ecological models (Peterson et al., 1997; Williams and Brown, 2014; see also Chapter 5). Until recently, there was little formal assessment of the potential consequences of changes to climate, such as sea-level rise, warmer temperatures, and changing precipitation patterns. However, the increasing availability of climate projections for South Florida and advances in climate change planning outside of the CERP offer potential to consider climate change within the CERP. This chapter addresses four opportunities to better incorporate information about certain aspects of climate change and associated tools into the CERP:

1. use of temperature and precipitation scenarios for restoration planning,
2. use of dynamic sediment accretion modeling to consider the effects of sea-level rise,
3. improved consideration of climate in ecological models, and
4. systems operating manual updates as a form of adaptive management in light of climate change.

This chapter builds on the last committee’s extensive discussions on climate change in NASEM (2023), which addressed changing sea level, precipitation, and temperature and U.S. Army Corps of Engineering (USACE) processes to incorporate this information into CERP planning, operations, and management, and a broader review of climate change and sea-level rise from that in NRC (2014). Interested readers can also consult other recent committee reports for additional in-depth discussions on saltwater intrusion and peat collapse (NASEM, 2018) and effects on estuaries and coastal systems (NASEM, 2018, 2021).

CLIMATE SCENARIOS FOR RESTORATION PLANNING AND MANAGEMENT

The success of Everglades restoration at least partially depends on the future precipitation and evaporation to be experienced in South Florida over the next decades. The extreme precipitation events of 2023 and 2024 illustrate the vulnerability of infrastructure to climate extremes that have not been previously experienced. All climate projections for South Florida indicate higher temperatures, which will in turn drive increases in potential evapotranspiration (USGCRP, 2023) and have potential to modify hydrologic processes in the Everglades. In addition, when the moisture-holding capacity of the atmosphere increases, more moisture is available to precipitate in each storm (when conditions are favorable), and precipitation is anticipated to be more extreme. Observations suggest this effect is extant in extreme precipitation events in the southeast United States (Kunkel et al., 2020; USGCRP, 2023).

For this reason, there is a clear imperative to ask whether restoration objectives will be achieved given the possible effects of a warmer climate on the

hydrology of South Florida. Answering this question is difficult under any circumstances and is especially so considering the inherent uncertainties with climate change projections and the multiple objectives, constituencies, and constraints that restoration faces. However, the modeling tools developed for the CERP provide a basis for answering this question, and methodological advances provide guidance for conducting such analyses. CERP planners appear to be well positioned to begin incorporation of climate change into their analyses. The framing of these analyses will have a large influence on the insights and utility for decision making gained from the results.

Outside of the CERP, efforts in Florida aim to create climate change scenarios from the latest climate models (General Circulation Models [GCMs] used in the sixth Coupled Model Intercomparison Project [CMIP6]). In one example, the U.S. Geological Survey (USGS) and Florida Flood Hub for Applied Research and Innovation¹ developed an ensemble of changes in extreme precipitation for South Florida categorized in terms of the duration of the storm event (e.g., 1 hour to 1 day) and the frequency of occurrence (or return period) (Irizarry-Ortiz and Stamm, 2022; Irizarry-Ortiz et al., 2022). The relationships between the depth of precipitation over a given storm duration and the associated frequency are known as Depth-Duration-Frequency values and are used in stormwater management planning. The Florida Flood Hub study comprehensively addresses many of the sources of uncertainty that are pertinent in climate change studies, including unknown future greenhouse gas emissions, the particular GCMs used, the downscaling methods used, and even the observed data used as the baseline. Change factors, which are multipliers applied to historical design variables (e.g., the 1-hour design storm precipitation depth) to create “climate-impacted” future design variables, were derived from a large ensemble of CMIP5 and CMIP6 (latest generation) GCMs that were downscaled using multiple statistical and dynamical methods and bias corrected with several historical datasets. The median change factors range from a change of 5 to 20 percent in near-term projections (10 years into the future) to 10 to 60 percent in the long-term future (100 and 200 years), while some change factors suggest greater than 500 percent increases. The comprehensive approach to addressing uncertainty ensures that the true range of uncertainty in climate projections is reasonably characterized and protects against planning based on a single or arbitrary sample of climate projections, which may lead to overestimating or, worse, underestimating risk. However, because results display a wide range, translating the results to actionable information is challenging.

