Situating Green Infrastructure in Context: A Framework for Adaptive Socio-Hydrology in Cities

Abstract

Management of urban hydrologic processes using green infrastructure (GI) has largely focused on stormwater management. Thus, design and implementation of GI usually rely on physical site characteristics and local rainfall patterns, and do not typically account for human or social dimensions. This traditional approach leads to highly centralized stormwater management in a disconnected urban landscape, and can deemphasize additional benefits that GI offers, such as increased property value, greenspace aesthetics, heat island amelioration, carbon sequestration, and habitat for biodiversity. We propose a Framework for Adaptive Socio-Hydrology (FrASH) in which GI planning and implementation moves from a purely hydrology-driven perspective to an integrated socio-hydrological approach. This allows for an iterative, multifaceted decision-making process that would enable a network of stakeholders to collaboratively set a dynamic, context-guided project plan for the installation of GI, rather than a ‘one-size-fits-all’ installation. We explain how different sectors (e.g., governance, non-governmental organizations, academia, and industry) can create a connected network of organizations that work towards a common goal. Through a graphical Chambered Nautilus model, FrASH is experimentally applied to contrasting GI case studies and shows that this multi-stakeholder, connected, de-centralized network with a co-evolving decision-making project plan results in enhanced multi-functionality, potentially allowing for the management of resilience in urban systems at multiple scales.

1. Introduction

Human society shapes all components of the water cycle [Pataki et al., 2011; Sivapalan et al., 2012; Vogel et al., 2015], and especially in cities. Impervious surface cover is the predominant influence on the partitioning of precipitation inputs in the urban hydrologic cycle [Rodriguez-Iturbe, 2000]. The direct runoff from hard surfaces during rainfall events led to the need for centralized drainage systems to mitigate and otherwise control urban flooding. Intentional reorganization of the water cycle as a part of city building is a centuries old practice. As early as 600 BC, drainage ditches along houses and streets moved combined runoff and sewage to centralized collection systems like the Cloaca Maxima built in Rome [Delleur, 2003]. Until the Middle Ages in Paris, France, drinking water was taken from the Seine River, while sewage was conveyed to the city’s fringes and infiltrated into rural areas, where the local water cycle was closed with the Seine River recharged with baseflow that had been filtered through the streambank soils [Delleur, 2003]. For more than 100 years, the City of Berlin, Germany pumped freshwater from the Spree River, pulling streamflow through riverbanks for filtration [Jaramillo, 2012]. In the last century, the increasing recognition that unsealed urban surfaces, soil and vegetation can provide hydrological benefits to cities, has fueled the interest in novel stormwater management approaches, such as best management practices (BMPs), low-impact development (LID), nature based solutions, and water-sensitive urban design (WSUD) among others [Brown et al., 2009; Fletcher et al., 2015]. These terms reference the set of technologies and practices that we now collectively refer to as green infrastructure (GI). Green infrastructure, can be broadly defined as green space that promotes access to recreational space, can preserve biodiversity, and leverages continuity of ecosystem processes and functions of ecological systems to regulate and manage technical problems like stormwater quality and quantity, as well as urban heat [Sandström, 2002; Tzoulas et al., 2007]. Green infrastructure in many urban planning schemes is, however, principally about stormwater management, whether this revolves around flood mitigation, maintaining wastewater system capacity, or water quality improvements to surface waters [Jarden et al., 2016; Shuster et al., 2017]. To manage stormwater, GI enhances retention, infiltration, and evapotranspiration through amendment of soil and plant systems in the urban landscape to redirect water away from wastewater pipes or constructed channels (i.e., gray infrastructure) [Berland et al., 2017; Shuster et al., 2017]. Thus, GI acts as an alternative or a complement to centralized gray infrastructure [Fletcher et al., 2015] practices by offering these typically overburdened collection systems additional detention capacity. This additional capacity is key to maintaining system hydraulic capacity throughout its various conveyances and storages. When sub-system capacity is exceeded, for example in a “sewershed” (a catchment delineated by extent of sewer drainage), this condition can lead to local or systemic wastewater system malfunctions, which are usually observed as combined- or septic-sewer overflows into receiving waters.

A focus on planning GI for stormwater management will typically site and engineer a GI installation so as to optimize hydrologic outcomes. This process often employs urban rainfall-runoff models that represent common hydrologic processes as inputs and losses to the site of interest. There is a litany of site physical variables (e.g., slope, percent impervious area, soil conditions, and catchment area) that affect the perception of the suitability of the site, which is distinct from how the GI practice is situated in the social and economic contexts. Green infrastructure, though, can be a versatile and effective means of managing landscape changes that occur due to human influences and can provide benefits to individuals and communities more broadly. The versatility of GI and its potential interest to many stakeholders stems from its capacity to provide a range of ecosystem services [Hardin and Jensen, 2007; Yuan and Bauer, 2007; Chaparro and Terradas, 2009; Gómez-Baggethun and Barton, 2013; Andersson et al., 2014; Berland and Hopton, 2014; Coutts and Hahn, 2015]. An expanded accounting of ecosystem services is based on the agenda provided by the Millennium Ecosystem Assessment [2005], which include: supporting (nutrient cycling-uptake, soil formation); regulating (as temperature regulation [Scalenghe and Marsan, 2009] and evaporative cooling [Song and Wang, 2015]); provisioning (pollinator habitat, food, water); cultural (recreational experiences and greenspace, improvements in public safety [Wolch et al., 2014; Coutts and Hahn, 2015]); increased property values [Schilling and Logan, 2008; Dunn, 2010]; and overall re-integration of urban spaces with multi-functional landscaping practices [Green et al., 2012; Berland and Hopton, 2014; Kremer et al., 2016]. From this perspective, GI motivated by stormwater management tends to be optimized for a narrow range of ecosystem services, such as, for example, stormwater volume reduction to reduce combined sewer overflow discharges, or the reduction of runoff pollutant loads to receiving water bodies. To move GI from a specialized functionality, such as runoff collection, to intentional multi-functionality, which aims to provide several ecosystem services it must account for how human activity, social organization, and governance interrelate and can jointly broaden the scope of urban design and retrofit redevelopment [Meerow and Newell, 2017]. Through its general integration of social, economic, and environmental objectives, GI has the potential to address multi-sector challenges; though to realize this potential, participating municipal organizations must engage in and commit to structured coordination.

