A Model for Future Development Scenario Planning to Address Population Change and Sea Level Rise

Abstract

Population growth and land use change often have significant environmental impacts, affecting biodiversity, water supply, agricultural production, and other resources. Future scenario models can provide a better understanding of these changes, helping planners and the public understand the consequences of choices regarding development density, land use, and conservation. This study presents a model that has been used to identify alternative future scenarios for Florida considering future population growth and land use. It includes two scenarios: a “Sprawl” scenario reflecting a continuation of current development patterns and a “Conservation” scenario with higher densities, redevelopment, and more land protection. The study incorporates sea level rise scenarios for both 2040 and 2070. Results show that the Sprawl scenario could lead to 3.5 million acres of new developed land and 1.8 million acres of lost agricultural land by 2070 in Florida. In contrast, the Conservation scenario for 2070 results in 1.3 million fewer acres of developed land and 5 million more acres of protected natural land, showing that it is possible to accommodate future population growth while reducing impacts to agricultural and conservation priorities in Florida. Although this is by no means a “prediction” of future Florida, it has been useful as a tool for evaluating potential future land use scenarios and is a model that may be more broadly applied by other locations and users.

1. Introduction

The use of geographic information systems (GISs) has accelerated in recent years in what is referred to as a geospatial revolution. Increased quality of geospatial information, lower software and hardware costs, and a growing use of maps and spatial data have provided benefits to environmental and land use planning [1]. Easily accessible, high-quality data is often available from local, state, and federal governments and can be utilized by planners to develop scenarios to assist in future scenario planning, including to evaluate impacts from population change and future development. Future development or land use scenario models can assist in achieving planning goals by helping users to explore the potential outcomes of planning and policy decisions. Comparisons of multiple scenarios can aid planners and other decision makers in creating future land use plans and policies that consider the needs of both humans and the natural environment [2]. With an increased focus on sustainability and climate change, there is a growing need for future development and land use change models [3,4]. Coupled with the greater availability of land use and land cover data, socio-economic data, and computer software and hardware advances, there has been an increase in the use of these models [5].

Land use change represents one of the most significant drivers of environmental damage in the world [2]. The United States lost or compromised approximately 2000 acres of farmland per day between 2001 and 2016 [6].The environmental impacts from changes such as these can be wide-ranging and include loss of biodiversity, pollution, increases in nutrient loading and runoff, changes in the hydrologic cycle, and exacerbation of climate change impacts [1,7]. Ineffective urban plans are often blamed for many of the environmental and economic issues facing the United States [1]. Sound land use planning is particularly important in Florida, with a net population growth of approximately 849 people per day [8] that has spurred extensive suburban and urban development—coupled with natural and agricultural landscapes that provide food, fiber, and other conservation, cultural, and recreational values.

Sound land use planning is particularly important in Florida, a state with a net population growth of approximately 849 people per day [8] that has spurred extensive suburban and urban development—yet still possesses extensive natural and agricultural landscapes that provide food, fiber, and other conservation, cultural, and recreational values.

This study, entitled Sea Level 2040/2070, sought to expand on an existing methodology for future development scenario modeling in Florida, which has been used to better understand the impacts of future land use decisions, including to agricultural lands and natural resources. The models explore the role of density, redevelopment, sea level rise, and suitability in determining potential future land use and development patterns and have been tested and used extensively for planning and public engagement applications within the state. A suitability-based approach was used to identify potential areas of future development, incorporating county-level population projections and future sea level rise scenarios. The work was based on the Land Use Conflict Identification Strategy (LUCIS) method developed at the University of Florida by Dr. Paul Zwick and Professor Peggy Carr, and the 2006 Florida 2060 [9] and 2016 Florida 2070 models [10]. Updates to these models were incorporated, including sea level rise, and these additions are further described herein.

The LUCIS methodology uses GISs to analyze the suitability of land for various uses and identify conflict between competing land uses [10]. The methodology was further developed into LUCIS plus which added more detailed allocations including employment for uses such as commercial, retail, service, industrial, and institutional [11]. The LUCIS methodology is adaptable to a wide range of scales and needs, and not all aspects of the methodology are implemented on every project. More detailed analyses such as suitability layers for non-residential uses, employment allocations, and conflict identification sometimes used at county-level scales are also discussed as part of the LUCIS model but were not included in Sea Level 2040/2070 models.

The most recent iteration of these models described herein was titled Sea Level 2040/2070 and released in 2023. In Sea Level 2040/2070, two scenarios were developed to investigate potential alternatives: a Sprawl Scenario and a Conservation Scenario. The Sprawl Scenario is a “business as usual” or trend scenario, representing a continuation of existing development patterns. The Conservation Scenario avoids development of land important for conservation and increases development densities. Both scenarios were modeled for two future time periods—the years 2040 and 2070. Scenario results were subsequently used as part of a public outreach campaign to engage planners and citizens to better understand the implications from future development planning decisions on Florida land, agriculture, and conservation resources.

A variety of future development scenario models exist. The models for future growth for this study used a deterministic process which provides the same results each time the model is run as long as there are no changes made in the data or methodology. Many land use and land change models are stochastic, which introduce variability and produce a range of results. These models have gained popularity due to their ability to account for variability [12]. Notable examples include the Farms Under Threat report published in 2023 by the American Farmlands Trust (AFC) [6]; the Peninsular Florida Landscape Conservation Cooperative [13] model developed by GeoAdaptive; and FUTURES, developed by North Carolina State University [14]. The AFC and FUTURES studies modeled larger regions of the country that often lack consistent publicly available data.

Research by Pickard et al. [12] on the difference between multiple land change models found substantial variability in results with differing inputs and parameters, affecting the quantity of land use change, landscape configuration and spatial allocation. The models and methodologies used for the Sea Level 2040/2070 study differ from the previously mentioned models in terms of specific methodologies, data sources, and ease of use. The Sea Level 2040/2070 model used standard ESRI ArcGIS tools combined with publicly available data, with the goal of producing a relatively simple and transparent approach for modeling future development scenarios, which can be easily communicated to the public and policymakers.

Using the results from the Sea Level 2040/2070 models, this study helped to expand the discussion around future growth in Florida. The process can be thought of as a series of “what if” questions. What if we continue the current patterns and density of growth? What if future populations are accommodated with denser, more compact development? What if we can accommodate growth while also protecting high priority conservation and agricultural lands? The methodology in this study describes one tool to investigate the outcomes from decisions about future growth before the impacts are irreversible. The methodology is also broad enough to be adapted for other locations where future development scenarios are needed to support land use planning activities by decision makers and citizens.