___________________

¹ See https://www.usf.edu/marine-science/research/florida-flood-hub-for-applied-research-and-innovation.

---

A recent example from South Florida illustrates a pragmatic attempt to translate science into practice. The South Florida Water Management District (SFWMD) produced a summary of the USGS change factor study for adoption in South Florida flood management practices (Figure 4-1; SFWMD, 2022) and maintains a Resilience Metrics Hub that provides summaries of recent trends in precipitation as well as other variables. Broward County and the South Florida Regional Climate Compact have adopted a 20 percent increase in design storm events for land development planning to manage flood risks (Irizarry-Ortiz and Stamm, 2022).² The approach recognizes the need to address the risks of extreme events in the face of uncertainty and provides developers and planners certainty for design. By incorporating additional stormwater capacity in their development designs, planners will be providing protection against increases in precipitation extremes at a cost that is likely to be less than the damages that could be incurred if no action were taken (Rosner et al., 2014).

___________________

² See https://www.broward.org/resilience/Planning/Pages/FutureConditions100YearFloodElevation.aspx.

efforts have an important role to play in mitigating these effects by enhancing conditions that support sediment accretion, thereby enabling more gradual vegetation transgressions inland as sea level rises rather than sudden land loss. Ultimately, survival of coastal wetlands with sea-level rise will largely depend on their ability to accrete vertically and migrate inland. For mangroves in Biscayne Bay and the southern Everglades, accretion is almost entirely dependent on organic matter accretion from the accumulation of leaf and root biomass in the soil due to the relatively low inputs of mineral sediment (Breithaupt et al., 2017). An increase in freshwater and nutrient inputs through restoration could increase accretion through enhanced mangrove productivity in areas such as Biscayne Bay and the southern Everglades where anthropogenic impacts to hydrology have reduced freshwater flow and increased salinities. Additionally, accretion rate will also likely be affected by the rate of organic matter decomposition in the soil, which has been shown to increase with less flooding and greater soil phosphorus concentration (Poret et al., 2007). Sediment accretion models are used globally to estimate the vertical change in wetland elevation from a combination of organic matter and sediment accumulation in response to changes in flooding (including sea-level rise) and other factors such as sediment supply and wetland plant productivity.

A desired outcome of the Biscayne Bay and Southeastern Everglades Ecosystem Restoration (BBSEER) project is an enhanced accretion rate in coastal wetland habitats in response to sea-level rise with greater freshwater input. It is assumed that wetland plants within the project area will have greater productivity with a greater hydroperiod and water depth and lower salinity than under current conditions (Sklar et al., 2021). However, plant productivity will decline when optimal flooding and/or salinity conditions are exceeded. This constraint is important because the coastal environments of Biscayne Bay and the southern Everglades are relatively sediment limited and thus reliant on organic plant-based accretion (Breithaupt et al., 2017). Therefore, the survival of these important ecosystems depends on vegetation productivity for vertical accretion and/or the ability to migrate, but landward migration does not always offset marsh loss at the seaward edge (Osland et al., 2022). Sediment accretion models are being used in CERP planning for BBSEER to understand the potential outcomes of restoration alternatives. The sections that follow review the latest tool being applied in the BBSEER project and its use and limitations for informing decision making on restoration investments.

BBSEER Adaptive Foundational Resilience

Sediment accretion models are being used in the BBSEER planning process as the Adaptive Foundational Resilience (AFR) performance measure (see Box 4-1 for details), which is one of nine performance measures used to assess project

wetlands; Figure 4-2). This sets conditions similarly for all future comparisons but may not be at all realistic as a starting point.

