Building a Green Infrastructure Plan for Worcester County, Maryland A three part suitability analysis
Dana Coelho Submitted to The University of Maryland Graduate Program in Sustainable Development and Conservation Biology In partial fulfillment of the Masters of Science Degree May 2007
green infrastructure
Table of Contents Executive Summary .................................................................4 Green Infrastructure................................................................8 Principles of good design ........................................................................................... 9 In context ............................................................................................................ 10 Green infrastructure and suitability modeling ................................................................. 11
Reviewing the plans .............................................................. 12 Worcester County Plans............................................................................................ 12 Worcester County Comprehensive Plan ...................................................................... 12 Worcester County Land Preservation and Recreation Plan ............................................... 12 Berlin Comprehensive Plan..................................................................................... 13 Ocean City Comprehensive Plan............................................................................... 13 Ocean Pines Comprehensive Plan ............................................................................. 15 Pocomoke City Comprehensive Plan.......................................................................... 15 Snow Hill Comprehensive Plan ................................................................................ 16 Isle of Wight Watershed Restoration Action Strategy (WRAS) ............................................ 17 Coastal Bays Comprehensive Conservation Management Plan (CCMP) .................................. 17 Wicomico County Land Preservation and Recreation Plan ................................................... 18 Somerset County Land Preservation and Recreation Plan.................................................... 18 Sussex County Comprehensive Plan.............................................................................. 19 Accomack County Comprehensive Plan ......................................................................... 20
Creating a suitability model using GIS ......................................... 21 Establish project goals ............................................................................................. 21 Define analysis criteria ............................................................................................ 22 Identify data needs and collect data ............................................................................ 23 Identify GIS operations to be performed........................................................................ 23 Create a model...................................................................................................... 25 Run the model....................................................................................................... 26 Interpret results .................................................................................................... 29 Refine the model as needed ...................................................................................... 29
Next steps: from a model to a plan ............................................ 32 References ......................................................................... 33
2
Appendix 1: Model criteria and workflow diagrams ......................... 36 Conservation Suitability Model Criteria ........................................................................367 Agricultural Preservation Suitability Model Criteria..........................................................368 Recreation Suitability Model Criteria ...........................................................................369 Conservation Suitability Model Diagram ........................................................................ 40 Agricultural Preservation Suitability Model Diagram .......................................................... 41 Recreation Suitability Model Diagram ........................................................................... 42
Appendix 2: Summary of model runs........................................... 43 General Discussion.................................................................................................. 44 Conservation Submodel ............................................................................................ 48 Agricultural Preservation Submodel ............................................................................. 51 Recreation Submodel............................................................................................... 55
Appendix 3: Editing the model.................................................. 57 Appendix 4: Case studies ........................................................ 58 Cuyahoga County Greenspace Plan .............................................................................. 59 Florida Statewide Greenways Planning Project................................................................ 65 Greenprint for King County, Washington........................................................................ 93 City of Kinston—Lenoir County, North Carolina ...............................................................119
3
Executive Summary Green infrastructure planning in Worcester County Maryland’s Eastern Shore, and Worcester County in particular, is an ecologically and culturally rich place as well as an area committed to strategic conservation and smart growth. Worcester County is home to some of the nation’s earliest coastal settlements as well as some of her most important and sensitive habitats. Worcester County is also the second fastest growing county in Maryland. In 2000, the Maryland Department of Natural Resources (DNR) identified over 60 percent of the county as belonging to the state’s green infrastructure network. Green infrastructure planning incorporates conservation goals for the maintenance of functioning ecosystems as well as healthy social systems and seeks to create a connected network of forests, wetlands, working lands and other critical habitats and culturally relevant spaces. As stated in the literature: “Green infrastructure planning represents a strategic approach to conservation that combines the efforts of previous conservation planning methodologies and practices in the United States into a systematic framework that can encompass larger landscapes and broader planning goals” (McDonald 2005).
Methods and results This report documents the application of green infrastructure planning concepts at the local scale to Worcester County, Maryland, through the design and execution of three suitability models identifying priority conservation, agricultural preservation and recreation opportunities within the northern half of the County. Together, the results from these models form the foundation of a green infrastructure plan. The methodology outlined below may be refined and repeated at smaller and larger scales to meet the conservation planning needs of the County and larger region.
Establish project goals It is the ultimate goal of this model to identify areas that support the overall goals of a larger green infrastructure plan. By including goals and objectives voiced in county and municipal plans (in and surrounding Worcester County) this project aims to agree with the citizen input and professional opinions that shaped those plans. The model seeks to identify priority areas in service of the following goals: • • • • •
Conserve biological diversity and the productive capacity of ecosystems through the creation of an interconnected network of forests, wetlands and working lands. Create and sustain long-term socioeconomic benefits through the preservation of valuable natural and historic resources. Provide equitable and easy access to outdoor recreation. Complement county, municipal and state smart growth and conservation plans and goals. Improve quality of life for county residents and visitors.
4
Define analysis criteria Analysis criteria were based on a review of Worcester and neighboring county plans as well as consultation with staff during model development. Since the reasoning behind protecting ecologically significant areas, working farms, and places appropriate for recreation differ, this model operates as three distinct submodels that can later be combined or overlaid. The model criteria include policy variables (land use and zoning), land cover characteristics (soils, forests, etc.), protected areas (MD green infrastructure network, Coastal Bays critical areas, etc.), sensitive habitats, and water resources (floodplain, streams, groundwater recharge, etc.), as well as more socially oriented variables such as underserved communities, tourist destinations and access/connectivity that were important to the recreation submodel. The study area is defined by subwatershed boundaries and encompasses the county’s five northern Coastal Bays subwatersheds.
Identify data needs and collect data The criteria listed above demanded a great deal of high resolution spatial data. These data were obtained from Worcester County and the Maryland Department of Natural Resources. All data included are the most up to date files available to the county, but efforts should be made when updating the model to search for the most recent data that may be newly available.
Identify GIS operations to be performed The principal actions taken in the model are conversions of vector (point, line, polygon) data to raster (grid) data. Layers were combined in a weighted overlay in order to compute the final suitability surface for each submodel, conservation, agriculture and recreation.
