112-4 Improving Total Water Storage Estimates by Fusion of GPS-Derived Surface Deformation and Hydrological Models across the Contiguous United States

Session: Geophysics in Investigating and Exploring for Mineral, Energy and Groundwater Resources (Posters)

Presenting Author:

Authors:

Abstract:

Abstract Total Water Storage (TWS), which includes all forms of terrestrial water such as groundwater, soil moisture, and surface water, is critical for understanding hydrological dynamics and managing water resources. However, in-situ observations of individual components of TWS, especially groundwater, are spatially sparse, often limited to isolated well measurements that do not capture regional variability or deeper aquifer responses. This makes it difficult to assess system-wide water availability, especially under stress from extraction or climatic extremes. While the Gravity Recovery and Climate Experiment (GRACE) and Gravity Recovery and Climate Experiment Follow-On (GRACE-FO) satellite missions have transformed TWS monitoring through gravity-based mass change estimates, their coarse spatial resolution (~300 km) and temporal data gaps limit their applicability for regional-scale decision-making.  There is a need for high-resolution, continuous TWS estimates that are both physically grounded and temporally complete, particularly in regions experiencing rapid water storage changes. Existing methods have yet to fully integrate dense geodetic observations with model-based storage estimates to address this challenge. To address these limitations, we present an inversion-fusion framework that leverages deformation data and hydrological models. We first estimate TWS anomalies from GNSS-derived vertical land motion (VLM), filtered to isolate elastic responses to hydrological loading. After correcting for long-term tectonic and postglacial rebound signals, we invert VLM data from ~2,000 GPS stations across the contiguous United States. The results reveal strong spatial agreement of inverted TWS signals with those from hydrological models and GRACE in regions of known hydrological variability, such as California and the Southern Plains. Finally, the inverted TWS estimates are fused with outputs from the GLDAS hydrological model using wavelet multiresolution analysis, which preserves low-frequency mass change from GNSS while enhancing spatial detail using model components. The fused product demonstrates improved spatial resolution and temporal completeness relative to GRACE, offering enhanced utility for water monitoring during data outages GRACE. This research provides a scalable, observation-driven framework for augmenting satellite-based TWS records and supporting long-term hydrological assessments.Bottom of Form

Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-11117

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.