Second, as currently being utilized for the AFR performance measure, the relationships between accretion rate and flow, hydroperiod, and salinity are loosely based on data but with significant smoothing based on conceptual understanding (Figure 4-2). The assumed relationships seem reasonable, but the specifics are critical (e.g., maximum and minimum rates, slopes, intercepts). For example, the maximum accretion value for mangroves (8–10 mm/yr) is slightly above the upper threshold rate of 7 mm/yr quantified from a comprehensive review of mangrove accretion associated with Holocene sea-level rise (Saintilan et al., 2020). Furthermore, accretion is highly site specific, depending on local hydrology, supply of sediment, salinity, nutrients, and other factors. Thus, the relationships used for estimating accretion may also have large associated errors not included in the modeling effort. To clarify the reliability of estimates provided by the AFR performance measure, relationships between hydrology and salinity and habitat-specific accretion rates should clearly show and reference relevant data from Everglades ecosystems or similar geomorphic settings (e.g., Breithaupt et al., 2017; Craft and Richardson, 1998; Feher et al., 2020; Lynch et al., 1989; Reddy et al., 1993; Sklar et al., 2021; Smith et al., 2009; Smoak et al., 2013). Placing existing data on curves representing relationships between peat accretion and hydrologic drivers will lend confidence to relationships and outcomes.

Third, simply averaging accretion rates derived from the three drivers (salinity, sheet flow, and depth-duration) is unrealistic; therefore, the model should incorporate some interplay among these drivers. For some habitats (e.g., sawgrass) salinity will likely be a more important constraint on productivity and accretion than flooding. But even for mangroves, which are more salt tolerant, there is little scientific support for the averaging of accretion outcomes in response to different abiotic conditions. Considering that plant productivity is limited by the most constraining abiotic factor, a limiting factors approach or a weighting approach may be a more appropriate way to combine multiple factors. Applied to the accretion response curves in the most simplistic way, this would be the minimum accretion rate of the three factors. Ultimately, sediment accretion modeling is being used in a relatively limited scope (only for the AFR performance measure) to compare project alternatives, and not as a predictive model of future conditions. There is the potential value of accretion modeling to be used in project cost-benefit analysis, planning, monitoring, and adaptive management. Modeling the effects of sea-level rise on coastal habitats is challenging because of the complexity of wetland morphodynamics; however, realistic quantitative predictive models exist and could be utilized (see Fagherazzi et al., 2020, for a comprehensive review). For example, multiple predictive models of marsh evolution, such as the marsh equilibrium model (MEM) and

the WARMER model, which are point-based models of elevation and sediment accumulation related to sea-level rise, could be applied to BBSEER habitats (e.g., Fagherazzi et al., 2020; Morris et al., 2002; Schile et al., 2014). A recent steady-state model of soil accretion rates in the Everglades with sea-level rise has been developed (Chambers et al., 2021), which may be a good example or starting point. Dynamic modeling could be conducted as a parallel effort to habitat unit evaluation and can serve as the basis for comparing future monitoring efforts.

In evaluating the value of the AFR accretion modeling effort to gauging resiliency of coastal wetlands to sea-level rise within the CERP, the question is “What are the consequences of unrealistic predictions of sea-level rise effects on coastal habitats for BBSEER restoration planning and management?” Without more confidence in the assumed accretion rates under all scenarios, it is difficult to assess whether restoration investments are likely to deliver anticipated returns. To increase confidence in predicted future accretion rates, more accurate models should be incorporated, which would also have the added benefit of enhancing the outputs of ecological modeling through better predictions of the productivity and distribution of vegetation communities. Uncertainties in the distribution and health of the foundational ecosystems will compound the uncertainties of ecological models at higher trophic levels. Lastly, the AFR is one of nine performance measures (see Chapter 2) being used in the evaluation of project alternatives. All performance measures are weighted equally in the evaluation of alternatives, and with sea-level rise resiliency just one performance measure out of nine, the relative importance of this metric in the alternative evaluation exercise is very low despite the fact that it will largely determine future submergence and distributions of coastal ecosystems. Overall, while the AFR performance measure is a relative measure of project performance of alternative actions, there remains large uncertainty about the ability of the ecosystems to adjust to future rates of sea-level rise, and these uncertainties merit further use of sediment accretion data and modeling tools to inform future investments.

Restoration Planning Needs

More accurate accretion estimates necessitate a model that begins with current measured accretion rates in representative habitats under present conditions of sea-level rise, hydrology, and salinity. Such a modeling approach would use the current digital elevation model and integrated model inputs of hydrology and salinity from project alternatives and projected rates of sea-level rise to estimate annual accretion rates. By considering feedback between processes and/ or limiting factors in annual accretion rate calculations, more realistic estimates of changes in wetland elevation from present day to 2085 can be determined (e.g., Schile et al., 2014). Additionally, the relationships used for model-based

accretion predictions should undergo rigorous calibration to the extent possible using field data and existing literature values (see citations provided above), moving beyond the conceptual relationships that are currently used. Improved dynamic modeling approaches would build on the current utility of the AFR modeling for comparing potential restoration alternatives to provide more realistic evaluation of wetland responses and project outcomes under a range of future climate scenarios.