Create a model I used the model builder tool within the ArcGIS Spatial Analyst extension to create the model in order to best document the data and processes used and results obtained. Model builder works by pulling tools into a workflow diagram, which can be used to run the model in its entirety or to run discrete segments. Each submodel followed the same basic steps outlined below: 1. Convert. A majority of the data layers were in vector format and first needed to be converted into raster format for the cell-based analysis. 2. Analyze. Calculations were made regarding the type and density of features or proximity to features. 3. Reclassify. Information from each analysis were reclassified onto an ordinal scale to be used in the final weighted overlay. Future users of the model may change the criteria to meet different purposes. 4. Overlay. All reclassified raster images were overlaid to see where the highest priority areas emerged. The data and processes are brought into the model via the workflow diagram, which links data (blue oval) to processes (yellow rectangle) to achieve an output (green oval). Outputs may be used as inputs to other processes, as was frequently the case in each of the green infrastructure submodels.
5
Model Builder workflow diagram Input feature (1)
Convert to raster
Output raster (1)
Reclassify (1)
Output raster (3) Weighted overlay
Input feature (2)
Analyze
Output raster (2)
Reclassify (2)
Output raster (5)
Output raster (4)
Run the model I ran the model numerous times to test its sensitivity to different assumptions and variable weights. The final “runs” chosen to best reflect the goals and conditions within the county yielded significant conservation opportunities throughout the northern half of the county. Areas defined as moderately suitable for conservation are shown in light green, areas of greater priority in dark green. The yellow areas along the Pocomoke River (the western boundary of the study area) scored the highest for conservation suitability. The areas in brown and red indicate agricultural preservation priorities, the red areas being those identified as highly suitable for preservation. Finally, the areas in magenta were identified as most highly suitable (values of 4 and 5) for recreational activities. The image also shows major roads, municipalities and villages (red outline) and growth areas (orange outline) for context.
Interpret results Under nearly all scenarios a good portion of the county – particularly surrounding major water resources – came back as relatively suitable for conservation. Agricultural areas were included as highly or moderately suitable in all three submodels, reflecting the importance of these areas to conservation, economic and cultural priorities. These results indicate that development should be limited where feasible, but always conducted in as responsible a manner as possible. This result is also reflective of the County’s progressive zoning laws that are already in place and have acted to preserve much of the agricultural land and forested areas within the County. The model identified priority areas both inside and outside of the green infrastructure network identified by the State of Maryland. Areas identified within this network help to confirm the relevance of the state network and Worcester County’s contribution to statelevel ecological sustainability and to set priorities for land acquisition within the Maryland Green Infrastructure. Areas identified outside of the network can be interpreted as opportunities to extend and/or modify the state network to reflect local conditions and goals more accurately.
6
Composite green infrastructure vision from county modeling
Model runs and the MD green infrastructure network
Highest recreation suitability
Highest agriculture preservation suitability
Highest recreation suitability
Highest agriculture preservation suitability
Highest conservation suitability
High agriculture preservation suitability
Highest conservation suitability
High agriculture preservation suitability
High conservation suitability
Unsuitable for conservation or agricultural preservation
High conservation suitability
Unsuitable for conservation or agricultural preservation
Moderate conservation suitability
Existing municipalities and developed areas
Moderate conservation suitability
Maryland Green Infrastructure
Low conservation suitability
Planned growth areas
Low conservation suitability
Refine the model as needed Many edits were made to the model as it approached its final state presented here. This process ought to continue in the future, one of the primary reasons it was created using Model Builder. In particular, there are opportunities to refine the spatial scale of the models in order to take a closer look at suitable and priority conservation areas within growth areas and inside municipalities. This approach could be used to identify restoration projects as well.
Looking to the future Creating and running this series of conservation, agricultural preservation and recreation suitability models, while resulting in a valuable product, does not constitute a green infrastructure plan. In order for this vision (or an amended vision developed by the County as new data become available and better understood) to become a reality it must be coupled with the County’s existing comprehensive planning efforts and linked with specific goals and objectives, as well as conservation tools and funding sources. Fortunately, many of these tools exist and are being used in Worcester County already. These include restrictive conservation and agricultural zoning, critical areas legislation and development restrictions, and agricultural/conservation easements. Thousands of acres are also under protection through US Department of Agriculture (USDA) programs, including the Conservation Reserve Enhancement Program (CREP), the Wetlands Reserve Program (WRP) and the Northern Bobwhite Quail Habitat Initiative. There are also opportunities to link conservation with a transfer of development rights program as well as forest and wetland mitigation banks currently under consideration and development within the County.
