Development of an integrated, watershed-scale, planning tool

for stormwater management in Vermont

Breck Bowden and Julie Foley

Problem or Need: Many streams in Vermont’s urbanizing environments are significantly impacted by stormwater runoff. A TMDL approach to stormwater management is problematic and so an alternative approach that is simple to employ, objective, quantitative, and defensible is desirable if possible. Significant progress toward such an approach was achieved by the ‘Assessment Methodology Working Group’ that formed as a part of the recently concluded investigative docket on stormwater, initiated by the Vermont Water Resources Board (Nicholls et al. 2004). The draft methodology outlines an innovative and potentially effective new approach to stormwater management on a watershed scale. The specific need is to transform this draft method into a fully functional planning tool.

This project will address, primarily, the watershed-based planning category of eligible activities in the Vermont 319 NPS project RFP. In part, it will also address the technical component of the technical/financial assistance category. The project has verbal support from a number of the participants in the WRB Stormwater Assessment Working Group. The project will address to some degree all three of Vermont DEC’s priority management concerns for 2004, most notably the ‘improvement of stormwater management within the Lake Champlain drainage from developed areas or areas undergoing development”. This project is specifically designed to help restore all waters in Vermont listed as impaired by stormwater (303.d, Vermont DEC 2002). The overall project will provide a means to estimate the amount by which water and sediment loading have to be reduced to attain Vermont Water Quality Standards in these listed streams.

Project Purpose and Objectives: To develop a practical stormwater management protocol that integrates a hydrologic method to assess targets for stormwater reduction at a watershed scale with a risk assessment method to assess the priority locations for best management practices at a site scale.

Key objectives are to:

1.  Develop a methodology to employ ‘synthetic’ flow duration curves (FDC) derived from simple stormwater hydrologic models, to identify watershed-level targets for stormwater runoff reduction.

2.  Refine a suggested multi-component, cumulative risk assessment methodology to identify priority areas to implement stormwater best management practices in Stormwater Impact Areas (SIA).

3.  Develop and field test a practical framework to link the watershed-level targets for stormwater reduction identified by the FDC analysis with the locale-specific targets for best management practices identified by the SIA analysis.

4.  Second year objective: Explore alternatives to link the FDC and SIA analyses with assessments of the potential for ‘wash-off’ sediment loading and ‘in-stream’ bank erosion and sediment resuspension. (See the Specific Project Workplan.)

General Project Plan: The general plan for this project is to employ a graduate student in the Natural Resources Master’s of Science program at the Rubenstein School of Environment and Natural Resources to conduct the project, under the guidance of Dr. Bowden (1 month), Dr. Watzin (1 week) and at least one other UVM faculty member to be determined. The graduate student would be devoted to this project alone. Both Dr. Bowden and Dr. Watzin will donate their time to advise the student, providing the required match to the federal dollars. Ideally, the student would become an ad hoc member of the working group that will be developing the new approach to stormwater management in Vermont and will essentially become an integral research resource to the working group.

A graduate student has been identified (Julie Foley, Colgate University, Geology, 1999) and will be on board at UVM by late August 2004. Ideally this will provide time for the student to interact with the working group in meetings that are tentatively planned for mid- to late-September and that might coincide with visits from Dr. Bruce Cleland, with whom we would expect to interact.

The student would be expected to accomplish objectives 1-3 in draft form by May 2005 and in final form by June 2005. Some consideration of sediment ‘wash off’ dynamics will be included in this effort. However, a major portion of the working group discussion about the proposed new stormwater assessment methodology focused on questions about whether and how to include sediment loading estimates. We anticipate that a thorough consideration of this question leading to practical recommendations will require additional time, study, consultation with the working group, and perhaps field work. Thus, objective 4 in this project is proposed as a ‘potential second year objective’ that will require additional funding.

Specific Project Plan:

1.  Develop a methodology to employ ‘synthetic’ flow duration curves (FDC) derived from simple stormwater hydrologic models, to identify watershed-level targets for stormwater runoff reduction.

