QUANTIFYING JOINT PROBABILITIES OF SALINITY AND LITHOLOGY USING AIRBORNE ELECTROMAGNETICS

In modelling distributions of groundwater salinity, the spatial continuity and coverage of airborne electromagnetics (AEM) is indispensable if one considers bulk resistivity to be a function of salinity. However, interpreting bulk resistivity as a proxy for groundwater salinity requires assumptions about the other contributing parameters of resistivity- namely rock matrix resistivity (i.e., lithology). In practice, the lithologic framework can be diverse and cannot be known with complete certainty, particularly in geologically complex basins such as California’s Central Valley. Petrophysically derived salinity distributions informed by AEM often underrepresent this uncertainty and non-uniqueness of the interpretation.

To leverage AEM data in mapping salinity distributions more accurately and comprehensively, we propose a Bayesian approach that incorporates AEM data along with borehole logs, groundwater observations, and any existing prior information to produce continuous probability distributions of lithology and salinity, with measures of uncertainty, in a three-dimensional model domain. By quantifying both primary drivers of the observed bulk resistivity, we can more appropriately leverage AEM to derive three-dimensional probability maps of groundwater salinity. We exemplify this workflow in a variable-salinity environment near the southern coast of Louisiana using a recently collected AEM dataset.