MODELS FOR ASSESSING ECOLOGICAL RESPONSES TO CLIMATE CHANGE

Evaluating responses of biota and ecosystems to changes in hydrology in the Everglades is a key component of measuring restoration progress and success under the CERP. The USGS-led Joint Ecosystem Modeling (JEM) collaboration updates and applies 20 ecological models spanning a number of key taxa (e.g., birds, fishes, crayfish, herpetofauna) and their habitats to a range of restoration alternatives and environmental scenarios including sea-level rise and increased salinity due to climate change. These models use as input a variety of hydrological variables from the Everglades Depth Estimation Network (EDEN), the Regional Simulation Model (RSM), and more recently BISECT, as well as data and information from monitoring of species, their habitats, and other ecosystem variables. A few models feed output into other ecological models; for example, the Snail Kite model (EverKite) requires as input population densities estimated from the Florida apple snail model (EverSnail), and the wading bird models use fish density and biomass as input (Beerens et al., 2015; D’Acunto et al., 2021; Shinde et al., 2014). Although most of the ecological models output spatially explicit habitat suitability indices that integrate environmental predictors of species presence, a small number of them estimate or project spatially explicit population abundances, densities, or biomass founded on life history characteristics of the target species (e.g., Small Fish Density Model, Prey Fish Biomass Model, EverKite, EverSnail, and Wader Distribution Evaluation Modeling), and one model estimates amphibian community richness.³

The models under the JEM collaboration have the capacity to incorporate the climate change impacts of altered hydrology, sea-level rise, and increased salinity where relevant and thus can be (and in some cases have been) used to project the ecological effects of hydrologic change from restoration and climate change when data are available (Perez et al., 2017). For example, Catano et al. (2015) and Romañach et al. (2023) use climate change scenarios and projected hydrological responses, including sea-level rise, as input into ecological models to predict changes in fish density, alligator habitat, wading bird distribution, apple

___________________

³ See https://www.jem.gov/Modeling.

snail density, amphibian occurrence, and Cape Sable seaside sparrow probability of presence. Results from these studies indicated a decrease in suitable habitat and fish and apple snail density under climate change scenarios that project lower precipitation and higher evapotranspiration, while an increased rainfall scenario benefited fish densities and alligator habitat. Both studies emphasize the importance of freshwater flow into the southern Everglades to counteract reduced rainfall, protect habitat, and forestall the impacts of sea-level rise under climate change. These studies show that ecological models can be powerful tools to explore the effects of different climate change scenarios and restoration strategies on the wildlife species that are indicators of Everglades restoration success. When they integrate both habitat and population responses to environmental change, these tools can clarify the aspects of climate change that are likely to be the most destructive to ecosystems and the restoration activities that can have the greatest potential of mitigation. They are one of the few means available for forecasting the effects of climate change on individual species distributions and populations across large spatial scales and time periods (Wilsey et al., 2013). Despite the sophistication, longevity, and ground truthing of these models, they suffer from limitations when applied to the ecological impacts of climate change, in part because almost all of them do not explicitly include relevant effects of temperature increases on the target species. Only the EverSnail model, which projects the abundance and distribution of Florida apple snail using demographic data and environmental variables, explicitly includes the effects of temperature on reproductive rates. EverSnail output is input as prey into the EverKite model to project the movement, reproductive success, and mortality of the endangered Everglades snail kite (Darby et al., 2015).

