A data-driven approach to assessing the impact of water harvesting structures on regional water storage in East Africa

Abstract

A small-scale water harvesting structure known as a sand dam has gained popularity across East Africa due to the efforts of non-governmental organizations. A sand dam is a subsurface water reservoir that stores water between sand grains. Stored thus, the water is filtered and protected from evaporation. This study uses remotely sensed data to investigate the impact of these structures on water storage and vegetative growth. The relationship between sand dams and water storage was modeled using a binary sand dam factor, climate data from the Famine Early Warning Systems Network Land Data Assimilation System (FLDAS), and water storage data measured by the Gravity Recovery and Climate Experiment (GRACE) twin satellites. The analysis revealed that GRACE largely fails to detect a statistically significant impact of sand dams on regional water storage. However, analysis of the Normalized Difference Vegetation Index (NDVI) indicated that sand dams have a significant impact on regional vegetation. Vegetative growth is correlated with groundwater levels, indicating that sand dams have a positive impact on water storage albeit on a smaller scale than GRACE can regularly detect. Significantly, this study shows that NDVI data can be used effectively to study small-scale, regional changes in vegetation and water storage. HIGHLIGHTS The impact of small-scale water harvesting structures on regional water resources is explored through spatial generalized linear models of remote sensing products.; Sand dams increase regional vegetation, which suggests groundwater levels are also positively impacted.; Sand dams may have a smaller impact on groundwater levels than believed.; NDVI and, to a lesser extent, GRACE record the impact of sand dams.;

Keywords

• generalized linear model
• principal component analysis
• remote sensing
• sand dam
• spatial autocorrelation