Abstract
As part of the ongoing recovery from catastrophic flooding associated with Hurricanes Harvey and Florence, communities along the U.S. Atlantic and Gulf Coasts must plan for adaptations that integrate resilience into the redevelopment of community spaces. In designing resilient community spaces, particular consideration must be paid to socially vulnerable communities. Community visioning and neighborhood-scaled design is the first step in the development process. However, it can be a hurdle for communities that lack the capacity to conduct and evaluate such work collaboratively. Service-learning projects implemented through university–community partnerships can help communities increase resilience by developing master plans. The products generated through service-learning projects often conclude at the conceptual level, with no evaluation of the feasibility of their implementation. This study examines the use of landscape performance models in evaluating proposed master plan parameters. It is situated in Manchester, a community in Houston that is physically and socially vulnerable to flooding. Findings demonstrate that landscape performance models can objectively evaluate the costs and performance measures of service-learning products rooted in local conditions and community feedback.
INTRODUCTION
Cities around the world face the challenge of becoming more resilient in light of the combined effects of environmental degradation, deteriorating infrastructure, and increased frequency and intensity of climate-related hazards (Ahern, Cilliers, & Niemelä 2014; Reja et al. 2017). Developing a unified agenda to address these challenges and their effects on the built environment requires interdisciplinary methods. One approach to developing effective solutions focuses on collaborations between universities and communities through service-learning projects (Cerra 2016). However, modifications to the existing built environment are rarely easy or inexpensive, especially for communities already living at the social, economic, and political margins of society. Environmental justice focuses on how certain communities or social groups are unequally protected by environmental development, regulation, policy, and design because of their racial composition, ethnicity, national origin, and economic status (Bullard & Wright 2009; Wilson et al. 2008; Wolch, Byrne, & Newell 2014; see also Chang2019 in this issue). Social vulnerability focuses on how disasters differentially affect individuals and groups who are socially stratified based on race, ethnicity, income, disability, sex, age, nationality, or housing tenure and disadvantaged in their ability to resist, prepare for, respond to, and recover from hazards (Cutter, Boruff, & Shirley 2003; Peacock, Morrow, & Gladwin 1997; Van Zandt et al. 2012).
Flooding is a major issue faced by many socially vulnerable urban communities (Oulahen et al. 2015; Rufat et al. 2015). Socially vulnerable neighborhoods tend to experience greater damage during flood events than do other communities (Highfield, Norman, & Brody 2013). Many of the root causes of inequitable flood damage outcomes are linked to environmental injustices and differences in the built environment and across communities. Social forces and settlement patterns can lead to development on cheaper land, the isolation of specific demographic groups, depopulation/abandonment of neighborhoods, deflated tax bases, and built environments in disrepair (Hirsch et al. 2016; Mohai & Saha 2006; Van Zandt 2007). These development patterns are typically coupled with neglect, disinvestment, and a failure to update and retrofit neighborhood infrastructure (Isaacson et al. 2014; W. Liu, Chen, & Peng 2014).
Historically, planning for built environments has been done by expert-driven groups without community involvement (Ahern, Cilliers, & Niemelä 2014; Opdam et al. 2015). An engaged, participatory, and collaborative approach to urban planning and design can improve the development of knowledge and understanding, attain more effective solutions, and address issues related to equity, justice, and inclusion (Hendricks et al. 2018; Masterson et al. 2019; Meyer et al. 2018). For these reasons, there has been an increase in the use of service learning, a pedagogy fostering interactive learning wherein students engage in providing a service to communities through the integration of academic objectives and community activities (Helms et al. 2015). There is a need for more of these efforts, particularly in disadvantaged communities.
Service-learning projects can be critical to planning redevelopment of the built environment in underserved communities. In many cases, universities collaborate with resource-constrained communities to help them develop designs and growth plans that address neighborhood-scaled issues. Often, in academic settings, designs produced through service-learning efforts conclude at the conceptual level and limit attention to issues associated with implementation, such as liability, licensure and permitting, and competition with professionals in the planning and design fields. Furthermore, there is usually little or no measurement or evaluation of the effects of the implementation of these projects. Although it is not currently standard in the academic practice of landscape architecture and urban planning, there is a need to evaluate the potential outcomes of the product(s) generated by service-learning projects.
The Landscape Architecture Foundation (LAF) has developed a series of Landscape Performance Tools (LPTs) to measure the effectiveness with which designed/planned solutions fulfill their intended purpose and help designers and planners assess proposed community conditions (Yang, Shujuan, & Binder 2016). The LAF’s Benefits Toolkit offers a broad range of calculators to analyze social, hydrologic, and economic effects of design and plan parameters. Practicing professionals use these tools to determine the performance impacts of built projects, but to our knowledge, they have not been used to evaluate service-learning projects at the conceptual level. Because a majority of these calculators use area inputs in formulae/algorithms to calculate impact outputs, they are useful in evaluating conceptual designs.
