摘要
自2007年来,格陵兰地区逐步建立了50多个GNSS连续观测站用以监测冰川质量变化导致的地壳回弹,至今已累计了超过15 a的坐标时间序列,这使得确定坐标时间序列的噪声特性成为可能。本文选取了格陵兰53个GNSS观测站自2008-2018年共10 a的坐标时间序列,利用功率谱分析和极大似然估计对其噪声特性进行了分析,并基于最优噪声组合估计了格陵兰测站的位移速度及其不确定度。利用贝叶斯信息准则作为评估依据的结果表明,闪烁噪声和白噪声+闪烁噪声的组合可以解释33%的测站噪声,幂律噪声可以解释52%的测站噪声,其余测站可用随机游走或一阶高斯马尔科夫噪声来解释。在利用极大似然估计分析测站坐标时间序列时,对于对97%的测站位移分量的速度估值而言,不同噪声模型的影响小于0.1 mm/a,最大为0.31mm/a,但不同的噪声模型会导致测站位移速度的不确定度的最大估值和最小估值之间存在1.3~7.1倍的差异。因此准确地确定测站坐标序列的噪声模型对于准确估计测站位移速度及其不确定度是十分重要的。
Since 2007,more than 50 GNSS continuous observation reference stations have been established in Greenland to monitor crustal rebound caused by changes in glacier quality,and has accumulated coordinate time series for more than 15 years,which makes it possible to determine the noise characteristics of coordinate time series.In this paper,the coordinate time series of 53 Greenland GNSS stations from 2008 to 2018 are selected,their noise characteristics are analyzed by power spectrum analysis and maximum likelihood estimation,and the velocity and its uncertainty of Greenland stations are estimated based on the optimal noise combination.The results based on Bayesian information criterion show that the combination of flicker noise and white noise plus flicker noise can explain 33%of station noise,while power law noise can explain 52%of station noise,and other stations can be explained by random walk or first-order Gaussian Markov noise.When using maximum likelihood estimation to analyze station coordinate time series,although the choice of noise model has little influence on the estimation of velocity of station component,for the estimated velocity of 97%of the station displacement components,the influence of different noise models is less than 0.1 mm/a,and the maximum is 0.31 mm/a,but different noise models will lead to the difference between the maximum station displacement velocity uncertainty is 1.3 to 7.1 times.Therefore,it is very important to accurately determine the noise model of the station coordinate sequence for accurately estimating the station displacement velocity and its uncertainty.
作者
范峥研
姚宜斌
张豹
FAN Zhengyan;YAO Yibin;ZHANG Bao(School of Geodesy and Geomatics,Wuhan University,Wuhan 430079,China)
出处
《测绘地理信息》
CSCD
2024年第1期42-49,共8页
Journal of Geomatics
基金
国家自然科学基金(42074035)
湖北省自然科学基金(2021CFB319)
中央高校基本科研业务费专项资金(2042022kf1202)。