Hydrologic Signals in GNSS Geodesy and Their Implications for Advancing Hydrologic Models

Studies show that GNSS geodetic measurements offer rich information in terrestrial hydrologic storage dynamics, especially signals in deep soil and groundwater that are commonly missing in land surface models (LSMs). In the meanwhile, the growing GNSS geodesy networks in the U.S. have enabled the tracking and estimation of terrestrial storage signals over large domains at a fine spatial/temporal resolution. In this study, we are motivated by two questions: (1) where long-term changes in retrieved mass anomaly data (inverted from geophysical models) at watershed scales come from, and whether the signals are captured by simple groundwater models in LSMs; (2) what added value GNSS geodesy can provide for hydrologic modeling/forecasting at fine temporal/spatial scales (i.e. at daily and watershed scales) especially for extreme events. To answer these questions, we perform high-resolution offline hydrologic model simulations over the Western U.S. with the WRF-Hydro/National Water Model (NWM) for 2006-2020 when there are enough GNSS measurements. We focus on three river basins across different climate regimes: the Russian River (rain dominated), the Upper San Joaquin (transient between rain and snow), and the Selway-Lochsa River (snow dominated). For the first question, we perform comparisons among vertical displacements, different inversion products, model-simulated and observed water storage components, as well as GRACE satellite measurements. For the second one, we examine whether GNSS data provide useful antecedent information at daily scales for flood events, and how the information can be used to improve hydrologic modeling/forecasting.