UPDATES is a monthly email newsletter on stormwater management, assessment (including monitoring), and maintenance research at St. Anthony Falls Laboratory and the University of Minnesota.
Or Text SWUPDATES to 22828 to Join. Message and data rates may apply.
The Minnehaha Creek Watershed Stormwater Adaptation Study
March 2014 (volume 9 - issue 1)
Contributed by Trisha Moore, Assistant Professor, Kansas State University (formerly Research Associate, St. Anthony Falls Laboratory, University of Minnesota).
Latham J. Stack, Co-Principal investigator, Syntechtic International, LLC
Michael H. Simpson, Co-Principal Investigator, Antioch University
Jim S. Gruber, Project leader of community education, outreach, and participatory decision-making, Antioch University
Joel B. Smith, Advisor to climate modeling and adaptation policy, Stratus Consulting
John S. Gulliver, Project leader of hydrologic modeling, stormwater modeling, and development of non-pipe alternatives for stormwater management, University of Minnesota
John L. Nieber, Advisor to hydrologic modeling, stormwater modeling, and development of non-pipe alternatives for stormwater management, University of Minnesota
Bruce N. Wilson, Advisor to hydrologic modeling, stormwater modeling, and development of non-pipe alternatives for stormwater management, University of Minnesota
Telly Mamayek and Leslie Yetka, On-site leaders for community education and outreach, Minnehaha Creek Watershed District
Lois Eberhart, Assists leadership of on-site community education and outreach, City of Minneapolis
Stormwater infrastructure – including pipes, ponds, and rain gardens – is typically designed to convey or capture runoff flows associated with a design storm, the magnitude of which is based on a probability distribution of observed rainfall events and a community’s tolerance for risk. One of the underlying assumptions of this design approach is that the rainfall probability distribution is static. However, recent climate trends across much of the country indicate large events are occurring with greater frequency, casting doubt on the notions that (1) rainfall distributions are static in time and (2) stormwater infrastructure designed by our current design storm approach can be expected to provide the intended level of service throughout its lifetime.
Minnesota and the upper Midwest are among the regions in which the frequency of large events is projected to increase. Because detection of a climate change signal in observed rainfall events may not occur for 30-50 years, SAFL researchers teamed with the Minnehaha Creek Watershed District, the Cities of Minneapolis and Victoria, and climate researchers from Antioch University and Syntectic International to develop alternative forecasts and evaluate potential impacts to stormwater infrastructure in the Twin Cities Metro Area due to climate change. Two study sites, both located in the Minnehaha Creek watershed, were selected for the study (Fig. 1). The first, located in South Minneapolis, is Pipeshed 76-010 (Fig. 1, upper right), representative of a fully developed watershed while the second, located in the growing community of Victoria (Fig. 1, upper left), represents a rural watershed in transition to urban land uses. Among the key research questions addressed were:
- Using forecasts derived from downscaled general circulation models (GCMs), how might various design storms change in the Twin Cities area?
- Despite the high level of uncertainty in forecasts, what can we anticipate about the capacity of existing stormwater infrastructure to convey future events? To what degree might flooding increase?
- What are some options to consider for adapting stormwater infrastructure to manage climactic changes? What might be the cost implications of examined options?
The answers to these questions are discussed in the following sections.
Figure 1. Impacts of climate change on stormwater infrastructure in a fully developed (South Minneapolis; 50% impervious area; top right) and developing (Victoria; impervious surface coverage expected to increase from 14% to 29% by 2030; top left) watershed, both of which lie within the Minnehaha Creek Watershed (bottom, yellow outline).
Changing design storms? Changes in the magnitude of the 10-year probability storm (upon which current storm sewer design in Minneapolis and Victoria is based) by the mid-21st century were obtained by statistically downscaling the outputs of several different GCMs. By considering model outputs from a variety of different climate models and emissions scenarios, we intended to capture the uncertainty associated with climate prognostications. The magnitude of model outputs ranged along a continuum associated with potential mid-21st century greenhouse gas concentrations from “optimistic” (e.g., A1b emissions scenario) to “pessimistic” (the upper 95th confidence interval of A1fi emissions scenario) scenarios. Under an optimistic future, the 10-year design storm, currently 4.1 inches by Technical Paper 40 (1961), increased less than 4%. However, a 150% increase was expected under the pessimistic emissions scenario. A moderate emissions scenario resulted in a 25% increase in the 10-yr design storm depth. Model outputs indicate that an increase in the current design storm is indeed likely.