7
Green Infrastructure Rapid and poorly planned low density residential development (sprawl) is one of the most significant threats to natural resource lands, working farms, and communities. Its impact on wildlife, particularly bird species, is well documented (Merenlender 2005, McDonald 2005). Sprawl development fragments and disconnects habitat, making it inhospitable or less valuable to many species. However, poorly planned conservation projects can also be damaging, or at least fall short of meeting size and connectivity thresholds to support viable populations and ensure the functioning of ecological systems (Noss 1987 in McDonald 2005). Green infrastructure planning incorporates conservation goals for the maintenance of functioning ecosystems as well as healthy social systems and seeks to create a connected network of forests, wetlands, working lands and other critical habitats and culturally relevant spaces. “Green infrastructure planning represents a strategic approach to conservation that combines the efforts of previous conservation planning methodologies and practices in the United States into a systematic framework that can encompass larger landscapes and broader planning goals” (McDonald 2005). By definition, green infrastructure is “an interconnected network of protected land and water that supports native species, maintains natural ecological processes, sustains air and water resources and contributes to the health and quality of life for… communities and people” (Conservation Fund 2004). A system focusing only on ecological goals is an isolated conservation strategy. One focusing only on benefits to people is a “greenways” system. A greenways system does provide a link between people and nature, especially in urban areas, where green spaces and recreation areas are hard to come by (Searns 2003), but they do not necessarily place great weight on ecology. Green infrastructure, in contrast, incorporates and balances human and ecological goals, with priority given to the preservation of ecosystem function. As such, green infrastructure planning presents an opportunity to unite environmental and social goals into one comprehensive strategy. The process of planning for an interconnected network of ecologically and culturally significant lands and waterways is also a scientific and data rich approach that is more flexible and can better adapt to uncertain or unanticipated changes. It is an interdisciplinary approach to solving a complex problem with interactions between the natural and built environment, decision makers and citizens, land use patterns, species diversity, flood hazard, and other concerns (Meenar 2006). Planning for green infrastructure means valuing the existence of a connected network of healthy open spaces, capable of supporting the ecological health of an area as well as the social and economic wellbeing of residents. There are four main steps in green infrastructure planning: 1. Goal setting: a necessary first step – at any scale, county, regional, state or national – involving significant public and stakeholder input 2. Analysis: the setting of network design criteria based on ecological and land use planning theories at the landscape level 3. Synthesis: the creation of a planning framework that includes protection and risk ranking as well as the means to find network gaps in order to develop preservation and restoration priorities 4. Implementation: beyond a scientifically defensible plan, implementation requires public and political support, as well as specific tools and viable funding sources (McDonald 2005)
8
To answer the question of why plan for and protect green infrastructure there are numerous responses. Preserving forests helps to sustain healthy and diverse plant and animal communities, protect threatened and endangered species, provide carbon storage (which will help reduce the negative impacts of global climate change including increased storm frequency and intensity and sea level rise), and provide open space and recreational opportunities. The protection of wetlands and waterways supports diverse aquatic communities, creates recreation opportunities, sustains a natural flood protection network, and provides potable water and natural flood control services. Significant loss of life, property and income can be avoided by maintaining coastal wetlands and forested areas. However, wetland conservation is important not only in tidal areas, but in the headwaters as well, where projects often show the greatest potential for reduction of sediments, nutrient and toxic pollutants that degrade water quality (NOAA 2002). Conserving working farms helps to prevent sprawl (unplanned and usually damaging residential or commercial development), protect open space, retain natural lands, control public costs (farms are net tax contributors), support local economies and promote rural heritage.
Principles of good design An effective green infrastructure network is composed of hubs, large areas of intact natural habitat, and corridors, smaller linear elements that connect hubs to one another. Core areas (within hubs) should be strongly protected since they represent the most sensitive habitat elements, but buffers around the core and corridors should be open to human use. In fact, recreation should be encouraged in these areas as a way to connect people to their natural environment and to provide financial support for conservation. Focusing on the following ideals for green infrastructure elements will contribute to the overall success of the network.
Identifying hubs •
• • • • • •
Bigger is better. Larger intact areas can better support sensitive plant and animal communities. However, the size of hubs is also dependent on the scale of analysis. The State of Florida requires that hubs be at least 5000 acres, while the State of Maryland identified hubs of 250 acres or more in its statewide green infrastructure assessment. County or regional assessments may also consider smaller areas important at the local level. Greater diversity is desirable. More diverse ecological communities are more resilient – capable of withstanding or adapting to change – and healthier. Naturalness, naturally: Areas less impacted by human activities are higher priority conservation targets because ecological communities and their associated functions are more intact. Representation required: Characteristic but rare ecosystem types not currently protected should be priorities in order to maintain overall health and diversity at the landscape scale. Focus on rare and fragile systems: Sites that contain rare or particularly fragile plant and/or animal communities should be priorities. It is also important to “keep common species common” and maintain viable population sizes and densities. Demand historical data: Real, accurate data on species presence and population numbers is preferable to guesswork in defending a scientific approach to conservation. Value landscape position and potential: Connected habitats are the most valuable as are those which show significant potential for restoration.
9
•
Identify intrinsic appeal: Use culturally important plant and animal species or visually striking ecosystems to garner public support for conservation. (list adapted from Benedict and McMahon 2006)
Identifying corridors and linking hubs Connecting elements may be identified by hands and eyes in the field or through GIS analysis but should be based on a suitability analysis taking into account the following: • • •
Physical characteristics: habitat size, quality and sensitivity, Proximity to: other hubs and corridors (taking a “stepping stone” approach that makes use of habitat patches between hubs is desirable), urban areas and roads, and Policy characteristics: land management, protection and/or stewardship.
In context Maryland’s Eastern Shore, and Worcester County in particular, is an ecologically and culturally rich place as well as an area committed to strategic conservation and smart growth. Worcester County is home to some of the nation’s earliest coastal settlements as well as some of her most important and sensitive habitats. The MD Department of Natural Resources (DNR) identified over 60 percent (approximately 182,000 acres) of the county as belonging to the state’s green infrastructure network. Worcester County is also the second fastest growing county in Maryland (Calvert County, another coastal community along the western shore of the Chesapeake Bay, is the fastest). The population, now 46,543 has doubled since 1950, and continues to grow (US Census Bureau 2000, Worcester County Comprehensive Plan 2006). Most of this growth (53 percent by 2020) is anticipated in the northern half of the county, where roughly half of the current population resides (approximately 24,800 in Berlin, Ocean Pines, West Ocean City and Ocean City). The local economy depends upon agriculture, retail/services, construction/manufacturing and tourism, especially along the coast. Coordinated conservation and development planning in the region will help sustain a landscape and an economy that is historically, environmentally and culturally responsible. The state’s green infrastructure network is presently being used to identify priority properties for acquisition through the Maryland Agricultural Land Preservation Foundation (MALPF) program via a desktop parcel evaluation tool created by DNR (Weber 2003). A more detailed look into green infrastructure was also called for in the 2006 Worcester county Comprehensive Plan as well as the Land Preservation and Recreation Plan. Given this support and the opportunity to serve multiple goals, green infrastructure planning in Worcester County should be considered a flexible and inclusive concept, one that embraces conservation of valuable natural resources, preservation of economically and historically important working farms, and provision of sufficient and high quality outdoor recreation areas for county residents and visitors. It can and should work well with the existing Comprehensive Plan, Land Recreation and Preservation Plan and the Hazard Mitigation Plan. A green infrastructure plan for Worcester County will address several needs and threats the community faces, including a relative lack of protected areas and outdoor recreation in the northern part of the county. There are significant threats from global climate change predicted, including more intense and/or frequent storms and sea level rise that may cause flooding, damage to habitat and developed areas and saltwater intrusion into aquifers, that
10
strategic conservation may successfully address (Worcester County Hazard Mitigation Plan Draft). Maryland does not currently suffer from the worst of the hurricanes that hit the east coast (Nor’easters are more common), and storm control is not considered an economic threat (Ocean City Comprehensive Plan), but this is not reason to ignore conservation of coastal wetlands as a source of natural flood protection.