Although highly-desirable, a multi-faceted dimensionality and multi-disciplinary engagement is not straightforward to incorporate into GI planning and design, particularly in urban systems, which are recognized as being highly complex [McPhearson et al., 2016]. Instead, it is often represented through centralized, narrowly-focused institutions (e.g., municipal sewer district) that do not necessarily reflect the diversity and dynamics of organizations and individuals that are interested or involved [Srinivasan et al., 2012; Carballo-Penela and Castromán-Diz, 2015; Vogel et al., 2015]. Fostering a more participatory and inclusive interaction between potential stakeholders though, has the potential to achieve great impact on the ability of local governance to provide services and improve urban hydrologic outcomes by expanding engagement and the range of GI installations into which it is incorporated. For example, GI practices can be expanded to be included in the yards of homeowners if supporting programs exist. This approach at once decentralizes and extends coverage of GI over a greater area and increases stormwater detention, while engaging citizens [Mayer et al., 2012].

A participatory, inclusive approach facilitates locally-specific, situationally-relevant concepts and ideas to be incorporated into projects. However, it is a more involved approach to GI that may require new tools to enable the process. To support the process, we introduce the concept and approach of Situating GI (or Situated GI). Situating GI is contrasted with Siting GI, which encompasses physical characteristics for making decisions on hydrologic objectives. Situating GI is an approach to urban landscape management in which the nexus of several contexts defines the placement and design of a GI installation. Situating GI assumes that multiple functions of the system interact synergistically in sharing a physical place [Lovell and Taylor, 2013]. In contrast, Siting GI for hydrologic performance treats other desired outcomes as undefined or coincidental synergies.

In this paper, we discuss three integrated concepts for improved GI in urban systems. However, these concepts can be applied to other environmental management plans and projects. Our intent is to present these concepts in an exploratory manner, in order to spur additional research and scholarship. First, we present Situated GI as a green infrastructure implementation approach and compare it to other GI implementation archetypes based on organizational network structure, capacity to adapt, and impact on a specific objective. We then introduce a framework called the Framework for Adaptive Socio-Hydrology (FrASH) for a participatory and inclusive approach to the development of hydrologic infrastructures that can support multiple ecosystem services. Based on FrASH, we offer a heuristic model – the Chambered Nautilus – which is intended to visualize and facilitate the participatory and inclusive nature of a Situated GI approach to green infrastructure planning. Together, Situated GI and the Chambered Nautilus model formalize FrASH. Finally, we explain how the concepts of FrASH can be applied and visualized using the Chambered Nautilus model with two urban GI projects as case studies.

2. Comparing situating GI in context to other GI implementation approaches and strategies

To exemplify how Situated GI can be differentiated from other types of GI installations and projects, we first define four archetypes of approaches to GI implementation: Lone Ranger, Situated GI, The Collective, and Specialized. The archetypes were determined by a combination of the authors’ extensive observations as GI researchers and a binning process based on concepts the authors determined were meaningful GI planning and implementation approaches. These four archetypes are differentiated by three concepts: network structure of participating organization(s), combination of GI versatility and participation/inclusiveness of organizations/objectives, and tradeoff in capacity for adaptation and environmental impact. Importantly, the archetypes and their differences (as they are described) are simplifications of a much more complex system. The goal here is to create a useful breakdown of the range of GI approaches and strategies for understanding Situated GI in relative terms. Before detailing the concepts that differentiate the four archetypes of GI installation, we define and exemplify each archetype.

• Lone Ranger. An individual or organization acting independently of others. For example, an individual that installs any or all of the following GI practices: a rain barrel and green roof to capture runoff from their roof to irrigate their vegetable garden and plants that support pollinators.

• Situated GI. A group of individuals or organizations with unique interests and contributions collaborating on a joint environmental project that has multifunctionality to serve the diverse interests of the groups involved. An example is a park designed to include a variety of ecosystem services (e.g., pollinator habitat, stormwater retention, community agriculture, recreational space).

• The Collective. A group of individuals or organizations connected through a main organization that encourages and fosters collaborative action on a shared objective. An example of this is an organization that attempts to involve homeowners (in a neighborhood or city) in a de-centralized stormwater management project. The group offers to install rain gardens and/or distributes rain barrels to expand runoff detention capacity in a watershed and relies on citizen involvement.

• Specialized. An arrangement among a few individuals or organizations that has a single objective, often formalized through policy. An example is a sewer district mandated to manage stormwater by government policy installing a detention pond engineered to maximize reduction of stormwater runoff. This archetype is most relevant to the traditional approach of Siting GI.

In the following sections, we relate these four archetypes to the three concepts of network structure of participating organizations, dimensions describing GI versatility and participation/inclusiveness of organizations, and tradeoff in capacity for adaptation and environmental impact (Figure 1).

Figure 1.

Theoretical concepts that define the four archetypes of GI implementation approaches and strategies: a) network structure of participating organizations, b) dimensions describing GI versatility and participation/inclusiveness of organizations, and c) tradeoffs between capacity for adaptation and environmental impact.

Network structure of participating organizations

Instrumental to defining the four GI archetypes is how they are categorized into different network types along dimensions of connectivity and centrality (Figure 1a). In a GI implementation network, nodes are individuals or organizations with capital in GI (natural, human/social, or financial/manufactured), and links exist when two nodes are connected via a relationship. Relationship links could be legal, financial, or social, among others. Here we define connectivity as the number of linkages among nodes in the network; higher connectivity networks have greater linkages among nodes such that removal of a link does not disconnect nodes (i.e., no linked pathway to a node or a networked group of nodes). Centrality is the degree to which one or a few nodes serve as a link to many nodes. In general, we are considering the perceived measure of centrality: ‘betweenness centrality’. Betweenness centrality is to what degree connections between nodes are mediated through a few nodes [Freeman, 1977]. If most nodes in the network are connected to each other through a highly connected node, the network has a high degree of centrality.

