Abstract
Climate adaptation presents some new forms of planning uncertainty. We identified thirteen types of climate change uncertainty and grouped these into four categories. Next, we summarized eleven planning techniques, noting that only six of these techniques reflect an adapt and monitor approach that actively engages uncertainty. We then evaluated the types of uncertainty and planning techniques identified in forty-four US local climate adaptation plans. We found no communities used scenario planning or robust strategies despite the emphasis placed on these techniques in the literature.
Introduction
Climate adaptation planning, the process of planning for actual or expected changes in climate and its effects, presents some unique challenges relative to other types of planning efforts that also involve managing uncertainty. The climate change literature concisely states these unique challenges are due to its complex, stochastic, and nonlinear nature. However, additional dimensions complicate local planning efforts in the United States, including the symbolism that now surrounds climate change (Hulme 2009; Leiserowitz 2006), resulting “evidence trap,” and role of uncertainty. The risk literature states that some public debates take on symbolic meaning and the adversarial American policy arena exacerbates this conflict. Cultural risk theorists, such as Douglas (1985), believe that risk perception is a mediated social process. “Groups and social context, not individual cognition, plays the primary role in the selection and response to risk” (Krimisky and Golding 1992, 20). A similar debate occurred thirty years ago around nuclear power. “Groups in this arena are not only concerned about risks of nuclear power, but view the debate over nuclear power as a surrogate for larger policy questions about desired lifestyles, political structure, and institutional power” (Renn 1992, 191). In the United States, climate change has assumed a symbolic quality: It has become a polarized debate about economics versus environmental quality and the role of government versus the market.
A second challenge to climate adaptation planning is that in the United States, this symbolic debate occurs within an adversarial policy arena (Renn 2008). In an adversarial policy arena, the “evidence trap” frequently occurs. The evidence trap occurs when all groups provide their own evidence but the content or quality of the evidence becomes secondary to factors of power, influence, and moral values. For planners working in adversarial environments where the concept of climate change has political connotations, challenging anti–climate change positions with evidence will generally have little effect. However, the risk literature notes that experiences can shift opinions. “If a group claims that a particular hazard is actually benign, repetitive occurrences of accidents with negative consequences will finally trigger a revision of this claim” (Renn 1992, 187). In many Michigan communities, for example, planners are redirecting contentious climate debates away from the larger symbolic debate to focus on extreme weather events in their own communities. In this region of the country, many municipalities or their nearby neighbors have experienced extreme rain events that have overwhelmed their existing infrastructure systems and caused significant flooding (Abt Associates 2016). In these situations, climate adaptation planning conversations often begin with a discussion on how existing infrastructure systems are failing to adequately manage current conditions and what could be done to improve this unacceptable situation (Fischer et al. 2012; UNEP 2012). This shifts conversation toward solving “acknowledged” problems, discussing acceptable risk levels, and strategizing about what combination of hard and soft adaptation techniques might be most effective (McEvoy, Fuenfgeld, and Bosomworth 2013; Renn 2008).
A third related challenge to climate adaptation planning, and the focus of this paper, concerns the issue of uncertainty. While dealing with uncertainty is an acknowledged part of the planning profession (Abbott 2005, 2012; Abunnasr, Hamin, and Brabec 2013; Chakraborty et al. 2011; Christensen 1985; Friend and Jessop 1969; Hopkins and Zapata 2007; Kartez and Lindell 1987; Mack 1971; Quay 2010; Walker, Haasnoot, and Kwakkel 2013), climate adaption planning efforts can be derailed by the presence of new kinds of uncertainty. The purpose of this article is to explore uncertainty relative to climate adaptation planning. In the first part of the article, we begin by identifying thirteen types of uncertainty that appear in the climate change literature, which we have organized into four categories. This organization highlights two categories of uncertainties that are within local planning influence and two categories of uncertainties that are beyond local planning’s influence. In the second section of this paper, we identify eleven common planning techniques that are recommended for use when addressing climate uncertainty and differentiate those that reflect a conventional predict and plan approach from those that embody an adapt and monitor approach. Those techniques that favor an adapt and monitor approach consciously highlight uncertainty. In the third and final section of this paper, we analyze the content of forty-four local US climate adaptation plans to understand what types of uncertainty are identified within the plans and summarize the planning techniques used, concluding that the majority of these climate adaptation plans are not using the techniques that best address uncertainty.
Risk and Uncertainty
Within the climate change literature, a significant amount of attention has been devoted to identifying sources of climate-related uncertainty. To help condense these sources of uncertainty, we have organized the most commonly identified sources into four overarching categories with twelve subtypes (Table 1). In the following text, we briefly summarize and define what we mean by (1) uncertainty in future climate conditions, (2) external political and behavioral uncertainty, (3) uncertainty in the local community’s coping capacity, and (4) uncertainty in what will constitute an effective local response.
Types of Climate Change Uncertainty.
Uncertainty in Future Climate Conditions
The most frequent type of uncertainty discussed in the climate change literature relates to not knowing what future climate and weather conditions will be (Cai et al. 2011; Dessai and Hulme 2007; Hallegatte 2009). To help provide granularity, we have broken this category into nine subtypes of interrelated uncertainty: (1) uncertainty related to future greenhouse gas emissions, which is related to (2) uncertainty in what future climate conditions will be; (3) uncertainty in the direction of change for climate conditions, especially for precipitation (i.e., more or less); (4) uncertainty related to the intensity and severity of change; (5) uncertainty related to when changes will occur, including uncertainty related to the return frequency of changes (Abunnasr, Hamin, and Brabec 2013; Mearns and Norton 2010; van Aalst, Cannon, and Burton 2008); (6) uncertainty related to where changes will occur (Dessai and Hulme 2007; Mearns and Norton 2010); (7) uncertainty related to the ability of global climate models to replicate the climate system and adequately project future conditions; (8) uncertainty inherent in the climate system, including uncertainty related to climate variability; and (9) uncertainty that relates to what future climate impacts will be at the local or regional level. Combined, these sources of uncertainty address the temporal and spatial nature of future climate-related changes.