Both hydrological and temperature changes are expected to have profound effects on biodiversity in myriad ways (Bellard et al., 2012; Mantyka-Pringle et al., 2012; McLaughlin et al., 2002). Hydrology has been the central focus of Everglades restoration; hence, much is known about the ecosystem effects of altered flow, water depth, fluctuations in precipitation, and sea-level rise. With perhaps the exception of the influence of temperature on algal blooms in Lake Okeechobee and Florida Bay (Havens et al., 1994; Koch et al., 2007), very little consideration has been given to the effects of increased temperature trends on the Everglades. The extreme surface ocean temperatures in coastal southern Florida in 2023 underscore the urgency of better understanding the effects of increased temperature on the success of ecological restoration in the Everglades. Moreover, under all emissions scenarios, projected increases in temperature are more certain than precipitation projections, which predict high variability in rainfall (Kunkel et al., 2020). Therefore, a focus on the effects of increased temperature on Everglades biota could better illuminate the effects of climate change on ecological restoration success than hydrology alone (Grieger et al., 2020).

ecological models are based. Some of these models could readily incorporate explicit effects of increased temperature in addition to changes in hydrology (e.g., breeding phenology [Shinde et al., 2014]). Indeed, Shinde et al. (2014) acknowledge, “Alligators nest earlier following warmer springs and delay nesting following colder springs (Joanen and McNease, 1989; Kushlan and Jacobsen, 1990). Effect of temperature on nesting will be an option to include as more data become available.” Other endeavors to explicitly incorporate the effects of temperature changes would require adaptations of the existing models. For instance, a stage-structured population model of American alligators (Alligator mississippiensis) linked to the Alligator Production Suitability Index model may be necessary (and feasible) to investigate the effects of male-skewed sex ratios due to climate change (Brook et al., 2000; Gerber and White, 2014; Visintin et al., 2020).

Additionally, consideration should be given to the development of mechanistic niche models that link physiology with spatial data to project species ranges under current conditions and climate change scenarios (Kearney and Porter, 2009) to supplement the JEM models (see Fordham et al., 2012; Franklin et al., 2014). Mechanistic niche models can have greater relevance to climate change applications than purely correlative niche models and the types of hydrology-based mechanistic habitat suitability models used in JEM models. Mechanistic niche models can link thermal tolerances (among other physiological responses) with environmental variables and spatial occurrences to constrain species distributions to more realistic ranges that track physiological responses to the environment (Evans et al., 2015). This should currently be feasible for crocodilians, for which much is known about thermal tolerances, behavior, and responses to environmental conditions (e.g., Fujisaki et al., 2014; Lawson et al., 2018; Mazzotti, 1989; Smith, 1979). Moreover, there exists a foundation of species distribution models developed in the region on which to base adaptations to include physiological tolerances to climatic variables (Bucklin et al., 2015; D’Acunto et al., 2021). Like all models, mechanistic niche models are subject to uncertainties in input parameters and model uncertainty (Elith et al., 2002; Regan et al., 2002), which compounds under climate change scenarios. However, a greater suite of tools in the climate change modeling toolbox that can shed light on projected biotic responses to climate change, and the inherent uncertainties, can better equip managers to plan for vulnerabilities, responses, and activities to reduce uncertainty.

Any effort to explicitly incorporate the effects of temperature changes into species- and ecosystem- level ecological models will require long-term monitoring data and field experiments, to measure the response of organisms to novel environmental conditions, and spatially explicit forecasts of air, water, and soil

temperatures under climate change and restoration. Such an endeavor can and should be tied to adaptive management and monitoring—the ecological models can be used to support project management decisions by comparing expectations to outcomes (see Chapter 5) and to highlight monitoring needs. Monitoring and adaptive management will provide data and information on how models could be updated and applied (NASEM, 2021), although care needs to be taken to account for non-stationary processes associated with climate change (Austin, 2007; Koons et al., 2016; Wolkovich et al., 2014). Such an undertaking requires long-term commitment and careful coordination and communication among teams overseeing models, management, monitoring, and data collection, although databases on species’ thermal traits (e.g., GlobTherm, TRAD) have been compiled that can support some data needs (Bennett et al., 2018; DuBose et al., 2024; Lancaster and Humphreys, 2020; Pottier et al., 2022). Machine learning methods also offer opportunities to gain insights into complex ecological systems under uncertainty and non-stationarity (Jones et al., 2023; Peters et al., 2014; Stupariu et al., 2022). These methods span hypothesis-data-driven, data-intensive, and deep learning approaches and include random forests, neural networks, support vector machines, regression trees, and Bayesian networks, among many other methods that continue to be developed and refined as the capacity of computing and artificial intelligence expands (Stupariu et al., 2022).