The Collective and Specialized have high centrality networks where nodes are connected through a key node, while this centralized network does not define the Lone Rangers and Situated GI. In Situated GI there is no requirement for centrality because any node can potentially play a significant organizing role, dependent on context and connectedness, while the Lone Ranger is carrying out a GI installation out of their own motivation. Lone Rangers and Specialized fall into the space of having low connectivity. In the case of Lone Rangers, in fact no linkages are required so the connectivity can actually be undefined. Specialized rely on key linkages, which if removed would isolate nodes. In contrast, high connectivity networks are used by The Collective and Situated GI, where cooperation among groups is vital to carrying out a project. While the two are similar, a Situated GI approach relies on a mesh-style network which has a low degree of centrality; specifically, it does not have a key node through which most nodes are linked, which is the case in The Collective.

Participation/inclusion and GI versatility

The second set of axes distinguishing the archetypes is the potential for stakeholder participation/inclusion and GI versatility (Figure 1b). We treat participation and inclusion as separate aspects of stakeholder engagement. Participation is the number and diversity of individuals and groups that are involved in some aspect of GI, whereas inclusion is the number and diversity of GI goals of stakeholders that are actively shaping the GI project (for more detail, see FrASH: Framework for Adaptive Socio-Hydrology section below). Versatility is the potential for the organizers of the GI project to take advantage of multifunctionality of natural systems to provide a range of ecosystem services. In a very simplified view, the four archetypes can be arranged into quadrants of a 2x2 table with high and low values for each axis (Figure 1b).

Situated GI and The Collective can use greater levels of participation and, especially in the case of Situated GI, inclusion compared to the Specialized and Lone Rangers archetypes (Figure 1b). Specialized can have many nodes in the total network suggesting high potential for participation, but it generally relies on a few representative organizations (e.g., city government and stormwater management agency) to navigate the GI project. Lone Rangers is self-sufficient approach to carrying out GI projects, accomplishing GI projects independent of others.

Situated GI and Lone Rangers offer greater versatility than Specialized or The Collective archetypes. Lone Rangers are driven by personal ideology and Situated GI encompasses multiple objectives. Therefore, these archetypes take on variations of GI that can serve multiple ecosystem services. Specialized and The Collective, on the other hand, are more limited in their versatility. Specialized focuses on a given type of GI functioning and The Collective exists to accomplish particular objectives.

Impact versus capacity for adaptation tradeoff

Impact describes the degree to which a GI project accomplishes a specific function. As the number of functions increases for any one project, it is likely to reduce the magnitude of any single function. Capacity for adaptation is the ability to shift functions or participants in different contexts across time and space, mainly related to governance. These two axes represent a trade-off that GI implementation approaches face; having greater impact comes at the cost of lower capacity for adaptation and vice versa (Figure 1c).

At the high impact end, the focus on a particular objective indicates Specialized can have a great impact on ensuring that success of the target ecosystem function is met. However, multifunctionality is lost at the expense of maximizing single benefits. In the Specialized approach, capacity for adaptation is most limited. This is because GI approaches following this archetype are carefully sited and engineered to maximize the objective outcome and under a specific organization arrangement. It will be limited to the cases where a narrow set of conditions are satisfied and cannot easily adapt to different places or to changes through time. Lone Rangers are the opposite end of this trade-off spectrum, low impact but high capacity to adapt. Because Lone Rangers are acting out of their own motivation the capacity to adapt to changing circumstances exists more so than in any of the other archetypes. However, because the geographical footprint of the Lone Rangers is generally smaller than that of any of the other archetypes, the impact is also expected to be smaller.

The Collective and Situated GI approaches have elements of both impact and capacity for adaptation. Of these two, The Collective is more invested in impact and Situated GI is more invested in capacity for adaptation. Groups with a common interest (The Collective) are likely to have a strong impact when working together; however, the centralized network structure of The Collective is inherently more vulnerable to the viability of the central organization and changing circumstances that could jeopardize The Collective’s overall functionality. Because Situated GI relies on various involved organizations to bring individual expertise and objectives to a project, a large impact is not made on a single objective; rather, more objectives are intended to be met through tradeoffs in functioning. However, should an involved organization lose funding or shift priorities, it can be replaced, the project can downsize, or the organization’s overall objective and role in the GI project can change. Because the organizational aspects of Situated GI are decentralized, it is also more likely to be able to adapt to differences in social-ecological systems such as laws, cultural norms, and biophysical parameters.

Here, we argue that in a sustainable city planning approach that aims at providing multiple ecosystem services, an approach that mimics Situating GI should be implemented. In some circumstances, such as runoff detention next to highways or other transportation arteries, a Specialized approach is more relevant. Similarly, cities or non-profit organizations should not discontinue their approach of stormwater management that falls under the archetype of The Collective, as each of these initiatives still assist in targeting individual aspects of urban environmental management. Finally, Lone Rangers should also continue to implement GI, especially if it falls on private property, and the Lone Ranger homeowner wants to make a difference in stormwater management on their land. We do, however, encourage stormwater utilities and consulting agencies to consider using the proposed approach of Situating GI and its benefits. In the next few sections we discuss using Situating GI over other implementation strategies, to highlight this approach to GI installation. We recognize that this method of GI installation is not a panacea for sustainable city design and integrated water resources management, but hope it encourages other scholars and practitioners to build upon this model.

