Estimation of specific yield and its variability by electrical resistivity tomography

Abstract

Specific yield (Sy) is defined as the average volume of water that can be drained, per unit surface of aquifer per unit drop of head. This parameter is of critical relevance for groundwater resources assessment, but its estimation suffers from numerous difficulties, including spatial variability, hysteresis, dependence on depth to water table, or delayed drainage. As a result, no widely accepted method is available for its estimation. Here we show that most of these difficulties can be overcome by using time-lapse electrical resistivity tomography (ERT) to map water content variations in response to phreatic level fluctuations. We applied the method to a synthetic example and to a real site in Argentina. Results demonstrate that the approach is robust and provides a picture of how, how much and where water is being released from the soil when aquifer heads drop, which is the ultimate goal of Sy concept. ERT produced some oscillations in estimated water contents, which we attribute to inversion artifacts. Still, the error analysis and the synthetic example suggest that the impact of spurious oscillations tends to cancel out during estimation of Sy, which supports the robustness of the method. Estimated Sy is sensitive to the petrophysical law. If the slope of this law is inadequate, calculated values of Sy may be biased, but their variability patterns are well captured by the method. What is clear is that Sy is a spatially and temporally variable parameter, whose estimation is difficult. ERT represents a possible avenue to evaluate Sy and its variability.