)InSAR形变数据对深部滑动反演的探测性相对较高,但略低于数量充足的高精度GPS三维形变(每一维施加3mm随机噪声)对深部滑动反演的可探测性,表明大地测量形变数据的精度相比于维度对深部滑动探测能力的影响相对更大。
[Abstract]:The earthquake is a direct characterization of fault movement, destructive earthquake, disaster widely, which makes the exploration of the earthquake and earthquake disaster reduction mechanism has become one of the important research target.InSAR and GPS satellite geodesy technology, earth science is becoming an effective means of Monitoring Crustal deformation, deformation observation can provide rich data for seismic research. Especially combined with the successful application of geodetic and seismic fault dislocation theory, the geodetic deformation source model inversion constraint became an important means of scientific understanding of earthquake mechanism. However, the use of land seismic deformation inversion of fault slip measurement still exist, such as precision and reliability. Based on the LOS direction of InSAR single dimensional deformation data on the accuracy of inversion of fault slip distribution, fault InSAR near field interference loss phenomenon of slip distribution inversion Influence of different dimensions and different station distribution density inversion of fault slip distribution GPS deformation data based on precision, as well as geodetic deformation data for sensitivity constrained inversion of fault slip distribution on deep sliding. Therefore, this paper take typical slip faults, normal faults, thrust faults as the experimental object, to carry out experiments quantitative and systematic analysis of the above problems. LOS direction verification of InSAR single dimensional deformation observation precision on the reliability of inversion of fault slip distribution inversion, the different noise level of LOS dimensional deformation display results for distribution, sliding constraint, the noise intensity reaches 60mm, the residual slip inversion as acceptable residual 30% to 90%, and the North, East, three direction vertical deformation field explanation degree is greater than 95%, indicating the observation accuracy in conventional InSAR conditions for nearly Mw7.0 and above the earthquake, with a single LOS dimensional shape Data can be estimated by inversion of fault slip distribution reliability. Analysis of fault InSAR near field interference decorrelation effect on inversion of fault slip distribution, based on the LOS dimensional absolute variable level interference loss related deformation field inversion slip distribution gradient construction in different degree. The results show that when the InSAR LOS direction is greater than 20% in absolute maximum deformation inversion, inversion of residual sliding is less than 1/10 the average absolute slip, can obtain the slip distribution reliable. In order to evaluate the different dimensions of inversion of fault deformation data of GPS sliding precision, this paper carried out using GPS three-dimensional deformation, the North East and two-dimensional deformation, single one-dimensional deformation slip distribution inversion to the five set of experiments. The results show that the use of North East Precision two-dimensional deformation inversion is slightly higher than the retrieval precision of GPS 3D deformation, were higher than the accuracy of one-dimensional deformation inversion alone. The GPS level of the 3mm measurement accuracy, vertical visual The accuracy of 10mm for reference, the use of a sufficient number of any dimension GPS deformation data can be derived from the fault slip distribution reliability. For the analysis of GPS station distribution density on the inversion of fault slip distribution inversion, the sliding grid under the condition of different GPS density distribution. The results show that considering the Laplasse smoothing constraints when the number of discrete sub fault GPS stations evenly distributed and the number of not less than 1/7, can obtain reliable fault slip distribution. With the 2011 earthquake in Japan as an example, the inversion fault slip distribution of the earthquake in different GPS stations distribution density conditions, verify the results consistent with the simulation results is very high. For the analysis of InSAR and GPS geodetic inversion of fault slip distribution measurement of deformation data on deep sliding of the sensitivity of the inverted slip distribution center sliding deep changes along the depth direction. The results show In the conventional InSAR observation accuracy (30mm random noise) detection of deformation data of deep slip inversion is relatively low, and high precision (3mm random noise) InSAR detection of deformation data of deep slip inversion is relatively high, but slightly lower than the three-dimensional deformation of high precision GPS sufficient number (each Weishijia random 3mm the noise on the detectability) deep slip inversion, show that the geodetic deformation data accuracy compared to the dimensions of the influence of deep sliding detection ability is relatively large.

【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P315.7

【参考文献】