3. How the situating in context framework moves us to meaningful action

Collaborative efforts can result in trust, effective conflict resolution, and more normative decisions that can lead to sustainable outcomes compared to efforts in which only a single perspective is represented [Pearson et al., 2010; Chaffin et al., 2016a; Mott Lacroix and Megdal, 2016; Chini et al., 2017]. Evidence of this has been published throughout the years, citing situations where sometimes groups with opposing views have been able to come to a workable solution through conflict resolution, trust, and an overall collaborative approach. [Weber, 1998; Poff et al., 2003; Bidwell and Ryan, 2006; Ison et al., 2007; Flynn and Davidson, 2016]. A recent study presented how communication between stakeholders (even those with differing incentives) can lead to a more beneficial outcome for planned projects on various scales [Mott Lacroix and Megdal, 2016]. Here, the authors investigated three water management approaches that delineated sustainable use of water for communities, businesses, and agriculture at different scales in Arizona that ranged from town, to watershed, and the state of Arizona. The authors attributed the successful development of the management plans to the “stakeholder engagement wheel,” which involves a highly collaborative and interactive approach to environmental planning [Mott Lacroix and Megdal, 2016]. This builds on findings from the social learning literature, which emphasizes the importance of inclusion and interaction between stakeholders [Mayer et al., 2012; Scott et al., 2012] in addition to flexibility in reaching objectives, some of which may be motivated by regulations [Shuster and Garmestani, 2015].

A number of frameworks for urban watershed management that incorporate landscape and socio-economic components have been proposed and included foci such as sustainability [Hellström et al., 2000], ecosystem services [Derkzen et al., 2017], valuation [Wild et al., 2017], and flood control [Schubert et al., 2017] to name just a few. Nonetheless, understanding these complex systems is far from resolved and there still is a call for other frameworks to deal with the problem of managing stormwater in urban watersheds [McPhearson et al., 2016; Liu et al., 2017]. Here we present an adaptive framework that focuses on Situating GI in context, in combination with a graphical model applied to two case studies to emphasize that climate, socio-economic settings, and target ecosystem services drive the outcome of the resulting planning and implementation processes.

4. FrASH: Framework for Adaptive Socio-Hydrology

Situating GI is an approach for the organization and implementation of green infrastructure. One key question to its usefulness is: How can Situating GI be translated into practice? To address this, we propose a framework called FrASH – Framework for Adaptive Socio-Hydrology – to organize elements of its implementation. FrASH is intended to address green infrastructure/environmental projects where 1) there is versatility in potential outcomes introduced by the range of possible ecosystem structures and functions and 2) the involved individuals and organizations are connected through a mesh-style network. FrASH works under the assumptions that different types of capital can contribute to a project, that multifunctionality is possible, and there is capacity for interdisciplinary collaboration. The stated goal of FrASH is to facilitate participatory and inclusive organized action on a green infrastructure/environmental project in a sustainable manner. The three components of FrASH are adaptive governance, participation, and inclusion.

Adaptive governance

Adaptive governance recognizes the complexity of social-ecological systems such as GI, and allows the governance team to respond to changing circumstances through changes in structure, thereby increasing the ability to maintain functionality of social-ecological systems [Folke et al., 2005]. Adaptive management can be a powerful tool nested within an adaptive governance project, but adaptive management projects generally do not fit a “one size fits all” model [Green et al., 2016]. Thus, adaptive management is best suited to projects with high controllability and high uncertainty in order for management to reduce uncertainty and increase learning [Williams and Brown, 2014; 2016]. Further, controllability is a critical aspect for adaptive management to operate in the manner in which it was intended [Holling, 1978]. The project leader, in particular, must have control of the management applications that are made to GI, in order to accelerate learning and improve environmental management (e.g., stormwater management) [Allen and Garmestani, 2015]. An interdisciplinary group comprised of governmental (e.g., federal, regional, state and/or local), non-governmental, community organizations, academia, and industry, for example, should be involved in planning GI, both economically and intellectually. Finally, an interdisciplinary governance team should be organized in a way that preserves functionality of the social-ecological system.

Participation and inclusion

Participation and inclusion represent two aspects of stakeholder engagement in a GI project. Following Quick and Feldman [2011], participation refers to those who provided input during planning whereas inclusion refers to the co-shaping of the GI project such that those stakeholders’ goals are part of the project outcome. In practice, this can be simplified, as participation is having a voice and inclusion is having desired objectives of a participant incorporated into the GI project. Situating GI requires the participation of multiple organizations and individuals, which bring more interests, ideas, and capitals to the project [Kusters et al., 2017]. Situating GI also uses the inclusion of multiple goals of the various participants in the design and execution of a project [Flynn and Davidson, 2016], which in part, is a necessary criterion to motivate the continued engagement and benefit many organizations [Quick and Feldman, 2011].

5. The Chambered Nautilus visualization model

We built a heuristic model based on FrASH, called the Chambered Nautilus, which is meant to be a novel approach to visualizing how FrASH can be applied to a project. The visualization uses the shell of a chambered nautilus (marine mollusks from the family Nautilidae) which forms a spiral consisting of chambers that decrease in size as it spirals inward. This symbolism implies that all chambers of the Chambered Nautilus represent the entirety of the GI project. There are two types of nautilus shells in the visualization model: the organizational shell and themed shell, each of which communicate different components of a GI project. In the organizational nautilus shell, the chambers represent the individuals and organizations involved in a project. The Chambered Nautilus visualizes projects as having a main organization, a few major participants, several intermediate participants, and the potential for many minor participants (Figure 2). Participants are arranged in the nautilus by their relative contribution to, or role in the project. This can be defined by the overall effort put toward the project in the form of financial support, guidance, or vision. The organization acting as the principal driver of the project makes up the largest compartment or chamber of the nautilus, while the other organizations populate the remainder of the nautilus based on their decreasing contribution. While each chamber represents a specific individual or organization (i.e., participant), these participants can be grouped into classes or types of organizations to visualize the makeup of contributors (Figure 2). In the case of a GI project that is being implemented, this visualization could be useful in ensuring that aside from the main organization (e.g., a sewer district), other organizations are included in the planning. These could include non-profit organizations, non-governmental organizations, or foundations that act as community outreach partners or advocates for additional ecosystem services.

