Predicting environmental factors governing regional groundwater uranium contamination

Abstract:

Groundwater serves as a critical resource to fulfill domestic, agricultural, and industrial needs in semi-arid and arid regions, like the Central Valley, California. Aquifer use will be further exacerbated by climate change, and naturally occurring contaminants, like uranium (U), threaten groundwater quality. In this work, we seek to improve our predictive capacity by integrating statistical learning with chemical equilibria-based modeling to study the biogeochemical controls on groundwater U concentrations within the Central Valley. Twenty-three environmental parameters from a statewide groundwater geochemical database and publicly available maps of soil and aquifer physicochemical properties were used in a random forest regression to predict groundwater U concentrations. After 5-times repeated, 10-fold cross validation, the final model explained 71% of the total variance in the validation dataset. The effects of single and paired predictors on groundwater U content was visualized using partial dependence and accumulated local effects plots. We find that, in order of importance, groundwater calcium, nitrate, and sulfate concentrations, soil pH and clay content (weighted average between 0-2m depth) are the most important predictors of groundwater U concentrations. Importantly, we find that regional groundwater U exceedances of 30 µg L^-1 are strongly dependent on the abundance of uranyl-calcium-carbonato species, which is further supported by modelled aqueous U speciation and surface complexation outputs. These uranyl-calcium-carbonato complexes likely form in the vadose zone and are transported to shallow aquifer zones with recharge water, which illustrates the critical role of water infiltration and recharge conditions on groundwater quality.