云南GNSS时间序列共模误差提取与站点形变分析  

Common Mode Error Extraction and Deformation Analysis of GNSS Time Series in Yunnan

作　　者：徐良叶 邵德盛[1] 吴学群[2] 牛甜 韩逍 XU Liang-ye;SHAO De-sheng;WU Xue-qun;NIU Tian;HAN Xiao(Yunnan Earthguake Agency,Yunnan Kunming 650200,China;Faculty of Land Resources Engineering,Kunming University of Science and Technology,Yunnan Kunming 650093,China)

机构地区：[1]云南省地震局,云南昆明650200 [2]昆明理工大学国土资源工程学院,云南昆明650093

出　　处：《防灾减灾学报》2022年第2期57-65,共9页Journal of Disaster Prevention And Reduction

基　　金：GNSS数据分析与预报创新团队;云南省地震局地震科技专项基金(2021ZX02)。

摘　　要：在提取GNSS时间序列共模误差之前要剔除本地效应较强的站点,怎么剔除这些站点,并没有一个定量的准则,针对这一问题本文提出根据PCA结果与KEL结果对应的空间响应差值的2倍标准差来剔除异常站点;其次这些被剔除的站点时间序列里仍然包含有用的构造运动信息,而如何从本地效应强的站点时间序列里提取信息方面的研究还较少,因此本文通过KLE方法来去除本地效应的影响,然后利用小波变换和自适应卡尔曼滤波对去除本地效应后的坐标时间序列进行去噪处理,结合2018年通海5.0级地震,分析地震前站点形变信息。实验结果表明:本文中的方法为更准确地从GNSS时间序列中去除共模误差提供了参考,对更好地提取GNSS时间序列中的信息有借鉴意义。Sites with strong local effect should be eliminated before extracting common mode error of GNSS time series.There is no quantitative criterion for how to eliminate these sites.To solve this problem,this paper proposes to eliminate abnormal sites according to the 2 times of the standard deviation of the difference of the spatial response PC1 between PCA results and KEL results.Second time series are eliminated the site still contains information useful tectonic movement,on how to extract from the effect of strong local site time series information of fewer problems,In this paper,local effects are removed by KLE method,and then wavelet transform and adaptive Kalman filter are used to denoise the coordinate time series after local effects are removed,combined with Ms 5 earthquake in 2018,analysis of deformation anomalies before the earthquake site information.The experimental results show that the method presented in this paper provides a reference for more accurate removal of common mode errors from GNSS time series,and It has reference significance for better extracting information in GNSS time series.

关 键 词：GNSS时间序列 共模误差 去噪 形变 

分 类 号：P315.72[天文地球—地震学] P228.4[天文地球—固体地球物理学]