Grantee Research Project Results
Final Report: Center for Green Infrastructure and Stormwater Management
EPA Grant Number: R835142Center: Center for Integrated Multi‐scale Nutrient Pollution Solutions
Center Director: Shortle, James S.
Title: Center for Green Infrastructure and Stormwater Management
Investigators: Echols, Stuart Patton , Orland, Brian A , Royer, Matthew B , Ready, Richard C , Clark, Shirley E , Gray, Barbara L , Shortle, James S. , Saacke-Blunk, Kristen , Wagener, Thorsten
Institution: Pennsylvania State University
EPA Project Officer: Packard, Benjamin H
Project Period: March 1, 2012 through February 28, 2018
Project Amount: $2,173,026
RFA: Sustainable Chesapeake: A Collaborative Approach to Urban Stormwater Management (2011) RFA Text | Recipients Lists
Research Category: Congressionally Mandated Center , Sustainable and Healthy Communities , Water
Objective:
Project 1
This research project sought to understand the cognitive and institutional barriers that currently prevent the adoption of innovative green infrastructure solutions for stormwater management, and to identify ways in which those barriers can be overcome.
Project 2
-
- Develop visual/verbal “dashboards” that communicate to local decision makers (municipal governments and school boards) the information necessary to portray the implications of stormwater management plans for their facilities via data-coupled 3-D visualization;
Dashboards available via https://sites.psu.edu/cgiswm/project-2-green-infrastructure-design-and-visualization/
Descriptive paper via http://sites.psu.edu/cgiswm/project-2-green-infrastructure-design-and-visualization/project_2_report_page/shared_communication_through_stories_and_games/
-
- Investigate the effects of introducing data-coupled visualization into the deliberation of stormwater management plans addressing perceived improvements in decision-making processes and timing, and confidence in the outcomes;
Available via Community Outreach reports at https://sites.psu.edu/cgiswm/community-outreach/
-
- Develop visualization “sets” representing alternative development scenarios and calibrated to the needs of and linkages with other projects.
Completed and integral to Project 4. See http://sites.psu.edu/cgiswm/project-2-green-infrastructure-design-and-visualization/project_2_report_page/.
Descriptive paper via http://sites.psu.edu/cgiswm/project-2-green-infrastructure-design-and-visualization/project_2_report_page/calibrated-images-to-explore-barriers-to-green-infrastructure-practices/
Project 3
The major objective of this research is to implement a high resolution, spatially explicit watershed model that resolves the local effects of current and projected stormwater practices at a watershed within the Chesapeake Bay Watershed region and to test the relative performance of green infrastructure practices as a function of surrounding land cover, position in the watershed, soil type and soil conditions. Our approach to modeling of green infrastructure design involves a spatially explicit, multi-process strategy for assessing and simulating water quantity and quality impacts of urban stormwater flooding, soil degradation and the role of altered residence times of contaminants and flooding on the urban landscape as well as the larger watershed in which it exists. The project will evaluate 3 testbeds in the Lancaster, PA area.
Project 4
The objective of this project is to measure residents' attitudes toward and preferences over non-hydrological aspects of green stormwater management, including the various ways that stormwater management can affect the built and natural landscape, and to measure residents' willingness to pay for changes in attributes of the landscape that can be affected by green stormwater management approaches.
Project 5
Project 5 the public engagement and outreach plan for the Center. As such is seeks to engage target region communities and fully involve decision makers faced with green infrastructure and stormwater management choices at multiple levels from state to local so that the research conducted in Projects 1-4 is located in and has relevance to the target region. This ensured that research outcomes have greater likelihood to inform actual decision makers and achieve the fundamental shift in whole scale adoption of sustainable stormwater management practices in Pennsylvania communities in the Chesapeake Bay watershed.