2. Materials and Methods

2.1. Study Area

This project encompassed all 67 counties in the state of Florida, United States (Figure 1). The state is ecologically diverse and includes both sub-tropical and tropical climates, with a wide range of natural communities including upland forests, dry and wet prairies, swamps, and freshwater marshes among many others. The state is primarily flat with the northern portions reaching 100 m above sea level while elevations in the southern part of the state are rarely over 30 m [15]. Areas such as the Big Bend of Florida are low and flat, with this shallow gradient extending into the Gulf. These areas are already being affected by changes in sea levels [16].

The biodiversity of the state is exceptionally rich and the state is home to over 16,000 invertebrates, wildlife, and fish species [17]. There are over 600 springs, 7800 freshwater lakes, 11 million acres of wetlands, and 1700 rivers and streams [18].

In addition, approximately a third of the 36.6 million acres in Florida are in agriculture including ranching, silviculture, field crops, nurseries, citrus, and other commodities [19]. Florida has over 44,000 farms composed of over 9.7 million acres. Agriculture, forestry, and fishing brought in over 11 billion dollars in 2022 [20].

Sea level rise is a significant concern for Florida in part due to its low topography. High density coastal populations will likely be impacted by sea level rise complicating flood management and mitigation strategies. Higher sea level rise can also increase groundwater levels which can lengthen the duration of flooding events and impact salinity levels and groundwater storage [21].

Florida is also experiencing rapid growth in population. Four of the five fastest growing metropolitan statistical areas in the United States were in the state from 2022-2023. [22]. This development has degraded the form and function of ecosystems and has also had economic impacts on the multi-billion dollar income from nature-based tourism [23].

2.2. Data

This study used a GIS-based model to produce future development scenarios for the state of Florida that incorporate population growth projections, assumptions about development patterns, and future sea level rise for a 2040 and 2070 time horizon. The study utilized publicly available data from a variety of state and academic sources. In some cases, older population projections or other data were used in the 2070 models to maintain consistency with the 2016 Florida 2070 study [10]. The 2040 models used the most currently available data to better facilitate analyses using the shorter timeframe.

The majority of the input data was either raster data or vector data that had been converted to raster format. Raster data is particularly well-suited for analyzing multiple data layers, which is essential for conducting suitability analyses [24]. Vector data was mainly employed to create maps and define boundaries in the analysis of individual counties, such as urban density calculations and statistical assessments.

The data used in the analysis was obtained from publicly available sources, primarily the Florida Geographic Data Library (FGDL), Florida Natural Areas Inventory (FNAI), and the Florida Department of Agriculture and Consumer Services (FDACS). Vector data was converted to raster format with a 30 m resolution and using the Albers Conical Equal Area projection.

This study relied on several data sources to identify existing and potential conservation lands (Figure 2). The Florida Managed Areas database (FLMA) developed by the Florida Natural Areas Inventory [25], includes boundary and statistical information provided directly by the managing agencies for federal, state, local, and privately managed lands. These are existing already protected lands conserved by fee simple, easement, or other conservation mechanisms. Florida Forever is one of the state’s two primary land acquisition programs, focused on acquisition of lands for conservation and nature-based recreation. Approximately 3.8 million acres of unprotected Florida Forever project lands were used in this project as one of two datasets to identify conservation priorities [26]. The Florida Wildlife Corridor (FWC) (Priorities 1–3 of the Florida Ecological Greenways Network (FEGN)) [27], consists of a network of ecologically significant, functionally connected public and private lands with approximately 7.7 million acres of unprotected lands, which were used as a second dataset for identifying high conservation priorities. This network plays a crucial role in guiding federal and regional land acquisition efforts by providing information on priorities for maintaining critical, intact landscapes and ecological corridors [28]. While the FWC and Florida Forever projects were permitted to develop in the Sprawl Scenario, they were excluded from the Conservation Scenario as a means of evaluating whether future development might be accommodated while avoiding impacts to high priority conservation areas.

2.3. Methods

The concept for the model used in this study was to use an “apples to apples” approach with the previously described 2016 model, known as the Florida 2070 study [10], to maintain as much consistency as possible and so that the primary difference would be sea level rise. Consistency was a lesser goal for the 2040 model, allowing for further differences in inputs and methodology. The results of the 2070 model are not an extension of the 2040 results. Areas shown as developed in the 2040 model may not be projected to be developed in the 2070 results due to the differences in data and methodology. Work was conducted in three main steps as shown in (Figure 3): (1) develop model assumptions and scenarios and identify areas not available for future development (masks); (2) conduct a suitability analysis to determine the most likely locations for future development based on model assumptions; and (3) distribute projected future population based on the suitability model and development density to produce the final future development scenarios.

2.3.1. Scenarios

Five scenarios were developed to explore the patterns and potential impacts of future development, population growth and sea level rise in Florida: Baseline, Sprawl 2040, Sprawl 2070, Conservation 2040, and Conservation 2070. Each of the future scenarios includes assumptions regarding population growth, sea level rise impacts to Florida, and which areas are the most likely or “suitable” locations for future development. The term “suitability” is used in this paper to refer to the areas assumed most likely for future development based on the modeling parameters used. However, this should not be interpreted as a statement of where future development should go.

The baseline scenario identifies the locations of “current” development using 2019 Florida Department of Revenue Land Use Codes (DORUC) from statewide parcel data and was used as a means of measuring future change. A detailed listing of land use codes and assumptions to identify existing development is included in Appendix A. The Sprawl Scenarios represent a “business as usual” approach to development based on existing (2010) development patterns and densities. The Conservation Scenarios, as the name suggests, aimed to accommodate the same future population as the Sprawl Scenario but at higher densities and with increased land protection.

2.3.2. Assumptions

Florida’s population is expected to grow by approximately 5 million residents by 2040 and 12.2 million residents by 2070, increasing the population by 23% and 57%, respectively, above the 2019 census population. These population projections were sourced from the Bureau of Economic Business Research (BEBR) at the University of Florida. The 2040 model used the most recent medium projections available at the time of the study [29]. The 2070 population projections were based on the projections used in the 2016 Florida 2070 model to maintain consistency between the studies [10].

The 2040 scenarios used a 0.25 m sea level rise (SLR) (

Table 1. 2040 and 2070 scenario assumptions.

