摘要
大地测量反演利用多种高精度地壳形变数据,对地震事件进行参数计算和机制解释.随着基于统计理论的贝叶斯方法受到越来越多学者的关注,贝叶斯联合反演已成为大地测量反演领域的一个研究热点.本文以2016年中国台湾美浓地震为研究对象,提出了一种将非常快速模拟退火算法(Very Fast Simulated Annealing,VFSA)与基于单位权中误差无偏估计值的计算公式为目标函数相结合的新反演方法,用于反演美浓地震的最优断层几何参数,并以此为约束,使用von Karman正则化的贝叶斯滑动分布反演方法,利用GNSS和InSAR数据联合反演了兼具走滑和倾滑的美浓地震滑动分布,进一步丰富了该方法在不同类型地震中的应用.研究结果显示,发震断层兼具左旋走滑和逆冲倾滑的特征,主要滑动分布在地下9~16 km之间,最大滑移量达到0.54 m,反演得到的最大矩震级为M_(W)6.47.本研究采用的方法和模型在保证结果质量的同时显著提高了参数反演的计算效率,并且采用的无偏估计量更能反映观测数据的真实质量,避免了有偏估计所导致的高估模型拟合效果的问题,易于推广到其他非线性智能搜索算法中使用.研究结果不仅对于深入理解该地区的应力积累和释放机制具有重要意义,而且为进一步研究该地震的成因机制和评估未来地震风险提供了重要的参考依据.
Geodetic inversion employs high-precision crustal deformation data for the parameter estimation and mechanism interpretation of seismic events.As Bayesian methods based on statistical theory garner increasing attention,Bayesian joint inversion has emerged as a focal point in this field.This study investigates the 2016 Meinong earthquake in Taiwan,China.We propose a novel method to determine the optimal fault geometry,which integrates the Very Fast Simulated Annealing(VFSA)algorithm with an objective function based on an unbiased estimator for the variance of unit weight.Constrained by these geometric parameters,we then jointly invert for the coseismic slip distribution of the Meinong earthquake—resolving both its strike-slip and dip-slip components—using GNSS and InSAR data via a Bayesian framework with von Karman regularization.This application to an earthquake with a mixed-faulting mechanism further enriches the utility of the methodology.The results reveal a seismogenic fault characterized by both left-lateral strike-slip and reverse dip-slip motion.The primary slip is concentrated at a depth of 9~16 km,with a maximum slip of 0.54 m and a moment magnitude of M_(W)6.47.The proposed model significantly enhances the computational efficiency of parameter inversion while ensuring robust results.Furthermore,the use of an unbiased estimator more accurately reflects the true quality of the observational data,avoiding the overestimation of model fit commonly caused by biased estimators,and can be readily generalized to other nonlinear search algorithms.These findings are vital not only for advancing the understanding of regional stress accumulation and release but also for providing a crucial reference for investigating the seismogenic mechanics and assessing future seismic hazards.
作者
孙漳林
王乐洋
席灿
黄英杰
SUN ZhangLin;WANG LeYang;XI Can;HUANG YingJie(School of Surveying and Geoinformation Engineering,East China University of Technology,Nanchang 330013,China;Key Laboratory of Mine Environmental Monitoring and Improving around Poyang Lake of Ministry of Natural Resources,East China University of Technology,Nanchang 330013,China)
出处
《地球物理学报》
北大核心
2025年第8期3085-3102,共18页
Chinese Journal of Geophysics
基金
国家自然科学基金项目(42174011)
东华理工大学研究生创新基金(DHYC-202402)联合资助。