Green infrastructure and suitability modeling Suitability modeling involves an analysis of spatial data across a landscape in order to determine the appropriate location(s) of a project. This approach has been widely used in site planning and in determining the location of development projects, growth areas and parks. The use of computer models both to predict suburban growth and to set conservation priorities is also becoming increasingly popular. There are numerous examples of this type of advanced planning at multiple scales, from statewide initiatives down to individual communities (Merenlender 2005, Carr 2005, additional case studies included in Appendix 4). More widespread use of Geographic Information Systems (GIS) technology in local planning departments is making this kind of planning feasible on an even wider scale. Suitability analysis is highly appropriate for modeling green infrastructure. As McDonald (2005) points out: “The network design should be created by conducting a suitability analysis or using a similar method (based on determined network criteria) to calculate a range of resource values for the study area.” This is precisely how the Maryland Green Infrastructure Assessment was conducted in 2000, and how the northern Worcester County project to which this report pertains was carried out.
11
Reviewing the plans This section summarizes Worcester County, municipal and neighboring county comprehensive, preservation and recreation plans as well as other regional and subregional watershed and conservation plans prepared by state and county governments or independent conservation organizations (such as the Coastal Bays Program). All plans involved significant public participation and are taken to reflect shared community goals. This review served as the basis for creating the goals and criteria used in the green infrastructure suitability model. It is important to coordinate conservation with neighbors because habitats and other species frequently traverse political boundaries. Several common themes emerged, which support the work on this model and the creation of a green infrastructure plan for Worcester County, including: • • • • • •
Conserving natural resources and unique wildlife habitats, Preserving working agricultural lands, Protecting water quality, Preserving historic and cultural amenities (historic sites, town/rural character), Providing recreation opportunities, and Coordinating the goals listed above.
Worcester County Plans Worcester County Comprehensive Plan The 2006 Worcester County Comprehensive Plan is the primary document guiding land use decision making for the county. The plan makes extensive mention of natural resource conservation and agricultural preservation, as well as green infrastructure in particular. It points out a need to focus future forest conservation and reforestation efforts on the subwatersheds of the northern coastal bays in order to mitigate impacts from proposed development. With regard to green infrastructure specifically, the plan outlines four components of a system that provides spaces for both conservation and recreation: • • • •
Ecological corridors: purchased/protected primarily for natural resource protection or wildlife corridors, although they often contain trails or other amenities aimed at serving the human population. Recreational corridors: purchased/protected primarily for recreation, although these corridors contain at least a minimal natural buffer affording some ecological and/or habitat benefits. Connectors: walkways or on-road routes in built-up environments that provide key connections between or within recreational corridors; these have little, if any, ecological benefits. Water Trails: recognized water routes with access points, resting places, and destination spots along marine and inland waterways, designed for appropriate watercraft in accordance with the natural characteristics of each specific area.
Worcester County Land Preservation and Recreation Plan The 2006 Worcester County Land Preservation and Recreation Plan provides further details on the county’s priorities and plans for conservation and development of recreational
12
amenities. While there are sufficient recreation facilities in the county as a whole to satisfy state standards, there are no county-owned water trails (for canoeing and kayaking) or bike trails, and a relative lack of outdoor recreation opportunities in the northern part of the county. The Nassawango Creek Preserve (9100 acres owned by The Nature Conservancy) is the only privately owned canoeing/kayaking trail, and there are no privately owned bike trails. Assateague Island National Seashore (8000 acres owned and managed by the National Park Service) has canoeing/kayaking and bike trails. Northside Park (57 acres) in Ocean City has the only municipality-owned bike trail. Milburn Landing (370 acres owned by the state of Maryland) in Snow Hill has biking trails.
Berlin Comprehensive Plan The 2005 Berlin Comprehensive Plan includes significant elements aimed at 1) maintaining Berlin’s small town character, 2) encouraging tourism, and 3) preserving open space, all in support of the creation of a balanced community. Specific recommendations for the preservation of open space include 1) enforcing a 50’ buffer around all streams, 2) restricting development in the 100-year floodplain, 3) encouraging acquisition of wetlands and other areas, 4) directing growth into designated areas only, and 5) exploring opportunities for bike and walking trails. The preservation and restoration of natural wetlands as part of the town’s green infrastructure will support Berlin’s desire to maintain an adequate stormwater system. Stated goals relevant to green infrastructure • • • • • •
Recognize the assets, as well as the limitations, of Berlin's natural environment. Assure that appropriate land is maintained for open space, conservation and recreational uses. Capitalize on the tourist industry. Continue to enhance and improve present standards for property maintenance. Protect environmentally sensitive areas within and around the town. Create walkable communities.
Stated objectives relevant to green infrastructure • • • •
Maintain Berlin’s small town character by limiting denser development to designated areas and promoting mixed use development. Provide adequate stormwater drainage. Encourage land use policies supportive of a “balanced community.” Preserve historic and natural features to support a “balanced living and working environment.”
Ocean City Comprehensive Plan Due to the highly developed nature of Ocean City (only 5.5 percent of land remains vacant) the focus of the plan is on redevelopment, preservation and rehabilitation of existing structures and spaces. Significant development is anticipated west of Ocean City. In order to reduce the environmental impacts of development and redevelopment, the plan encourages 1) the avoidance of sensitive areas (such as wetlands and dune habitat) through clustered development, 2) the use of low-impact development (LID) strategies to minimize and improve the quality of stormwater runoff, and 3) the use of native landscaping plants to provide wildlife habitat. The city also seeks to foster a legible pattern of land use which accommodates variety in development type and scale appropriate to distinct neighborhoods 13
or districts within the city and which meets the residential, commercial and cultural needs of the community. The city contains 15 parks, occupying about 80 acres. With the exception of Northside Park, these are mostly small parks supporting active recreation (tennis, fitness centers, skate parks). The plan recommends acquiring additional parkland and the preservation of existing parks. Stated objectives relevant to green infrastructure • • • •
Encourage infill and redevelopment of existing vacant or underutilized sites. Encourage new development and re-development to minimize the impacts of strip commercial development by encouraging clustering of commercial uses and activity at optimal locations. Minimize the impacts of all development and re-development to insure pollution does not adversely affect the Coastal Bays and Ocean Dunes and beaches. Support the continuing improvement to the Boardwalk and the development of an inlet and bayside boardwalk to increase opportunities for water vistas.
