Comprehensive Analysis of Mass Loading Effects on GPS Station Coordinate Time Series Using Different Hydrological Loading Models

This paper comprehensively analyzes the effect of surface mass loading on GPS height time series in terms of weighted root mean square (WRMS), annual timescale, velocity and noise properties. Considering the complexity of the hydrological process, we adopt the same atmospheric loading and nontidal ocean loading estimated from the TUGO-m barotropic model (ATMMO) and different hydrological loadings estimated from the Global Land Data Assimilation System (GLDAS) and the Global Land Surface Discharge Model (LSDM). The results indicate that the combined ATMMO & GLDAS outperforms the combined ATMMO & LSDM in reducing the WRMS of GPS height times series due to the annual phase asynchrony between GPS height and ATMMO & LSDM. Furthermore, cross-wavelet transform-based (XWT-based) semblance analysis is adopted to quantitatively express the annual phase relationship between GPS height time series and different loading model combinations. The results show that (1) mass loading is the main driving force for the annual fluctuation in GPS height, and (2) the annual term of ATMMO & GLDAS better matches the GPS height time series than ATMMO & LSDM. Consequently, it is concluded that the LSDM may not work well in Eurasia with its current accuracy. Finally, we investigate the impacts of the mass loading correction on site velocity and noise properties. For most sites, the velocity change can reach 0.1 mm/year after the loading correction, indicating that the influence of mass loading should not be ignored in the construction of millimeter-level earth reference frames. However, for some sites, the velocity uncertainty is increased due to undesirable changes in noise properties, suggesting that using the mass loading model to correct GPS height time series is potentially feasible but not sufficient.