	Sprawl 2040	Conservation 2040	Sprawl 2070	Conservation 2070
Development Densities	New population to be accommodated is allocated at each county’s 2019 gross density	New population is allocated at a development density 30% greater than that used for Sprawl 2040.	New population to be accommodated is allocated at each county’s 2010 gross urban density	The remaining new population to be allocated after the redevelopment population is removed is allocated at a development density 20% greater than that used for Sprawl 2070.
Redevelopment Areas	No new population is accommodated within existing urban areas	No new population is accommodated within existing urban areas	No new population is accommodated within existing urban areas	In all counties, some of the 2070 population growth is accommodated through an increase in the densities of existing urban areas
Protected lands	2021 Florida Managed Areas are included as protected	In addition to 2021 Florida Managed Areas, 2021 Florida Forever project lands, and Florida Ecological Greenways Network Priorities 1–3 (the Florida Wildlife Corridor) are protected.	2021 Florida Managed Areas are included as protected	In addition to 2021 Florida Managed Areas, 2021 Florida Forever project lands, and Florida Ecological Greenways Network Priorities 1–3 (the Florida Wildlife Corridor) are protected.
Agricultural lands	No agricultural lands are excluded from population allocation	All irrigated agricultural lands (using the 2045 projection) on good soils (USDA/NRCS Capability Rating Excellent – Marginal) from FSAID, 2020 are excluded from population allocation	No agricultural lands are excluded from population allocation	All irrigated agricultural lands (using the 2045 projection) on good soils (USDA/NRCS Capability Rating Excellent – Marginal) from FSAID, 2020 are excluded from population allocation
Sea Level Rise	50% of population impacted by sea level rise is reallocated within the county affected. The other 50% are assumed to move out of the state. Areas between 0.25–0.75 m in elevation are assumed to be less likely to develop.	50% of population impacted by sea level rise is reallocated within the county affected. The other 50% are assumed to move out of the state. Areas between 0.25–0.75 m in elevation are not allowed to develop.	50% of population impacted by sea level rise is reallocated within the county affected. The other 50% are assumed to move out of the state. Areas between 0.9–1.4 m in elevation are assumed to be less likely to develop.	50% of population impacted by sea level rise is reallocated within the county affected. The other 50% are assumed to move out of the state. Areas between 0.9–1.4 m in elevation are not allowed to develop.

), based on the National Oceanic and Atmospheric Administration’s (NOAA) 2022 report [30] “Intermediate projection of SLR in Key West, Florida by 2040”. The 2070 scenarios used a 0.9 m SLR based on the 2017 NOAA Intermediate-High projection. The 2017 NOAA data [31] was the most recent available at the time the 2070 project was conducted (Table 1). The 2040 study followed later and was able to use the 2022 NOAA data.

Areas inundated by SLR were considered undevelopable (unsuitable for future development) and added to masks outlining places where future development would not be permitted. The model assumes that residents in these locations would need to relocate due to permanent or tidal flooding. A basic assumption was made that 50% of the population in areas projected to be inundated by a 0.9 m sea level rise was reallocated within the same county based on the assumption that existing residents would want to move to locations close to their existing employment, social networks, and residence. The remaining 50% was assumed to relocate outside of the state. The number of people expected to be impacted by sea level rise was identified by applying the population percentage impacted by a 0.9 m sea level rise in 2100 [32] to BEBR’s medium 2070 population projection for each county where population impacts from sea level rise would occur. The 2040 model followed the same approach but was modified for the shorter timeframe.

The likelihood of future land development was assumed to vary depending on location and land characteristics. For example, areas closer to major roads (for access) and waterbodies (for desirability) were assumed to be more likely to be developed, while land in wetlands was deemed less likely due to higher development costs. These assumptions were developed based on expert opinion, and where available supported by literature on land use development patterns and characteristics [33,34].

2.3.3. Suitable Development Areas

Key to the scenario process was determining the most suitable or likely locations of future development. Masks were used to delineate areas available for future development, with masks created for each of the four future scenarios (but not the baseline). Each mask contained areas deemed undevelopable, which were excluded from receiving future development through the allocation process. The masks excluded areas such as existing urban development, major roads right of ways, existing protected lands, mitigation banks, sea level rise, open water, various easements such as Natural Resources Conservation Service (NRCS) easements, and certain specific areas such as the Miccosukee Indian Reservation in Broward County. The only difference between the 2040 and 2070 masks was the level of sea level rise.

In the Conservation Scenarios, additional areas were added to the masks. As noted, these included Florida Forever project lands and FEGN priorities 1 to 3, consistent with methods used in the 2016 Florida 2070 study.

2.3.4. Suitability Analysis

A GIS-based suitability model was used to identify areas of potential future development, based on the assumptions described earlier, and referencing parameters and assumptions used in the 2016 Florida 2070 model. Suitability modeling is used in a variety of future land use models to measure the fitness of a piece of land for a particular purpose [24,35]. The GIS analysis in this study used ArcGIS Desktop 10 and ArcGIS Pro 3 from the Environmental Systems Research Institute (ESRI). The majority of this work was conducted using raster data in ESRI’s GRID format. The pixel size used for this analysis was 30 m × 30 m (approximately ¼ acres), a common resolution for statewide analysis that also matches the data used and created for previous iterations of the model [9,10].

Nine suitability layers were used (

Table 2. Suitability criterion and weights.

Urban Suitability Criterion	Rational for Use	Weight
Proximity to Large Urban Areas (over 50,000 people)	Major urban areas tend to accommodate more additional population than do smaller urban areas.	5%
Proximity to All Urban Areas (over 2500 people)	New urban development tends to occur in close proximity to existing urban development.	27%
Proximity to coastline	The coast has historically been an attractor for urban development. However future sea level rise is expected to make these areas less attractive in 2070.	9%
Preliminary Development Approvals	Areas within approved DRIs and DSAPs are highly likely to develop. The only DSAP that was used, however, was West Bay in Bay County, because the other existing DSAPs fell in the path and pattern of new urban development and their boundaries did not affect the pattern or timing of new urban development.	8%
Road density	New urban development tends to occur in areas of relatively higher road density.	12%
Presence/absence of wetlands	Urban development on lands without wetlands is often less costly than lands with wetlands.	16%
Proximity to open water	Access to the view of water has historically been an attractor for development.	2%
Proximity to major roads	Roads facilitate new urban development.	5%
Absence of USDA/NRCS Soils within FSAID 2045 Projected Irrigated Agricultural Lands	There is an economic incentive to convert poorer agricultural soils to urban development before good agricultural soils.	16%

) to determine where future growth is more likely to occur. The suitability layers included proximity to urban areas, road density, proximity to major roads, the influence of existing cities and towns, absence of wetlands, proximity to coastline, developments of regional impact (DRI), and proximity to open water. The majority of the suitability layers were chosen to maintain consistency with the 2016 Florida 2070 study to allow for meaningful comparisons between the projects. Some suitability layers were developed using similar methods to the prior study, while updated methods were employed for other layers [10]. The nine suitability layers were assigned values from 1.00 to 9.00, with 1.00 representing areas least suitable for future development and 9.00 being the most suitable.