Much of the literature on the benefits of design/planning based service-learning projects has focused on student learning outcomes, rather than the benefits that accrue to the community or the performance of the deliverables. Little research has measured the impact of plans/designs created through service learning. The potential for actual implementation of these projects is not well understood. Case studies applying the LAF’s Toolkit primarily focus on implemented projects, with few examples of the application of these tools in socially vulnerable communities (Newman et al. 2018). We seek to address these gaps by using LPTs to evaluate green infrastructure (GI) adaptations for a service-learning project in a flood-prone marginalized community in Houston, Texas.
LITERATURE REVIEW
Service Learning and Marginalized Communities
The implementation of service-learning projects in marginalized communities has increased in recent years, partly to address the problem of urban green space equity (Pearce & Manion 2016). Service learning is an alternative teaching model (Cameron et al. 2001) that goes beyond traditional lecture based teaching to generate learning through engagement with social groups (Hill 2005). Through experiential learning, service learning provides students with real-world opportunities to develop and integrate professional knowledge and skills into a design process in an actual community setting (Hou 2014). Marginalized and underserved communities can work with universities to develop designs to help solve neighborhood issues (Chou 2018). Through service-learning studios, students interact with local, underserved communities, organizations, and stakeholders to design space cooperatively with the involvement of community members who will use the spaces (McNally 2011).
Service-learning projects permeate a wide range of courses in higher education, including ecosystem services, urban planning and design, landscape architecture, construction science, sustainability, environmental justice, public health, and many other STEM (science, technology, engineering, and mathematics) and professional disciplines. They typically involve a series of collaborative actions with communities, including consultation, information sessions, data gathering, critical feedback, and guidance for enacting change (Glackin & Dionisio 2016). Service-learning studios help landscape architecture and urban planning students learn and apply the skills to work in an increasingly complex and multicultural context (Forsyth, Lu, & McGirr 1999). Criticism of university–community partnerships often occurs when they meet only the needs of academic units and fail to provide mutual benefits for nonacademic partners (Nixon & Salazar 2015). Hence, service-learning courses must respond to community-identified needs and ensure the delivery of valuable and helpful products that meaningfully evaluate community impact to ensure that the projects focus on more than just student benefits (Rinaldo, Davis, & Borunda 2015). In addition, a combination of student and faculty work enhances the quality of the planning and design products generated through service learning that provide useful design services to neighborhoods (Forsyth, Lu, & McGirr 2000). Service learning also allows future professionals (i.e., planning and design students) to develop vital skills for working with communities.
Design is a problem-solving venture, and service learning involves problem-based learning (O’Grady & Alwis 2012). Despite the positive benefits of incorporating service learning in design fields, objective evaluation of service-learning products is limited. Evaluating the community benefits of service-learning projects requires a holistic assessment method (Serow 2014). Calls for evidence-based practices in higher education for service learning have increased (U.S. Department of Education 2006). Methods of assessment must move beyond capabilities of assessing student learning into new models that also consider the products and effects in the communities within which the service-learning efforts are based (Steinke & Fitch 2003). The LPTs may provide part of the foundation for this holistic assessment. LPTs’ capability of evaluating the effects of GI could open up the possibility of creating concrete data to evaluate the potential impacts of conceptual designs created through service learning.
Impacts of Green Infrastructure
Natural hazards intensified by climate change will increasingly affect communities. Environmental mitigation strategies enable the development of sustainable and resilient communities (Newman, Brody, & Smith 2017; Tyler 2016). Although this approach is promising, there are many unknowns, both about implementation and about how communities might effectively design, plan, and pay for the installation of these features (Zellner et al. 2016). As one type of environmental mitigation strategy, GI practices improve stormwater management by reducing runoff volumes and impervious cover, decreasing and delaying peak discharge, preventing pollution, and recharging groundwater (Thiagarajan, Newman, & Van Zandt 2018). For example, a study done by Schubert et al. (2017) at the watershed scale showed that a full implementation of GI could reduce downstream flooding by an average of 91 percent across a variety of rainfall and storm events. GI can also reduce energy needs and costs and increase individual and community well-being by improving human health, neighborhood aesthetics, air quality, recreational opportunities, and property values. Increased exposure to trees, vegetation, nature, or green space in communities provides specific health benefits, such as reduced mortality, morbidity, stress, and mental fatigue (Kondo et al. 2015). Kardan and colleagues (2015) found that the siting of trees can positively affect cardio-metabolic conditions similar to that of an increase in annual personal income, moving to a wealthier neighborhood, or being younger.
Low-impact development (LID), a method of design and development that seeks to treat stormwater at its source, uses GI as a means of generating reductions in stormwater runoff (Martin-Mikle et al. 2015). LID reduces impervious cover and retains natural areas to reduce runoff volume and peak runoff rate, control flow frequency/duration, and improve water quality. For example, Damodaram et al. (2010) demonstrated that the use of GI in LID yielded significant stormwater control for small events. When combined with best management practices, GI practices provided runoff control for flood and frequent rainfall events. Likewise, Zahmatkesh et al. (2014) conducted a climate change impact study on urban stormwater runoff in the Bronx River watershed in New York City. Using models and simulations, they found that while the average increase in historical annual runoff volume under climate change impacts were approximately 48 percent, implementation of GI measures provided an average reduction of 41 percent in annual runoff volume. Application of GI also reduced peak flow rates by an average of 8 to 13 percent. A study completed by Sparkman et al. (2017) predicted that a LID watershed annually would remove 78 kg more nitrogen, 3 kg more phosphorus, and 1,592 kg more sediment per square kilometer compared with a watershed without LID practice implementation.
