全球GPS基准站坐标序列噪声模型估计及其对站速度的影响分析

Global GPS reference station coordinate time series noise model estimation and analysis of its influence on station velocity

针对全球定位系统(GPS)基准站坐标序列噪声模型的复杂性及其对站速度估计的不利影响,选取全球范围内的1443个GPS基准站,采用3种常用组合噪声模型并增加广义高斯-马尔科夫(GGM)与白噪声(WN)组合模型来估计GPS站速度参数,并利用改进贝叶斯信息量准则精确识别最优噪声模型。对不同噪声模型假设下的站速度参数进行分析的结果表明:GPS基准站的噪声特性存在多样性,3种常用组合噪声模型都有一定占比,GGM+WN模型占比虽小,但该模型背景下站速度精度较高;而随机游走特性(RW)则会放大站速度不确定度的影响,精确识别两种噪声特性对于站速度精确估计具有重要意义;此外,噪声模型的改变对站速度影响显著,在E、N方向的影响较大,U方向的影响较小。

In view of the complexity of the Global Positioning System(GPS)reference station noise model and its adverse effects on station velocity estimation,1443 GPS reference stations worldwide are selected.The four noise models,including 3 commonly used combined noise models and the Generalised Gauss-Markov(GGM)plus White Noise(WN),are used to estimate the velocity parameters of GPS stations.Finally,the improved Bayesian information criterion is used to accurately identify the optimal noise model.Analyzing the station velocity parameters under different noise model assumptions,the results show that:The noise characteristics of GPS reference stations are diverse,and the three common combined noise models all have a certain proportion.Although the proportion of GGM+WN model is small,the speed accuracy of GPS reference stations under the background of this model is higher.The Random Walk(RW)characteristic will amplify the influence of the station speed uncertainty,so it is important to identify the two kinds of noise characteristics accurately for the accurate estimation of the station speed;In addition,the change of the noise model has a significant impact on the station speed,especially in the E and N directions,and less in the U direction.