2.3.5. Methods for Developing Individual Suitability Layers

The Proximity to Large Urban Areas layer was created using urban area data based on TIGER/Line Files from the United States Census Bureau [36]. The data delineates urban areas including residential, commercial, and nonresidential urban land uses and categorizes areas as urbanized (50,000 or more people) or as urban clusters (2500–50,000 people). Distance from the urbanized boundaries was used to determine suitability values based on findings from the Brookings Institute for selected cities, where areas closer to or within existing urbanized areas are deemed to be more suitable [37]. For cities not listed in the Brookings data, the average of all cities studied in the report for Florida was used.

The Proximity to Small to Medium Urban Areas layer was assessed similarly to large urban areas. A layer was created for urban clusters with populations between 2500 and 50,000. Areas within or nearest to urban areas received the highest suitability ratings, while those farthest from development received the lowest suitability value of 1. Suitability values were assigned based on the Brookings data as was done for large urban areas.

Being close to the coast is often considered desirable for development. However, with the potential for loss of land due to sea level rise, the Proximity to Coastline layer reflects the potential hazards of developing close to the shore. Areas expected to be inundated were removed by development masks and were not available for future population allocations. Areas between the projected sea level rise and an additional 0.5 m of sea level were given a value of 1 while all other areas received a suitability value of 9 for the Sprawl Scenario. This allowed development in higher risk areas but at a lower suitability therefore directing development to areas away from potentially less suitable locations. The Conservation Scenario removed the areas between the projected sea level rise and 0.5 m from consideration for future population allocations.

The Preliminary Development Approvals layer focused on DRIs [38]. DRIs are projects in Florida that impact citizens in more than one county due to the location, magnitude, or character of the project. Examples include airports, power plants, post-secondary campuses, and residential developments [39]. These projects require coordination with many reviewing agencies. Statutory changes in Florida law in 2018 eliminated developments from needing review in the DRI program [40]. To maintain consistency with the 2016 study, DRIs were included in this study. Areas within DRI’s were considered more likely to develop and were given a suitability value of 9 while areas outside of DRI’s were assigned a value of 1.

In recognition of the role roads play in urban development, a Road Density layer was used to identify development suitability, where areas with denser road networks were deemed to be more suitable. The ArcGIS line density tool was used on the TIGER (Topologically Integrated Geographic Encoding and Referencing)/Line data with higher weights given to larger roads.

The Proximity to Major Roads layer encouraged population allocation in areas of higher urban density. Areas within the mean distance from all commercial and residential properties from major roads from the TIGER/Line data were assigned a value of 9 and areas greater than the mean plus two standard deviations received a 1 [11].

The Presence/Absence of Wetlands layer reflects the increased cost of developing in wetland areas. Wetlands were given a value of 1 with all other areas receiving a value of 9. This allowed allocation of new population in wetland areas but encouraged development outside of wetlands.

The Proximity to Open Water layer categorized the distance from open water, with areas closest to water assigned a value of 9 and reducing in equal intervals to a value of 1 at half a mile distance. Various studies show that the economic impacts of open water on real estate can be identified within a range of 100 m to ¾ of a mile from water bodies [36,41].

Areas with high-quality soils and projected future irrigated areas were given a low suitability value based on expert opinions and consistent with the 2016 Florida 2070 study. The data is included in the Florida Statewide Agricultural Demand project (FSAID). The Florida Department of Agriculture and Consumer Services (FDACS) develops statewide estimates of agricultural water demand. The FSAID data also projects future irrigated areas and includes soils data using the NRCS Land Capability Classification [42].

2.3.6. Weighting of the Suitability Layers

The nine suitability layers were weighted to reflect their varying importance to overall suitability for development and then combined to create the final suitability layer. The weights reflect the assumption that future urban development is more likely to occur close to existing urban development and roads. The weights and selection of suitability layers in this study were informed by expert opinion and match those used in the 2006 Florida 2060 and 2016 Florida 2070 models. Roughly half of the weighted values of the suitability analysis were based on roads and proximity to existing urban development. Hence, the results of the suitability analysis show much of the projected development occurring in close proximity to existing urban development.

2.3.7. Development and Population Allocation

The population allocation process determined where new urban growth may occur and included two factors: where growth is likely to occur and how much growth will occur. The suitability analysis was used to determine where future development may occur, with land under a value of 3 out of 9 on the suitability scale excluded from future development to maintain consistency with the previous 2016 study. Population projections and the density of development were then used to model the amount of land to be developed.

This study used a gross development density (GDD) metric to determine the density of projected development based on existing county development densities. The GDD was calculated by dividing each county’s current population by the number of developed acres in each county. Developed lands included residential, commercial, industrial, institutional, and governmental parcels. The amount of land needed to accommodate future residents was calculated by dividing the projected population by the GDD. Allocations assumed future development included the same categories of land use as those used for calculating the GDD. In the 2070 Sprawl model, the 2010 GDD for each county, along with the 2015 population data, was utilized to facilitate a more direct comparison with the methodology and findings of the 2016 study. In contrast, the 2040 Sprawl model aims to assist communities in understanding potential growth over the short term; therefore, it employed 2019 population estimates and 2019 parcel data to calculate the GDD. Additionally, in the Conservation Scenarios, a higher (GDD) was used for each county to represent denser development patterns than the Sprawl scenarios (30% higher in 2040 and 20% higher in 2070).

The Conservation Scenario also assumed a percentage of the new population to be allocated would be in redevelopment. Redevelopment was based on the ratio of the projected population for each county divided by the land available for development. Counties were grouped into six categories reflecting development densities which included 50% redevelopment (6 counties), 30% redevelopment, 15% redevelopment, 3% redevelopment, and 0% redevelopment [10]. Counties with less available land for development were given a higher redevelopment rate than counties with abundant land for development (

Table A3. Redevelopment percentages by county. These were used in the Conservation 2070 scenario only.