Although LID attempts to reduce increases in impervious surfaces that exacerbate flooding, GI can be a complementary technique, increasing the amount of nondeveloped space to allow for more stormwater absorption or infiltration (Pyke et al. 2011). There are numerous documented benefits of GI in cities, and it particularly benefits flood risk management and climate-change adaptation (Carter et al. 2017). For example, a national study of cities participating in the National Flood Insurance Program Community Rating System demonstrated that on average they save approximately $200,000 per municipality per year in flood-related losses by protecting open space in the 100-year floodplain (Brody & Highfield 2013). GI is also an effective approach for reducing stormwater runoff and pollutant contamination in bodies of water proximate to flooding (Martin-Mikle et al. 2015).
Social Equity of GI
Most residents show a strong willingness to implement GI in their neighborhoods. Key factors affecting citizens’ willingness toward implementation include efficacy, aesthetics, and cost (Baptiste, Foley, & Smardon 2015). Marginalized populations are likely to support GI to improve the overall quality and aesthetic of their community, but only if their current financial obligations would not be strained (Young 2011). Several recent studies have addressed these concerns about cost, but only at a regional scale. For example, a comparison of the costs of preventing floodplain development versus the costs of mitigating flood damages in the East River Watershed of Wisconsin’s Lower Fox River demonstrated that the costs of preventing projected floodplain development exceeded flood damage mitigation benefits; that is, the economic benefits of mitigation were less than the losses from flooding (Kousky et al. 2013). Y. Liu et al. (2016) found that to reach the same runoff volume and pollutant loads for projected 2050 land use patterns as in historic (2001) land use patterns, runoff volume had to decrease by an average of 12.5 percent. The corresponding average annual cost for implementing this GI in the watershed was $1.4 million. Although these costs may seem quite high, they are significantly less than that of gray infrastructure (e.g., contribution of levees and other flood control and protection devices).
The financing and cost reductions of GI tend to exceed implementation costs over time (Jennings, Johnson Gaither, & Gragg 2012). For example, Clements et al. (2013) found that the cumulative economic value of GI benefits can be millions of dollars over the long term, with benefits at both the individual structure and neighborhood scales. Over a forty-year period, the implementation costs are much less than the potential stormwater utility fee savings. In this case, direct benefits included increased rents and property values, retail sales, energy savings, stormwater fee credits, and increased mental health and worker productivity, while reducing infrastructure costs, costs associated with flooding, water bills, and crime (Clements et al. 2013).
Evidence suggests positive associations between GI, improved stormwater management, and public health (Clements et al. 2013; Kardan et al. 2015; Schubert, Burns, & Fletcher 2017). The environmental justice literature, however, demonstrates that benefits, including GI, are not always equitably distributed in society (Kweon, Marans, & Yi 2016). Disadvantaged communities are unequally protected by GI (Wolch, Byrne, & Newell 2014). For example, census tracts with higher poverty or greater percentages of minority populations had a lower availability of green spaces (Wen et al. 2013). Even in minority neighborhoods where green spaces and parks are available, Latinos, African Americans, and low-income groups in the Los Angeles metropolitan region were more likely to live close to parks with higher congestion (Sister, Wolch, & Wilson 2010). Similarly, in Kansas City, Missouri, low-income census tracts had significantly more parks but more quality concerns per park (Vaughan et al. 2013). Predominately African American and Latino census tracts have more parks with basketball courts, but fewer parks with trails and other ecological features. In Miami–Dade County, Florida, African American neighborhoods had the lowest tree density and leaf area index, tree and shrub cover, and tree and shrub diversity (Flocks et al. 2011). These neighborhoods received the least amount of ecosystem services in terms of air pollution removal and energy savings in comparison with neighborhoods populated by white and Latino groups. Public right-of-ways in neighborhoods with a higher proportion of African Americans, low-income residents, and renters in Tampa, Florida, had lower proportion of tree cover (Landry & Chakraborty 2009). B. Wright (2011) also demonstrated that in New Orleans after Hurricane Katrina, changes in infrastructure designed to protect areas from extreme flooding and storm surge were closely related to neighborhood racial composition. Findings reveal that in white and affluent areas, in contrast to African American and working-class areas, there was an increased levee height of 5.5 feet for flood protection.
RESEARCH OBJECTIVE
This article seeks to explain the development of a conceptual master plan produced through a service-learning process and project cost and performance measures for the design parameters using landscape performance models. To demonstrate the methods, we (1) describe a process by which service learning is used as a pedagogical approach in a landscape architecture studio to develop a master plan, and (2) use landscape performance modeling to evaluate the GI provisions as part of the overall master plan design. Manchester, a marginalized community located in southeast Houston, Texas, was the case study site (Table 1).
