UVM Theses and Dissertations
Format:
Print
Author:
Foley, Julie
Title:
Dept./Program:
Rubenstein School of Environment and Natural Resources
Year:
2008
Degree:
M.S.
Abstract:
An Investigative Docket process was initiated in 2004 by the Vermont Water Resources Board to explore options for designing and implementing effective cleanup plans for stormwater-impaired waters. This Docket concluded that the primary cause of stream impairment by stormwater is excessive runoff from impervious surfaces. The Docket also decided that the primary objective of stormwater management should be to return the hydrologic characteristics of impaired streams to a regime that closely parallels the hydrologic characteristics of streams that are not currently impaired. However, there was no agreed protocol or framework to identify these targets or to prioritize specific locations where stormwater reductions would provide the greatest benefits. This project was initiated to develop a framework that could be used to address these needs. We used a set of 12 stormwater-impaired streams and 15 unimpaired (or "attainment) streams. Only a few of these streams had flow-gauging records and so we estimated flow with a simple stormwater hydrologic model (P-8) and used the output from this model to generate 'synthetic' flow duration curves (FDC) for comparison.
Statistical clustering methods were used to identify groupings of stormwater impaired and attainment watersheds. A hierarchical cluster analysis of inherent watershed characteristics identified watershed groupings that included both impaired and attainment streams. For these groupings, we used the mean one-day estimated flow values for the attainment watersheds as flow targets for the corresponding impairment watersheds. These flow values satisfy the target setting requirements for assessments of Total Maximum Daily Loads (TMDLs). A risk assessment methodology was developed to address the TMDL pollutant load allocation requirement. Binary logistic regression methods were used to estimate the probability that a watershed is impaired as a function of watershed characteristics. The areas with higher impairment probabilities can be targeted for priority management actions. This approach utilizes readily available data, employs simple models, is acceptable to a wide array of stakeholders and is amenable to adaptive management.
Statistical clustering methods were used to identify groupings of stormwater impaired and attainment watersheds. A hierarchical cluster analysis of inherent watershed characteristics identified watershed groupings that included both impaired and attainment streams. For these groupings, we used the mean one-day estimated flow values for the attainment watersheds as flow targets for the corresponding impairment watersheds. These flow values satisfy the target setting requirements for assessments of Total Maximum Daily Loads (TMDLs). A risk assessment methodology was developed to address the TMDL pollutant load allocation requirement. Binary logistic regression methods were used to estimate the probability that a watershed is impaired as a function of watershed characteristics. The areas with higher impairment probabilities can be targeted for priority management actions. This approach utilizes readily available data, employs simple models, is acceptable to a wide array of stakeholders and is amenable to adaptive management.