Analysis and Discussion on the Optimal Noise Model of Global GNSS Long-Term Coordinate Series Considering Hydrological Loading

Yuefan He,Guigen Nie,Shuguang Wu,Haiyang Li

doi:10.3390/rs13030431

Abstract

The displacement of Global Navigation Satellite System (GNSS) station contains the information of surface elastic deformation caused by the variation of land water reserves. This paper selects the long-term coordinate series data of 671 International GNSS Service (IGS) reference stations distributed globally under the framework of World Geodetic System 1984 (WGS84) from 2000 to 2021. Different noise model combinations are used for noise analysis, and the optimal noise model for each station before and after hydrologic loading correction is calculated. The results show that the noise models of global IGS reference stations are diverse, and each component has different optimal noise model characteristics, mainly white noise + flicker noise (WN+FN), generalized Gauss–Markov noise (GGM) and white noise + power law noise (WN+PL). Through specific analysis between the optimal noise model and the time series velocity of the station, it is found that the maximum influence value of the vertical velocity can reach 1.8 mm when hydrological loading is considered. Different complex noise models also have a certain influence on the linear velocity and velocity uncertainty of the station. Among them, the influence of white noise + random walking noise is relatively obvious, and its maximum influence value in the elevation direction can reach over 2 mm/year. When studying the impact of hydrological loading correction on the periodicity of the coordinate series, it is concluded whether the hydrological loading is calculated or not, and the GNSS long-term coordinate series has obvious annual and semi-annual amplitude changes, which are most obvious in the vertical direction, up to 16.48 mm.

Highlights

The global International Global Navigation Satellite System (GNSS) Service (IGS) base station location time series accumulated in the past 20 years have provided valuable basic data for geodesy and geodynamic research
In the above formula, y0 is the intercept; v is the linear trend in unit per year, where a year is defined as 365.25 days; ti is the epoch of the GNSS time series; ak, and bk are the amplitude of periodic signals and f k is the corresponding frequency; g j and Tg j is the offset and corresponding epoch, respectively; r refers to the residual time series, which can be various combinations of noise models
As far as the global GNSS single-day solution time series are concerned, if no measures are taken to reduce the spatial correlation of the station, it can be concluded that WN + FN

Summary

Introduction

The global IGS base station location time series accumulated in the past 20 years have provided valuable basic data for geodesy and geodynamic research. Many scholars have launched a lot of research on it, including periodic signal analysis, stochastic model research, spatial filtering method and surface loading model research, etc., in GNSS coordinate time series [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16] Analyzing it can get the precise movement trend of the station, so as to further study the internal driving mechanism of the station movement. A large number of researchers generally believe that the characteristics of GNSS coordinate series noise model are WN+FN [18,19,20]

Methods

Results

Discussion

Conclusion