Capacity to accommodate future climate change? The capacity of the existing Pipeshed 76-010 and Victoria stormwater drainage networks to convey runoff associated with the range of mid-21st century 10-year storms was assessed using the EPA Stormwater Management Model (SWMM). Both systems were overwhelmed under future conditions, resulting in increased flooding (Figure 2), though the Victoria system was adequate up to a moderate emissions scenario. The volume of over-curb flooding increased 1,500% for the pessimistic precipitation scenario over existing conditions in the 76-010 pipeshed. Undersized pipes associated with hypothetical surface flooding in the Minneapolis pipeshed are highlighted in Figure 3. While the volume of flooding also increased substantially in Victoria, projected flood waters were contained within streets or public open spaces such as recreational areas and preserved wetland buffers (Figure 4). This was true even for future development scenarios in which the total impervious surface area of the Victoria watershed was doubled from 14% to 29%. Compared to an older, fully built area such as Pipeshed 76-010, Victoria’s stormwater network, which is dependent on a system of constructed stormwater ponds and natural lake and wetland systems, displayed a relatively high degree of resiliency to climate change.
Figure 2. Response to climate change scenarios in the Minneapolis (Mpls; dark gray, filled circles) and Victoria (light gray, open circles) study sites in terms of % components undersized (solid lines) and the volume of surface flooding (dashed lines). The Minneapolis study site is 1100 acres (445 Ha) and the Victoria study site is 1145 acres (463 Ha) in size.
Figure 3. Flood vulnerability for the existing (left), moderate mid-century (middle; 6.56 inches) and pessimistic mid-century (right; 10.1 inches) 10-year storm scenarios as indicated by undersized pipes in the Minneapolis pipeshed. Individual pipe segments are highlighted according to their classification as either (1) adequately sized (green), (2) surcharged but not associated with surface flooding (yellow), (3) surcharged and associated with downstream street flooding (blue), or (4) surcharged and associated with downstream over-curb flooding with the potential to contribute to property damage (red).
Figure 4. Vulnerability mapping of Victoria’s stormwater system under future climate and build-out (areas projected to be developed by 2030 are highlighted in red) for the existing (left), moderate mid-century (middle; 6.56 inches) and pessimistic mid-century (right; 10.1 inches) 10-year storm scenarios. Constructed stormwater ponds are highlighted according to remaining volume for stormwater storage: adequate (green), adequate but less than 10% storage volume remaining (yellow), or volume exceeded such that pond overtops (red). Areas of the landscape where flood volumes are expected to accumulate include streets (pink), public open spaces (orange), and naturally occurring lakes, wetlands, and stream (blue).
How can we adapt? SWMM was also used to test the efficacy of various adaptation pathways with the goal of maintaining a similar level of service in either pipeshed. Adaptation options included (1) enlarging pipe diameters to accommodate predicted increases in runoff peak flows, (2) increasing surface storage and infiltration through the implementation of LID, and (3) utilizing over-curb surface in areas where structures would not be impacted. In SWMM, LID was modeled by defining a bioinfiltration unit sized to capture 1-inch of runoff from impervious surfaces. Multiple LID scenarios were then modeled by assigning these bioinfiltration units to 10%, 15%, 20%, 25%, and 100% of impervious surfaces to represent a low to high intensity application of LID.
Figure 5 depicts the impact of adaptation options on flood volume over the range of mid-century precipitation scenarios developed for the project. In the Minneapolis pipeshed, flooding in excess of that allowed in streets and over curbs could be completely eliminated through pipe-upsizing up to a storm depth of about 5.7 inches, a storm depth that is less than the projected moderate scenario for the mid-21st century. However, backwater effects in the lower end of the watershed precluded further pipe enlargement to convey peak flows associated with larger storms. While flooding was not eliminated through LID for any precipitation scenario, modest reductions in flood volume were predicted. The majority of flood reduction benefits were achieved through the 10% LID scenario (reductions of 55% and 35% for the optimistic and moderate precipitation scenarios, respectively; 7% reduction under the most pessimistic scenario). Due to the lack of available land at the surface for additional large detention facilities, underground storage practices would likely be needed to manage the remaining excess flood volume in the Minneapolis watershed. In Victoria, flooding could be eliminated completely through pipe upsizing. An alternative for consideration would be over-curb storage, which model results indicated would also be sufficient to manage increases in surface runoff across the range of precipitation scenarios. Modeled LID scenarios had minimal impact on flooding, likely due to the high clay content and low hydraulic conductivity of the soils (0.17 inches/hour) modeled at this site.