Figure 2.

Planning and implementation trajectory of the FrASH model. The chambered nautilus model can accommodate a number of organizations from various sectors (e.g., government, non-governmental, community organizations, academia, or industry) represented by different colored (sectors) chambers (organizations) in the organizational nautilus. The themed nautilus can represent each of the involved organizations’ primary objective or theme.

The organizational nautilus can be complemented with themed nautili. Themes are attributes of the organizations (i.e., chambers) with regard to the project. Examples of themes include the type of capital contributed or the ecosystem service of interest. To illustrate the interdisciplinary nature of a GI project and applying FrASH, the chambers can be categorized to reflect different sectors involved in the project. For example, a project that is implemented by 26 different participants (chambers) could have three different organization types represented that can be categorized into four themes (Figure 2). In the case of GI, the idea of maximizing ecosystem services comes to mind. Here the four categories of ecosystem services (regulating, provisioning, supporting, and cultural) could be used as attributes that each organization represents. If the idea of a project is to ensure a diversity of ecosystem services, using a themed nautilus can assist in assuring multi-functionality is included in the planning process. Therefore, the nautilus model is a visual cue (conceptual) to give project partners a chance to evaluate the variation in involved organizations and proposed target objectives.

The strength of the nautilus model is that the socio-hydrology of a given green infrastructure intervention is easily deconstructed into sectors or themes (as colors) or organizations (chambers), and is sensitive to the changes (e.g., governance, new information) in project plan that may take place over time. The fluidity of the model allows new organizations to be added to the project plan, or dropped if interests have changed, or funding has run out, and the overall direction of the spiraling trajectory does not change. In a situation where a conflict or disagreement arises about a specific attribute of the GI project, all stakeholders are asked to weigh in and provide guidance to come up with a solution that results in a compromise. This characteristic embraces the qualities of Situating GI, in which the associated network structure emphasizes the decentralized connectedness of participating organization. Finally, the completed conceptual model falls into the ecosystem services matrix where the project’s realized social, hydrological, and ecological benefits are represented. While this visualization is currently only qualitative, a formalized network analysis between the participating organizations could lead to a quantitative model, which could be used to first plan the project and later on evaluate project success in terms of ecosystem services generated. This framework could then be expanded for use in monitoring studies that evaluate GI functionality and community perception over time.

Applying FrASH with the Chambered Nautilus model encourages practitioners that the project is designed with several goals in mind to benefit the greater community in more ways than had only one organization taken the lead in project development. In scenarios where this type of adaptive framework is not being applied, such changes and shifts in governance through loss of stakeholder participation could lead to project failure [Chaffin et al., 2016a]. However, an adaptive approach allows for the remaining organizations to fill in the gaps and continue the project [Chaffin et al., 2016a]. Because FrASH involves a multitude of organizations that have relatable goals, the project is more likely to maximize several benefits (e.g., stormwater management, increased shade, greenspace provision, improved air quality, pollinator habitat) and therefore falls within the center of the ecosystem services matrix. Further, due to the involvement of a multitude of organizations, costs of implementation, operation, and management can be shared. The collaboration of the involved organizations can also result in more effective planning on related projects and prevent the “reinvention of the wheel” that smaller, disconnected organizations (Lone Rangers) may face when working alone. Finally, the concept of FrASH encourages collaboration between various organizations that have a stake in the local community, which could generate local jobs and contribute to a more sustainable city model.

6. Exemplifying FrASH and specifying the Chambered Nautilus in Situated GI projects

Two GI case studies where a Situating GI approach was use are located in Cleveland, OH, and Atlanta, GA. While project implementation in both case studies was the result of a consent decree due to Clean Water Act violations, the projects were located in different climate regions as well as socio-economic settings in the respective cities. Cleveland, OH is located in a continental climate region on the southern shore of Lake Erie, that is experiencing long-term, population loss and decline in economic activity (i.e., it is a shrinking city [Herrmann et al., 2016]). Here, the GI project focused mainly on stormwater runoff detention with rain gardens. In contrast, Atlanta, GA is located in the Sun Belt, a subtropical region in the U.S. south of 36^th parallel, and is experiencing an increase in population. In this case, the GI project focused on creating cultural ecosystem services in the form of a community space in a neighborhood that had experienced economic downturn. Both projects were focused on providing ecosystem services from re-purposed vacant lots to prevent blight. While FrASH and the Chambered Nautilus were applied post-hoc to these projects, these examples emphasize that Situating GI in context can be applied to projects in various climatic regions, different socio-economic circumstances, and can span various scales in terms of the number of organizations involved.

6.1 Adaptive governance guides a GI project in the Slavic Village neighborhood in Cleveland, OH

The Slavic Village project in Cleveland, Ohio, demonstrated that FrASH can be useful for installing rain gardens in a neighborhood. The original project objective was to install several highly-landscaped rain gardens with a focus on stormwater runoff regulation (e.g., combined sewer overflow reduction). Though the development of these rain gardens appeared straightforward, social and economic barriers developed over time, which changed the project objective and outcome [Chaffin et al., 2016a]. The application of adaptive and transformative governance allowed the project to continue under a new organizational structure that led to the development of expensive, highly landscaped and inexpensive, low-tech rain gardens.

In this example, a complex set of organizations, institutions and their associated jurisdictions, as well as competing priorities of project partners, contributed to the rain garden installation (reflecting inclusiveness). The shift in GI objectives and organizational structure splits the project timeline into two distinct phases. The visual representation of the chambered nautilus allows us to illustrate the ensuing shift in governance by showing the distinct levels of involved organizations (number of chambers) and their sectors (categories of representing organizations) for each of the two phases (Figure 3). The capacity for adaptation in this project is reflected through a change in how the nautilus is structured, which was also the major driver in the shift in outcome. To visualize the changes in the nautilus framing, chambers in the organizational nautilus are categorized based on the type of organization (e.g., federal agency, regional agency, public utility, community development corporation, non-profit organization, and university). On the other hand, the thematic nautilus reflects the ecosystem services that each organization prioritized.

