Integrating GNSS and GRACE Observations to Investigate Water Storage Variations Across Different Climatic Regions of China

China has a vast territory, complex hydrogeological structures, and a diverse and variable climate. However, the sparse distribution of Global Navigation Satellite System (GNSS) stations presents a challenge to ensuring the accuracy of hydrological inversions across such a large area using GNSS observations alone. To address this, we integrate GNSS and Gravity Recovery and Climate Experiment (GRACE) observations to jointly invert changes in terrestrial water storage (TWS) across different climate zones in China. Using a checkerboard load mass source, we first test the feasibility of the joint inversion method. The results show that, compared to the GNSS-only inversion, the root mean square error (RMSE) of the joint inversion is reduced by 23.25% (~18.64 mm). This demonstrates that the joint inversion method captures more detailed spatiotemporal variations of TWS, confirming its feasibility and advantages. Moreover, the synthetic test results of the chessboard load mass sources with different station distributions reveal that as the average station spacing within the study area decreases, the TWS variations obtained from the joint inversion method become more reliable. We then apply this method to actual GNSS vertical displacements, initially focusing on the seasonal variations in TWS across six climatic zones in China. The results indicate that the joint inversion method more effectively captures regional TWS variations, particularly in areas with limited precipitation (e.g., northwest China). However, due to the limitations in station coverage, the improvements are somewhat constrained, although significant potential remains. In contrast, in regions with abundant precipitation (e.g., southwest China), TWS variations obtained through different data exhibit greater consistency. Finally, we analyze the relationship between TWS variations and hydrometeorological data (i.e., precipitation and temperature) across the six climate zones. The results show that there is a significant correlation between TWS variations and hydrometeorological data, and the former shows a time lag of 1-2 months relative to the latter, a phenomenon that may be attributed to the complex transmission mechanism of TWS. This research highlights the potential of space geodetic techniques (e.g., GNSS and GRACE) in hydrological climatology and hydrogeodesy, providing valuable insights into the interactions between TWS variations and climate dynamics.