County	Percent of New Population to be Accommodated Through Redevelopment	County	Percent of New Population to be Accommodated Through Redevelopment
ALACHUA	13	LEE	40
BAKER	10	LEON	13
BAY	10	LEVY	10
BRADFORD	10	LIBERTY	10
BREVARD	25	MADISON	10
BROWARD	60	MANATEE	13
CALHOUN	10	MARION	13
CHARLOTTE	40	MARTIN	13
CITRUS	40	MIAMI-DADE	40
CLAY	13	MONROE	60
COLLIER	13	NASSAU	10
COLUMBIA	10	OKALOOSA	13
DESOTO	10	OKEECHOBEE	10
DIXIE	10	ORANGE	40
DUVAL	40	OSCEOLA	10
ESCAMBIA	13	PALM BEACH	13
FLAGLER	10	PASCO	25
FRANKLIN	10	PINELLAS	60
GADSDEN	10	POLK	13
GILCHRIST	10	PUTNAM	13
GLADES	10	ST JOHNS	10
GULF	10	ST LUCIE	10
HAMILTON	10	SANTA ROSA	10
HARDEE	10	SARASOTA	25
HENDRY	10	SEMINOLE	40
HERNANDO	25	SUMTER	10
HIGHLANDS	13	SUWANNEE	10
HILLSBOROUGH	40	TAYLOR	10
HOLMES	10	UNION	10
INDIAN RIVER	13	VOLUSIA	25
JACKSON	10	WAKULLA	10
JEFFERSON	10	WALTON	10
LAFAYETTE	10	WASHINGTON	10
LAKE	25

). Redevelopment was not included in the 2040 Conservation Scenarios, as the impact of redevelopment on future development patterns by 2040 was assumed to be lower.

2.3.8. Population Allocation Process

A process of “allocating” future population (Figure 4) was used to identify areas of potential future development based on the density assumptions described above. An infill/greenfield mask (

Table 3. Infill/Greenfield DORUC categories.

CAMPS	GRAZING LAND SOIL CLASS 4
IMPROVED AGRICULTURE	GRAZING LAND SOIL CLASS 5
CROPLAND SOIL CLASS 1	GRAZING LAND SOIL CLASS 6
CROPLAND SOIL CLASS 2	ORCHARD, GROVES, CITRUS
CROPLAND SOIL CLASS 3	POULTRY, BEES, TROPICAL FISH, RABBITS, ETC
TIMBERLAND	DAIRIES, FEED LOTS
TIMBERLAND	ORNAMENTALS, MISC. AGRICULTURE
TIMBERLAND	MINING, PETROLEUM, AND GAS LANDS
TIMBERLAND	ACREAGE NOT ZONED AGRICULTURE
TIMBERLAND	VACANT RESIDENTIAL
TIMBERLAND	VACANT COMMERCIAL
GRAZING LAND SOIL CLASS 1	VACANT INDUSTRIAL
GRAZING LAND SOIL CLASS 2	VACANT INSTITUTIONAL
GRAZING LAND SOIL CLASS 3

) was created to determine areas available for development that were not already developed or in open water, existing protected lands, major road rights of way, mitigation banks, or other easements. The first step in this process was to allocate population to DRIs. If the area within a county’s DRI had sufficient land with a suitability value of 3 or higher to accommodate future population, the allocation process was complete. However, the majority of counties did not have sufficient acres within DRIs to accommodate all population. Some counties have no DRIs, while others with DRI’s often had too little land area to accommodate future population needs. Following the DRI allocations, the remaining acres needed to satisfy the population growth requirements were identified.

The allocation process used a criteria evaluation matrix (CEM) (

Table 4. Example of a Criteria Evaluation Matrix (CEM) used to select areas that have the highest suitability values and acres needed to support the projected population. The Value column is assigned automatically with each value representing a unique combination of the suitability layer, DRI, parcel ID, county and land use. The Count is the number of cells that share the same combination of these values. The LUCIS_ID is assigned during the conversion of parcel data from a vector format to a raster format to allow data to be analyzed based on the parcel designation. The Suitability column is the value derived from the suitability layer. The DRI column used 0 and 1 to identify if the area was in a DRI (value of 1). The Acres column shows the land available for population allocations. ALLYR was used to record cells allocated for new population. The DORUC column provided the Department of Revenue Land Use Code for the cells which identifies the land use based on the parcel data.

Value	Count	LUCIS_ID	SUITABILITY	DRI	COUNTY	ACRES	ALLYR	DORUC
776092	1	27661	6.34	0	ALACHUA	0.22	0	53
776244	2	8001265	6.34	0	PUTNAM	0.44	2070	0
776457	10	27332	6.34	0	ALACHUA	2.22	0	61
776739	9	2078908	6.34	0	DIXIE	2.00	0	63
777217	5	8004729	6.34	0	PUTNAM	1.11	2070	99
778031	5	7929387	6.34	0	PUTNAM	1.11	2070	51
778117	2	2365932	6.34	0	GILCHRIST	0.44	0	60
778523	2	7939736	6.34	0	PUTNAM	0.44	2070	99
778736	3	2079670	6.34	0	DIXIE	0.67	0	56
778767	6	2075842	6.34	0	DIXIE	1.33	0	0
779113	10	2078874	6.34	0	DIXIE	2.22	0	56
779457	2	2076295	6.34	0	DIXIE	0.44	0	63
779470	12	2078875	6.34	0	DIXIE	2.67	0	56
779491	3	2370576	6.34	0	GILCHRIST	0.67	0	59
779592	4	7932144	6.34	0	PUTNAM	0.89	2070	99

) to determine where future development would occur in each scenario. A CEM is a raster containing information needed for allocations—such as suitability value, acres, county, allocation year, developments of regional impact (DRI), and other criteria. A CEM for a more detailed county level analysis might also include urban service areas, conflict values for agriculture, conservation, and urban uses, and suitability values for single family residential, multi-family residential, commercial, industrial, and other uses [10]. The CEM table allows queries to be made that determine where allocations are projected to occur based on the suitability value and number of acres needed to accommodate future population increases. For example, if a county was projected to grow by 100,000 residents in a county with a GDD of 10, then 10,000 acres would be needed to accommodate the increase. The CEM would then be used to find 10,000 acres with the highest suitability.

During the allocation process, if the population projections for a county could not be accommodated, the remaining population was allowed to ‘spill over’ into neighboring counties. The spillover was allocated using the GDD of each adjacent county that received the additional population. The additional population was redistributed among the adjacent counties in proportion to the sum of the existing counties’ population. For example, in one scenario, after all available land with a suitability of 3 or higher was allocated, Lee County still needed to accommodate 373,060 residents. The additional residents were allocated as spillover population to each adjacent counties: Charlotte, Glades, Hendy, and Collier (Figure 5). The percentage of the 373,060 Lee County residents to be allocated to each adjacent county was based on each adjacent county’s population divided by the total population of the spillover counties. Charlotte County had a population of 189,000 residents based on the data used in the study. The total population of the adjacent counties was approximately 629,000 residents. Charlotte County had 30% of the adjacent counties’ population (189,000 ÷ 629,000) and therefore received 30% of the spillover from Lee County.

