The effects of dykes and faults on groundwater flow in an arid land: the Red Sea Hills, Sudan

Abstract

In this study the focus is on a part of the Red Sea Hills of Sudan, an area which suffers from a severe shortage of groundwater. This shortage is partly because the precipitation in this area is very small, from a maximum of only 164 mm year −1 to a minimum of 36 mm year −1. Partly, however, the shortage is related to the generally low permeability of the (mostly Precambrian but partly Phanerozoic) bedrock. The bedrock is, however, dissected by numerous lineaments, mostly faults and basaltic dykes, some of which transport groundwater to the surface in springs and wells. We made field studies of 107 dykes, complemented by Landsat ETM and SPOT image studies of 1419 lineaments interpreted as dykes. Additionally, we made image studies of 1707 lineaments interpreted as faults, fractures and shear zones many of which meet with the dykes at nearly right angles. Many of the dykes are of dense, low-permeability basalt and range in thickness up to 14 m and in length up to several kilometres. The dominant dyke strike is NNW, roughly parallel with the coast of the Red Sea and perpendicular to the topographic slope and the trends of many of the lineaments interpreted as faults. Using the field and image data, as well as a new digital elevation model of the study area, we propose a conceptual model to explain the relationship between faults, dykes and groundwater in the area. In this model the NNW-trending dykes, particularly the long and thick low-permeability dykes, act as barriers for much of the topography-driven groundwater flow. The groundwater collected by these dykes is transported along their margins towards the topographic depressions occupied by the (comparatively) high-permeability E–W trending fault zones. Because these fault zones trend parallel with the inferred hydraulic gradient in the area the faults also tend to collect groundwater. In terms of the model groundwater is thus driven along both dykes and faults to their near-orthogonal intersections. These intersections normally have relatively high fracture-related permeability, along which groundwater is transported towards the surface. This model thus predicts that water wells and springs would be expected at dyke–fault intersections, which is in agreement with the available data indicating that the majority of the springs occur at such intersections.