长期地心非线性运动反演与时间序列分析  

Long-term nonlinear geocenter motion inversion and time series analysis

作　　者：王鑫林 王虎[2] 任营营 焦静 党亚民[1,2] WANG Xinlin;WANG Hu;REN Yingying;JIAO Jing;DANG Yamin(College of Geodesy and Geomatics,Shandong University of Science and Technology,Qingdao,Shandong 266590,China;Chinese Academy of Surveying and Mapping,Beijing 100830,China;College of Surveying and Geo-Informatics,Tongji University,Shanghai 200092,China)

机构地区：[1]山东科技大学测绘与空间信息学院,山东青岛266590 [2]中国测绘科学研究院,北京100830 [3]同济大学测绘与地理信息学院,上海200092

出　　处：《导航定位学报》2024年第4期198-210,共13页Journal of Navigation and Positioning

基　　金：科技部重点研发计划项目(2022YFB3903902,2022YFB3903900);国家自然科学基金项目(42274044);地理信息工程国家重点实验室、自然资源部测绘科学与地球信息技术重点实验室联合资助基金项目(2023-01-01);中国测绘科学研究院科研业务费项目(AR2403,AR2202)。

摘　　要：地心运动会对以国际地球参考框架(ITRF)为基准进行的大地测量工作产生影响,是推进毫米级参考框架建立和维持的关键。基于国际全球导航卫星系统服务组织(IGS)第三次重处理(Repro3)精化结果,通过网平移法反演了2012—2022年长期的地心运动,为解决地心运动时间序列中高频噪声混淆的问题,采用自适应噪声完备集合经验模态分解(CEEMDAN)方法进行降噪重构。研究结果表明:Repro3数据可取得更优的反演结果,与美国德克萨斯大学空间研究中心(CSR)发布的结果相比,X、Y、Z方向的误差分别为1.50、1.45和3.97 mm;CEEMDAN方法在保留原始时间序列内在特性的前提下,能更有效地削弱时间序列中的噪声,相比于经验模态分解(EMD)方法,3个方向序列经过该方法处理后信噪比分别提升了23.92%、4.36%和4.74%。另外,对长达十年的反演结果进行分析,发现了一些非季节性周期项与不同程度的长期周期项;同时X、Y、Z 3个方向的时间序列均具有明显的年周期项与半年周期项,其中年周期项振幅最大,分别为1.31、3.39和2.55 mm。Geocenter motion has an impact on geodesy work based on the international Earth reference frame(ITRF),which is the key to promote the establishment and maintenance of the millimeter-level reference frame.Based on the refinement results of the third reprocessing(Repro3)of the International Global Navigation Satellite System Service(IGS),the long-term geocentric motion from 2012 to 2022 was inversed by the net translation method.In order to solve the problem of high-frequency noise confusion in the geocentric motion time series,the complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)method was used for noise reduction reconstruction.The results showed that the Repro3 data could achieve better inversion results.Compared with the results published by the Center for Space Research(CSR)of the University of Texas,the errors in the X,Y and Z directions were 1.50,1.45 and 3.97 mm,respectively.The CEEMDAN method could effectively weaken the noise in the time series while retaining the intrinsic characteristics of the original time series.The signal-to-noise ratios of the three directions were increased by 23.92%,4.36%and 4.74%,respectively,compared with the sequences processed by the empirical mode decomposition(EMD).In addition,some non-seasonal periodic terms and varying degrees of long-term periodic terms were found by analyzing the inversion results of up to ten years.At the same time,the time series in the directions of X,Y and Z all had obvious annual and semi-annual periodic terms,among which the annual periodic terms had the largest amplitude,which were 1.31,3.39 and 2.55 mm,respectively.

关 键 词：第三次重处理(Repro3) 地心运动 经验模态分解(EMD) 时序分析 交叉验证 

分 类 号：P228[天文地球—大地测量学与测量工程]