Improved GNSS‐Based Estimates of Terrestrial Water Storage in California Through Suppression of Non‐Elastic Loading Signals

Abstract

Accurate estimation of terrestrial water storage (TWS) variations is important for understanding climate change and managing water resources. Satellite gravimetry has been a unique tool for this purpose, but has inherently low spatial resolution. Dense networks of the Global Navigation Satellite System (GNSS) receivers may provide a higher resolution alternative by observing crustal deformation due to changing water mass loads. However, GNSS observations contain noise and other residual signals from various sources that are unrelated to elastic loading, which is troublesome in TWS applications. In this study, we estimated monthly TWS changes on a 0.25° grid in California from January 2019 to December 2023 using GNSS data, after developing a method to separate elastic loading signals through a Combined Empirical Orthogonal Function (CEOF) analysis with Gravity Recovery and Climate Experiment Follow‐On (GRACE‐FO). The corrected GNSS data were used to estimate TWS change via a regularized linear inversion, revealing clear seasonal variations and long‐term decline from January 2019 to October 2022, followed by an abrupt increase associated with atmospheric river events from October 2022 to March 2023. GNSS‐based estimates are validated with in situ lake water elevation data, and provide additional spatial detail compared to GRACE‐FO mass concentration solutions and hydrological model estimates. These results improve the potential of GNSS‐based TWS estimates in regional hydrological analyses, water resource management, and climate change assessment.