Figure 3.

Chambered nautilus shells that describe the project trajectory of the GI project in the Slavic Village neighborhood in Cleveland, OH during two distinct project phases. Due to a change in organizational structure the number of organizations involved changed along with the project objective.

In phase one, researchers from US EPA Office of Research and Development (ORD) (interested in understanding the potential of rain gardens to reduce combined sewer overflows) were the drivers of the project, with key partners such as the Northeast Ohio Regional Sewer District (NEORSD), the Slavic Village Community Development Corporation, USGS, EPA-Region 5, Ohio State University, Cleveland Botanical Garden, Department of Justice (DOJ), Emory University, The Wildlife Society, and the Cleveland and Cuyahoga County Land Banks (Figure 3). The Department of Justice was also involved in the consent decree process that preceded the project, and was therefore a key player in Phase 1. Emory University and The Wildlife Society were key players in the development of the adaptive management plan that was implemented in phase one. After approximately two years of the project, shifting priorities amongst the stakeholders created a new organizational arrangement in phase two. Most of the organizations in phase one remained part of the project (except for DOJ, Emory University and The Wildlife Society). ORD remained the driver of the project, however the role of the Cleveland Botanical Garden changed from a supportive role to an essential role in the success of the project. Using the chambered nautilus shell as a visual representation, this shift is evident when comparing phase panels.

While the shift in the project was unexpected, the project stakeholders adapted and proceeded to accommodate changes in the social-ecological system that arose due to the change in the project organizational network. Despite these setbacks, organizational redundancy allowed the project to continue, as the role of Cleveland Botanical Garden evolved from a secondary participant to a key player and leader of the project. Even though the number of organizations changed in phase two, the adaptive component of the project remained, in the form of adaptive governance [Green et al., 2016]. Over time, adaptive governance evolved into transformative governance as the governance network sought to transform the social-ecological system, in this case via a GI project [Chaffin et al., 2016b]. Here, due to the innovation and creativity of the project stakeholders, the GI project continued, albeit with a different GI focus, reflecting high versatility and some tradeoffs on impact. The shift in the project proceeded with a mix of three higher-cost, highly-landscaped and eight lower-cost, lower-input rain gardens that were now being managed for multiple ecosystem services, as opposed to hydrology as the sole focus. Of the three highly-landscaped rain gardens, two are sited on land that is favorable for receiving stormwater runoff from adjacent impervious surface, with high hydraulic efficiency. On the other hand, the third and largest highly-landscaped rain garden was installed adjacent to a recreational trail, with a disproportionately small catchment area. While the rain garden installation enhances the aesthetics of the recreational trail and contributes to provisioning ecosystem services through pollinator habitat, few other regulating services (e.g., stormwater management) are gained. Here, the situational context demanded high aesthetic appeal in a landscape that offered few hydrologic benefits from the siting of this rain garden. The change in the main driving organization from NEORSD to the Cleveland Botanical Garden also drove a change in objectives. The Cleveland Botanical Garden was cognizant of native flora and local pollinator habitat in addition to regulating stormwater management through retention and detention. The continued involvement of NEORSD and strong support of the base organization (ORD) ensured that hydrologic benefits (regulating services) were integrated into the design of the low-tech rain gardens. Even though the strategy in constructing the low-tech rain gardens was intentional in integrating GI design guidelines, the final outcome had limited impact and functionality because curb cuts (which provide a hydraulic connection between impervious surface and the inlet to green infrastructure) were not installed and therefore limited the rain gardens to only capturing direct precipitation input. While the two rain garden systems have yet to be compared based on their hydrologic effectiveness (reduction in volume of total runoff), provisioning of multiple ecosystem services, and overall cost-effectiveness of GI versus gray infrastructure, the shift in organizational structure required an adaptive framework that allowed for changes in carrying out the project installations.

Ultimately, ORD and its capacity for organization of the GI intervention was responsible for maintaining the project trajectory. In a typical scenario of highly-centralized and hierarchical organization that does not accommodate capacity for adaptation, a shift in organizational structure could have led to project failure. The project instead succeeded with a tradeoff that moved from specialized, single impact (stormwater management) to an adaptive strategy with multiple ecosystem services, resulting in Situated GI. The analysis of this GI intervention stresses that capacities for adaptation, inclusiveness, creation of incentives, building of trust, and proper communication between stakeholders are the primary reasons behind project success [Chaffin et al., 2016a].

6.2 Local community involvement brings green space to Lindsay Street in Atlanta, GA

The Lindsay Street Park project (Proctor Creek Watershed, Atlanta, GA) originated out of local interests in the development of accessible, neighborhood-level green space from what was formerly a blighted area of the neighborhood. Even though Atlanta is overall a growing city, there are neighborhoods that have experienced population loss and an emergence of abandoned houses and vacant lots created by housing demolitions. This is the case in the English Avenue neighborhood, where the presence of vacant lots was taken as an opportunity to develop a GI project that would provide usable green space to the community. Through a collaborative land purchase by the City of Atlanta’s Parks and Recreation and two Atlanta-focused NGOs (Park Pride, Waterfall Foundation), the neighborhood acquired ownership of six vacant lot parcels, all on one side of the street, comprising a substantial 1.5 ac (0.6 ha) land area suitable for a park. These vacant lots were the outcome of city-level blight control efforts, and the initial motivation for creating Lindsay Street Park was to provide accessible greenspace, such as a common recreational-play area and public, common space in an aesthetically pleasing setting. The benefits to the neighborhood provided by these green-spaces represent cultural ecosystem services. Due to ongoing sewer overflows in the municipal collection system and residential flooding, there was an opportunity to build in regulating ecosystem services that were seen as an ancillary benefit (http://www.conservationfund.org/projects/lindsay-street-park, accessed on 2.2.2017). To achieve such stormwater management benefits runoff was re-routed from the residential and street areas downslope into the headwaters of a stream channel.

