Hydrostratigraphic analysis of the Darling River valley (Australia) using electromagnetic induction data and a spatially constrained algorithm for quasi-three-dimensional electrical conductivity imaging

Abstract

Efforts are being made to improve the irrigation efficiency in the Murray-Darling River Basin, Australia, to deal with predicted rainfall decline and to reduce the incidence of secondary soil and water salinization. The latter commonly occurs as a result of locating water reservoirs upon relic drainage channels. To better manage irrigation, information is required about the spatial distribution of soil type and the stratigraphic features capable of redistributing deep-draining water. In previous research, electromagnetic (EM) induction instruments (e.g. EM38 and EM34) have been used to map the distribution of soil type, hydrological processes (e.g. deep drainage) and vadose-zone features. The aim of this research is to demonstrate how a joint inversion of EM38 and EM34 data, using a one-dimensional spatially constrained algorithm for quasi three-dimensional (quasi-3D) electrical conductivity imaging, can be used to infer the areal distribution of soil types and physiographic and hydrogeological units. The quasi-3D modeling of true electrical conductivity provides a framework for future environmental monitoring and management to mitigate the hydrological processes that drive localized secondary salinization in the study area.