Earth observation and land surface model applications for improved groundwater resources monitoring over East Africa

Groundwater resources are an important supplier of fresh water to East Africa’s communities. Resource degradation, resulting from climate variability and population growth, places many livelihoods at risk of water shortages. Monitoring of regional groundwater behaviors and associated climate and anthropogenic controls remains a limited practice over the region, a hindrance to effective resource management. In this thesis, I apply remote sensing and land surface modeling approaches to increase the state of knowledge of groundwater behaviors and resource monitoring over East Africa. In Chapter 1, a methodology is developed aimed at improving the estimation of groundwater storage changes from the Gravity Recovery and Climate Experiment (GRACE) satellites over East Africa (EA), a region characterized by significant surface water variability. Results document strong correlation between GRACE groundwater estimates and in-situ groundwater observations (Spearman’s ρ = 0.6), after systematic data processing that limits the effect of spatial leakage due to surface water bodies. Additional analysis, comparing groundwater storage behaviors with satellite precipitation, permits a regional view of groundwater characteristics as well as regional recharge behaviors, which can enhance effective resource management. In Chapter 2, a multi-objective calibration approach (using GRACE total water storage anomalies and streamflow) is applied to improve land surface model parameterization and simulation of baseflow and groundwater depth within the Community Land Model (CLM)). Model simulation of groundwater depth and baseflow is shown to significantly improve, formulating a basis for further hydrologic experiments, including anthropogenic impacts. In Chapter 3, I use the calibrated model to examine the hydrologic impacts of groundwater abstraction. Specifically, the CLM groundwater generation scheme is modified to include ‘human’ abstraction fluxes. Results show a decline in baseflow and groundwater depth with increased groundwater abstraction, while evaporation, especially during dry periods, increases. The results suggest that a deeper groundwater table reduces capillary influences, limiting the potential to affect soil moisture and Hortonian runoff, while causing induced recharge from soil layers. Results show that induced recharge does not exceed abstractions with increased groundwater withdrawals, suggesting unsustainable use of the aquifers