Demographic Characteristics of Manchester Neighborhood, Houston, and the State of Texas
Figure 1 shows the location and context of the Manchester neighborhood, locating it from the national and state scales to the county/city, and site scale.
Manchester neighborhood study site.
This study used two tools for measuring landscape performance: the National Green Values Calculator (GVC) and the Value of Green Infrastructure Calculator (VGI) (Jayasooriya & Ng 2014) to project the performance, costs, and benefits of GI for a conceptual master plan from a design produced in a service-learning studio (see Figure 2). Because most service-learning projects involving academic institutions conclude at the conceptual level, the performance calculations evaluated GI proposals embedded in a conceptual master plan, not a built project.
Conceptual framework that outlines the process and research questions for the study.
METHODS
Study Area
Over the past fifty years, Houston has had one of the largest number of flood-related fatalities in the United States (Highfield, Norman, & Brody 2013). It is frequently affected by coastal storms (Berke et al. 2015). Houston is not only physically vulnerable to storms, it has many communities that are socially vulnerable. According to the U.S. Census (2016), Houston’s Manchester neighborhood is 82 percent Hispanic, with 40 percent of its households making less than $25,000 annually (the poverty threshold in Houston is $24,339 a year for a family of four). Built before 1960, 64 percent of the housing stock has a median house value of $75,000, and 44 percent of the neighborhood’s residents have no high school diploma. Transportation infrastructure and industry surrounds Manchester, causing high river impairment/low water quality in the Buffalo Bayou. An impaired body of water fails to meet one or more water quality standards per the U.S. Clean Water Act, which protects water bodies from bacteria or pollutant loads before it is no longer drinkable, swimmable, fishable, or usable in other beneficial ways. Manchester is within a mile of twenty-one industrial facilities that must report to the Toxic Release Inventory of the U.S. Environmental Protection Agency because they release toxic chemicals to the air, water, or land (EPA n.d.). Texas Environmental Justice Advocacy Services (t.e.j.a.s.), a local community organization, provides Toxic Tours for outsiders to illustrate the neighborhood’s adverse environmental conditions. These tours include documentation of issues related to flood vulnerability, drainage, and air and water pollution. Sixty-four percent of the neighborhood’s surface is impervious, and only 2 percent contains park/open space. Compared to occupied parcels that are entirely used, overall land use is characterized by an excess amount of vacant parcels (16 percent) (Figure 3).
Manchester existing site conditions.
Engagement Process
Enhancing flood resilience in socially vulnerable populations is most effective with local community participation (Stevens, Berke, & Song 2010). Working in conjunction with local community groups in Manchester, including t.e.j.a.s, Charity Productions, and Furr High School’s Green Ambassadors, the participatory process connected local citizens with university partners to integrate educational and community needs through feedback loops supporting climate justice–based planning and design. Climate justice is an expansion of environmental justice that seeks fair application of environmental laws and strategies in direct response to climate mitigation and hazard adaptation, ensuring equitable and just transitions regardless of race, ethnicity, class, or national origin. A series of community engagement sessions developed urban growth plans with GI provisions. An initial session allowed residents to voice community issues and help spatially locate areas of concern. Subsequent discussions solicited resident feedback and focused on informing residents about how GI application might alleviate community issues.
A second meeting involved presenting an inventory and analysis of existing conditions to the community. Feedback from the community provided valuable insight on hyperlocal conditions that would normally require “boots on the ground” and informed creation of a portfolio of community design ideas worked into a conceptual master plan. A series of follow-up meetings incorporated a feedback loop between community members and the design team, which allowed for the presentation of numerous design scenarios and critiques by stakeholders and community members. Feedback from each design scenario guided development of a synthesized master plan, which was a condensed version of relevant ideas extracted from each scenario (Figure 4).
Manchester master plan developed through the service-learning process.
The process for producing the master plan used a “redevelopment” approach that allowed residents to give their opinions on existing conventional development they wanted to keep, existing conventional development they wanted to remove, or new conventional or new GI development they wanted to add. Conventional development refers to traditional “gray” development or features not considered GI and used as such throughout this article. Recommendations from faculty and students on what existing features seemed to work well and the opinions of the community facilitated decision making on what to retain. Displacement was a major concern of the community, resulting in a decision to retain most of the existing single-family housing. Adding new conventional or GI development focused on the land left over from vacant properties, underused parcels, and removal of unwanted land uses and blighted properties. Evaluation of the new master plan involved placing both conventional and GI development and design parameters into the landscape performance models. They were evaluated simultaneously as a complete master plan design. Therefore, the cost and performance outputs include the total costs and performance of conventional and green features for the final plan for the whole site. The landscape architecture studio faculty, graduate research assistants, and students in the course completed most of the landscape performance modeling. Although the projected costs and performances were not incorporated into the final master plan design, we identify opportunities for future research and community feedback to make adjustments to the design presented here.