Reducing these sources of uncertainty has been a main focus of much scientific work to date. Underlying this work is what Walker (2012, 956) calls a “supposition that the[se sources of] uncertainties result from a lack of information,” thereby leading researchers to invest in the generation of more knowledge and better models capable of projecting future climate conditions with greater specificity. Despite significant efforts to reduce uncertainty in future climate conditions, “exactly” what future climate conditions will be remains an area of contention. According to Hallegatte (2009), there are two reasons why this scientific-based form of uncertainty continues to stymie climate action. The first reason is a scalar mismatch between what climate models can provide and what information decision makers need. The second is a general lack of understanding on the side of end-users about the inherent uncertainties associated with modeling future climate conditions. Despite these limitations, climate modeling has progressed significantly in the past several years, meaning that while we may never be able to know exactly what will happen in the future, we have a strong sense of projected changes, especially at global and regional scales (Berkhout et al. 2014; Ylhäisi et al. 2014).
Uncertainty in Climate-Related Behaviors and Political Decisions External to Municipality
As with the first category, the second category of uncertainty also relates to uncertainty beyond or external to the planning process. Within the climate literature, external conditions are generally actions, behaviors, or policies enacted by higher levels of government that will directly or indirectly affect the ability of local communities to adapt (Dessai and Hulme 2007; Intergovernmental Panel on Climate Change 2012; Mearns and Norton 2010; Refsgaard et al. 2013; Stone, Vargo, and Habeeb 2012). Uncertainty related to external factors can broadly be grouped into three subcategories: (1) uncertainty related to human behavior and responses to weather and climate conditions, specifically greenhouse gas emissions and technological adaptation; (2) uncertainty related to the actions of other entities at different scales (e.g., neighboring municipalities); and (3) uncertainty related to state, national, or international policies, funding programs, or agreements. This type of uncertainty includes policies established at state and national levels that impact local-level decision making as well as decisions made in the private and nonprofit sectors that will directly or indirectly influence local circumstances. Inherent in this category of uncertainty are nuances related to municipal funding, which often flow from state and federal entities to local communities for specific projects or program work. However, the impact of these conditions on the local level is far greater than the influence of the local level on shaping decisions or behaviors at this scale.
Uncertainty in Local Climate-Related Coping Capacity
A third category of uncertainty relates to a community’s current capacity to cope or adapt to changing climate conditions (Chakraborty et al. 2011; Kates, Travis, and Wilbank 2012; Quay 2010). This source of uncertainty deals with a lack of knowledge regarding how a community has responded or been affected by historic weather/climate impacts, thereby informing what type of coping capacity a community may have for future climate impacts (Dessai and Hulme 2007; Lau 2015). Further complicating this type of uncertainty is a lack of understanding or agreement on what constitutes coping. Nelson, Adger, and Brown (2007, 397) refer to coping as the “pre-conditions necessary to enable adaptation, including social and physical elements, and the ability to mobilize these elements.” Knowing the preconditions necessary to enable adaptation sheds insight into whether or not a community and its residents have the resources necessary to adapt to changing climate conditions (Brown and Westaway 2011; Engle 2011).
Uncertainty in Effective Local Responses
The fourth type of uncertainty concerns what the most effective response strategies to lessen the negative local impacts of climate change are. Unlike climate mitigation, where energy efficiency is a common starting point, climate adaptation efforts are inherently local and not easily transferable across locations. This means that what is a good practice in one location may be less relevant or appropriate in another. For example, the outcome achieved by a grey infrastructure upgrade in one community may be more effectively achieved through a green infrastructure or land use strategy in another community.
While these four categories of uncertainties are related, the uncertainties related to future climate conditions and the uncertainties related to climate-related behaviors and political decisions external to the municipality are largely beyond the planner’s direct influence. The uncertainties around the local coping capacity and effective local responses are within the planner’s influence (Figure 1).
The types of climate change uncertainty and their interactions.
Climate Adaptation Planning Techniques
In light of the significant role uncertainty plays in planning for climate change, a number of uncertainty-reducing techniques have emerged within the peer-reviewed and grey literature. After reviewing the literature, we selected the most common techniques and divided them into one of two categories: (1) those that support a predict and plan approach to planning and (2) those that support an adapt and monitor approach to planning (Table 2). In total, we identified eleven techniques: five techniques that are either fully grounded or partially grounded in the predict and plan model of planning and six that were fully or partially grounded in the adapt and monitor model of planning. In general, all techniques fall somewhere on the continuum between these two approaches, and our effort at categorization required some subjective decision making.
Techniques Identified within the Planning and Climate Literatures for Addressing Climate-Related Uncertainty.
Predict and Plan
The predict and plan model of planning emphasizes the use of predictions of the future as a baseline from which to plan (Quay 2010). As Quay (2010, 498) notes, predict and plan is based on “forecasts [of a] future trend or a future desired state and then identif[ying] the infrastructure needed to serve or create this future.” This model of planning has its roots in population and employment forecasts done in the 1960s (Chapin 1965; Kent 1964). Today, it continues to be commonly used in all areas of planning (e.g., transportation, water system, economic development, sustainability). The predict and plan model of planning works well when social and environmental systems are “stable and predictable over short periods of time” (Quay 2010, 468).
As shown in Table 2, we identified five approaches as fully or partially grounded in a predict and plan approach to planning: (1) conducting a vulnerability assessment, (2) using multiple climate change scenarios, (3) downscaling global or regional climate scenarios, (4) using no-regrets and low-regrets strategies, and (5) planning for multiple timeframes.
Vulnerability Assessments
A vulnerability assessment is a technique to assess how climate change could affect a local community (Fussel 2007b; Luers et al. 2003). To conduct a climate vulnerability assessment, a local community compiles information on projected changes in climate, explores what these changes could mean in terms of local impacts, and then assesses the community’s sensitivity to these changes as well as their capacity to adapt (adaptive capacity) (Intergovernmental Panel on Climate Change 2014). A number of qualitative and quantitative vulnerability assessments techniques exist, each emphasizing different levels of data granularity.
