Spatio-temporal characteristics of background noise of GPS continuous stations over Antarctica and Greenland

The vertical land motion from GPS observation is of great significance for studying the changes in glacier mass in polar regions. However, there is significant uncertainty when using GPS to derive the vertical velocity field due to the spatial variability of background noise within GPS time series, which may be induced by surface environmental loads. In this study, we evaluate the influence of the spatial characteristics of background noise on the velocity field and its uncertainty in the GPS continuous observations of Antarctica and Greenland. GPS time series are corrected using surface environmental loading models including atmospheric, non-tidal ocean, and hydrological loads. Additionally, the surface elastic deformation is modeled using gravity satellite observations from GRACE/GRACE-FO. Through the analysis of background noise characteristics in the GPS time series before and after correction, it is concluded that the correction of environmental load deformation can effectively reduce the background noise in GPS time series. The correction of atmospheric and non-tidal oceanic loads resulted in a reduction of the background noise spectrum index at most stations, with approximately 46.1% and 84.6% of the total GPS stations in Antarctica and Greenland, respectively. We find that the improvement in GPS velocity uncertainty estimation is not obvious. In addition, we analyze the spatial heterogeneity of background noise, which reveals that the spatial characteristics of background noise parameters and velocity uncertainties vary with latitude.