Simulating reactive transport of selenium coupled with nitrogen in a regional-scale irrigated groundwater system

•Selenium fate and transport is simulated in an agricultural groundwater system.•Fate and transport of nitrogen species also simulated.•Model is corroborated using extensive set of field data.•Presence of marine shale governs the extent of selenium contamination and loading.•Model can be used to explore remediation strategies for selenium contamination.

SummaryThe contamination or deficiency of Selenium (Se) has become a critical issue in watershed systems worldwide during the past half-century, with either elevated or deficient Se concentrations in groundwater, surface water, soils, and associated vegetation. Regardless of the reason for concern, there is a basic need for tools for use in assessing baseline concentrations and mass loadings in large-scale environmental systems and for exploring remediation strategies. This paper presents the application of UZF-RT3D, a multi-species variably-saturated reactive transport model for groundwater systems, and its recently developed Se agricultural and chemical reaction module to a regional study site (50,600 ha) along the irrigated Arkansas River Valley in southeastern Colorado. The study region has been monitored for Se contamination during the previous decade, with all segments of the Arkansas River impaired with respect to Se. The Se reaction module accounts for near-surface Se cycling due to agricultural practices, oxidation-reduction reactions, and sorption, and includes a nitrogen cycle and reaction module due to the dependence of Se transformation and speciation on the presence of nitrate (NO3). Of prime importance is the role NO3 plays in oxidation of residual Se from marine shale, which is prevalent throughout the study region, along with its inhibition of Se chemical reduction to less toxic forms. The model is corroborated against groundwater solute concentrations, mass loadings of solutes to the Arkansas River, relationships between solutes within the groundwater system, and overall regional statistics of groundwater solute concentration. Results indicate that the model is successful in replicating the major spatial and temporal patterns in Se and NO3 contamination and transport, hence providing a tool that can be used with confidence to explore remediation schemes.