Deliverable: An analysis of simple metrics that can be derived from FDC’s to quantify how much reduction in stormwater runoff is needed to reasonably assure that currently impaired (i.e. 303.d listed) streams are able to comply with Vermont Water Quality Standards.

By: January 2005 in interim report form

Explanation: Currently Region 1 EPA has contracted with TetraTech to provide FDCs for a series of impaired and ‘attainment’ streams in Vermont, largely in the Chittenden County area. It is anticipated that outputs from this contracted work will be available over the summer of 2004 and that work will begin on ways to interpret and use these data in late summer and early autumn. The selected graduate student would join this discussion in September 2004 and may be available earlier if this seems useful and can be arrange. TetraTech is only contracted to produce the FDCs and will not provide any substantive analysis of these data. The working group will do whatever analysis is deemed useful and necessary. We anticipate that this discussion about how to use the FDC data will generate a series of questions that would form the basis for this part of the graduate student’s effort. For example, what portions of the FDCs are most sensitive to land use change and there most useful to set reduction targets? What is the variance among the FDC estimates and what factors contribute most to this variance (sensitivity analysis)? Is there a way to reduce the variance among FDCs through multivariate analyses? How might we expect the FDCs to change as a consequence of various development or restoration scenarios? (This latter question is one that TetraTech is not likely to address in their contracted work.) What is the relationship between FDCs (for water) and ‘loading curves’ for washoff sediments? To what degree will a focus on stormwater water management address washoff sediment issues? (Note that the focus here is entirely on ‘washoff’ sediments; i.e. in-stream processes will not be addressed here. See Objective 4.) We expect results from this work will be available before the January 2005 milestone, in draft form for discussion. The milestone is a realistic target for a formal interim report.

2.  Refine a suggested multi-component, cumulative risk assessment methodology to identify priority areas for implementation of stormwater best management practices (Stormwater Impact Areas).

Deliverable: Analysis of appropriate weight factors and threshold levels to identify ‘stormwater impact areas’ with an application of the refined methodology to at least one impaired and one ‘attainment’ watershed.

By: April 2005

Explanation: The SIA analysis provides a simple and objective means to assess risks from a variety of environmental factors that can affect stormwater runoff of water, sediments and other materials of concern. The current approach has been developed as a prototype to demonstrate the feasibility of the approach. The purpose of this objective is to refine the methodology by considering whether there are additional data layers that could be used, examining the weighting factors used, evaluating alternative threshold values, and conducting sensitivity analysis based on the inherent errors in the component data layers used in the SIA analysis. Results from this analysis would be available in draft form before the indicated April 2005 milestone. The milestone is a realistic target for a formal interim report.

3.  Develop and field test a practical framework to link the watershed-level targets for stormwater reduction identified by the FDC analysis with the locale-specific targets for best management practices identified by the SIA analysis.
Deliverable: A workable draft framework for consideration by the stormwater working group with a plan for field testing in the 2004/2005 building season.

By: May 2005 in draft form

By: June 2005 in final form

Explanation: The purpose of this objective is to synthesize the two key components of the FDC analysis (Objective 1) with the SIA analysis (Objective 2) to create an integrated management plan. The draft plan should evolve quickly from the January and April 2005 interim reports, for discussion by the working group. This framework will incorporate the nine required components of watershed-based plans as per EPA 2004 319 Guidance. The final report would be delivered in June. Effectively this would conclude the deliverables of the first year project. If additional funds were available, the student would continue to objective 4.

4.  Potential second year objective: Explore alternatives to link the FDC and SIA analyses with assessments of the potential for ‘wash-off’ sediment loading and ‘in-stream’ bank erosion and sediment resuspension.

Deliverable: An analysis of methods that could be reasonably used to link sediment generation methods with watershed-level FDC and SIA analyses. This analysis will be documented in the form of a formal thesis submitted to the Graduate College of the University of Vermont and, if desired, in the form of a report to the Agency of Natural Resources in a format to be negotiated.