Recreation and open space objectives • • • • •
Provide sufficient park, boat launch, and other sport facilities to meet the needs of year-round residents and vacationers. Provide a variety of recreational outlets to meet the needs of all age groups. Allocate sufficient resources to plan for and implement necessary emergency management measures. Support city “greening” by creating pockets of open space, via commercial development standards (e.g., setbacks and landscaping). Encourage land acquisition and preservation of existing parks.
Sensitive areas/conservation • •
Stated goal: To protect the quality of the air, water and land from the adverse effects of development and growth and, where feasible, to enhance the quality of the natural environment and sensitive areas. Stated objectives relevant to green infrastructure: o Continue to inventory and evaluate the city’s natural resource base and establish policies to protect and preserve resources. o Maintain and enhance the quality of the Coastal Bays and the ocean. Continue to actively participate in and implement actions in the Maryland Coastal Bays Management Plan. Provide resources to implement actions. o Utilize development standards for the location and construction of structures to minimize the impacts of flooding and to mitigate major flood hazards. o Protect and preserve wetlands as valuable spawning areas and to maintain the benefits they provide for water quality, shoreline stabilization, and wildlife habitat. o Use best management practices, low impact development techniques, flexible development regulations and innovative site design and mitigation measures to protect and improve environmental quality o Require all forms of development and re-development to avoid sensitive areas whenever possible. o Use flexible development standards to protect sensitive areas when they can be demonstrated to better protect sensitive environmental resources than would result from applying standard restrictions/regulations.
14
Ocean Pines Comprehensive Plan Completed in 1998 and updated in 2002 (there is also a comprehensive plan update scheduled and contracted for 2007), the Ocean Pines Comprehensive Plan provides a strong vision for the community. The plan provides the means to integrate various goals and objectives that should work together to protect natural features, enhance park facilities, provide open space, maintain and enhance community facilities, improve connectivity between residents and facilities, and protect and enhance community character. It includes detailed sections on land use guidelines as well as recreation and open space, the primary goals and objectives of which are summarized below. Land use guidelines • •
Goal: Maintain a planned land use pattern of compatible utilization of land resources guiding future growth into efficient and serviceable form while protecting the character of the community. Objectives and policy recommendations: o Prevent development on sensitive soils, sensitive environmental resources. o Protect residential character. o Achieve balance between residential, commercial and community uses. o Encourage the county to direct the form of commercial development near Ocean Pines to promote a campus like form as opposed to a linear or corridor form along adjacent highways. o Provide facilities (stormwater management, recreation, open space, etc.) for future residents.
Recreation and open space • •
Goal: Provide a diversity of accessible open spaces, parks, playgrounds, and recreation areas in community-supported areas. Objectives and policy recommendations: o Coordinate the location of open space and recreation areas with other land uses for maximum benefit and minimum negative impact. o Improve connections between and access to pedestrian and biking trails by linking parks. Create additional trails and aim to separate vehicle traffic from trails. o Acquire open space to satisfy recreation needs and manage growth to protect community character and qualities. o Work with the County school system to insure general use of nearby school facilities.
Pocomoke City Comprehensive Plan The Comprehensive Plan for Pocomoke City has not been updated since 1980, but does include general community and open space goals relevant to the development of the county’s green infrastructure plan. Community goals • •
Encourage a variety of cultural and recreation activities for the enjoyment and betterment of all residents of the Pocomoke City area. Preserve the significant natural features of the areas through the establishment of parks and open spaces. 15
•
Preserve and protect the significant historic structures within the town as a visual expression of Pocomoke city’s heritage.
Open space and recreation • •
Encourage the development of sufficient new areas for community parks and recreation areas in areas where they will be needed the most or in areas where a natural resource area needs protection. Preserve the natural stream valleys within the planning area in an effort to protect the natural drainage system, reduce the potential danger and damage from periodic flooding and to preserve the scenic and natural beauty for these streams for future residents.
Snow Hill Comprehensive Plan Completed in 2004, the Snow Hill Comprehensive Plan lays out a vision for the town to develop as a “sustainable, growing rural community that provides a cherished quality of life for all residents and a model for others to follow.” This can be accomplished through a concerted effort to protect a healthy and diverse local economy, environment and history. Preservation of the town’s “character and vitality” is of utmost importance as the town and county continue to grow. Implementing this vision requires specific policy goals for land use, parks and open space, community revitalization and the protection of sensitive areas. Land use policies • •
Protect sensitive areas and water quality, and promote efficient use of land within the growth boundary. Coordinate development of public services, schools, etc. with growth area, and give priority to neighborhood, business and employment projects with revitalization potential.
Parks and open space •
• •
Maintain and improve recreation areas, adding additional amenities in current parks (playground equipment, amphitheater for summer concerts), establishing pedestrian access, cleaning up poorly maintained sites, and adding additional areas within walking distance of existing and planned development. Enlarge the high school playing field and encourage general use by town residents. Explore the development of a greenway, making use of undevelopable wetlands to create pedestrian and bike trails through the growth area that would effectively knit together natural and developed areas. “This network will require extensive investigation but, as development proposals are brought forth, its creation should become an important goal of the Town.” (Snow Hill Comprehensive Plan, p. 22)
Community revitalization • • •
Create a downtown cultural arts district to support the visual and performing arts, as well as environmental education. Encourage historic preservation and heritage tourism (walking tours and festivals). Focus on waterfront improvements along the Pocomoke River. This town asset should become a focal point for redevelopment, retail and tourism.
16
Sensitive Areas • •
Reduce risk of flooding, protect water quality in streams and groundwater, preserve productivity of agricultural and natural lands, protect habitat for rare, threatened and endangered species, and maintain scenic beauty. Encourage responsible floodplain management and the preservation of natural wetlands.