Figure 5. Effectiveness of adaptation strategies for reducing the volume of surface flooding associated with projected increases in the 10-year design storm in the Minneapolis (top) and Victoria (bottom) study sites.
What is the cost of adaptation? Cost curves were developed to reflect the upper and lower bounds of costs expected to install larger pipes or underground storage reservoirs using data from the City of Minneapolis. Costs associated with the construction and maintenance of bioinfiltration facilities were obtained from Weiss et al. (2007), and from actual cost data for recently completed projects by the City. The resulting costs to eliminate excess flooding through adaptation measures outlined in the previous section across the spectrum of mid-21st century climate scenarios is presented in Figure 6. In Minneapolis, adaptation costs for the moderate climate scenario ranged from $40 to $70 million across the 1100 ac pipeshed; under the most pessimistic scenario, costs could be as high as $140 million to eliminate surface flooding. Expected costs to construct and maintain bioinfiltration facilities were lower than pipe-upsizing and underground storage costs. Accordingly, inclusion of LID – peak flow and flood reductions by which offset some of the need for pipe-upsizing and underground storage – resulted in a 50% - 55% reduction in adaptation costs for the moderate mid-century climate scenario. In Victoria, pipe-upsizing costs to maintain current levels of service ranged from $16 to $30 million for the most pessimistic scenario. To reiterate, over-curb storage may be an alternative.
Figure 5. Upper and lower bounds of projected adaptation costs (in millions of dollars) to eliminate over-curb flooding in the Minneapolis pipeshed, by pipe-upsizing and underground storage only (Mpls – No LID) or with a combination of pipe-upsizing, underground storage, and bioinfiltration units for 10% of the watershed impervious area (Mpls – 10% LID). Costs to completely eliminate surface flooding in the Victoria study site (Victoria – No LID) are shown for comparison. LID had minimal impact on flooding and pipe-upsizing requirements in Victoria and is not shown.
While the magnitude of changes in climate patterns, particularly with respect to the frequency of large events, is uncertain, the frequency of very large events has already increased by 45% in the upper Midwest, and emissions trends favor a pessimistic future. We can be certain that these changes are and will continue to create challenges for stormwater management and the protection of property and life. The Minneapolis and Victoria case studies presented here demonstrate that addressing this challenge will likely require an integrated approach as opposed to simply upsizing infrastructure. The relative resiliency of Victoria’s stormwater system can be attributed to the City’s development policies such as buffer setbacks and wetland conservation. This case highlights the importance of preserving or restoring hydrologic functions of such ecosystems in adapting urban/suburban landscapes to climate change. These studies also show that most existing systems already have vulnerability not intended by original designs, but that even under pessimistic assumptions a significant proportion of existing systems should be adequate. For more information on this project, including information on stakeholder engagement and community workshops, please visit http://www.minnehahacreek.org/project/weather-extreme-trends.
- Hershfield, D.M. 1961. Rainfall frequency atlas of the Unitied States for durations from 30 minutes to 24 hours and return periods from 1 to 100 years. Technical Paper No. 40. U.S. Department of Commerce, Washington, D.C.
- Weiss, P.T., Gulliver, J.S., Erickson, A.J. 2007. Cost and pollutant removal of storm-water treatment practices. Journal of Water Resources Planning and Management 133(3): 218-229.
We want to hear from you!!!
Let us know your thoughts, experiences, and questions by posting a comment. To get you thinking, here are a few questions:
- Within your spheres, are projected climate changes perceived as a challenge to stormwater management? If so, are adaptation strategies in place or being considered?
- Uncertainty in climate projections is often cited as a hurdle too large to allow for adoption of adaptation policies. Are there other challenges you foresee or with which you've dealt? What information (or research) do you think is critical to overcoming these hurdles?