Once a vulnerability assessment is complete, stakeholders are able to produce a relative weighting of where or who within their community are the most vulnerable to changing climate conditions. This provides insights that can be used to select appropriate adaptation strategies. At their core, vulnerability assessments are based on using models or projections of future climate to determine future vulnerability. As such, vulnerability assessments fall into the traditional predict and plan model of planning and if used without being paired with techniques identified in the adapt and monitor category could lead to communities being underprepared for potential changes in climate, especially if future changes are outside the range used to guide the vulnerability assessment process.
Use of Multiple Climate Scenarios
In climate change science, projections of future climate change are based on an array of future greenhouse gas emissions scenarios. These scenarios include wide-ranging variables for future land use, demographic developments, socioeconomic developments, and technological change as well as assumptions about consumption patterns. The most commonly cited scenarios are those used by the Intergovernmental Panel on Climate Change (IPCC) (Füssel 2007b; Styczynski et al. 2014), which has created four different narrative storylines (RCP 2.5, RCP4.5, RCP6, and RCP8.5) to “describe consistently the relationships between emissions driving forces and their evolution . . . [with] each storyline represent[ing] different demographic, social, economic, technological, and environmental developments” (IPCC 2000, 3).
In climate change planning, practitioners are often encouraged to consider multiple climate scenarios. This frequently includes looking at projected impacts associated with a low or medium emissions scenario as well as looking at the impacts associated with a high emissions scenario. Using multiple emissions scenarios affords planners insight into a potential range of impacts associated with climate change. Despite their utility, however, no one scenario is likely to be fully predictive of future conditions, meaning that planning based explicitly on a single emissions scenario or a very small set of scenarios could lead to ill-prepared communities.
Downscaling Climate Information
Statistical and dynamic downscaling are two different techniques that take data from global climate models and use it to project future climate conditions at regional and local scales (National Research Council 2010). In dynamic downscaling, “empirical relationships between past observations of local- and regional-scale climate variations are used to translate large-scale projections from global climate models to smaller space scales and shorter time scales” (National Research Council 2010, 220). Statistical or empirical downscaling, in contrast, develops “statistical relationships that link the large-scale atmospheric variables with local/regional climate variables” (IPCC 2014, 10). While a significant quantity of literature in both the climate and planning domains has called for more downscaling, these techniques could actually enhance uncertainty due to factors such as spatial inhomogeneity between models, the more localized effect of aerosols on determining local impacts, and mismatches between regional models and the global models they draw from (IPCC 2012; Preston, Mustelin, and Maloney 2013).
No-Regrets and Low-Regrets Strategies
No-regrets strategies are those that “generate net social and/or economic benefits irrespective of whether or not anthropogenic/[human]-induced climate change occurs” (IPCC 2007, 878). No-regrets strategies are frequently touted as techniques for addressing uncertainty since, regardless of what the future holds, these strategies will provide value by building resilience (Biesbroek et al. 2010; Hallegatte 2009; Heltberg, Siegel, and Jorgensen 2009). Examples of no-regrets strategies include increasing green infrastructure, increasing public and nonmotorized transportation, and improving the livelihood of the poor and frontline communities (Abunnasr, Hamin, and Brabec 2013; Heltberg, Siegel, and Jorgensen 2009), all of which are well-established goals and practices in urban planning.
Low-regrets strategies are those that “provide benefits under current climate and a range of future climate change scenarios” (IPCC 2014). These strategies are generally low cost and have a high possibility for benefit, meaning that if they are later found to be unnecessary, the opportunity cost is minimal (Kettle and Dow 2014). Examples of low-regrets strategies include building in extra margins in infrastructure designs to allow for increases in precipitation, restricting development in floodplains, and creating additional water storage facilities. As noted by the United Kingdom Climate Impacts Programme (n.d., 15), no-regrets and low-regrets options are “particularly appropriate for the near term as they are more likely to be implemented (obvious and immediate benefits) and can provide experience on which to build further assessments of climate risks and adaptation measures.”
No-regrets and low-regrets strategies, however, do not imply that the strategies are cost-free (Wilby 2008). These techniques can also prove ineffective in the long term if stakeholders only embrace these strategies without also selecting more challenging strategies that may take more time to implement due to the need to generate political will and public support or identify appropriate financing. Additionally, if a strategy is identified as being no-regrets and low-regrets based on a single or small set of projected climate futures, the actions may prove to be largely ineffective if climate change is more severe than anticipated.
Multiple Timeframes
The final technique we identified as falling within the predict and plan model of planning is the selection of actions that cover multiple timeframes. By selecting strategies for immediate implementation as well as those that will be implemented five to ten years in the future, stakeholders can prepare for immediate concerns while also setting a foundation to address future threats and risks (Hallegatte 2009). Maintaining the momentum to implement longer-term strategies, however, can prove challenging, especially at the local level where political environments, public interests, and financial resources fluctuate constantly. To address this challenge, stakeholders can phase in longer-term projects by setting short-term benchmarks such as reducing water demand by 10 percent by 2020, 20 percent by 2025, 40 percent by 2035, and so on.
The use of multiple timeframes paired with a technique such as adaptive management could produce a solution that falls within the adapt and monitor category. However, we hypothesize that stakeholders do not regularly revisit strategies that are slated for implementation in future timeframes. As such, we expect that planners are using multiple timeframes as a technique within the traditional predict and plan model of planning—meaning that they are selecting strategies that cover multiple timeframes based on what they assume to be a relatively predictable set of future climate conditions.
Adapt and Monitor
Accepting that the future can no longer be predicted, the adapt and monitor model of planning emphasizes taking action that is adjustable as new information emerges (Quay 2010; Walker 2012). This approach is based on calls for flexible and iterative adaptation planning that intrinsically builds in mechanisms to evaluate progress and readjust activities as new information emerges or local situations and context evolve (Boyd and Juhola 2015; Hughes 2015; Mimura et al. 2014). Based on our review of the climate and planning literatures, we have identified six common climate-relevant planning techniques that are either fully grounded or partially grounded in the adapt and monitor model of planning (Table 2). These include (1) use of adaptive management, (2) scenario planning, (3) selecting robust strategies, (4) using incremental or flexible strategies, (5) using thresholds or tipping points, and (6) monitoring changing climate conditions.
