区域CORS站坐标时间序列特征分析

发布时间：2018-01-11 20:12

本文关键词：区域CORS站坐标时间序列特征分析　出处：《武汉大学》2017年硕士论文　论文类型：学位论文

【摘要】：近20余年来,全球IGS基准站以及各等级连续运行参考站(CORS)积累的GPS观测数据为大地测量学和地球动力学研究提供了宝贵的数据基础,特别是在区域地壳形变分析、强震位移场检测、全球板块运动模型分析以及冰期后回弹等领域发挥了其他手段难以媲美的重要作用。然而在通过常规GPS数据解算获得站点坐标时常常包含各类非构造噪声的影响,因此对此类噪声的特性、来源方面进行研究,对降低GPS解算的坐标序列中非构造噪声的影响,促进区域断层滑动的监测、地震预报等研究工作的展开,具有重要的科学和应用价值。本文主要针对深圳市CORS参考站的2010～2015年共6年数据,以GAMIT/GLOBK和PANDA两种解算软件所求得坐标时间序列为研究对象,在前人已做过的工作基础之上,本文采取多种手段,研究GPS坐标时间序列中非构造噪声的特性和来源。具体研究内容为:1.对GPS坐标时间序列分析原理和方法做了较为详细的介绍。叙述了各类常用的时间序列分析方法,包括功率谱分析法、最大似然估计法、区域叠加滤波法、主成分分析法(PCA)和Karhunen-Loeve展开法(KLE),并对各自的优缺点进行了对比分析。2.分别使用GAMIT/GLOBK10.5和PANDA处理了深圳市CORS参考站连续6年的观测数据。结果表明,两种静态数据处理软件所得单日解坐标精度均能达到mm级。详细分析了两种数据处理软件的原理、解算过程、精度评定,并对所得坐标时间序列进行了预处理。3.深圳CORS参考站时间序列特征分析。从多个角度利用多种方法对坐标序列展开了深入的研究,分析了不同噪声模型对序列各类参数估值的影响规律,利用主成分空间滤波法提取SZCORS站的共模误差,并对滤波前后坐标序列的参数估值进行了比较分析。计算了地表质量负载对坐标序列的影响大小,并分析了序列垂向分量周期项的起源。确定了深圳CORS站的最优噪声模型,在此基础上构建了参考站运动速度场,对深圳市地壳的形变模式进行了初步探究。4.结合不同机构(SOPAC、JPL、CMONOC)的时间序列产品分析了中国大陆及周边地区的IGS站的时序特征。结果表明,不同机构的结果之间略有差异,但总体处于同一水平。白噪声和闪烁噪声的分布具有较强的海陆差异,靠近海洋的测站比内陆测站具有较大的白噪声和闪烁噪声。经PCA滤波后有色噪声的含量有所下降,基于WN+RW+FN噪声模型的参数估值获得了中国大陆及周边地区IGS站水平和垂直速度分布图。
[Abstract]:The past 20 years, the global IGS base station and the level of continuously operating reference station (CORS) GPS accumulated data provides valuable research data base of Geodesy and geodynamics, especially in regional crustal deformation analysis, earthquake displacement field detection, global plate motion model and the analysis of postglacial rebound, play areas the important role of other means not comparable. However in GPS solution by conventional data often contain noise impact of various types of non structural site coordinates, so such noise characteristics, sources of research, to reduce the influence of non structural noise sequence of coordinates computation of GPS in the monitoring area promote the slip, such as earthquake prediction research, has important scientific and application value. In this paper, Shenzhen City, 2010 ~ CORS reference station data in 2015 a total of 6 years, to GAMIT/GLOB K and PANDA two kinds of software solution obtained by coordinate time series as the research object, on the basis of a has been done in the previous work, this paper adopts many kinds of methods, characteristics and sources of non structural noise of GPS coordinate time series. The main contents are as follows: 1. the GPS coordinate time series analysis principle and method of detailed introduction. Describes the commonly used time series various analysis methods, including power spectrum analysis method, the maximum likelihood estimation method, regional stacking method, principal component analysis (PCA) and Karhunen-Loeve expansion method (KLE), and their advantages and disadvantages were analyzed respectively using GAMIT/GLOBK10.5 and.2. PANDA in Shenzhen CORS reference station of 6 years of continuous observation data. The results show that two kinds of static data processing software the itssolution coordinate precision can reach mm level. A detailed analysis of the two principle of data processing software, solutions The calculation process, accuracy assessment, and the coordinate time series analysis pretreatment.3. Shenzhen CORS reference station time series feature. Using a variety of methods from various angles to research deeply on the coordinate sequence, analyze the influence law of different noise model of the sequence of all kinds of parameter estimation, extraction of SZCORS common mode error by the master station components of spatial filtering method, and the parameter estimation of coordinate sequences before and after filtering are analyzed. The surface quality of the impact load on the coordinate sequence size calculation and analysis of the sequence of the vertical component of periodic origin. To determine the optimal noise model of Shenzhen CORS station, constructed on the basis of reference station velocity field. The deformation mode on the crust in Shenzhen city conducted a preliminary inquiry combined with different.4. (SOPAC, JPL, CMONOC institutions) time series products of Chinese, and the surrounding areas of the IGS station The timing characteristics. The results show that the different results are slightly different, but generally at the same level. The distribution of white noise and flicker noise has a strong difference between land and sea, near the ocean stations than inland stations with large white noise and flicker noise filtered with PCA content of colored noise decreased parameter estimation of WN+RW+FN model based on the noise figure standing horizontal and vertical velocity distribution, China and surrounding area IGS.

【学位授予单位】：武汉大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P228.4

【相似文献】