Landscape Performance
The economic rationale of GI is emerging as an essential component of flood provision strategies (Jayasooriya & Ng 2014). For example, the Center for Neighborhood Technology’s (CNT) National Stormwater Management Calculator, also known as GVC, is an online tool developed to compare costs, benefits, and performance of GI compared with conventional stormwater management practices (Foster, Lowe, & Winkelman 2011). It allows one to assess the effectiveness of stormwater management practices on water quality and hydrology (Y. Liu, Bralts, & Engel 2015), construct adaptive capacity for flood proofing urban areas (Wen et al. 2013), predict green roof runoff capture (Carson et al. 2017), assess GI impacts in new housing developments (Wright et al. 2016), and evaluate stormwater runoff storage for urban community gardens (Gittleman et al. 2017).
The GVC includes a step-by-step procedure that allows users to specify runoff reduction goals for individual sites across the United States. The procedure involves entering area calculations for each GI provision. Users can input site-specific design parameters, including land cover distribution, soil type, runoff reduction, and types of GI (Jayasooriya & Ng 2014). Based on land use percentage inputs, the GVC calculates the volume of runoff. The procedure calculates the impact of GI on infiltration rates, evapotranspiration amounts, and stormwater runoff reuse by modeling the ability of each type of GI used to capture runoff (Kauffman et al. 2011). It estimates construction and maintenance costs for each type of GI in the design/plan, providing a total life cycle cost for a given project (benefits minus costs). The tool’s cost module allows users to analyze GI costs and benefits over 5-, 10-, 20-, 30-, 50-, and 100-year periods. Using current literature and applicable industry data, this tool estimates cost values and variations for GI maintenance construction (CNT 2016).
Another CNT tool, the VGI: A Guide to Recognizing its Economic, Environmental and Social Benefits, allows assessment of the economic benefits provided by GI (CNT 2010). The VGI helps decision makers understand, quantify, and assign economic value to GI practices and investments. It estimates economic benefits for water, energy, air quality, and climate change mitigation benefits for green roof, permeable pavement, and rainwater harvesting technologies (Shafique & Kim 2017; Zhan & Chui 2016).
This study used the GVC to measure projected stormwater volume control and to quantify construction and annual maintenance costs for the Manchester site in the service learning–generated scenarios. The GVC accurately assesses the specific volume control performance and effectiveness of the master plan scenario for the site, and it provides detailed construction and annual maintenance costs for streets, parking lots, stormwater storage, conventional and green roofs, permeable surfaces, and rain gardens. To measure the projected changes, we input existing site conditions followed by the newly developed master plan design parameters. We used the VGI to estimate these benefits because it provides detailed equations to calculate values from the specific types of GI recommended in the master plan, including green roofs, rain gardens, rainwater harvesting areas, and pervious surfaces. VGI assigns a monetary value to the benefits provided by the GI areas in the site.
FINDINGS
Projected Performance of Proposed GI Strategies
If implemented, the green practices in the master plan for the design site in Manchester will increase pervious surface area from 36 percent to 51 percent. The public park space area will increase from 2 percent to an estimated 13 percent, (nearly seven times its initial amount). The master plan added nine green roofs (103,938 square feet), three rain gardens (30,242 square feet), and seven areas to harvest rainwater (177,090 square feet) (Table 2).
Green Practices in Manchester Neighborhood Master Plan
In applying the GVC to the projected runoff volume control calculation for impermeable surfaces, this study applied the North Carolina Ordinance (NCG01, State of North Carolina Department of Environment and Natural Resources Division of Water Quality General Stormwater Permit) standard (1.5 inches of precipitation depth) as a runoff volume reduction goal. Application of the green stormwater best management practices (BMPs) in creating the post–master plan scenario decreased impermeable area on the site by 15 percent and captured 82.2 percent of the required runoff volume. The required volume captured from 1.5 inches over impermeable surfaces is 272,108 cubic feet. Table 3 presents the total runoff volume captured by current BMPs.
Projected Volume Control in the Master Plan as Projected by the GVC
Compared with conventional approaches (without the addition of GI), the green practices tested in this analysis decreased the total average annual rainfall runoff (and runoff volume) by 11 percent (from 16.02 inches and 6,180,442 cubic feet to 14.27 inches and 5,507,669 cubic feet). For a 90 percent storm event, the difference is much more dramatic, with the GI practices reducing runoff (and runoff volume) by 77 percent (from 0.32 inch and 124,385 cubic feet to 0.08 inch and 28,951 cubic feet) compared with conventional development without GI (Table 4).
Comparison of Rainfall Runoff Retention Performance for a 90 Percent Storm Event Between GI Practices and Conventional Development Without GI Practices
Calculated Benefits of Proposed GI Strategies
Calculating the benefits of implementing the four types of GI (green roofs, rain gardens, rainwater harvesting areas, and pervious surfaces) used VGI (CNT 2010). The publication contains algorithmic models for calculating estimated outputs based on mathematical manipulation of specified quantitative inputs. It further divides some of the GI benefits into subservice types. Calculating each GI benefit requires two steps: benefit quantification and benefit valuation. The product of these parameters is the annual benefit (in U.S. dollars) for each benefit subservice type. For example, the following paragraph explains operation of these models for estimating the GI benefits of green roofs.
