Resolving groundwater-driven surface deformation in upper east Ghana using downscaled GRACE and InSAR

Ghana's semi-arid Upper East Region (UER) is flooded annually due to rainfall and dam spillage. Yet the long-term impacts of hydroclimatic changes on groundwater and their corresponding effects on land surface stability in the region are unknown. We combine GRACE data and InSAR to study groundwater-deformation relationships in the region from June 2018 to August 2021, following two key preparatory steps. First, we assess the regional hydrological consistency of the GRACE signal over the small, data-sparse region (8,842 km2). Then, using Geographically Weighted Regression, we downscale GRACE-derived groundwater to a 10-km spatial resolution to enable a detailed spatial analysis of groundwater deformation relationships at five selected locations. Our results reveal increasing groundwater storage trends (ˆβ = 18.8 − 37.0 mm year−1, sˆβ = 4.19 − 5.75 mm year−1; sˆβ : standard error of the estimate) at the locations, corresponding to land uplift (ˆβ = 2.2 − 6.6 mm year−1, sˆβ = 0.1 mm year−1). The trends in the two variables are highly correlated (r = 0.80 − 0.95, p < 0.001, sˆβ = 0.007 − 0.021). Our study adopted a framework to combine Earth Observation (EO) methodologies for groundwater-deformation analysis in data-sparse regions at at a finer spatial scale than is possible with native GRACE observations alone, providing critical information that directly supports national policies (e.g., for Ghana, the One Village, One Dam initiative and the construction of the Pwalugu Multipurpose Dam). While our framework enables an EO only approach, we find that combining EO with ground-based measurements remains vital for comprehensive hydrological monitoring.