Date of Award

2016

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Natural Resources

First Advisor

William B. Bowden

Abstract

Despite decades of U.S. water quality management efforts, over half of assessed waterbody units were threatened or impaired for designated uses in the most recent assessments, with urban runoff being a leading contributor to those impairments. This cumulative research explores several aspects of urban runoff dynamics through a combination of field study and modeling.

Stormwater ponds are ubiquitous in developed landscapes due to their ability to provide multiple forms of treatment for stormwater runoff. However, evolving design goals have reduced the applicability of much of the early work that was done on pond effectiveness. In this study, we instrumented a recently constructed detention pond in Burlington, VT, USA. Flow gaging demonstrated that the pond achieved a 93% reduction in event peak flow rates over the monitoring period. Storm sampling showed that the pond significantly reduced total (TN) (1.45 mg/L median influent, 0.93 mg/L median effluent, p < 0.001) and total phosphorus (TP) (0.498 mg/L median influent, 0.106 mg/L median effluent, p < 0.001) concentrations over the events sampled. A loading analysis estimated the TN and TP removal efficiencies for the pond to be 23% and 77% respectively. Lastly, temperature data collected from the pond showed that during the summer the pond accumulates considerable heat energy. This study adds to the body of literature on detention pond performance, and raises concerns about the extensive use of stormwater ponds in watersheds where thermal stress is a concern.

EPA SWMM is a widely used urban hydrologic, hydraulic and water quality model, though its application can be limited due to its deterministic nature, high dimensional parameter space, and the resulting implications for modelling uncertainty. In this work, I applied a global sensitivity analysis (SA) and evolutionary strategies (ES) calibration to SWMM to produce model predictions that account for parameter uncertainty in a headwater tributary case study in South Burlington, VT, USA. Parameter sensitivity was found to differ based on model structure, and the ES approach was generally successful at calibrating selected parameters, although less so as the number of concurrently varying parameters increased. A watershed water quality analysis using the calibrated model suggested that for different events in the record, the stream channel was alternately a source and a sink for sediment and nutrients, based on the predicted washoff loads and the measured loads from the stream sampling stations. These results add to the previous work on SWMM SA, auto-calibration, and parameter uncertainty assessment.

Lastly, given the extent of eutrophication impairment in the U.S., I compared TN and TP data collected in these original works with national and regional datasets. TN concentrations sampled in this work were generally commensurate with values reported elsewhere, however TP data were not. Drainage area attributes and an event based rainfall runoff analysis of the study catchments provided circumstantial support for the idea that runoff from lawns is driving the high TP loads in Englesby Brook. The role of pet wastes is considered as a potentially fruitful area for further research.

Language

en

Number of Pages

261 p.