Fracture aquifers identification in the Zou basin (West Africa) using Remote Sensing and GIS

Abstract

The riparian communities of the Zou basin of West Africa rely heavily on drinking water supplied by fractured aquifer systems. This study aims to provide accurate fracture maps and derived products (fracture density, coincidence map and cross-points) using Landsat 8 (visible and infrared bands) and PALSAR DEM datasets with borehole data collected from the national integrated database. Digital image processing techniques through image enhancement and directional Sobel filters application were used to analyse Landsat 8 and PALSAR data. Results show that the main orientations of fractures are N10-N20, N90-N100 and N180 for the visible, 10-N20, N40-N50, N90, N140 and N180 for the infrared and for the PALSAR DEM, they are N10-N20, N90-N100 and N170-N180. The fracture length overall interval is 0.15 to 21.2 km and the cumulative length of each fracture map is adjusted to power law with characteristic of exponent respectively 0.86, 0.91 and 0.96. The chi square (χ^2) test of cumulative length versus class frequency reveals a significant relationship for the visible and PALSAR fracture maps. Although, the calculated χ^2 is greater than the theoretical chi square for the infrared, there is a strong and positive correlation for the three fracture maps (0.94, 0.99 and 0.98). More than 50% of the fracture cross-points (CP) occurs with the North-South (NS) orientations. The coincidence map analysis shows that the PALSAR fracture map has the highest ratio of 0.88, proving it should prioritize over the visible and infrared maps. High flow (Q> 5 cubic m/h) boreholes located near the N10, N20 and N180 main orientations are more than 50% confirming their significant contribution to aquifers’ productivity. The spatial and statistical analysis of these fractured aquifers are critical to improve drinking water access and water resources planning for the basin communities.