Despite its direct emphasis on building a park for local residents, the project relied on the participation of more than 20 organizations. To facilitate visualization, we categorized the organizations into three groups: those with a predominant focus on environmental stewardship, community-based, and others whose main role was providing financial support (Figure 4), while the themed nautilus again describes the primary ecosystem services each organization prioritized. The Lindsay Street Park effort was spearheaded by the local branch of the Conservation Fund, a national non-profit organization interested in environmental preservation and economic development in the United States. The roles of the Conservation Fund were stakeholder organization, project planning, and implementation. The rest of the organizations were entrained behind the Conservation Fund, who identified gaps and then channeled requests for support. Even though the organizational structure appears to be a top-down model with many funders being active at the national level, the social networks formed created a cohesive consortium of local organizations, sharing a common interest in serving their community, that can be described by a mesh network structure.

Figure 4.

Chambered nautilus shells that describe the project partners of the green space project in the English Avenue neighborhood in Atlanta, GA.

In addition to the land purchase, many local foundations with a spectrum of interests made financial contributions to the park’s development and programming. Multiple non-profit organizations specialized in contributions that addressed the construction or development of the park including: environmental education programming, site preparation, tree planting, and community-level job training programs. For example, the local nonprofit Community Improvement Association (CIA) was interested in greening underserved communities as a way to achieve stormwater management. CIA arranged a trip to Milwaukee, WI, for English Avenue residents, to tour GI in neighborhoods with similar demographics and judge for themselves the broader influence of GI on communities facing similar environmental and economic challenges. Through this trip and follow-up activities, CIA maintained the interest and sustained the involvement of the community in the stormwater management component. In this way, the CIA served as a bridging organization in setting up Lindsay Street Park to provide multiple ecosystem services (pers. comm. S. Lee, Conservation Fund). Because the park was primarily designed as a recreational space, appropriate flora as herbaceous plants and shade trees would promote aesthetic value, but also give relief from the urban heat during summer [Boone, 2015]. Both the Atlanta Botanical Garden and Trees Atlanta (city-level, non-profit organization oriented to maintenance of healthy tree canopy cover) were responsible for ensuring that provisioning services such as pollinator habitat and regulating services such as shaded areas were included in Lindsay Street Park. To do so, Trees Atlanta was successful in acquiring funding from Boise Paper to purchase trees which were planted through the efforts of resident volunteers and other NGOs. The project was completed in October 2015. The park is now a focal green space for the community, and a popular destination for neighbors and their children who enjoy the playground and picnic benches. Collectively, the many organizations and interests resulted in the implementation of GI providing multiple ecosystem services as well as being culturally appropriate to the place.

6.3 Using the basis of FrASH to compare green infrastructure interventions

The two highly-distinct case studies describing GI installations in Atlanta and Cleveland could be well described using FrASH and the Chambered Nautilus. Even though the two projects were different based on their motivations to install GI, the Chambered Nautilus could be used to define the initial conditions of the project. In Cleveland, the GI intervention was driven first by enforcement, then by the interests of local non-profit participants, while the Atlanta project originated from a constellation of local interests whose primary objective was transformation of a blighted area to accessible, neighborhood-level green space. The Chambered Nautilus outlined what kind of organizations were involved, and grouped by their respective objectives. Further, in both cases local organizations were involved in Situating GI, exemplifying that the concepts outlined in FrASH can foster community building and bring local economic and job benefits. Here the use of the Chambered Nautilus guides the reader and project partners through the project plan and acts as a reminder of project goals. In the Cleveland case study, the Chambered Nautilus illustrates the change in the organizational network, which responded to shifting project goals. In this way, ongoing charting of organizational engagement as the project progresses provides strong visual cues to the lead organization (or consortium). These diagrams may further improve ongoing accounting for diversity in stakeholders, identification of critical input (e.g., consulting soil scientist), and leading to completeness in representation. These circumstances may then promote and sustain the prospects for multifunctionality over the course of what are usually long-term projects with distinct and drawn-out planning and implementation phases.

In terms of the two projects falling into the Situated GI archetype of GI installation, both projects exhibited high network connectivity, while centrality was low, creating a mesh like network structure that differentiates Situated GI from the other archetypes of GI installation approaches. Participation and inclusiveness was high as both GI projects were carried out with involvement of multiple organizations that had various focal points. Because in both situations the potential for multifunctionality was recognized and pursued with intent, flexibility was recognized, valued, and ultimately integrated into the project outcome. Finally, each project had limitations with respect to the impact they could have had on regulating services, exemplifying that the multifunctionality that comes with Situating GI faces trade-offs. The projects were carried out relying heavily on adaptability, which allowed for successful installation of multifunctional GI, as is suggested by the Situating GI archetype.

The outcomes for each case study were borne of unique circumstances and yielded corresponding suites of target ecosystem services. Overall, FrASH and the Chambered Nautilus present a workable and complete accounting of influences on the definition and relationship between Situating GI and Siting GI where the results guide social-hydrological interaction. This approach to visualization is useful to illustrate socio-hydrological differences in contrasting GI interventions, as their respective networks evolve the objectives and organically adapt.

Aside from the types of organizations involved and different cities in which the two case studies took place, there are also variations in motivations, i.e. bottom-up vs. top-down approaches to which FrASH can be applied. In Cleveland, the sewer district was compelled by a top-down process (Clean Water Act enforcement action), which called for GI interventions under consent orders and initiated a change in culture with regard to wastewater management. This process required the sewer authority to make GI an additional and integrated component of its existing portfolio of wastewater collection and treatment infrastructure. Specifically, the enforcement action engendered a shift to decentralized, local management of stormwater inputs to a specific part of the centralized wastewater collection system. However, the scale of the effort remained at a demonstration-level, far short of achieving GI detention capacity required to maintain system capacity. Much like more involved social network analysis, the nautilus model clearly delineated the change from a municipal-regional use of green infrastructure to a combination of NEORSD’s larger-scale, capital intensive approach with a group of lower-input (and less expensive) retention strategies distributed throughout the neighborhood. What FrASH tells us is that this top-down process reached a limit as to suitability and integration of green infrastructure, even at a small-scale, demonstration level of investment. The adaptability of the project was characterized by the ascendance of an NGO (Cleveland Botanical Garden), which had local knowledge and prior experience in transforming vacant lots. This local knowledge was critical, and allowed the NGO to maximize the different types of provisioning, regulating, supporting, and cultural ecosystem services. The totality of the Cleveland observations is in direct contrast to the Atlanta GI intervention. The fundamental difference is shown clearly by comparison of the respective nautilus models (Figures 3, 4). The Lindsey Street model indicates a bottom-up initiative led by several local Atlanta watershed-level organizations, compared to the succession of top-down initiatives that drove the Cleveland work.