INTEGRATION

Understanding and managing for the effects of climate change on ecosystems requires an overarching framework to integrate ecological and physical models with monitoring data and climate projections, a consistent set of climate scenarios and assumptions across models, and mechanisms to adapt and update each of these models according to the information that each provides. Such an endeavor is necessary to understand, plan, and execute ecological restoration under a more comprehensive suite of the likely effects of climate change and restoration. However, when predicting potential effects of climate change, it is vital to understand the many factors, and their uncertainties, that might explain or impede restoration success.

The Everglades Vulnerability Analysis (EVA) model shows promise for integration across ecological and physical models, climate projections, and monitoring data that could provide feedback to adaptive management. This recently developed framework aims to integrate physical and ecological models and, with further development, could serve as a basis for a larger enterprise that links models to monitoring and adaptive management efforts. EVA evaluates and synthesizes the impacts of hydrology and salinity on indicators of Everglades

---

models,⁴ available data, and expert knowledge (Figure 4-5). Vegetation types are classified into six freshwater marsh and coastal communities—freshwater prairie, mangrove, mangrove scrub, open water, sawgrass, and upland—the results of which feed into peat accumulation and wading bird modules. The Bayesian network outputs probabilities of outcomes generated by influence diagrams and conditional probability tables, which are used to calculate a “vulnerability score” spatial surface defined as the distance from a predefined target site outcome. As such, the EVA does not assess vulnerability per se; rather, it characterizes the expected state of the system as it compares to a set of desirable outcomes (in a complementary way to the BBSEER performance measures), thus highlighting areas of the landscape that are furthest away from ecological restoration goals. The output appears to be limited to separate spatial visualizations of model output (e.g., maps of vegetation types; wading bird colony size; alligator nesting

___________________

⁴ See available models at https://jem.gov/Modeling.

sites in wet, dry, and median years), which can be useful information for targeted management of individual species or vegetation types but has limitations for understanding trade-offs across multiple management objectives.

A major advantage of the EVA is that it is based on a Bayesian network, which is equipped with a set of tools to evaluate model accuracy, sensitivity, and uncertainty that can be readily updated to incorporate additional modules and new data. This feature is particularly pertinent for adaptive management applications in which the learning process is a central feature (Landis et al., 2017). Management can be informed by the EVA, which in turn can be updated (with new input parameter values or modules) based on new information to further inform management and data collection. In this way, ecological modeling, management, and monitoring are integrated into an ongoing adaptive management cycle (Howes et al., 2010; Rumpff et al., 2011).

The EVA framework’s decision support utility could be enhanced by the adoption of a post-hoc tool that aggregates or optimizes across the ecological indicators and landscape features. This tool might involve multi-objective optimization approaches (Brias and Munch, 2021; Chadès et al., 2017; Williams and Kendall, 2017) or an approach such as EverForecast that examines trade-offs between all pairs of species (Romañach et al., 2022). Trade-offs exist in nearly all Everglades restoration decisions (e.g., between suitable hydrology for tree island restoration and Cape Sable seaside sparrow breeding habitat [see Chapter 3, Box 3-5]). In light of these trade-offs, decision making should be supported by an integrated framework based on ecological and physical models and informed by monitoring and continuous learning (NRC, 2010). Such a framework should enable weighing the effects of restoration projects on multiple ecosystem components in the context of climate change.

APPLICATION OF CLIMATE TOOLS FOR ADAPTATION OF OPERATIONS

Climate change and associated changes in the frequency and magnitude of extreme events complicate our ability to precisely predict the future performance of CERP infrastructure. Although infrastructure often is considered static in nature, the rules and guidelines by which the infrastructure is operated add a degree of flexibility (see also Chapter 5). Operation of control structures and pump use can be adjusted in response to changing conditions and improved understanding of the system performance. System Operating Manuals (SOMs), regional water control plans such as the Combined Operational Plan (COP), and Project Operating Manuals (POMs) are the policies by which CERP infrastructure systems and components, respectively, are operated. In the context of climate change, these operating plans offer the means to adapt the operations of CERP infrastructure to the specific changes and surprises that directly affect CERP performance.

Recognizing and utilizing this potential can improve the performance of the CERP in the face of a changing and variable climate.