3. Results

3.1. Scenario Population Allocations

As previously described, this study analyzed the impacts of future development patterns and densities on land use for four future scenarios. The study quantified the total land needed to accommodate future population growth and relocation as a result of sea level rise, along with potential impacts to conservation and agricultural lands (

Table 5. Statewide acreage comparison of 2040 development scenarios. Other land includes timberlands, mining lands, and other miscellaneous land uses not classified as agriculture, developed, protected, protected agriculture, or open water.

	Baseline	% of Total Acreage	Sprawl 2040	% of Total Acreage	Conservation 2040	% of Total Acreage
Developed	5,428,000	14.94%	6,374,000	17.54%	6,102,000	16.79%
Protected Natural Land	9,850,000	27.11%	8,997,000	24.76%	14,076,000	38.74%
Protected Agriculture	856,000	2.36%	854,000	2.35%	3,236,000	8.91%
Agriculture	6,418,000	17.66%	6,022,000	16.57%	3,751,000	10.32%
Other*	11,778,000	32.41%	11,028,000	30.35%	6,110,000	16.81%
2019 Open Water	2,006,000	5.52%	2,006,000	5.52%	2,006,000	5.52%
Sea Level Inundation: Protected Lands	0	0.00%	854,000	2.35%	928,000	2.55%
Sea Level Inundation: All Other Land Uses	0	0.00%	201,000	0.55%	127,000	0.35%
Total Acreage	36,337,000	100.00%	36,337,000	100.00%	36,337,000	100.00%
Total Land Acreage	34,330,000	94.48%	33,275,000	91.57%	33,275,000	91.57%
Total Sea Level Inundation	0	0.00%	1,055,000	2.90%	1,055,000	2.90%
Total Open Water including SLR	2,006,000	5.52%	3,062,000	8.43%	3,062,000	8.43%

and

Table 6. Statewide acreage comparison of 2070 development scenarios. Other land includes timberlands, mining lands, and other miscellaneous land uses not classified as agriculture, developed, protected, protected agriculture, or open water.

	2019	% of Total Acreage	Sprawl 2070	% of Total Acreage	Conservation 2070	% of Total Acreage
Developed	5,428,000	14.94%	8,881,000	24.44%	7,612,000	20.95%
Protected Natural Land	9,850,000	27.11%	8,404,000	23.13%	13,405,000	36.89%
Protected Agriculture	856,000	2.36%	852,000	2.34%	3,115,000	8.57%
Agriculture	6,418,000	17.66%	4,645,000	12.78%	2,931,000	8.07%
Other*	11,779,000	32.42%	9,867,000	27.15%	5,586,000	15.37%
2019 Open Water	2,006,000	5.52%	2,006,000	5.52%	2,006,000	5.52%
Sea Level Inundation: Protected Lands	0	0.00%	1,296,000	3.57%	1,432,000	3.94%
Sea Level Inundation: All Other Land Uses	0	0.00%	386,000	1.06%	250,000	0.69%
Total Acreage	36,337,000	100.00%	36,337,000	100.00%	36,337,000	100.00%
Total Land Acreage	34,330,000	94.48%	32,648,000	89.85%	32,648,000	89.85%
Total Sea Level Inundation	0	0.00%	1,682,000	4.63%	1,682,000	4.63%
Total Open Water including SLR	2,006,000	5.52%	3,772,000	10.38%	3,772,000	10.38%

).

The 2040 Sprawl and Conservation Scenarios (Figure 6) were both based on a projected 4.9 million residents moving to Florida in less than two decades. Under the Sprawl 2040 Scenario, nearly 1 million acres of additional land are projected to be lost to either development or sea level rise, including over 200,000 acres of the Florida Wildlife Corridor, 400,000 acres of agricultural land, and 750,000 acres of other land including silvicultural land uses. Under the Conservation 2040 Scenario, approximately 700,000 acres of land are projected to be developed. In the Conservation Scenario, lands within the FWC and Florida Forever projects were included in the development mask and were not available for future development allocations.

In contrast, under the 2040 Conservation Scenario, the acreage of protected lands would increase by over 5 million acres. While there are still land losses due to sea level rise, the protection of Florida Forever projects and Florida Wildlife Corridor lands from development significantly increased the amount of protected land from around 27% of the total land in Florida in the baseline to 47%.

By 2070, Florida is projected to grow by over 12 million new residents, a 57% increase over the 2019 population. Results of the Sprawl 2070 Scenario (Figure 7) show that approximately 3.5 million acres could be developed by 2070 with a loss of 1.2 million acres within the Florida Wildlife Corridor if current development patterns and densities continue. The priority natural lands protected in the Conservation 2070 scenario are similar to those in the Conservation 2040 scenario though with a greater loss of these lands to sea level rise.

3.2. Sea Level Rise

Based on the 0.25 m sea level rise scenario for 2040, 1 million acres of land would be lost with over 200,000 residents displaced (

Table 7. Land lost to sea level rise by 2040.

	Sprawl 2040 Acres Lost to SLR	% of Total Acreage	Conservation 2040 Acres Lost to SLR	% of Total Acreage
Developed Land	31,000	0.08%	31,000	0.08%
Protected Natural Lands	852,000	2.34%	915,000	2.52%
Protected Agricultural Lands	2000	0.01%	3000	0.01%
Unprotected Agriculture	3000	0.01%	3000	0.01%
All Other Land Uses	167,000	0.53%	103,000	0.35%
Total Sea Level Inundation	1,055,000	2.97%	1,055,000	2.97%

). While many of these areas are along the coast, the rise in sea level is expected to also impact waterbodies in inland counties connected to the coast.

The projected sea level rise of 0.9 m by 2070 would result in inundation of approximately 1.7 million acres of land, almost 5% of the total acreage of Florida. Approximately 900,000 people would need to relocate in the 2070 scenarios (

Table 8. Land lost to sea level rise by 2070.