By: January 2006 in draft form for review and comment

By: April 2006 in final form

Explanation: The water delivered to streams by rainfall on developed and undeveloped lands also caries sediment and other materials. The sources and mechanisms of sediment movement during storms are complex. In Objective 1 we will focus on ‘washoff’ sediment. Broadly, this is sediment that is literally washed off the surface of the land and delivered to streams. This is a true external loading of sediments to streams. However, washoff sediments are only one component of sediment movement during storms and may not even be the most important component. Furthermore, washoff sediments – at least as simulated in some models – may only include the sediments that arise from impervious surfaces in developed areas. Sources of sediments from disturbed construction sites and fallow agricultural areas may be poorly represented. More sophisticated models (e.g., the Soil Water Assessment Tool [SWAT] or Water Erosion Prediction Project [WEPP]) than will be used in the TetraTech project are necessary to capture the full range of land-based erosion processes. And even these models tend to ignore in-stream bank erosion and sediment resuspension, which together may account for over 50% of the sediment movement during storm flows in many streams.

A second complication is that ‘sediments’ are composed of a wide range of materials that may range from largely organic to largely mineral and that range in size from particles that can remain in suspension for some time, even in quiet water (e.g., clay) to particles that can only move as ‘bedload’, bouncing along the stream bottom (e.g., gravel and cobble). The processes that control the generation, transport, and deposition of these different materials are reasonably well known, but differ substantially. Simple hydrologic and sediment models do not capture this complexity.

Thus there are several challenges to incorporate realistic and reasonable guidelines for sediment control into a simple and widely-applicable stormwater management protocol. We propose to use more sophisticated models (either SWAT or WEPP) to explore the degree to which the measures suggested by the simpler models address the more complex array of sediment dynamics that arise in the stormwater impaired streams of Chittenden County during storm events. This information will be used to supplement and refine the recommendations of the June 2005 report. Based on prior experience, it will take a substantial effort to gather the necessary data, parameterize the models, produce the output, and summarize the results from this effort. It is unrealistic to propose that this be done in Phase 1 of this effort. However, if it is relatively assured that this work could be funded in year 2, preliminary work could begin so that the Phase 2 work could proceed efficiently.

Interagency Coordination, Roles, & Responsibilities: We ask that the Agency of Natural Resources be willing to include the student in key discussions related to the development of the new stormwater management assessment protocols and provide advice and assistance in locating information relevant to the project. Depending on the level of interest, we would invite the agency to nominate an individual from the staff who would become a formal member of the student’s academic advisory committee at UVM. (These advisory committees are required for each student.) This association would promote a strong and tangible link between the student and ANR staff. Normally, academic committees such as this meet once or twice per year.

Public Participation: It is expected (and intended) that the student funded by this project would become an ad hoc member of the working group who are developing the new stormwater management protocols arising from the Water Resource Board Investigation Docket. Results from this project would become a part of the material made available for public review and comment as the stormwater management discussion continues.

Expected Benefits: The primary expected benefit is protection of the health of streams that are still attaining Vermont Water Quality Standards and improvement of the health of streams that are currently identified as stormwater impaired in the Vermont 303(d) listing. This benefit will be realized through a stormwater management protocol that is scientifically defensible, objective, and sensitive to developers’ interests. This protocol will build on existing guidance manuals (Vermont ANR 2002 a and b) and will include the recommendations of the Water Resources Board investigative docket (Nicholls et al. 2004).

Citation of Project Need: The primary citation of need for this project can be found in the final report of the Water Resources Board investigative docket (Nicholls et al. 2004). Supplemental justifications can be found in the Vermont Stormwater Management manuals (Vermont ANR 2002), especially Volume 2, and in the Vermont 303(d) List of Waters (Vermont DEC 2002). Informal citations of need can be found in numerous newspaper articles over the last several years.