Adaptive Management/Adaptive Planning
Adaptive management, with its roots in adaptive policy, stems from the mid-1920s when John Dewey (1927,) proposed that “policies be treated as experiments, with the aim of promoting continual learning and adaptation in response to experience over time.” Since then, the concept of adaptive management has gained traction in the ecosystems management, planning, climate, and resilience fields (Engle and Lemos 2010; Holling 1973; Nelson, Adger, and Brown 2007). The concept of adaptive management or adaptive planning means being able to change course (e.g., policy, management practices, or planning approaches) based on changing and unforeseen future conditions (Walker, Haasnoot, and Kwakkel 2013). At the core of adaptive management is the concept of learning by doing in which actions are viewed as experiments that can be learned from, replicated, and improved on as needed (Miles 2013; Tompkins and Adger 2004; Walters 1997).
Traditional governmental structures, however, make adaptive management challenging to implement in practice. In a case study of water resource management in California, Booher and Innes (2010, 1) found that “bureaucratic procedures, the lengthy processes of legislative deliberation, and the often arbitrary nature of judicial decision making” pose direct challenges to adaptive management. The focus on experimentation within adaptive management also presents a challenge to existing governance and operational structures that focus heavily on strategies that have a proven record of success (Folke et al. 2002).
Scenario Planning
Scenario planning is a “process of positing several informed, plausible and imagined alternative future environments in which decisions about the future may be played out, for the purpose of changing current thinking, improving decision-making, enhancing human and organizational learning and improving performance” (Chermack 2004, 16). More generally, scenario planning is a process to construct possible narratives about what the future could be (Evans 2011) and then selecting options that are viable across all of these potential futures.
Scenario planning does not allow stakeholders to predict the future or select an optimal future to plan for (Varum and Melo 2010). Instead, scenario planning provides a tool to explore multiple plausible futures by bringing together diverse stakeholders to co-construct possible narratives about their future and then use these scenarios to assess “strategic options and capabilities” that will serve across a range of plausible futures (Evans 2011).
One of the greatest challenges with scenario planning is ensuring that stakeholders do not fall into the trap of selecting one scenario upon which to plan but instead plan based on multiple possible futures. Another challenge with scenario planning is that it requires intensive stakeholder engagement, technical assistance, and time, all of which may be challenging to generate. Despite these limitations, there is a growing emphasis in the literature that scenario planning is an important technique for addressing uncertainty by identifying multiple plausible futures and then selecting robust strategies (see the following) that perform well across these futures.
Robust Decision Making and Strategy Selection
Robust decision making uses multiple views of the future to identify strategies that “perform ‘well enough’ across a broad range of plausible futures but may not perform optimally in any single future” (Walker, Haasnoot, and Kwakkel 2013, 960). To generate strategies that are robust, planners must think through multiple different future scenarios (i.e., through scenario planning).
The challenge with robust planning and the selection of robust strategies is that decision makers tend to focus on optimization and not risk minimization. According to Walker, Haasnoot, and Kwakkel (2013, 968), optimization can be defined as “trying to find the best solution among a set of possible alternatives without violating certain constraints.” Given that future climate conditions are uncertain, trying to optimize decision making can prove elusive and be highly dangerous if actions are selected that are only appropriate for a single future that does not materialize. When considering climate change, Hallegatte (2009) argues that there is no optimal solution, and as such, the selection of robust strategies that are able to perform across multiple possible futures should be prioritized. Robust planning also necessitates reliable up-to-date information and input from affected stakeholders (Stczynski et al. 2014), which can be time and staff intensive. The other challenge with the selection of robust strategies is cost: “Typically, the more robust a strategy, the more expensive it is” (Lempert and Collins 2007).
Incremental and Flexible Strategies
Incremental strategies are those that can be phased in over time. Often seen as modular strategies that can be sequenced to minimize risk, incremental strategies are touted as being able to address immediate concerns while leaving options open to deal with changes in the magnitude and timing of climate impacts (Easterling, Hurd, and Smith 2004; Hallegatte 2009; Quay 2010). Incremental strategies allow stakeholders to spread out costs and reduce losses if the investment or adaptation action must be abandoned or proves unnecessary (Quay 2010).
Flexible adaptation strategies, also commonly described as reversible strategies, are those that are capable of being adjusted, tailored, or tweaked as circumstances change (Hallegatte 2009; Quay 2010). This approach reduces the chances and costs associated with being wrong about the extent, timing, or magnitude of future climate change (Hallegatte 2009). Examples of flexible and reversible adaptation actions include the modular building of a sea wall, climate-proofing buildings, or limiting development in potentially vulnerable areas (Hallegatte 2009). All of these approaches can be modified or forgone quickly if climate impacts prove to be less than expected.
Kates, Travis, and Wilbanks (2012) note that incremental and flexible strategies are frequently extensions of existing actions and behaviors that emphasize doing slightly more of what is already being done to deal with natural variation in climate. As such, the major limitation of incremental or flexible strategies is that the magnitude and intensity of climate change may exceed the capacity of these strategies to cope. Moreover, some strategies simply cannot be done incrementally or in a flexible manner, such as increasing the capacity to treat wastewater. Finally, incremental and flexible strategies necessitate the continual monitoring of climate conditions to ensure that one is ready to implement the next module or change course if needed. This constant monitoring requires financial and human capital, and the future implementation of flexible/incremental strategies necessitates a continual stream of public and political will: all of which can be challenging to maintain.