In addition, by engaging key decision makers at all levels, the Center’s outreach and engagement strategy will facilitate Center research outcomes related to understanding the barriers to widespread adoption of green infrastructure approaches to stormwater management in Pennsylvania. Wide dissemination of outcomes is proposed so that key stakeholders will be better able to influence decision makers in their communities across the Chesapeake Bay watershed.
Specifically, the project objectives are to:
- Ensure research involves decision makers and communities within the target region.
- Engage key community stakeholders within the target region in an advisory capacity to ensure regional relevance of project.
- Fully engage public decision makers and citizens within the target region through conferences that frame issues, barriers, and potential solutions, disseminate outcomes and research results, and make recommendations for influencing decision makers and developing tools to achieve widespread implementation of green infrastructure in Pennsylvania.
- Achieve broad external engagement and outreach of research results and project outcomes to a wider citizen audience through the Center’s website.
Summary/Accomplishments (Outputs/Outcomes):
Project 1
The final report for this project has already been submitted. At the February 2018 Scientific Advisors’ Meeting that, since the time the interviews were conducted and the recommendations were presented, DEP and other state agency partners have moved forward to implement policies and programs that address several of these recommendations. What is reported here are the original recommendations from Project 1 for overcoming cognitive and institutional barriers to decision making about stormwater management and green infrastructure use. After each recommendation recent developments that represent progress in addressing each recommendation are listed.
Recommendation #1: DEP should take a proactive stance in creating educational opportunities for township officials, planners, engineers, maintenance staff and landscapers to learn about green infrastructure solutions and their maintenance and to provide data about their effectiveness in reducing water conveyance and improving water quality and ultimately providing cost savings.
Recent developments:
- A new DEP permit was issued for 2018: It requires MS4 communities in the Chesapeake Bay watershed or that contain surface water impaired with certain pollutants to create & implement a pollutant reduction plan to reduce sediment discharge by 10% percent over the next 5 years.
- DEP also prepared a table clarifying which MS4 municipalities needed to prepare plans
- In 2017 DEP trained almost 2,000 people on the requirements of the new 2018 permit.
- DEP recommended that MS$ communities file a single plan for DEP and EPA.
Recommendation #2: Sponsor and participate in the development of regional watershed planning to address stormwater management.
Recent developments:
State agencies and other partners have led an overall upswing in green infrastructure related training on design, installation and maintenance in Pennsylvania. For example, see developments under Recommendation #4 below.
Recommendation #3: Townships and municipalities should establish dedicated fees or restricted taxes for stormwater management activities.
Recent developments:
- A template for creating local authorities that can assess stormwater fees has been developed by PennFuture. It can be found at: http://www.pennfuture.org/Files/Admin/PennFuture_StormwaterManual_web_3.20.17.pdf
- Several PA townships have explored creating authorities for this purpose (e.g., Hampden, Radnor, Meadville and York County)
- In Sept. 2016, Act 62 authorized second-class townships to assess a reasonable and uniform fee (on all properties) for storm water management activities and facilities, without the need to establish a municipal authority. It also authorized a second-class township to enact and enforce ordinances to govern and regulate the planning, management, implementation, construction and maintenance of storm water facilities.
- Similar Acts (913-916) for 1st class townships and 3rd class cities were under consideration in the PA Senate starting Jan. 30, 2018.
Recommendation #4: Revitalize DEP task force to explore opportunities for municipalities to meet stormwater obligations through investments in riparian buffer restoration in rural, suburban and urban areas.
Recent developments:
- Responsibility for this activity now lies with DCNR’s Bureau of Recreation and Conservation.
- A Riparian Forest Buffer Advisory Committee was created in 2015.
- In 2017 PA received a $750,000 grant for stream buffers from the National Fish & Wildlife Foundation.
- In 2017, $1M from PENNVEST was set aside for grants to landowners who sign 25 yr. land agreements to create multifunctional forest riparian buffers (also includes cash crops)
- DEP added riparian buffers to stormwater BMP list.
- A Riparian Forrest Buffer Summit was held in State Colelge from 2/28-3/1 2018.
