Improved assessment of nitrogen and phosphorus fate and transport for intensively managed irrigated stream-aquifer systems
Date
2019
Authors
Wei, Xiaolu, author
Bailey, Ryan T., advisor
Arabi, Mazdak, committee member
Gates, Timothy K., committee member
Covino, Timothy, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Nitrogen (N) and Phosphorus (P) are essential elements for animal nutrition and plant growth. However, over the previous decades, excessive loading of fertilizers in agricultural activities has led to elevated concentrations of N and P contaminations in surface waters and groundwater worldwide and associated eutrophication. Therefore, precisely understanding and representation of water movement and fate and transport of N and P within a complex dynamic groundwater-surface water system affected by agricultural practices is of essential importance for sustaining ecological health of the stream-aquifer environment while maintaining high agricultural productivity. Modeling tools often are used to assess N and P contamination and evaluate the impact of management practices. Such models include land surface-based watershed models such SWAT, and aquifer-based models that simulate spatially-distributed groundwater flow. However, SWAT simulates groundwater flow in a simplistic fashion and therefore is not suited for watersheds with complex groundwater flow patterns and groundwater-surface interactions, whereas groundwater models do not simulate land surface processes. This dissertation establishes the modeling capacity for assessing the movement, transformation, and storage of nitrate (NO₃) and soluble P in intensively managed irrigated stream-aquifer systems. This is accomplished by (1) developing a method to apply the SWAT model to such a system, and includes: designating each cultivated field as an individual hydrologic response unit (HRU), crop rotations to simulate the impact of changing crop types for each cultivated field, including N and P mass in irrigation water, and seepage from earthen irrigation canals into the aquifer; (2) simulating land surface hydrology, groundwater flow, and groundwater-surface water interactions in the system using the coupled flow model SWAT-MODFLOW, with the enhanced capability of linkage between SWAT groundwater irrigation HRUs and MODFLOW pumping cells, and the use of MODFLOW's EVT package to simulate groundwater evapotranspiration; and (3) linking RT3D, a widely used groundwater reactive solute transport model, to SWAT-MODFLOW to credibly represent of NO₃-N and soluble P fate and transport processes in irrigated agroecosystems to evaluate best management practices for nutrient contamination. This last phase will also address the uncertainty in system output (in-stream nutrient loads and concentrations, groundwater nutrient concentrations model predictions). Each modeling phase is applied to a 734 km² study region in the Lower Arkansas River Valley (LARV), an alluvial valley in Colorado, USA, which has been intensively irrigated for over 130 years and is threatened by shallow water tables and nutrient contamination. Multiple best management practices (BMPs) are investigated to analyze the effectiveness in reducing NO₃-N and soluble P contamination in the LARV. These strategies are related to irrigation management, nutrient management, water conveyance efficiency, and tillage operations. The most effective individual BMP in most areas is to decrease fertilizer by 30%, resulting in average NO₃-N and soluble P concentrations within the region could be reduced by 14% and 9%, respectively. This individual BMP could lower the average NO₃-N concentrations by 19% and soluble P concentrations by 2%. Combinations of using 30% irrigation reduction, 30% fertilization reduction, 60% canal seepage, and conservation tillage are predicted to have the greatest overall impact that can not only provide a decrease of groundwater concentration in NO₃-N up to 41% and soluble P concentration up to 8%, but also reduce the median of the in-stream NO₃-N and soluble P to meet the Colorado interim standard. As nutrient conditions within the Lower Arkansas River Valley are typical of those in many other intensively irrigated regions, the results of this dissertation and the developed modeling tools can be applied to other watersheds worldwide.