Tipping Points or Thresholds
Defined as the “boundary conditions where acceptable technical, environmental, societal, or economic standards may be compromised” (Haasnoot et al. 2013, 371), the identification of adaptation tipping points represents an alternative mechanism for dealing with uncertainty. Contingent only on magnitude, not time (Gersonius et al. 2014), adaptation tipping points represent when the magnitude of climate change exceeds current management strategies, thereby necessitating that new strategies be implemented (Abunnusar, Hamin, and Brabec 2013; Walker and Salt 2006). Once identified, thresholds or tipping points can be modeled to give an estimate of the likelihood of that threshold being exceeded in the future or monitored so that stakeholders can be ready to act when a threshold is reached. If stakeholders rely solely on modeling thresholds and using those modeling results as a basis for decision making, then this approach would fall within the predict and plan model of planning. However, if used as a means to monitor changing climate conditions, thresholds and tipping points can be useful within the adapt and monitor model of planning.
The use of adaptation tipping points could be alluring for stakeholders in conservative communities where discussions about climate change are not viable due to political or public objection. The challenges of this approach are both knowing what existing or future thresholds or tipping points might be, which may be easy to decipher in physical or engineered systems but more challenging in social or natural systems, and ensuring that processes exist to continually revisit tipping points as climate change pushes systems outside of their management thresholds. For example, a project in Miami, Oklahoma, investigated the various elevations at which extreme precipitation would lead to riverine flooding. Climate scientists at the University of Oklahoma and Texas Tech University used climate models to determine how likely and frequently these critical thresholds could occur or be exceeded in the future. This type of modeling work does not necessitate belief in climate change, only an awareness of potential thresholds for key systems and operations.
Monitoring Climate Conditions
The final technique within the adapt and monitor approach is the regular monitoring of climate conditions. In this case, signposts or triggers can be set that indicate when additional action is needed such as revisiting a plan, scaling up a promising practice, or changing policy (Quay 2010; Walker, Haasnoot, and Kwakkel 2013). Generally, monitoring of climate conditions is promoted in two situations: (1) a plan exists for how best to prepare once a signpost or trigger is hit and one is just monitoring to determine when to implement that plan or (2) no action or plan is needed yet but one would be warranted once a certain signpost or trigger is met.
If a stakeholder only identified monitoring as their sole action, there is a risk that the stakeholder will either miss critical pieces of information that would indicate other types of climate action are warranted or may not have the ability to act whenever the signpost or trigger is hit. Setting up flexible and nimble systems that allow for quick and agile responses is needed if a stakeholder chooses to only monitor climate conditions as a way of preparing for an uncertain future. This, however, can prove challenging given the bureaucratic realities facing local governments (discussed previously) (Booher and Innes 2010). When selected as one strategy among a suite of others, monitoring climate conditions can ensure that stakeholders have the information they need to continually adjust to changing climate conditions.
While each of these eleven planning techniques offers value, those identified as adapt and monitor approaches actively recognize and address the uncertainties that surround climate adaptation.
Examining Uncertainty in Climate Adaptation Plans
In the previous section, we have summarized the different types of uncertainty associated with climate adaptation and the commonly recommended planning techniques to reduce these sources of uncertainty. In an effort to understand how uncertainty is being dealt with in planning practice, we turn to local climate adaptation plans. We identified the most common types of uncertainty cited within forty-four US local climate adaptation plans and investigated whether the number and type of uncertainties vary by city size (small, medium, or large). Next, we compared the types of planning techniques used to reduce uncertainty identified from the literature against those contained within the plans. Guiding our research were two questions:
Research Question 1: What are the types of uncertainty that US local communities are identifying in their climate adaptation planning?
Research Question 2: What techniques are local communities using to address uncertainty in their climate adaptation planning?
We hypothesized that larger communities will identify more sources of uncertainty than medium and small communities and use more adapt and monitor planning techniques to reduce uncertainty due to their greater planning capacity relative to medium and smaller communities.
Methods
Sample Selection
We used three criteria to select plans for our sample: (1) the central topic of the plan, as defined by the author, was adaptation, resilience, or preparedness; (2) the plan was written by or for a US city, town, or county government; and (3) the plan took a comprehensive approach to adaptation (i.e., sector-based adaptation plans were excluded). These criteria excluded plans that attempted to integrate or mainstream climate adaptation into other planning processes (e.g., sustainability or master plans) and regional or multi-jurisdictional plans (e.g., San Diego Bay, Metropolitan Washington Council of Governments).
We attempted to evaluate all plans in the United States. released before 2015 that met these criteria. We developed our sample based on a search of three adaptation clearinghouse websites: the Georgetown Climate Center, the Climate Adaptation Knowledge Exchange, and the Center for Climate and Energy Solutions. In addition, we collected plans through three 100-page Google searches for the terms local adaptation plan, local resilience plan, and local preparedness plan. In total, we collected eighty-five plans, forty-four of which met our evaluation criteria (Figure 2) (for a detailed description of the sample selection process, see Woodruff and Stults 2016).
Communities with plans included in analysis.
Coding Protocol and Procedures
We developed a coding protocol to identify the presence of uncertainty as well as techniques for addressing uncertainty in the local adaptation plans in our sample. In total, we developed twenty-one codes to assess uncertainty: Two codes centered on the identification of uncertainty, and nineteen codes focused on identifying specific techniques to overcome uncertainty (see Appendix). For strategies, each had to be explicitly labeled as being no-regrets, low-regrets, incremental, flexible, or robust in the plan to receive these codes: We did not infer whether strategies could be considered as meeting any of these criteria. We pretested the twenty-one codes on eight local adaptation plans from Europe and Australia to ensure that the codes captured the intended concepts and subsequently refined the coding scheme.
Each plan in the sample was coded independently by two trained coders in line with recommendations from the communications literature on content analysis and methodological recommendations from the plan evaluation literature (Krippendorff 2013). Before coding plans within the sample, intercoder reliability was calculated to ensure that the coders fell within an appropriate range of intercoder agreement (0.80 or greater).
Coders used NVivo version 10 qualitative analysis software package to link coding items with the content of plans. Once coders completed a plan, we compared coders’ quantitative data to identify disagreements on a metric-by-metric basis. All disagreements were discussed and reconciled by referring to the qualitative plan content, and final agreed-on codes were integrated into a master data set.