Isle of Wight Watershed Restoration Action Strategy (WRAS) Already degraded by development, The Isle of Wight Bay Subwatershed will likely continue to experience substantial growth. The goal of the WRAS is to “accommodate this growth in a way that minimizes its impact and enables the watershed to continue as an excellent place to live, play, and earn a living.” (NOAA 2002, p. 18) It also recommends that efforts be made to improve water quality and wildlife habitat through preservation of healthy and restoration of degraded wetlands. More specifically: • •
•
Water quality should meet or exceed state standards, and biological sampling methods should be used to monitor conditions in the subwatershed. Natural habitat should be able to support “healthy and diverse populations of native plants, animals and aquatic organisms: Natural terrestrial and aquatic habitat corridors should be established, maintained and restored in order to provide protected pathways for wildlife.” Responsible development practices should be followed within the subwatershed, including soft shorelines, cluster development, avoidance of on-site septic systems.
Coastal Bays Comprehensive Conservation Management Plan (CCMP) Completed in 1999, the CCMP is a multi-partner implementation plan detailing strategies to improve water quality, fish and wildlife habitat, recreation and navigation, and community and economic development. Goals from the specific program areas include: Water quality • •
Decrease nutrient and toxic inputs to the Coastal Bays and groundwater from point and non-point (residential, agricultural and commercial landuses, stormwater runoff, and the atmosphere) sources. Decrease sediment inputs (soil erosion) to improve water clarity.
Fish and wildlife • •
Protect and enhance forest and wetland habitats to benefit diverse wildlife populations (including threatened and endangered species) and aquatic resources. Limit impacts to native plant and animal species from non-native invasive species.
Recreation and navigation •
Improve and diversify access to water-based recreation while maintaining balance between natural resource protection and recreational uses.
17
Community and economic development • • •
Enable sustainable land use decision making. Manage the Coastal Bays watershed to maximize economic benefits and minimize negative environmental impacts. Create a community-supported vision of the Coastal Bays region as a “vacation destination, farming region, resource protection area, and retirement community, while protecting and conserving the coastal bays.”
Wicomico County Land Preservation and Recreation Plan Current as of March 2006, the Wicomico County Land Preservation and Recreation Plan includes numerous specific objectives in agreement with or relevant to Worcester County goals and the development of the green infrastructure plan. The contents of the plan can be summarized as follows: Recreation, parks and open space • •
Create an accessible network of high quality recreation areas to improve the mental and physical well-being of residents and to make the county a more desirable place to live, work and visit. Develop a county greenway system that “blend[s] the current popularity of trails with recreation, heritage tourism, and resource preservation.”
Agricultural and natural resource preservation • • • •
Strive for sustainable development in the county that supports a strong regional economy, preserves natural resources and rural character, and accommodates the diverse needs of existing and future residents. Encourage growth in a manner allowing preservation of significant natural features through proper planning and design. Direct growth into well-defined and manageable growth areas and promote the conservation of large natural resource and prime agricultural areas. Preserve environmentally sensitive areas and promote stewardship along critical waterways and the Chesapeake Bay.
Somerset County Land Preservation and Recreation Plan Somerset County’s Land Preservation and Recreation Plan (1998) established “primary growth areas along US Route 13, targets for 80 percent of new development occurring in the county. The plan also identified “secondary growth areas” to accommodate additional infill and bayfront development, and named the conservation of natural areas and the county’s agrarian character the primary objective for rural areas. The plan further set protection goals for water, wetlands, agricultural land, forests and historic resources, as well as for recreation, which are summarized below: •
Preserve the quality of Chesapeake Bay waterfront and waterway corridors by: avoiding excessive soil disturbance, stressing good site design principles, encouraging use of best management practices on farms, protecting natural drainage-ways and keeping storm drainage systems in working order, planting
18
•
•
•
•
•
riparian buffers to filter runoff and provide shade, continuing to support floodplain protection activities of county planners as well as efforts of the state and region to protect the Chesapeake Bay, and improving public awareness of the need to protect critical waterways. Avoid disturbing wetlands and their critical environmental functions by: completing maps of both tidal and non-tidal wetlands, enforcing Chesapeake Bay Critical Area regulations, encouraging developers to protect wetlands during construction, establishing new wetlands when existing wetlands are destroyed or degraded, and supporting wildland protection efforts of the state. Support local farming and foster the county’s rural heritage by: continuing to preserve farmland through the purchase of agricultural easements, considering the creation of a TDR program, encouraging farmers who must sell their land to sell to other farmers, avoid the extension of water and sewer into prime agricultural areas, participating in the state’s Rural Legacy Program. Protect the economic, recreational and ecological quality of woodlands by: recommending cluster development and conservation-oriented design in new subdivisions, continuing to encourage landowners to conserve and manage their forest resources. Enhance residents’ quality of life by conserving and enhancing historic resources and the county’s cultural heritage by: seeking official recognition of historic sites, adopting an official map, and supporting efforts within the Lower Eastern Shore Heritage Program. Improve residents’ access to recreational opportunities by: promoting heritage and ecotourism, making use of existing recreation areas including school grounds, creating additional recreation areas to serve population centers far from existing facilities, and supporting the state greenways concept.
Sussex County Comprehensive Plan Worcester County’s neighbor to the north, Sussex County, Delaware, completed its most recent Comprehensive Plan in 2003. The plan lays out goals for more responsible growth, natural resource conservation and recreation. Though not categorically excluding development anywhere, the county seeks to: • • •
•
Balance growth with the natural environment’s ability to support that growth; Conserve land by encouraging higher density development in appropriate growth areas, and allowing cluster development; Minimize adverse impacts on water quality that result from surface water runoff carrying nutrients from agricultural land and other pollutants from developed lands, especially within designated “conservation districts” and “environmentally sensitive developing areas”; and Preserve and maintain open space for recreational use and environmental conservation, particularly in designated “public and private resource districts”.
Conservation element The conservation element of the Comprehensive Plan further outlines objectives and strategies to maintain and enhance the county’s open spaces and natural resource lands. There are five state parks in Sussex County: Cape Henlopen, Delaware Seashore, Holts Landing, Fenwick Island and Trap Pond. The Prime Hook National Wildlife Refuge (north of the town of Lewes) covers 9000 acres and is managed by the US Fish and Wildlife Service.