The strength of FrASH, from a post-hoc perspective, is that project structure and gaps are readily identifiable from two highly-distinct case studies. In this way, the analysis clearly illustrates areas of process strength and weaknesses which require redress. For example, in Atlanta, instead of having a sole focus on management of stormwater, the impact on regulating services was reduced but also addressed the social and cultural concerns, and vision toward transforming blight into recreational space. Overall, the involvement of the local community was essential in seeking buy-in (gathering social capital), and attracting other important capitals (e.g., financial, labor training, education). On the other hand, other organizations saw the potential for hydrologic functions to bolster the rationale for redevelopment as a park. In this way, the eventual redevelopment as a park increased stormwater infiltration opportunities and likewise retention capacity to prevent runoff. Yet, this socio-hydrologic forum focused almost entirely on recreational development, with hydrologic functions as an afterthought (Figure 4). The use of low permeability fill soils, instead of topsoil, may be typical for park construction, but relegated the park to be a net generator of runoff, and a sub-optimal setting for establishing flora.

In contrast, the Cleveland example featured a set of highly-engineered rain gardens where improvements in hydrology were engineered into an otherwise marginally suitable setting. These engineered gardens were complemented by a set of gardens that were better integrated into the residential landscape, though administrative and financial barriers prevented their connection to sources of runoff, resulting in an overall lower impact on regulating services.

For both examples, FrASH suggests there were missing points of view or data that could have aligned green infrastructure design with available resources. In the case of Cleveland, it was impossible to find a sponsor for curb cuts to be installed and maintained into perpetuity. This gap meant that the basic rain gardens would catch only direct rainfall and run off from adjacent properties; a significant limitation to their effectiveness in storm water volume control. For the Atlanta project, early consultation on soils and their placement would have enhanced multi-functionality of the park for recreation and stormwater management functions. In these two examples, City of Cleveland governance was ultimately the limiting factor, while the Atlanta group moved forward without a basic soil hydrologic perspective. The difference here is that Cleveland lost additional sewershed detention capacity by including a suite of smaller rain gardens (with small drainage areas) that were disconnected from major sources of runoff volume. Alternately, the Atlanta project was focused on creating a common space for its residents using locally-common landscaping practices that did not emphasize yielding regulating ecosystem services. Although positioned at different points in the life of each project, the gaps in realizing the full potential of generally good intentions points to risk knowledge and perception. Since GI integrates across social (citizens, governance), economic (land valuation, redevelopment), and environmental (ecosystem services) processes, future projects should be planned with free and open forums for discourse. It is only in such forums that the constellation of ideas and visions can be heard. At this direct interface with the field setting, prospects for adaptation become clearer.

7. Conclusion

FrASH brings together ideas from socio-hydrology, the capacity for adaptation, participation and inclusiveness, and organized action. The overarching theme of socio-hydrology recognizes that humans play a significant role in the hydrologic cycle that needs to be accounted for. The capacity for adaptation in governance requires organizations that are involved in project planning and implementation to respond to changes in circumstances. Because GI are social-ecological systems, whether by design or chance, they benefit communities and overall urban landscape connectivity. Therefore, the planning of these systems should follow principles of an interdisciplinary approach that is implemented using organized action. With this approach (nested within integrated urban water resources management), the concept of FrASH can act as one component in a move towards sustainable city development.

Adaptive management (and adaptive governance) is well suited for integration of GI into urban systems (if conditions are favorable), as these types of projects generally do not follow the “one size fits all” approach (capacity for adaptation). Because FrASH is responsive to shifts in organizational structure without establishing a strict top-down hierarchy, and instead is organized as a mesh network structure (i.e., high connectivity, low centrality), it can be applied at various scales, ranging from a neighborhood to a city-wide application and beyond.

FrASH brings together organizations (inclusiveness) to establish interdisciplinary perspectives that can benefit each other and help implement successful GI projects that recognize multifunctionality (versatility). The two case studies show that FrASH could have multiple entry points into each design process, making it a progressive framework for current and future city planning. The importance of integrating community organizations in the planning process increases stakeholder involvement and can improve perceived outcomes of GI projects in communities. This can result in further community engagement, resulting in buy-in and support for additional projects. The gained trust of the community and collaborating organizations instills a form of social equity and sustainability that can find appreciation in both growing and shrinking cities.

Finally, FrASH supports the idea that GI can benefit multiple aspects of municipalities, sewer districts, public works departments, multiple organizations and people with vested interests in the community. The connected network of organizations is also more likely to work towards a project outcome that has more benefit than the sum of its parts. A connected, yet de-centralized effort such as Situating GI in context, can result in providing more ecosystem services than a project carried out by organizations whose aim is to create a highly centralized system (sewer networks) or an organization working towards a single goal (collecting rainwater in a rain barrel for irrigation). Finally, we hope that the concepts of FrASH and Situating GI are useful to other scholars and practitioners and that this conceptual idea can further the development of sustainable city planning.

Key Points.

• New conceptual framework combining socio-hydrology and capacity for adaptation

• Encourages a connected, interdisciplinary network approach to green infrastructure with multiple ecosystem services

• Two case studies demonstrate relevance at various spatial and temporal scales