Recommendation #5: Streamline the storm water management review and permitting processes at the local level to: 1) provide greater clarity about compliance for townships and municipalities, and 2) ensure that developers don’t incur a penalty for trying new green infrastructure methods.
Recent developments:
- DEP has streamed and clarified the regulatory requirements related to MS4 permits as described above under Recommendation #1.
Recommendation #6: Greater publicity about stormwater management and its importance is needed by federal, state and local officials as are resources to provide the necessary training for local officials to effectively implement existing regulations and ensure proper maintenance of BMPs over time.
Recent developments:
- Stormwater fees permitted by Act 62 can factor in cost of maintenance of BMPs installed by 2nd class townships.
- The PA Senate is considering passage of Acts 913-6 that would allow 1st class townships and 3rd class cities to do the same.
Project 3
Calibration and testing of the Conestoga watershed model in south-central PA has been the focus of this effort. The watershed includes agricultural, forest, suburban, urban and urbanizing landuse-land cover types. Initial simulations for the Conestoga watershed utilized climate Reanalysis period (hourly data for 1979-Present). We developed a multi-state variable calibration that utilizes streamflow, groundwater level records as well as a strategy to assimilate wetland and pond surface elevations into the calibration procedure.
Task 3a. Determine the state-of-the-practice for stormwater engineering in the Chesapeake Bay Basin. This included reviewing barriers to implementation to several common green infrastructure practices. The second goal was to query both the design and regulatory communities to identify, by community, what they perceived as the non-technical barriers to green infrastructure adoption.
Task 3c. This task focused on whether improvements in design criteria could be made based on a statistical analysis of the International BMP Database (bmpdatabase.org). The complete database was downloaded and the water quality results were linked to any existing design information. Correlation analyses were conducted to determine the relationship between treatment system effluent quality and design parameter selection. These design factors included loading rate, media depth, pond surface area, and others.
Task 3a. Update to State of the Practice
Results:
- For site level design, the 10-year storm is the typical drainage design storm size.
- The NRCS TR-55 methods is the primary (and nearly only) method used to calculate time of concentration.
- The frequency of computer use in drainage design has increased since the last survey; however, even with the improvement in computing power, the use of continuous simulation in design is still limited.
- Single-event design storms are still the common design criteria used in stormwater management design in the Bay Basin.
- The current definition of drainage system failure includes the older criteria of manholes popping or curb overtopping from the list, and added several other concerns, including basement flooding and system overflows.
Task 3a: Water Quality Concerns Addressed in Current Design.
Results:
- Water quality concerns were more prevalent than they were in prior surveys.
- Sediment control was a high concern of all respondents.
- Water quality concerns are assumed to be addressed through compliance with the state or municipal regulatory guidance.
- Few respondents were obtaining water-quality treatability information from the databases sponsored by the US EPA (International BMP Database and the EPA green infrastructure database) and CALTRANS.
Task 3a: Specific Green Infrastructure Techniques.
Results:
- Infiltration generally is an accepted practice throughout the watershed.
- If infiltration is not allowed or used, it is at the site level and it typically is due to specific site characteristics such as soils and/or prior contamination.
- There were concerns also expressed about maintenance which may limit client acceptance of infiltration, unless mandated by regulations.
- Rainwater harvesting is not widely applied in the Bay Basin.
- The respondents noted that it was allowed for at least landscape irrigation.
- The barriers included concerns about both upfront costs and maintenance costs of the system.
- Respondents noted that there were no or insufficient credits for adoption.
- Cluster development is not widely adopted in the Bay Basin.
- Responses were similar to those for rainwater harvesting.
- If it was not allowed, it often was due to conflicting regulations, such as zoning regulations.
- Concerns were expressed about a lack of credit for the practice.
Task 3a. Adoption and Barriers to Adoption (Non-technical) to Green Infrastructure Adoption.