	Sprawl 2070 Acres Lost to SLR	% of Total Acreage	Conservation 2070 Acres Lost to SLR	% of Total Acreage
Developed Land	94,000	0.26%	94,000	0.26%
Protected Natural Lands	1,291,000	3.55%	1,403,000	3.86%
Protected Agricultural Lands	4000	0.01%	4000	0.01%
Unprotected Agriculture	10,000	0.03%	10,000	0.03%
All Other Land Uses	283,000	0.78%	171,000	0.47%
Total Sea Level Inundation	1,682,000	4.63%	1,682,000	4.63%

). Similarly to the scenarios for 2040, inundation of interior areas of the state will be higher especially along the St. Johns River, though much more pronounced. Protected natural lands make up 77% of all land lost to sea level rise in the Sprawl Scenario and 83% in the Conservation Scenario due to the increased amount of protected land.

3.3. Natural and Agricultural Lands

Many of Florida’s currently protected lands are located at low elevations and are some of the first areas projected to be inundated by 2040. The 2040 Sprawl Scenario projected a loss of 850,000 acres of protected land while the Conservation 2040 would see a loss of 915,000 acres due to sea level rise. Again, the loss of protected land due to sea level rise was higher for the Conservation Scenario than the Sprawl Scenario due to the additional areas protected from development in the Conservation Scenario. There are many areas along the coast in existing conservation or else that are current priorities for conservation that are at low elevations and may be inundated by sea level rise.

Agricultural land is projected to drop from 20% of total state land in 2019 to 15% in 2070 for the Sprawl Scenario, a loss of 1.7 million acres to future development and sea level rise. The Conservation Scenario projects 16% of total land in Florida to be agricultural with a decrease of around 1.25 million acres. Protected agricultural land, however, would rise from 850,000 acres to 3.1 million acres in 2070 due to the increase in protection of Florida Forever projects and the Florida Wildlife Corridor.

The Sprawl 2070 Scenario projects a loss of 1.5 million acres of protected natural lands, primarily due to sea level rise, while the 2070 Conservation Scenario shows a gain of almost 3.5 million acres due to assumptions about more proactive conservation.

4. Discussion

4.1. The Need for Scenario-Based Models

The Sea Level 2040/2070 project provides a preview of potential outcomes from future development and land use change in Florida. Florida stands to lose 190 acres of land to development every day under the Sprawl 2070 scenario between 2022 and 2070. This includes almost 1.5 million acres of land that are conservation priorities, and 1.8 million acres of agricultural land important for fiber, food, and a variety of other ecosystem services. In addition, using the NOAA 2017 projections for the year 2070, over 33,000 acres of land could be lost to sea level rise every year over the same period. These statistics underscore the importance of making future development decisions that minimize the negative impacts from land use change on agricultural and natural landscapes. These drivers of change are also by no means unique to Florida, with many locations within the U.S. and globally at risk from either sea level rise, land use conversion, or both [43,44,45].

As previously noted, scenario-based models have a significant role in helping communicate the implications from land use decisions made today to the public and decision makers [46,47,48,49]. However, not all local governments or communities have the capacity to conduct these studies [50,51]. There is a need for relatively simple forward casting models that can be developed or used by staff with moderate GIS experience, and with minimal amounts of funding. The model described herein relies on a basic set of parameters and assumptions but provides a template for other regions or states with a need to examine and communicate future development trends. The steps and processes used in this model do not require extensive GIS expertise and can be replicated, with consideration of data sources and location-specific climate, demographic, or other parameters.

4.2. Model Assumptions and Inputs

Model assumptions have a significant impact on results, in ways that should be evaluated to ensure they accurately reflect the model’s intent [24]. As an example, the Conservation Scenarios avoided allocation of development to high priority conservation lands. These lands are sometimes located in close proximity to existing development thereby causing “leapfrog” development in order to accommodate new population, inconsistent with the goal of reducing sprawl. This shouldn’t be construed as an argument against conservation—but as reinforcing the need for careful urban planning, infill, and redevelopment as necessary compliments to land conservation. The details of how assumptions are defined (e.g., what is a “conservation priority”) as well as the implications of incorporating that assumption into the model are highly important and must be thought about critically in terms of defensibility, communicability, and adherence to the model and project goals.

It is also important to note that there are differences in model assumptions between the 2040 and 2070 models. The Sea Level 2070 model was intended to be an update with newer data and the inclusion of sea level rise that would be comparable to the previous study. The Sea Level 2040 study utilized a shorter timeframe that could be useful to communities making planning decisions in the near term that could benefit from the study. The 2040 models used different assumptions for redevelopment and gross development density. The six redevelopment categories used in the Sea Level 2070 model were considered too generalized to be accurate for a shorter timeframe but useful for long range models. Another change in assumptions between the models was the use of 30% new development density in the 2040 model compared to 20% in the 2070 model based on expert opinions.

Our goal of maintaining consistency between the 2016 study and the current study was intended for overall trends and development patterns but not specific results. The results were developed for regional analysis to reflect broader potential development trends. The baseline, for example, was developed from parcel data which was updated since the 2016 Florida 2070 study. Parcel descriptions can change from residential to non-residential or vice versa easily between datasets without any substantial changes in land use or land cover.

Although we believe that the Sea Level 2040/2070 model assumptions and results have been useful and defensible, there are ways that the model could be revised. There are potential changes to the methodology for allocating future population that may result in an improved model.

The methodology used for allocating spillover population may benefit from adjustments in gross development density. In receiving counties with lower current development densities than the source county, the future development resulted in higher amounts of land being needed to accommodate the spillover development. The Conservation 2070 scenario, which illustrates this issue more than the other scenarios, allocated 1.4 million acres to spillover development. Adjusting the GDD for spillover allocations based on the proximity to the originating county or by using a percentage of the originating county’s GDD could help reduce excessive allocations to less dense counties.

The GDD was calculated using parcel data from 2010 for the 2070 scenarios and 2019 for the 2040 scenarios. Inaccuracies from older parcel data may have an effect on the determination of GDD. Studies such as the American Farmland Trust’s Farms Under Threat study used United States Geological Survey (USGS) National Land Cover Database (NLCD). Other databases of land cover derived from satellite imagery such as the Florida Land Cover Classification System (FLUCCS) data may allow for more accurate calculations. The delineation of urban areas used for the baseline scenario could also utilize the same land cover data for consistency within the model. However, parcel data was chosen based on its use in the previous 2016 study and to maximize consistency of methods.

An additional consideration is the use of redevelopment in the 2070 scenarios. Because of the assumption that some portion of new residents would be accommodated through redevelopment in existing urban areas rather than requiring new greenfield development, this parameter has an impact on the overall spatial footprint and acreage of development in each scenario. The methodology for determining redevelopment percentages in the current 2040/2070 model is at minimum hard to communicate, but it is also important to consider the most accurate means of approximating future redevelopment trends. The LUCIS methodology on which the scenarios are partially based provides one alternative—identifying redevelopment based on historical trends for each county using changes in parcel data over time. This method of calculating redevelopment based on historical trends may be more transparent, and more accurately reflects the redevelopment potential of each county.