After all codes had been reconciled, we qualitative analyzed all of the text associated with the type or source of uncertainty codes, grouping similar types of uncertainty together. We also sorted the plans into groups by population size and whether the community was coastal or inland.
Results
Types of Uncertainty in Climate Adaptation Planning
Roughly three-quarters of the plans in our sample identify the presence of uncertainty when planning for climate change (32/44 plans, 73 percent). Just over one-quarter (12/44, 27 percent) of the plans did not mention any type of uncertainty. Within the 44 plans, uncertainty was independently mentioned sixty times (Figure 3).
Types of uncertainties identified within the climate adaptation plans.
We found that the most common type of uncertainty was associated with knowing what the future climate will be (47/60). Boulder County’s plan provides a summary of this type of uncertainty. “Adding to the complexity of managing the impacts of climate change is the uncertainty inherent in climate science. Although scientists are reasonably confident in the direction of temperature changes, the magnitude of change remains uncertain. For other changes, like precipitation, both the direction and magnitude of change are uncertain. This means that Boulder County and its municipalities are facing an uncertain range of possible future climate conditions with ensuing complications in identifying proactive management responses that are robust yet cost-effective across a potentially wide range of future climate conditions and design requirements” (Stratus Consulting 2012, 13).
In aggregate, the most common type of uncertainty identified within our sample was uncertainty related to future climate conditions (47/60, 78 percent) (Figure 3). The least common type of uncertainty identified in plans related to external uncertainty, specifically how external political decisions and/or larger societal behaviors may change. Only Austin, Texax; Boulder County, Colorado; and Los Angeles, California, identified this type of uncertainty. In all three cases, this external uncertainty was framed around lack of information regarding how humans will respond to changing climate conditions.
Only 17 percent of the identified uncertainties were within what we deemed the communities’ “direct local influence” (10/60). Also infrequent in plans in our sample was uncertainty related to effective local responses to dealing with climate-related impacts (4/44 plans). Only Boulder County, Colorado; Los Angeles, California; New York City, New York; and Punta Gorda, Florida, identified this source of uncertainty. Uncertainty in appropriate responses commonly dealt with lack of clarity about either what to do to prepare for changes in climate or uncertainty pertaining to the effectiveness of proposed actions. For example, Punta Gorda’s plan identifies the uncertainty associated with the effectiveness and potential side effects of large-scale beach nourishment: “As with statewide seawall construction, beach nourishment on this scale would be a mammoth engineering project, with uncertain environmental impacts of its own” (City of Punta Gorda 2015, 214). Uncertainty related to the coping capacity of a community was also infrequent in our sample, appearing in only five out of the forty-four plans analyzed.
On average, each of the thirty-two plans that acknowledged uncertainty identified 1.9 types of uncertainty. Three plans, Austin, Texas; Boulder County, Colorado; and Los Angeles, California, identified three of the four major types of uncertainty. We also grouped plans by community size to see if that influenced the number of identified types of uncertainty. Plans from small communities (i.e., populations under 50,000) identified, on average, less than one type of uncertainty each. Medium-sized communities (populations between 50,001 and 250,000) and large communities (population over 250,001) identified, on average, 1.6 and 1.7 types of uncertainty in their adaptation plans, respectively (Table 3).
Number and Percent of Plans in Sample That Contained Each Category of Uncertainty, Grouped by Population Size.
With fewer than five cases in some cells, we were unable to use chi-square techniques to statistically analyze our results. However, the results shown in Table 3 suggest that larger communities acknowledged more types of uncertainty.
Techniques for Addressing Uncertainty
The two most common uncertainty-reducing techniques were both in the predict and plan category of planning and included the use of multiple greenhouse gas emissions scenarios (used in 34/44 plans) and the conducting of vulnerability assessments (discussed in 32/44 plans) (Table 4). In regard to the use of multiple climate scenarios, there was wide variation in the types and numbers of scenarios used during the planning process, but nearly every community used greenhouse emissions scenarios developed by the IPCC. Four communities, Baltimore, Maryland; Los Angeles, California; Jamestown S’Klallam Tribe; and Lafourche, Louisiana, developed their own suite of scenarios. In the case of Baltimore, Los Angeles, and Jamestown, these scenarios were informed by the IPCC. For Lafourche, it was unclear what role, if any, IPCC scenarios played in their generation of place-based scenarios. The relatively common use of multiple climate scenarios is not surprising given the value this technique has for reducing uncertainty associated with what future climate conditions will be (the most common type of uncertainty identified in our sample). Importantly, the use of multiple climate scenarios did not translate into scenario planning: Instead, plans in our sample used multiple scenarios to provide a range of potential future climate conditions, and then communities chose one, possibly two, forecasts (often a high and low emissions scenario) upon which to base their planning.
Techniques Used to Address Uncertainty in Local Climate Adaptation Plans and the Percentage of Plans in Our Sample That Used Each Technique.
Nearly three-quarters of the communities in our sample (32/44 plans) used a vulnerability assessment to help inform their planning process. Vulnerability within the built environment, natural systems, and public health were frequently evaluated across all plans (for more details, see Woodruff and Stults 2016). Twelve out of the forty-four plans did not mention a vulnerability assessment during their local planning process. This could mean that a few additional communities conducted a vulnerability assessment as part of their planning but did not detail that process in their adaptation plan.
The popularity of vulnerability assessments in our sample is not surprising given the amount of emphasis placed on this technique in the peer-reviewed and grey literature. One organization that helped champion the importance of local adaptation planning, ICLEI-Local Governments for Sustainability, strongly encourages all local communities to conduct climate vulnerability assessments. While we were unable to determine which communities were members of ICLEI during their adaptation planning process, consultation with membership staff at ICLEI determined that twenty-four out of the forty-four communities in our sample either are or were within the last five years ICLEI members.
