Characterization of water quality pollution in mixed land use watersheds
Date
2020
Authors
Ludwig, Madeline, author
Arabi, Mazdak, advisor
De Long, Susan, committee member
Wilkins, Michael, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Anthropogenic sources of pollution often lead to degraded surface water quality in urban and agricultural streams. The Clean Water Act was developed to mitigate the negative effects of urbanization on water quality through the development of water quality targets and the Total Maximum Daily Load program. In this study, a probabilistic framework was developed to quantitatively assess how indicators of human influence impact vulnerability to E. coli impairment and nutrient concentrations in mixed land use watersheds across the state of Colorado. The models derived using this method can be used to predict instream pollutant concentrations and help regulatory agencies create sampling programs for at risk waterbodies. Specifically, the first part of this study explores vulnerability to E. coli impairment under varying levels of upstream anthropogenic influences and develops a probabilistic method for assessing E. coli pollution based on the regulatory monitoring program. In this study, vulnerability is defined as the probability that ambient instream pollutant concentrations exceed numeric water quality standards. The study objective was examined for 28 sites along the Cache la Poudre River and its tributaries including: Boxelder Creek, Fossil Creek, and Spring Creek in northern Colorado. Indicators of urban influence include land use, wastewater treatment plant discharge capacity, combined animal feeding operation capacity, and population. Multiple linear regressions analysis between anthropogenic indicators, E. coli concentrations and vulnerability provide significant (p < 0.05) and strong (R2 > 0.7) relationships. In general, land use predictor variables were able to accurately predict E. coli load, however the most important indicator of human influence differed between E. coli concentration response variables. Additionally, the second part of this study expands upon the multiple linear regression framework to develop regression models that can predict base level nutrient concentrations for stream segments in three regions of Colorado. Regression models were developed using data from 89 sampling locations upstream of wastewater treatment plants and 84 sampling locations downstream of wastewater treatment plants. An initial analysis of gaged sampling locations showed that flow was a significantly influenced instream nutrient concentrations. Area and slope of the contributing sub watershed were then analyzed in a regression analysis and were found to be a surrogate for streamflow. Strong (R2 > 0.7) and significant (p < 0.05) regression models for upstream and downstream locations were developed using area and slope, hydrologic, point, and non-point source predictor variables. The models showed that agricultural and urban activity significantly impacted instream baseline nutrient concentrations. The methodology developed in this study can be used to predict instream pollutant concentration and assist in the development of monitoring programs for at risk waterbodies.
Description
Rights Access
Subject
geospatial analysis
nutrient pollution
water quality
land use
E. coli
regression model