The first question of this section asked for a prioritization of the potential reasons for investing in green infrastructure. The results were as follows in rank order:
- Water quantity and quality (reduce treatment volume, reduce stream degradation, improve water quality, increase groundwater recharge, increase streamflow)
- Voluntary compliance with regulations (environmental standards, permit requirements)
- Reduce flooding
- Positive economic benefits (increase job opportunities, increase property value, improve investment in adjacent areas, reduce energy costs)
- Financial savings (reduce overall cost of project, reduce cost of drainage system, maintenance cost, cost of mitigating stormwater-related overflows)
- Increase conservation and green space (increase wildlife habitat, improve park and green space access, increase land conservation)
- Comply with court order/consent decree
- Improve community relations and connections (improve community aesthetics, serve low-income or minority communities, promote public education, development partnerships)
- Climate concerns (ameliorate heat island effect, reduce carbon dioxide emissions, improve air quality, mitigate and adapt to climate change impacts)
The rationale for investing in green infrastructure appear to primarily be related to regulatory and nuisance concerns. The ancillary benefits did not rank as high as the regulatory compliance and nuisance abatement benefits. The reason that these ancillary benefits did not rank as highly as did the others is unknown.
Project 4
Preferences over landscape attributes
Households are willing to pay $52.65 to change biodiversity from the baseline “low” level to the medium level, and $65.50 to change it from low to high. For the water presence attribute the baseline level was “never present”. The average household disliked water presence, and must be compensated $27.42 to put up with intermittent water and $118.09 to put up with permanent water. For the percent mowed attribute the baseline level as 100% mowed. Here, the average respondent exhibited an inverse u-shaped preference function, with positive willingness to pay for intermediate levels of mowing, but no statistically significant willingness to pay to go from 100% mowed to 0% mowed. Respondents did not have statistically significant preferences over the geometry of plantings.
The statistical model used to interpret choice responses also allows estimates of the heterogeneity in household preferences over each attribute. The model showed that there is statistically significant levels of heterogeneity in preferences over all attributes. For example, while the average respondent has a strong negative preference for permanent water, 14% of respondents have a positive preference for that landscape attribute. While respondents are fairly evenly split regarding preferences over 100% vs 0% mowing, the large majority (70-89%) prefers intermediate levels of mowing.
Preference Motivation
Choice responses were not significantly different between the baseline version and the no-stormwater version, suggesting that preferences are motivated by non-hydrological concerns, though it is possible that the language used to explain downstream impacts was not prominent enough to trigger changes in preferences.
Information Effects
A focus of this project has been close cooperation between Project 2 and Project 6 in order to investigate the role that visualizations (computer-generated graphical depictions of landscapes) have in preference formation and revelation. Visual representations such as photographs and computer-generated images have been widely used to illustrate choice questions and ensure respondents understand the scenarios in stated preference studies. However, there are very few studies that have explored whether the use of visual representations affects the results of stated preference methods compared to using other methods (e.g., text descriptions) of communicating the attributes that respondents are being asked to value.
Statistical comparison of repsonses to the text only, picture only and text plus picture survey versions suggest that images alone are as effective as text alone at conveying landscape attributes with high visual salience, such as variety of plant species and the presence of water. However, when respondents were presented with both text and images, they exhibited stronger preferences for visually salient landscape attributes and were less likely to ignore those attributes than when presented with either only images or only text. These results suggest that providing both images and text, as was done in our base version, helps respondents better understand choice questions and more accurately state their preferences for given scenarios.
Although people pay more attention to individual attributes in choice questions with the provision of visual representations, the extra visual stimuli also increase choice randomness (measured by a scale parameter), which is indicative that respondents are less certain whether they picked their preferred options for each choice question. This may be attributed to the cognitive overload, caused by paying more attention to the questions, or it may be due to different respondents focusing on different specific details of visual representations.