The third and fourth most common techniques are both examples within the adapt and monitor category of planning: monitoring changing climate conditions and adaptive management. Twenty-four out of the forty-four plans in our sample included content about continuing to monitor how the climate was changing and the impacts associated with that change. Half of the plans analyzed (22/44) mention the importance of adaptive management in planning for climate change. Terms such as learning, iterative, evolving, and dynamic were commonly used to describe the concept of adaptive management. While over half of the plans included language emphasizing the importance of adaptive management, only eight plans explained how they would adaptively manage the plan and strategies included in the plan (Anne Arundel, MD; Boulder County, CO; Confederated Salish and Kootenai Tribe; Guildford, CT; Lee County, FL; Los Angeles, CA; Santa Cruz, CA; and Satellite Beach, FL).
No plans in the sample identified specific actions as being robust, incremental, low-regrets, or no-regrets. Only two plans used strategies labeled as flexible (Punta Gorda, FL, and New York City, NY). Also uncommon was any discussion about the importance or value of low-regrets strategies (3/44), robust strategies (5/44), or incremental strategies (6/44).
Finally, despite the importance of scenario planning in the peer-reviewed literature, no plan in our sample used a scenario planning process to inform their planning and strategy selection. This suggests an important disconnect between what is suggested as a promising practice in theory and what is viable in practice.
The average number of techniques to address uncertainty in each plan in our sample was 4.5 out of nineteen: 2.3 from the predict and plan category and 2.2 from the adapt and monitor category. Small (those with a population under 50,000) and medium-sized communities (population between 50,001 and 250,000) used roughly 3.9 to 4 uncertainty-reducing techniques each. Large communities (population over 250,001) used, on average, 5.4 techniques per plan for addressing uncertainty and used a larger percentage of adapt and monitor techniques that foreground uncertainty. This suggests that large communities may have more capacity to undertake uncertainty-reducing planning techniques compared to smaller communities. Table 5 shows the total number of uncertainty-reducing techniques grouped by community population size.
Average Number of Uncertainty Reducing Techniques Used in Each Plan in Our Sample, Grouped by Community Population Size.
Discussion
In the first section of this paper, we grouped all uncertainties into four types related to climate adaptation planning in a way that highlights the uncertainties beyond local influence and the uncertainties within local influence. By examining the content of forty-four climate adaptation plans, we categorized 83 percent of identified sources of uncertainty as beyond local influence while only 17 percent related to local issues of coping capacity and effective responses. We believe that while local planners should acknowledge the many types of uncertainties, in the end, their time would be best spent by applying techniques from the adapt and monitor category to help minimize uncertainty associated with enhancing their coping capacity and identifying effective local responses—things they have the ability to directly influence.
In addition to being more focused and efficient, we believe this shift in planners’ thinking can avoid another challenge inherent in climate adaptation planning: defining what is climate adaptation planning success. In many cases, successful adaptation planning results in the lack of a bad outcome. On some level, proposing planning efforts where nothing happens makes it difficult to argue for action and resources. Much like our problems with the neglected repair of unseen infrastructure, community decision makers may have little motivation to direct limited resources toward adaptation if they can’t immediately show the resulting changes. For some, discussing the importance of building resilience to disruptions in continuity of service has addressed this outcome void. For example, in Grand Rapids, Michigan, the planning department is undertaking a process of asset management to reframe the need for infrastructure investments. In their explanations of this concept, planners begin by noting that most valued assets require maintenance. They use common examples: Houses require new roofs, and cars need regular oil changes. From this “good housekeeping” position, they explain how their city needs to actively care for its assets. This requires evaluating the condition of their existing infrastructure systems and setting incremental goals for system maintenance, upgrades, and expansion. In this way, they are advancing their understanding of coping capacity within an asset management approach that produces tangible outcomes that satisfy both politicians and citizens.
In many cases, municipalities have a sense for small or modest things they can do to begin preparing for climate change but are struggling with finding solutions to the larger, more challenging issues they face (Shi, Chu, and Debats 2015). In cases such as these, city-to-city networks, such as the Urban Sustainability Directors Network or the National League of Cities, provide forums where local governments can share promising practices with one another—sharing both successes and lessons learned (Moser and Pike 2015), among other benefits (Shi, Chu, and Debats 2015). Additionally, professional societies such as the American Society of Adaptation Professionals, American Planning Association, and American Society of Landscape Architects provide another venue by which individuals can share ideas, concerns, and promising practice with their peers (Moser and Pike 2015). These learning networks have repeatedly been shown to be instrumental in helping local communities start and advance their climate adaptation efforts. It is quite possible that these networks could also play a significant role in helping municipalities devise effective solutions for managing and reducing uncertainty while still taking steps to build resilience.
In the second section of our paper, we summarized planning techniques that are commonly identified for climate adaption planning and categorized these techniques as either traditional predict and plan or adapt and monitor. Overall, we identified the mention of 198 planning techniques for addressing uncertainty in the forty-four sample plans: 103 from the predict and plan approach to planning and 95 from the adapt and monitor approach. Each plan in our sample included an average of four of the nineteen uncertainty-reducing techniques. Counter to our hypotheses, we found that the number of uncertainties identified in a given plan did not correlate with the number of uncertainty-reducing techniques used during the planning process. For example, the plan by the City and County of Denver, Colorado, only identified one type of uncertainty but included nine uncertainty-reducing techniques. This suggests that other extraneous factors (e.g., municipal capacity, political support for adaptation) may influence how many and which uncertainty-reducing techniques are employed during adaptation planning.
The two most common uncertainty-reducing techniques were the use of multiple climate scenarios (present in 74 percent of plans) and vulnerability assessments (present in 73 percent of plans), both of which we categorized as predict and plan approaches. The two most popular approaches from the adapt and monitor approach were monitoring changing climate conditions (present in 55 percent of plans) and a discussion regarding the need for adaptive management (present in 50 percent of plans). Despite the slight dominance of predict and plan approaches, it is promising that thirty-five out of the forty-four plans include at least one uncertainty-reducing technique from the adapt and monitor framework. These techniques, however, tend to disproportionately focus on monitoring or discussing the value of adaptive management as opposed to changing planning processes or selecting adaptation strategies that will perform well across multiple possible futures. Consistent with our hypothesis, larger communities were more likely than smaller or medium-sized communities to use planning techniques from the adapt and monitor category to engage the topic of uncertainty.
