Towards a Simple and Pragmatic Approach to Deriving Groundwater Salinity Mapping From Airborne Electromagnetics

Based on Archies Law, there is not a direct relationship between AEM bulk conductivity and groundwater salinity as other variables such as porosity (lithology), moisture content and cementation come into play. These variables tend to be poorly characterised at the catchment scale, making any derivation of groundwater salinity mapping from AEM problematic. For a study area in the Darling River floodplain of western New South Wales, Australia, the relevant datasets of AEM depth slice conductivity, downhole induction conductivity, pore fluid salinity and pumped groundwater salinity were compared, with the goal of deriving simple AEM conductivity surrogate values for groundwater salinity thresholds. The comparison between pore fluid salinity and downhole induction conductivity highlighted the well-known effect of increased bulk conductivity due to increased clay content. Also, the pore fluid salinity from many shallow samples was relatively high compared to the measured downhole (and AEM) conductivity, due to low moisture contents. The best relationship was between AEM depth slice conductivity and the average pore fluid salinity for the depth slice. For the depth slices dominated by the target aquifer (22-61m), the R² for the linear regression is reasonable (~0.67-0.83). A percentile ranking approach was then used to derive from these datasets the AEM surrogates for groundwater salinity thresholds. The pumped groundwater salinity data was not necessarily directly comparable with the AEM-derived salinity mapping. This was because the screened intervals could encompass multiple AEM depth slices, different groundwater sampling methods were used and the pumped salinities are biased to the fresh groundwater zones in the target aquifer. It is argued that the simple mapping approach gives inherently conservative estimates of groundwater salinity from the AEM data.