The scenarios in this study used a single infill/greenfield mask to allocate population. The models could use separate masks for infill, greenfield, and redevelopment to allocate new single and multi-family development separately. The use of separate suitability layers could be developed for single family and multi-family to reflect differing requirements. This could offer a more nuanced and potentially more accurate approach for allocating new residential land use.

There are additional assumptions which were simplified or not fully incorporated, which could be addressed in a more detailed model or study. The scenarios used in the study only model where sea level rise is projected to occur. Other factors such as saltwater intrusion or storm surge, as well as economic or market factors such as property values and insurance rates that may also limit or affect future development patterns were not modeled. Detailed planning or policy factors at the county or municipal level were also not included. This was in part due to the fact that Sea Level 2040/2070 is a statewide model, with a limited scope and budget, and intended to provide a picture of potential land use futures for the state based on a set of simple and easy to communicate set of assumptions. It is also assumed that current land use policies and plans (such as future land use maps) may change. The intent of the model therefore is not to show future development patterns based on current policies and plans—but to do so based on a set of defensible spatial and environmental assumptions about suitability and likelihood of development. That said, the results from this “policy neutral” model may be overlaid with other data, such as future land use data, to further highlight the areas that may be at greater risk of land use conversion due to the alignment of both development suitability and current policies and plans. This can include the development of “combined threat” models, that show areas potentially at risk of change due to a combination of factors.

Finally, the accuracy of the Sea Level 2040 suitability model could potentially be improved with additional suitability layers. The LUCIS methodology includes proximity to many urban uses such as service amenities, shopping, and existing residential. These models assume that being close to existing uses is attractive to future growth. Separately, we are aware of the fact that some layers used in this model are several years old (such as the 2010 parcel data used to identify areas of existing development). The decision to use these layers was made either explicitly to maintain consistency with the prior iteration of the 2060 model, or because in some cases there were not newer statewide versions of datasets appropriate for the model. However, it is worth reiterating that in all GIS models, the best available data should be used that is consistent with the modeling objectives and goals.

Future research, building on this study could include testing and development of additional scenarios that have higher density development, varying amounts of redevelopment, and adjustment of other factors to better understand the impact these have on the model outputs. Adjustments to the suitability weights and their impact on the allocation results are another aspect that could further improve the model.

4.3. Detailed Planning Application

This model also does not directly attempt to show where people should locate—only where they may be likely to locate and the potential land use impacts from those decisions. The term “suitability” is used in this paper to refer to the areas assumed most likely for future development based on the modeling parameters used. However, this should not be interpreted as a statement of where future development should go. It should also be noted that model validation was not a part of this study but may be included in future iterations.

Therefore, it is important to state that the model described herein is not intended to be used as a basis for detailed land use planning decisions. Additional models at the municipal, county, or regional level could be used to build on these results, by incorporating local planning assumptions such as urban service and growth boundaries, zoning, and future land use, and complement/inform other existing documents such as comprehensive or future land use plans. The model used in this study did not use future land use or zoning since they are political factors that are potentially subject to change. The previous 2016 version of the models also did not include these factors as it is a pure suitability-based scenario model.

The 2040 and 2070 models were developed separately and do not build on each other. The Sea Level 2070 project was developed prior to the Sea Level 2040 study. The Sea Level 2040 study was therefore able to use newer data such as the updated NOAA sea level rise projections and included revisions to the methodology and assumptions. Areas shown as developed in 2040 may not be developed in the 2070 maps due to the differences in the models.

4.4. Sprawl and Conservation Scenario Comparisons

The Sprawl and Conservation Scenarios provide two potential outcomes of future development in Florida that result from differences in model parameters and assumptions. The removal of more land from population allocations in the Conservation Scenarios shifted development to other locations. The effects of decreased land available in some counties resulted in only modest changes with the Sprawl Scenario. However, in areas with limited land availability this had a cascading effect as counties adjacent to them were modeled to absorb these spillover allocations. The areas shown in blue in Figure A5 and Figure A6 illustrate these effects in the scenarios. Due to less projected development for 2040, this issue is not prevalent for the shorter time frame. The Conservation Scenarios also removed 0.5 m buffer above the projected sea level rise from potential development which further restricted where the models would project future population allocations.

4.5. Other Model Applications

The statewide model methodology has also been used for county-level analysis in several communities in Florida, considering local factors relevant to residents. The Pensacola and Perdido Bay Estuary Program study [52] looked at the impacts of development on water quality and quantity. The Future Development model and Event Mean Concentration (EMC) models for Escambia and Santa Rosa counties were combined to assess impacts from development patterns on water quality and quantity and water resource protection priorities.

This methodology was also used for Martin 2040: The Western Lands Study [53]. The conservation areas to be avoided for development went beyond the 2070 study. The study included the Comprehensive Everglades Restoration Plan (CERP) priorities, Preserve Area Management Plan priorities, wetlands, and existing agricultural lands in its conservation areas to protect. The model went through multiple iterations exploring the interactions of the urban service district, prime farmland protection, density, and the potential effects on conservation and future development.

5. Conclusions

The combination of existing development requirements, future development pressures and sea level rise present significant challenges for land use change and planning in many regions. Future scenario models are valuable tools that can be used to examine the impacts of changes in policy, land consumption patterns, economic factors, population dynamics, and other variables on future land use and resources. This information can be used to enhance public awareness of the potential impacts of future development and inform land use planning policies and decisions.

This paper describes a model being used in Florida to provide easily accessible information to the public and policymakers about potential future scenarios while supporting current land use and conservation planning needs. This replicable model can be applied in other locations to achieve similar objectives and can be adapted at smaller scales to incorporate locally important data and information on land use priorities and policy.

The results of the Sprawl 2040/2070 scenarios highlight the potential impacts of “business as usual” growth patterns and densities on other land uses, including natural and agricultural lands. The Conservation 2040/2070 scenarios indicate that accommodating future growth and protecting larger areas of natural resources is still possible (at least in Florida) and can be achieved through higher development densities, redevelopment, and strategic land protection. The results make it clear that the decisions we make today will have significant impacts on the world of tomorrow, with implications for urban design (walkability, transportation, and redevelopment), consumption (land and resources), human populations (resilience and climate migration and equity), food security (agricultural land protection), biodiversity, and the natural environment. The time to begin planning for a more sustainable future is now.