Scenario planning is highly praised in the peer-reviewed literature but was not used in any of our forty-four plans. In theory, scenario planning is a technique that can help reduce or ameliorate all the types of uncertainty identified in this paper. With its focus on “improv[ing] our understanding of the future through systematic analysis of available information and ideas while highlighting, through the presentation of multiple possible outcomes or scenarios, how open the future is and how limited our knowledge of it remains” (Rickards 2013, 34), scenario planning is uniquely positioned to help address the various types of uncertainty associated with planning for climate change. More to the point, scenario planning was designed specifically to address issues that do not lend themselves to simple prediction, issues such as climate change (Dessai et al. 2009; Rickards 2013). The absence of scenario planning approaches may reflect the cost, time, unclear benefits, or staffing commitment necessary to undergo a complete scenario planning process (Chakrobarty 2011; Quay 2010; Zapata and Kaza 2015). This disconnect between what theory says should be a readily embraced technique to address uncertainty and what local practitioners are using clearly speaks to a mismatch between theory and practice.
Similarly, no community in our sample labeled strategies as being robust, and only two labeled strategies as being flexible. Moreover, only five communities in our sample discussed the importance of robust strategies. However, these five communities used the term robust as a synonym for hard or strong. This differs fundamentally from how the term robust is used in the planning and climate literatures (Chakraborty et al. 2011), where it is meant as a technique that is viable across multiple different futures. If a community does not use scenario planning in their planning process, then it is unlikely they will identify robust strategies that are capable of performing across multiple different futures. Similarly, the small number of flexible strategies may be attributable to the rigid nature of local governance in which specific ideas and plans must be created to secure the financing needed for implementation. This rationale may also explain why so few communities identified incremental strategies in their planning processes.
In contrast, plans within our sample frequently used monitoring and adaptive management to address uncertainty, but only eight plans had specific adaptive management processes while twenty-two discussed the importance of the concept. This suggests that the concept of adaptive management is in the process of being translated from theory to practice. Additionally, sixteen plans in our sample discussed the importance of both no- to low-regrets strategies and flexible strategies as techniques for addressing uncertainty. Only one plan, however, labeled a strategy as being no or low regrets.
Our findings also showed that ten communities (nearly one-quarter) used downscaled climate data in their adaptation planning process. These data were almost always regional, not local, in nature, but it does suggest that more communities are looking for and using data at finer resolutions to guide their decision-making processes. The use of information at this scale is mirrored by a call for the production of more downscaled climate data within the scholarly literature (Dessai and Hulme 2007; Hallegatte 2009). Given the limitations of downscaling, especially its potential to increase certain types of uncertainty, we are reluctant to emphasize this approach as being important for reducing climate-related uncertainty.
Overall, our findings suggest that planners in general are relying on more planning techniques that fall within the traditional predict and plan model of planning. It appears that flexible approaches from the adapt and monitor framework are just starting to emerge but are underdeveloped. In 2001, Myers wrote that uncertainty and disagreement are the twin hazards of planning. While planning for the future always involves some uncertainties, climate adaptation efforts add an additional layer of complexity. To help planners more effectively plan for climate change, we need to ensure they are equipped with knowledge regarding the various sources of climate-related uncertainty as well as the techniques available to help them reduce that uncertainty. The categorization of climate change uncertainties provided in this paper can help. Since not every uncertainty-reducing technique will be appropriate, planners should carefully evaluate their local context and determine which techniques are most appropriate to address the uncertainties plaguing their adaptation planning process. More to the point, we encourage planners to focus their attention on reducing and addressing uncertainties that they have the power to influence, specifically those related to enhancing local coping capacity and designing effective local responses.
As Abbott (2005, 2012) argues, the planning field exists because of uncertainty; thus, uncertainty is something planners regularly deal with (Hallegatte 2009). However, the types and complexity of uncertainty associated with climate change require planners to rethink their approaches. Scenario planning, adaptive management, monitoring, and incremental and robust strategies are just some of the tools that needed to address climate adaptation. In addition, planners need assistance in presenting the problems of climate change in a way that minimizes the symbolic dimensions circling the climate debate, including avoiding the evidence trap, binding the potentially endless discussion of uncertainty through the application of uncertainty reducing techniques, and transitioning toward an iterative problem-solving process that enlarges the particular municipality’s coping capacity. It is our job as scholars and practitioners to ensure that uncertainty does not become an excuse for allowing our communities to remain vulnerable to climate change.
Footnotes
Appendix
Codes Used by Researchers to Identify Types of Uncertainty and Uncertainty-Reducing Techniques.
| Codes related to identification of uncertainty |
| If the plan acknowledges uncertainty |
| If the plan provides details about the types or sources of uncertainty that exist. |
| Codes related to techniques to overcome uncertainty |
| If the plan mentions the value of no-regrets strategies |
| If specific strategies are labeled as no-regrets strategies (no-regrets detailed) |
| If the plan mentions the value of low-regrets strategies |
| If specific strategies are labeled as low-regrets strategies (low-regrets strategies detailed) |
| If the plan mentions the value of incremental strategies |
| If specific strategies are labeled as incremental strategies (incremental strategies detailed); |
| If the plan mentions the value of flexible strategies |
| If specific strategies are labeled as flexible strategies (flexible strategies detailed) |
| If the plan mentions the value of robust strategies |
| If specific strategies are labeled as robust strategies (robust strategies detailed) |
| If the community undertook a vulnerability assessment |
| If actions are selected that span multiple timeframes |
| If the planning process involved scenario planning |
| If multiple climate scenarios were used |
| If the community used climate downscaling or downscaled data in their analysis |
| If monitoring changing climate conditions is included as a strategy in the plan |
| If the plan identifies thresholds or tipping points |
| If the plan mentions adaptive management |
| If the plan builds in an adaptive management approach (detailed adaptive management) |
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.