In CNT (2010), green roofs provide four types of subservices: reducing stormwater runoff, reducing energy use, improving air quality, and reducing atmospheric CO2. Benefit quantification for reducing stormwater runoff uses the annual stormwater retention performance equation (CNT 2010). Calculating benefit valuation for reduced stormwater runoff uses the value of annual avoided treatment cost equation (CNT 2010). Following the same pattern, benefit quantification for reducing energy use is calculated using the buildings’ annual cooling savings and the annual heating natural gas savings equations (Clark, Adriaens, & Talbot 2008). Benefit valuation for reducing energy use is calculated using the value of buildings’ annual cooling savings and the value of buildings’ annual heating natural gas savings equations (U.S. EIA 2010). Calculating benefit quantification for improving air quality uses the annual direct NO2 uptake and annual indirect reduction in NO2 equations (CNT 2010). Benefit valuation calculations for improving air quality use the value of total annual NO2 benefit equation (U.S. EIA 2010). Calculating benefit quantification for reducing atmospheric CO2 employs the total annual climate benefit equation (CNT 2010). Finally, benefit valuation for reducing atmospheric CO2 uses the value of total annual climate benefit equation (CNT 2010). Table 5 presents the equations used in estimating these benefits and their sources.
Green Infrastructure Benefit Quantification and Valuation Equations
The total annual benefit of green roofs is the sum of the four subservice type categories of benefits, which in this case is $15,975 (see Table 6). Annual benefits for the subservice types are as follows: reducing stormwater runoff =$4,621; reducing energy = $9,468; improving air quality =$786; and reducing atmospheric CO2 = $1,100. Following the same calculation process, the annual benefit of rain gardens, water harvesting, and pervious surfaces of the study area are $104,452, $8,630, and $96,775, respectively. The total annual benefit of the four types of GI proposed in the projected Manchester neighborhood site is $225,832 (Table 6).
Green Infrastructure Annual Benefits in the Study*
Construction and Maintenance Costs
Table 7 presents the estimated construction costs for implementing the final master plan developed through the service-learning process in the Manchester neighborhood. These expenses total $22,807,215. The difference between conventional development and GI development accounts for28 percent of the total cost ($6,573,881). Estimates of the total construction cost for conventional development in the Manchester site are $14,690,548, while the projected total construction cost with respect to the GI area is $8,116,667. Because the GVC defines conventional development as development with no GI, the model calculates GI costs as additional expenses, which are typically cheaper than gray infrastructure implementation. The construction costs for the GI components of the design are as follows: green roofs: $1,633,862; permeable pavement: $6,271,074; and rain gardens: $211,696 (Table 7).
Comparison of Annual Construction Cost Between Conventional and Green Development
Total annual maintenance cost for the proposed plan at the Manchester site is $294,322. Although the addition of all GI features in the master plan makes construction more expensive than conventional development, the plan realizes savings in life cycle costs. The maintenance cost for conventional development is $249,650, whereas the projected total annual maintenance costs for GI is only one-fifth of the conventional development total ($44,672). The annual maintenance cost for GI components are as follows: green roofs: $2,593; permeable pavement: $31,797; and rain gardens: $10,282 (Table 7).
The GI components in the Manchester site have a projected total construction cost of $8,116,667 and the projected total annual maintenance cost is $44,672. According to the analysis, the total annual benefit of the four types of GI in the site is $225,831. This results in a return period of forty-five years for the overall construction and yearly maintenance costs of GI, controlling for new property tax revenue and local economic benefits (equation 9 in Table 5).
The area for permeable pavement in the input data is composed of the three major arterial streets and their corresponding sidewalks. These areas defined the location of the primary flow paths and stormwater settling points. The master plan for the site transformed these streets into eco-boulevards and green streets, according to LID best practices. The reason maintenance costs of conventional roofs are much higher than green roofs is that the GVC takes into account existing and proposed roofs. There are currently 332 total roofs in the master plan (totaling 1,025,518 square feet) with only 9 green roofs proposed (totaling 103,938 square feet). This causes a large discrepancy in maintenance costs between the plans.
For the entire Manchester neighborhood (conventional and GI), the projected total construction cost is $22,807,215, while the projected total annual maintenance cost is $294,322. The total annual benefit is $225,831, all of which comes from GI. The maintenance cost is prohibitively high, such that there is no true rate of return or annual percentage return realized on an investment, which is adjusted for changes in prices due to inflation or other external factors. Nevertheless, controlling for new property tax revenue and local economic benefits, within 100 years for GI, the expected return will be 43.2 percent of the total construction and maintenance costs. This only accounts for returns for the GI itself, not any local revenue from new commerce or rents from new commercial and residential land uses. The return time for 50 percent of the construction and maintenance costs of the neighborhood is 145 years (see equation 10 in Table 5). The return describes the money made on the GI investment over this period of time, controlling for new property tax revenue and local economic benefits. This is important to mention because the master plan design parameters do not include an annual return on the investment of GI. However, over a longer period of time, a return is realized mainly from the initial upfront costs of GI—an important note for residents, designers, planners, and investors.
