Spatial characteristics of trends in hydrological load displacement at global GNSS reference stations

Global Navigation Satellite System (GNSS) reference station coordinate time series contain significant environmental loading signals, among which vertical deformations due to hydrological loading are particularly prominent. Many studies have focused on the nonlinear motions and noise characteristics of GNSS stations caused by hydrological loading, whereas the impact of long-term linear trends associated with hydrological loading on GNSS vertical motion has received less attention. In this paper, we analyze the spatial variability in the linear trends of vertical displacements due to hydrological loading, using time series from 1628 GNSS reference stations worldwide (2010–2024). We compare three hydrological loading models (GFZ, GGFC, and IMLS) and evaluate their performance in correcting GNSS station height time series. Our results indicate that the ERA5 (GGFC) and IMLS models produce similar vertical linear trends, both of which closely match trends observed in GNSS vertical displacements. In contrast, the GFZ model exhibits significant spatial discrepancies and shows a weaker correlation with GNSS-derived trends. For example, GFZ predicts substantially smaller uplift trends than ERA5 and IMLS in central North America, west-central Europe, northern Australia, and the Middle East. Conversely, GFZ predicts higher uplift trends than the other two models in the Amazon Basin, south-central Africa, and Southeast Asia. These discrepancies likely stem from differences in each model’s representation of terrestrial hydrological processes, data assimilation techniques, and the inclusion or omission of factors such as groundwater extraction. Overall, this study provides a basis for comparing and validating global hydrological loading models and underscores the importance of selecting an appropriate model in GNSS-based geodynamics research.