19
Recommendations for acquisition of additional areas are based on a rating of natural, cultural, recreation and location attributes, consideration of land use plans, and the feasibility of purchase options. Funding for land acquisition projects comes primarily from land and water conservation bonds, a portion of the realty transfer tax and legislative appropriations. The plan also supports creating a greenways system that makes use of parks, wildlife habitat areas, river and stream corridors, wetlands, floodplains, historic sites, business parks, urban sidewalks, abandoned rail lines, roads, beach areas, and vacant lands. Greenways will provide benefits like safe pedestrian, bicycling and equestrian routes for recreation and commuting, as well as habitat corridors connecting important biological reserves. Additional key actions recommended within the plan include: • • • •
Require environmentally sensitive development and a shoreline building setback line. Plant and manage forest/vegetated riparian buffers. Create a resource protection management plan and a habitat protection ordinance. Focus resource and farmland preservation activities in the Inland Bays Watershed and along the Nanticoke River.
Recreation and open space element The primary goal for the recreation and open space element of the Comprehensive Plan is to provide passive and active recreational facilities for the benefit of residents and visitors. Recreation needs are presently met by the existing network of State parks. Strategies being explored to enhance recreation opportunities and manage growth stress increased cooperation between county and state agencies; more balanced planning for growth, recreation and environmental conservation; and encouragement of making connections among existing and proposed conservation and recreation projects.
Accomack County Comprehensive Plan Accomack County, Virginia is Worcester County’s neighbor to the south. Following a detailed environmental inventory, the 1997 Accomack County Comprehensive Plan (updated in January 2007) outlines general recommendations, goals and objectives for development and resource conservation in the county. The plan acknowledges that unique local habitats should be conserved, ground water quality should be maintained and improved, agricultural land should remain intact, and the character of towns and villages should be preserved. Goals and objectives relevant to green infrastructure • •
Sustain a viable rural community (the “vegetable garden of Virginia”) proud of its history, diversity, natural resources, traditional industries and vision for the future. Create a balanced pattern of land use that protects agricultural land, forest, ground and surface water, wetlands, and other valuable resources as the basis for wildlife habitat, recreation, working farms, seafood industries and tourism. o Preserve open space by clustering new development. o Restore degraded and create new wildlife habitat. o Restore degraded water bodies. o Encourage responsible water use and waste disposal to reduce consumption and pollution.
20
Creating a suitability model using GIS Goals and objectives from county and municipal plans, as well as results and methods from the case studies discussed in Appendix 4, informed each step in the GIS suitability model. The methodology described below is adapted from two documents (North Carolina Division of Coastal Management 2005, Penfold et al. 2005) and includes the following eight steps: 1. 2. 3. 4. 5. 6. 7. 8.
Establish project goals. Define analysis criteria. Identify data needs and collect data. Identify GIS operations to be performed. Create a model. Run the model. Interpret results. Refine the model as needed.
The resulting model aims to capture the concept of “countywide significance” – identifying high priority areas within the county based on county data and analysis. This concept was used in other Maryland green infrastructure plans, namely Anne Arundel, Montgomery and Prince George’s Counties. As in any design process, this methodology is not so much a linear path but an iterative cycle. As new information and/or problems present themselves project goals can (and should) be revisited and analyses redone in order to create the most robust product for implementation.
Establish project goals It is important to keep in mind that the purpose of the eventual green infrastructure plan towards which the county is moving is different than the purpose of the green infrastructure model described in this document. The purpose of the model is not to set or direct specific policies, but to inform more detailed planning activities within the Department of Comprehensive Planning and to encourage a strategic and interdisciplinary approach to development, conservation and recreation planning. It is the ultimate goal of this model to identify areas that support the overall goals of the plan. By including goals and objectives voiced in county and municipal plans this project aims to agree with the citizen input and professional opinions that shaped those plans. The model seeks to identify priority areas in service of the following goals: • • • • •
Conserve biological diversity and the productive capacity of ecosystems through the creation of an interconnected network of forests, wetlands and working lands. Create and sustain long-term socioeconomic benefits through the preservation of valuable natural and historic resources. Provide equitable and easy access to outdoor recreation. Complement county, municipal and state smart growth and conservation plans and goals. Improve quality of life for county residents and visitors.
21
Define analysis criteria Analysis criteria – information included, buffer distances and the relative importance of layers – were based on the plans reviewed in the previous section as well as consultation with staff during model development. Because the scale of analysis (cell size) is in many cases larger than recommended distances from priority features, approximations were used. Since the reasoning behind protecting ecologically significant areas, working farms, and places appropriate for recreation differ, this model operates as three distinct submodels that can later be combined or overlaid. The study area is defined by subwatershed boundaries and encompasses the five northern Coastal Bays subwatersheds (Upper Pocomoke River, Isle of Wight Bay, Assawoman Bay, Newport Bay and Sinepuxent Bay). The specific elements Figure 1. Model study area defined by included in the models are listed below and subwatershed boundaries described in greater detail in Appendix 1. Policy variables • Land use plan*♣ • Zoning ordinance*♣ Land • • • •
cover characteristics Hydric soils* Prime agricultural soils♣ Forests and wetlands*♣® Road density*
Protected areas • Maryland Green Infrastructure* • Federal, State, County and privately owned lands (parks, reserves, easements) *♣ • Coastal Bays Critical Area* • CREP easements (agricultural best management practices in place) * Sensitive habitat • Threatened and endangered species habitat* • Forest interior dwelling bird species (FIDS) habitat* Water resources • 100-year floodplain*® • Streams (25’, 50’ and 100’ buffers) *® • Stream density (number of streams within a 500 m radius) * * conservation submodel ♣ agriculture submodel ® recreation submodel 22
• • •
Well head locations* Ground water recharge areas* Subwatersheds subject to TMDL regulations*
Underserved areas • Minority communities (Hispanic and African American) ® • High density areas® Destinations • National register historic places® • Existing recreation sites® Access and connections • Proximity to roads® • Land and water trails® • Maryland Greenways®
Identify data needs and collect data The criteria listed above demanded a great deal of high resolution spatial data. These data, in vector format (point, line and polygon shape files) were obtained from Worcester County and the Maryland Department of Natural Resources. All GIS shape files used are documented in Appendix 1. All data included are the most up to date files available to the county, but efforts should be made when updating the model to search for the most recent data as they become available. For example, the new high resolution land cover database and shape file being developed by Worcester County should be used to clarify the green infrastructure model.