CONCLUSIONS AND DISCUSSION
This case study sought to understand the capabilities of landscape performance models in objectively evaluating the potential impacts of products developed through a service-learning process. Findings suggest that service-learning processes can help in developing community-informed master plans that include conventional and GI applications. Landscape performance models can also be useful tools for providing objective assessments of the service-learning products generated, specifically in regards to the projected costs and benefits for proposed GI designs. This can be especially important to socially vulnerable neighborhoods that are often economically strained and contain deflated tax bases. Through established university–community partnerships, service learning can provide opportunities for a reciprocal learning process that results in an enhanced understanding of complex issues by all participants. It creates mechanisms to empower communities and enhance both community and student capacities (Shiel et al. 2016). Service-learning initiatives around community design and planning for GI can contribute to building more resilient communities, particularly in physically and socially vulnerable areas.
The VGI and GVC tools provide equations to calculate the specific benefits of GI components, including green roofs, permeable pavement, and water harvesting. In this case, the benefits of GI from overall performance and benefit quantification and valuation are consistent with the literature on GI. Increases in GI produce significant reductions in the adverse effects of flooding in areas with greater impervious surface amounts (Brody & Highfield 2013; Tyler 2016). The imperative is not just to address the value of GI but also to understand initial construction and ongoing maintenance costs. Maintenance expenditures are often not included in the GI conversation. However, they are particularly important for marginalized communities.
There are important limitations of this study and its findings. First, the validity of the reported results depends on the legitimacy of the performance tools used. As noted, most LPTs rely on area-based inputs specified in design proposals and existing conditions. Consequently, users of these tools are dependent on formulae and/or algorithms incorporated into each model. Therefore, using proven and peer-reviewed models such as the VGI and GVC increases accuracy and reliability of outputs. Because of the many tools available (and the number continues to grow), designer(s) (including professional designers and community groups, or some combination thereof) must choose the most appropriate calculator for their project based on its specific goals. For this project, because additional green space and flood resilience were needed in the community, a GI benefits–based calculator was the most suitable, as it allowed us to have a more detailed and robust understanding of the climate and hazard-related specific benefits of GI. Finally, the generalizability of this approach will be increased by more studies to verify our findings. Testing the impacts of changes in the landscape and different design strategies through comparisons of pre- and postconstruction of GI features and applications in differing locations will provide needed evidence for using LPTs as a component of conceptual service-learning projects.
It is also important to note that the model uses a black box of algorithms that process area-based input data (Table 5 provides a list of the equations used in this study) to provide outputs (Table 6). Continuing evaluation of the ability of these predictive models of landscape performance to generate accurate and reliable results in comparison with field-based performance measurements is essential.
The limited number of performance evaluation tools used in calculating GI values also limits the utility of this study. The breadth of available performance tools other than the GVC and the VGI affords rich opportunities for interdisciplinary collaboration around estimating and assessing the construction and maintenance costs of GI versus conventional project scenarios, particularly in the context of service learning. For example, future projects could include interdisciplinary student and faculty collaborators from other academic programs (such as business, real estate development, construction management, or engineering) to provide cost-benefit analyses and address financial issues arising from plans generated by service-learning projects. This especially applies to projects requiring substantial financial support, such as GI and LID (Newman et al., 2016; Sohn et al. 2014). This would help programs enhance student and faculty exposure to methods for calculating these costs and benefits. It would also provide a greater level of confidence in the validity of the results and give students experience in integrating GI concepts into the design of collaboratively developed community designs. Students would learn how to integrate the landscape performance assessment technology into the process of design development.
Another limitation of this study is that we did not complete a follow-up design that incorporated resident feedback based on the landscape performance model projections. Future research might include several iterations of designs that incorporate community and student responses to estimated costs and performances. This would be a significant contribution to understanding how communities ration GI costs versus benefits with the goal of effectively managing costs and maximizing benefits. Follow-up research must introduce the LPTs to influence the conversation between students and the community. This will significantly expand the value of landscape performance modeling and demonstrate its use in evaluating designs as they unfold.
Although the use of projects generated through service-learning design efforts coupled with GI performance evaluation modeling may allow communities to clear the first hurdle in the adaptation and development process, we also recognize that other hurdles exist, particularly in the form of financing implementation of the designs. Moving a project from design to implementation requires additional work to determine opportunities to affect the development process in disadvantaged communities. As noted earlier, university service-learning efforts often lack the capacity to move a project from design and planning phases into actual implementation.
Universities often purposefully avoid completing construction documents for community partners because of liability issues (e.g., students are not qualified to prepare a full set of construction documents for complex projects) and a belief that service-learning projects should not compete with design professionals in preparing community design proposals. This is a particular concern for socially vulnerable communities that have historically lacked access to green space and frequently do not have the economic resources required to move recommendations forward. In these situations, the “business as usual” urban growth trends continue and conditions tend to remain constant or deteriorate. Underserved communities must become aware of funding options that may be available to help implement GI plans/designs.
