Spatially refined global terrestrial water storage trends and annual cycles from GRACE and GRACE-FO

Satellite gravimetry has revolutionized the observation of shifts in terrestrial water storage (TWS) in reponse to climate and human activities. Robust detection and attribution of these changes remain a challenge because TWS exhibits strong seasonal variability and is traditionally observed at coarse spatial and temporal resolution. Recent studies have shown that direct regression of Level-1B observations (inter-satellite range data) from the Gravity Recovery and Climate Experiment (GRACE) and its Follow-On mission (GRACE-FO) can substantially improve effective spatial resolution of regression terms, compared to popular monthly mascon products. Applying this framework, we demonstrate that stacked Level-1B regression yields spatially refined estimates of both long-term TWS trends and seasonal amplitude, improving the ability to identify regions where human land and water use alter local freshwater availability. For trend analysis, the enhanced resolution strengthens attribution of storage change to anthropogenic drivers such as irrigation, groundwater extraction, reservoir operations, and land-use change at sub-basin scales. For seasonal characterization, we show that assuming simplified representations of the annual cycle, such as stationary, symmetric, or unimodal seasonality, can enable robust recovery of mean annual TWS amplitude with substantially reduced signal attenuation and leakage. Such refinements are particularly important for applications that depend on accurate annual water budgets, including water-balance-based evapotranspiration estimation and assessments of interannual hydroclimatic variability. The spatial scale at which GRACE satellites can independently observe water resources will continue to improve as additional years of measurements become available.