As CERP projects come online, they require operating plans, and those that cause downstream effects or depend on upstream projects require updates of the relevant SOM. These updates occur every 3 to 5 years, providing a means to incrementally adjust operations in accordance with prevailing and anticipated conditions. During that time increment, more information is revealed about external conditions and the performance of existing CERP projects, and improved tools are available to estimate expected future conditions; thus, uncertainty is reduced. To benefit from this opportunity for learning in the face of climate change, the system’s performance should be evaluated in light of these external factors (e.g., temperature, precipitation, sea-level rise trends) and factored into the updates of the POMs, regional water control plans, and SOMs, as well as the project design itself. As the Central Everglades Planning Project (CEPP) 1.0 comes online to replace the COP (see Chapter 2), there is an opportunity to standardize incorporation of the latest climate change considerations in operating plan revisions. Effectively doing so requires communication between modelers, monitoring staff, and climate experts as well as a process for learning to inform future CERP operations. It also requires updating of the historical record of weather and sea level that is used to develop operating manuals, ideally on the same frequency as the updating of the SOMs and POMs.

CONCLUSIONS AND RECOMMENDATIONS

Attempting to include climate change information in CERP planning could overwhelm even the most intrepid project planner, and CERP planners have been cautious in their attempts to do so. The amount of climate change information, the unknown credibility of the information, the number and complexity of models used in CERP planning, and the complications of incorporating climate information into models not designed for that purpose add up to a very challenging endeavor. Although there is danger in using climate projections without a carefully considered plan that is consistent across the diverse set of analysis and models involved in the CERP, the greatest danger is making no attempt to plan for climate change. The committee offers the following recommendations to advance the use of climate change information in CERP planning and operations.

A strategy to understand the impacts of climate change should be developed with a curated set of scenarios that are used consistently across all components of planning and restoration implementation. This set of climate change scenarios should represent the range of plausible changes based on review of the scientific literature and available climate projections and be used to assess project and system vulnerability to changes in temperature, precipitation, and

sea-level rise. Validated methods for the use of climate scenarios, of which there are several in the scientific literature, should be used to stress test restoration plans. These planning scenarios should be applied through existing hydrologic and ecological models to provide insights on the potential vulnerability of flora, fauna, and infrastructure to these plausible changes. This strategy can be implemented now, and the results can lead to better estimation of benefits of planning and ultimately better outcomes in the long run.

A dynamic model that predicts coastal wetland elevations through time informed by empirical data is needed to provide more accurate predictions of coastal restoration outcomes and guide investment decisions. Sediment accretion models are currently being used to compare alternative restoration plans for BBSEER in the context of sea-level rise. Because the models are based on approximated relationships between flooding and accretion and salinity and accretion, the predicted and even relative outcomes for each alternative may not reflect reality. Recommendations to improve the accuracy of accretion models, particularly if they are to be predictive of future conditions, are to (1) start models with elevation and accretion rates at present day with annual time steps that include projected sea-level rise and project alternative effects; and (2) ground truth relationships between flooding, salinity, and accretion using existing data from comparable environmental settings. A dynamic predictive model of accretion and wetland elevations over time will be more accurate and can serve as a guidepost for monitoring and adaptive management. Development of a dynamic model could provide helpful confirmation of the potential ecological return on the large, expected infrastructure investments in BBSEER and inform the pending planning of the Southern Everglades project.

Existing ecological models should be used to a greater extent and further developed to anticipate the effects of climate change, including temperature, on the wildlife indicators of Everglades restoration success. Because wildlife species and habitats are the ecological endpoints of Everglades restoration, the output of ecological models should be considered early in the process (when evaluating restoration plans) rather than late in the process (when preparing National Environmental Policy Act reports). Ecological models should be developed and applied to evaluate the effects of projected changes in precipitation and temperature on biotic indicators of restoration success. Because confidence in temperature projections is greater than that for precipitation, attention should be paid to the effects of increased temperature on life history, phenology, and physiology of wildlife species using tools such as mechanistic niche models. Furthermore, a more thorough accounting of uncertainty in ecological models with respect to climate change impacts should be undertaken. How changing climate is incorporated into models based on historical data that are unlikely to hold in the future as well as the reliability of tools under changing, or non-

This page intentionally left blank.