Regional selenium cycling in an irrigated agricultural groundwater system: Conceptualization, modeling, and mitigation.

摘要

Selenium (Se) is an element that occurs naturally as a trace constituent in geologic formations and associated soils and, although an essential nutrient for animals and humans, can prove detrimental to health at high concentrations. Over the previous decades, the presence of either deficient or elevated concentrations of Se in groundwater, surface water, and associated plants and cultivated crops has emerged as a serious issue in many regions of the world, including the United States, northern and western Europe, the Middle East, and East Asia.;Regardless of the nature of concern regarding Se, whether concentrations are deficient or elevated in water supplies and cultivated crops, there is a basic need for a thorough description of the movement and chemical processes of Se within a dynamic soil-aquifer system influenced by agricultural practices, and for the development of numerical simulation tools that allow these processes to be simulated in assessing baseline conditions and exploring remediation best-management practices (BMPs). While the individual processes controlling Se speciation, transformation, and movement within soil systems have been well documented, their synthesis into a comprehensive numerical model of Se fate and transport within an alluvial aquifer system influenced by agricultural practices has not yet been realized.;This dissertation presents the development of a numerical model that can simulate the fate and transport of Se species in irrigation-influenced agricultural soil and groundwater systems at a regional scale. The model was developed by first, linking RT3D, modified to handle multi-species reactive transport in variably-saturated porous media, to MODFLOW, which uses the UZF1 (Unsaturated Zone Flow) package to simulate groundwater flow in the unsaturated zone; and second, developing an Se reaction module for RT3D that accounts for the cycling, chemical activity, and transport of Se species in regional-scale agricultural soil and groundwater systems. The module also accounts for the influence of other chemical species such as dissolved oxygen and nitrate (NO 3). The resulting model, referred to as UZF-RT3DAG, is applied to a 50,600 ha regional site in the Lower Arkansas River Valley (LARV) in southeastern Colorado for the years 2006 through 2009. Using the calibrated model, multiple BMPs for remediation of Se contamination in the LARV are investigated. These strategies include decreasing annual loading of nitrogen fertilizer, decreasing species concentration in canal water, decreasing applied volume of irrigation water, and increasing chemical activity within riparian areas.;Research results are presented through a series of published and submitted articles and modeling results that outline the progression of model development and model application. Results of the BMP scenario testing indicate that alternative land-management practices can have a significant impact in decreasing the concentration of dissolved Se in groundwater by up to 5-8% as well as mass loadings of Se to the Arkansas River by as much as 20-30%. Practices also have a significant impact on decreasing NO3 concentrations and loadings by up to 50% and 45%, respectively. As the alluvial aquifer in the LARV is similar to other Se-contaminated aquifer systems, the results of this research are pertinent to the assessment and remediation of Se contamination world-wide.