Study of infiltration in a Sahelian gully erosion area using time-lapse resistivity mapping

Abstract

Observing that concentrated runoff destroys indurate and impermeable surface horizons to form gullies on Sahelian slopes, we investigated whether these gullies are preferential places for deep infiltration and groundwater recharge processes. The primary aim of this study is to determine if resistivity mapping is an appropriate method to use for locating recharge zones from the surface. The study area, in northern Burkina Faso, is a typical (1 ha) gully erosion area located at the outlet of an 82-ha catchment with solonetz soils and a crystalline basement. Taking advantage of a long dry season followed by a short rainy season, we made use of a time-lapse approach to carry out electrical resistivity mapping and monitor apparent resistivity variations that occurred in the soils during the rainy season, between June and September. We made nine apparent resistivity maps in the year 2000 and two in January and March 2001. To monitor expected infiltration and percolation to depths of 5 m or more, we laid out Wenner array profiles with an inter-electrode spacing of 5 m. The time-lapse mapping was also controlled with: (i) neutron probe measurements; (ii) resistivity measurements on outcrops during infiltration tests; (iii) electrical resistivity logging in auger holes. Geophysical results showed that the apparent resistivity parameter can either decrease (typical case) or increase (unexpected case) after a rain. Neutron probe measurements indicated that infiltration varies within a few decimeters even at the centre of the main gully. Using one dimensional (1D) modelling based on resistivity variations monitored during infiltration tests, we concluded that apparent resistivity variations are linked to the presence of carbonate in the soils. When soluble carbonates are present, the resistivity of the infiltrated layer varies from 220 Ω m (dry state) to less than 5 Ω m (wet state), bringing about a decrease in apparent resistivity value for the 5m spacing. In the absence of carbonate, resistivity varies from 1500 to 180 Ω m, but produces an increase of the apparent resistivity value for the same spacing. Consequently, we found time-lapse apparent resistivity mapping to be an efficient way to delineate certain soil properties. It also provided additional information about punctual observations. However, our results have led us to conclude that the 5-m inter-electrode spacing is too large to monitor this type of shallow infiltration phenomenon and that the effect of temperature on resistivity should be considered when comparing maps over the period of a few months. Furthermore, this type of survey should be controlled using electrical loggings in auger holes, or electrical soundings in order to get a better understanding of in-depth resistivity variations. Finally, this survey indicated that deep infiltration processes are not occurring below the gully situated on the slope. Further studies are required downstream to identify the location of groundwater recharge in Sahelian crystalline contexts.