Probabilistic Categorical Groundwater Salinity Mapping From Airborne Electromagnetic Data Adjacent to California's Lost Hills and Belridge Oil Fields

Abstract

Growing water stress has led to emerging interest in protecting fresh and brackish groundwater as a potential supplement to water supplies and raised questions about factors that could affect the future quality of fresh and brackish aquifers. Limited well infrastructure, particularly in regions where elevated salinity has led to limited historical groundwater development, hinders traditional mapping of salinity distributions through groundwater sampling. This paper presents a quantitative salinity mapping approach of the upper 300 m using high-resolution, regionally comprehensive resistivity models derived from Bayesian inversion of an airborne electromagnetic survey adjacent to the Lost Hills and Belridge oil fields in the southwestern San Joaquin Valley of California. Using local water quality observations as an interpretational foundation, a probabilistic approach yields maps of fresh, saline, and brackish groundwater while quantifying joint uncertainty inherited from the geophysical data and interpretational relations. Saline and fresh regions are mapped with relatively high confidence in many locations, while areas of lower confidence, particularly at depth, can be mapped as their most probable salinity category while reflecting the relative uncertainty in the interpretation. These maps identify a stratified salinity structure, where saline water commonly occurs in the surficial aquifer overlying fresher groundwater in the Tulare aquifer, separated by regional confining clay layers. Downgradient of unlined surface water diversions, recharge of imported surface water results in relatively fresh groundwater throughout the depth of investigation. Plain Language Summary Salinity, a measure of salt concentration, can impact the potential use of groundwater. Salinity can be difficult to map using wells alone, especially where groundwater development has been limited in the past. Dissolved salts are excellent conductors of electricity, and resistivity measurements are commonly used to estimate salinity where wells are not available. Uncertainty in estimating salinity from resistivity data can be introduced into the interpretation by two sources: (1) how well we can estimate salinity from resistivity and (2) how accurately we can measure resistivity. We present a method for mapping salinity that accounts for both sources of uncertainty using regional-scale, high-resolution airborne geophysical data, leading to spatially comprehensive 3-D maps of likely fresh (low salt) and saline (high salt) groundwater. The method is applied in the southwestern San Joaquin Valley of California adjacent to the Lost Hills and Belridge oil fields. Shallow groundwater is shown to have relatively high salinity in much of the study area, and a clay layer protects lower salinity water in the underlying Tulare aquifer. Leaking surface water canals have imported relatively fresh water into the area, resulting in fresher groundwater downgradient of the canals.

Stratigraphic and structural controls on groundwater salinity variations in the Poso Creek Oil Field, Kern County, California, USA

“…There is a focus in mapping TDS near San Joaquin Valley (SJV) oil fields, where oil production and agricultural water usage exist in proximity (Taylor et al 2014;Davis et al 2018). These efforts, which determine TDS concentration using a mix of geochemical and geophysical approaches, have focused both at the basin scale (Kang and Jackson 2016;Gillespie et al 2017;Metzger and Landon 2018;McMahon et al 2018;Kang et al 2020) and in specific oil fields (Gillespie et al 2019;Stephens et al 2019;Ball et al 2020).…”

Exploring the Model Space of Airborne Electromagnetic Data to Delineate Large‐Scale Structure and Heterogeneity Within an Aquifer System

“…The airborne electromagnetic (AEM) method can play an important role in the study and management of groundwater systems, by providing a model of the subsurface that captures the large‐scale variation in lithology or sediment type throughout a survey area. A typical workflow for the interpretation of AEM data involves the compilation and review of all relevant ancillary data from the survey area (e.g., driller's logs, geophysical logs, and geological cross‐sections); the initial processing of the acquired data, to remove low quality data (Auken et al., 2018); the inversion of the data, to recover the electrical resistivity distribution in the subsurface and define a resistivity model (Farquharson & Oldenburg, 1993; Viezzoli et al., 2008); and the transformation of the electrical resistivity model into a three‐dimensional (3D) model of lithology or sediment type with incorporation of available well data (Ball et al., 2020; Christensen et al., 2017; Foged et al., 2014; Gunnink & Siemon, 2015; He et al., 2014; Knight et al., 2018). The inversion step introduces significant uncertainty into the workflow due to the non‐uniqueness in AEM inverse problems (Christensen & Lawrie, 2012).…”

Watershed zonation approach for tractably quantifying above-and-belowground watershed heterogeneity and functions

“…Clair et al, 2015). Airborne geophysics-particularly airborne electromagnetic (EM) surveys-was originally developed for mineral exploration, but is now increasingly used for water-resources applications (e.g., Barfod et al, 2018;Ball et al, 2020). Despite these advancements, there are still challenges in associating watershed functions to heterogenous watershed patterns.…”