At the federal level, the U.S. EPA Environmental Justice Small Grants Program supports communities working on solutions to local environmental issues. Designed to help communities understand and address exposure to environmental risks and associated public health issues, this program funds project budgets under $30,000. The U.S. EPA also supports the Office of Sustainable Communities Greening America’s Communities Program, which helps cities and towns develop an implementable vision of environmentally friendly neighborhoods incorporating innovative GI practices. To succeed in securing funds from these grant programs, having design proposals (such as those produced by service-learning projects) in hand is a real asset for local communities in securing funding for implementation. Locally, community-based public–private partnerships and alternative market-based tools for using integrated GI also provide guidance for local governments on how to develop similar partnerships and use similar tools. The National Association of Flood and Stormwater Management Agencies’ Guidance for Municipal Stormwater Funding addresses procedural, legal, and financial aspects of creating viable funding approaches for local stormwater programs. It includes approaches for developing service, user, or utility fees to support integrating stormwater management into community design. Tools for facilitating this integration include taxes, fees, creation of utilities, credits and incentive programs, rebates and installation financing, bonds, grants, loans, and public-private partnerships.
To be most effective, GI must be grounded in local conditions and have the financial and social support of the community. These mandates imply that adopting GI practices into community design must also produce an increase in community capital, especially in socially vulnerable communities. Community capital includes the local knowledge used to establish sustainable community development as a democratic process representing a community’s unique characteristics and needs. Municipalities that support building capital in communities can use GI evaluation tools to balance budgets and allocate financial resources effectively, efficiently, and equitably. They must also recognize that financial investments in community capital might return in forms other than economic capital (Callaghan & Colton 2008).
Service-learning projects allow universities to provide design services at very reduced or no cost to underserved communities. The design products generated by university programs often have value that far exceeds the costs to support community partners. These products also have high value for community partners as they apply for grants, write proposals for professional design services, and eventually implement ideas rooted in the original student proposals.
The goal of environmental justice–based designs should be to make environmental protection and decision making more accessible to all sectors of the community. Accomplishing these objectives raises ethical and political questions such as who gets what, how, and in what amount. Addressing these questions requires bottom-up insight from the communities on local conditions and top-down design support services and partnerships between public agencies, private enterprises, local community groups and stakeholders, local government, and academic partners. The fields of urban planning and landscape architecture must evolve to balance creativity and costs to make large-scale GI projects possible.
The landscape performance modeling method demonstrated in this article provides an opportunity for conversations about economic feasibility. As service-learning projects and other university-community partnerships continue to grow in popularity, designing and evaluating GI adaptations is another example of the value of these pedagogical and methodological approaches, particularly in socially vulnerable communities.
AUTHOR CONTRIBUTION
Each of the authors included on this article made critical and substantive contributions to the development of the manuscript. Author Hendricks was the primary lead in initially engaging with the communities in which the site plan was developed for, building relationships with the communities, and identifying stakeholders that were involved in the service-learning component of the project. Furthermore, Author Hendricks specifically framed the research questions for this manuscript, completed the literature review, and helped to model and interpret the findings. Author Newman was the lead instructor for the service learning studio, the lead landscape architect on the project, and helped to identify the landscape performance tools. He also helped with writing the methods and other sections of the paper. Author Yu was a graduate student supporting with this work and helped to run the landscape performance models including extracting the design parameters and placing them in the model. Lastly, Author Horney was a public health expert on the project and helped to review the manuscript and make final edits and adjustments.
PEER REVIEW STATEMENT
This submission was peer-reviewed by four peer reviewers selected by the Editorial Office. Their contributions are gratefully acknowledged and appreciated.
ACKNOWLEDGMENTS
We thank the former and interim editors of Landscape Journal and the rest of the editorial board for their support throughout the publishing process. We also thank the reviewers for their thoughtful comments, which significantly improved the final manuscript.
Footnotes
Marccus D. Hendricks is an Assistant Professor of Urban Studies and Planning in the School of Architecture, Planning, and Preservation and a faculty affiliate with the Maryland Institute for Applied Environmental Health in the School of Public Health at the University of Maryland in College Park. He is also affiliated with the Clark School of Engineering’s Center for Disaster Resilience, the National Center for Smart Growth Research and Education, and the Environmental Finance Center. His research interests include infrastructure planning and management, hazard mitigation, sustainable development, public health and the built environment, and participatory action.
Galen Newman is an Associate Professor at Texas A&M University, where he also serves as Program Coordinator for the Bachelor of Science degree in Urban and Regional Planning, Associate Department Head for the Department of Landscape Architecture and Urban Planning, and Director for the Center for Housing and Urban Development.
Siyu Yu is a doctoral student in the Urban and Regional Sciences program at Texas A&M University. She serves as a Graduate Research Assistant and examines plan evaluation, ecological vulnerability, statistical analyses, and GIS-based analytical methods.
Jennifer Horney is a Professor of Epidemiology and a Disaster Research Center core faculty member at the University of Delaware. Her research focuses on measuring the health impacts of disasters, as well as the links between disaster planning and household actions related to preparedness, response, and recovery.
REFERENCES
This article requires a subscription to view the full text. If you have a subscription you may use the login form below to view the article. Access to this article can also be purchased.
Log in using your username and password
Purchase access
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.