Model Choice
A second methodological issue explored in this research is the appropriate choice of econometric model for analyzing data from a choice experiment. Several different approaches to incorporating preference heterogeneity have been developed recently, including latent class models, random parameter (mixed) logit models, and hybrid latent-class/random parameter models. Unfortunately, when analyzing stated choice data, we do not have external validity criteria to determine which model(s) best account for preference heterogeneity.
In our survey, we included an attribute level ranking exercise that provides such an external validity check. Prior to being asked the stated choice questions, the different levels of each attribute were described to respondents, and respondents were asked to rank the levels (from best to worst) for each attribute. Thus, we know, for each respondent, whether they like permanent water more or less than no water.
Project 5
- Developed a Community Partners Council to advise the Center in research approaches in the target area, and held periodic meetings of the CPC and research team leaders.
- Developed a Science Advisory Committee to advise the Center’s research team on programmatic matters and priorities for research integration, outreach and engagement, and held periodic meetings of the SAC and research team leaders.
- Two “all hands” meetings were held involving researchers, CPC members and SAC members: July 25, 2016 and February 23, 2018.
- Held “Green Infrastructure Forum: A Dialogue about Dealing with Stormwater in the Lower Susquehanna,” a kickoff conference for the Center’s research (June 26, 2013). Produced a forum report of proceedings.
- Developed Center website (https://sites.psu.edu/cgiswm/) to disseminate research results and tools.
- Held “Chesapeake Stormwater Summit: Overcoming Barriers to Green Infrastructure Solutions,” a final conference to disseminate research results of the Center and explore with stakeholders opportunities to inform Pennsylvania’s Phase 3 Watershed Implementation Plan (WIP) for the Chesapeake Bay Total Maximum Daily Load (TMDL) (October 27, 2017). Produced a summit report of proceedings.
Held six workshops/projects with community stakeholders in the target area to disseminate research results and implement recommendations.
Journal Articles: 3 Displayed | Download in RIS Format
Other center views: | All 22 publications | 5 publications in selected types | All 3 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Leonard L, Duffy CJ. Essential Terrestrial Variable data workflows for distributed water resources modeling. Environmental Modelling & Software 2013;50:85-96. |
R835142 (2013) R835142 (2015) R835142 (Final) |
Exit Exit Exit |
|
Yu X, Bhatt G, Duffy C, Shi Y. Parameterization for distributed watershed modeling using national data and evolutionary algorithm. Computers & Geosciences 2013:58;80-90. |
R835142 (2015) R835142 (Final) |
Exit Exit Exit |
|
Yu X, Duffy C, Zhang Y, Bhatt G, Shi Y. Virtual experiments guide calibration strategies for a real-world watershed application of coupled surface-subsurface modeling. Journal of Hydrologic Engineering 2016;21(11):04016043. |
R835142 (2016) |
Exit |
Supplemental Keywords:
Virtual environmental design, integrated models, geospatial data services, International BMP Database, Stormwater design, Green Infrastructure Barriers, Community; engagement; stakeholders; decision makers; watersheds; stormwater; water quality; nutrients; Chesapeake Bay; Susquehanna River; water; land; outreach; public participation; green infrastructure.Relevant Websites:
Penn State Center for Green Infrastructure and Stormwater Management Exit
The Penn State Integrated Hydrologic Model (PIHM) Exit
HydroTerre Consultants, INC Exit
Progress and Final Reports:
Original Abstract Subprojects under this Center: (EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
R835142C001 Decision Making – Cognitive and Institutional Barriers
R835142C002 Green Infrastructure Design and Visualization
R835142C003 Hydrologic and Water Quality Modeling for Green Infrastructure
R835142C004 Non-Hydrological Benefits and Citizen Preference
R835142C005 Public Engagement and Outreach
The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
- 2016 Progress Report
- 2015 Progress Report
- 2014 Progress Report
- 2013 Progress Report
- 2012 Progress Report
- Original Abstract
3 journal articles for this center