Identify GIS operations to be performed The principal actions taken in the model are conversions of vector (point, line, polygon) data to raster (grid) data. This is accomplished using two commands in the ArcGIS Spatial Analyst Toolbox: feature to raster and Euclidian distance (see Figures 2 and 3). Resulting grids were then reclassified using the reclassify tool in Spatial Analyst. In the case of the demographics layers (underserved Hispanic, African American and high density communities) I used the map algebra tool to target specific census blocks. These layers were combined in a weighted overlay in order to compute the final suitability surface for each submodel, conservation, agriculture and recreation. Other commands were considered (clip and buffer) but were abandoned because they were less appropriate means of creating the same grid layers.
* conservation submodel ♣ agriculture submodel ® recreation submodel 23
Feature to raster – Converts vector data to raster data. Information from a single field (zoning district or land use code) may accompany the conversion and be used to reclassify the grid. Figure 2. The feature to raster function used to convert zoning polygons
Euclidian distance – Measures the distance from a point, line or polygon feature. These distances can be reclassified to represent buffers around priority features. Directions can also be calculated. Figure 3. The Euclidian distance function used to create stream buffers
Map algebra – Applies a user-entered expression to two raster layers for comparison. Can generate a single output or multiple output raster images. Reclassify – Creates an ordinal scale based on feature information (the field accompanying the feature to raster command), calculated distance or density.
24
Figure 4. The reclassify function used to prioritize zoning categories
Weighted overlay – Considering reclassified raster layers as single utility analyses (SUA), the weighted overlay combines these layers using a common scale and assigns weights to layers. Cell values are multiplied by their weight (a percentage) and added together to produce the final multiple utility analysis (MUA). The final value is rounded to the nearest integer on the defined scale, in this analysis, 1 through 5. For example: ROUND [.10(landuse SUA) + .10(zoning SUA) + .30(landcover SUA) + .30(primeag SUA) + .20(protect SUA)] = ROUND [MUA] = MUA ROUND [.10(5) + .10(5) + .30(3) + .30(4) + .20(1)] = ROUND [5.32] = 5
Create a model I used the model builder tool within the ArcGIS Spatial Analyst extension to create the model in order to best document the data and processes used and results obtained. Model builder works by pulling tools into a workflow diagram (Figure 5), which can be used to run the model in its entirety or to run discrete segments. As mentioned previously, three separate suitability models were created, one each for the identification of conservation, agricultural preservation and recreation priorities (Figures 9, 10 and 11 in Appendix 1). Each model, however, followed the same basic steps outlined below: 5. Convert. A majority of the data layers were in vector format and first needed to be converted into raster format for the cell-based analysis. 6. Analyze. Calculations were made regarding the type and density of features or proximity to features. 7. Reclassify. Information from each analysis was reclassified onto an ordinal scale to be used in the final weighted overlay. Future users of the model may change the criteria to satisfy different purposes. 8. Overlay. All reclassified raster images were overlaid to see where the highest priority areas emerged.
25
The data and processes are brought into the model via the workflow diagram, which links data (blue oval) to processes (yellow rectangle) to achieve an output (green oval). Outputs may be used as inputs to other processes, as was frequently the case in each of the green infrastructure submodels. Figure 5. Model Builder workflow diagram Input feature (1)
Convert to raster
Output raster (1)
Reclassify (1)
Output raster (3) Weighted overlay
Input feature (2)
Analyze
Output raster (2)
Reclassify (2)
Output raster (5)
Output raster (4)
Run the model Basic results and model sensitivity A composite of the conservation, agricultural preservation, and recreation suitability submodels yields the results displayed in Figure 6 below.1 There are significant conservation opportunities throughout the northern half of the county. Areas defined as moderately suitable for conservation are shown in light green, areas of greater priority in dark green. The yellow areas along the Pocomoke River (the western boundary of the study area) scored the highest for conservation suitability. The areas in brown and red indicate agricultural preservation priorities, the red areas being those identified as most highly suitable for preservation. Finally, the areas in magenta were identified as highly suitable (values of 4 and 5) for recreational activities. Figure 6 also shows major roads, municipalities and villages (red outline) and growth areas (orange outline) for context. The model proved sensitive to differential weighting of variables, particularly the policy variables (land use and zoning). Several runs of each submodel and comparisons between runs are detailed in Appendix 2.
About running the model Run time varied between the three submodels based on their complexity. The entire agriculture submodel can be run in under an hour, the weighted overlay in 10–15 minutes. The recreation submodel can be run in about one hour, the weighted overlay in about 20 minutes. The conservation submodel takes considerably longer to run, the full model taking between two and three hours to complete. This is because it is the largest submodel and creates some of the initial raster layers that are reclassified for use in the agriculture and recreation submodel. Running the conservation weighted overlay only takes about 30 minutes. Map algebra comparisons (which were used to compare different model scenarios) ran quickly, taking less than five minutes each.1
1
The green infrastructure vision shown here is composed of the cons100_6m, agr100_1r and rec100_1m submodel runs, which are detailed in Appendix 2. 26
Figure 6. Composite green infrastructure vision
Highest recreation suitability
Highest agriculture preservation suitability
Highest conservation suitability
High agriculture preservation suitability
High conservation suitability
Unsuitable for conservation or agricultural preservation
Moderate conservation suitability
Existing municipalities and developed areas
Low conservation suitability
Planned growth areas
27
Figure 7. Model results and the Maryland Green Infrastructure Assessment
Highest recreation suitability
Highest agriculture preservation suitability
Highest conservation suitability
High agriculture preservation suitability
High conservation suitability
Unsuitable for conservation or agricultural preservation
Moderate conservation suitability
Maryland Green Infrastructure
Low conservation suitability
28
Interpret results Under nearly all scenarios a good portion of the county came back as relatively suitable for conservation, indicating that development should be limited where feasible, but always conducted in as responsible a manner as possible. This result is also reflective of the County’s progressive zoning laws that are already in place and have acted to preserve much of the agricultural land and forested areas within the County. Water resources, including the Pocomoke River and sensitive shorelines are critical pieces of a comprehensive green infrastructure vision for the County and were consistently identified as conservation priorities in the model. Agricultural areas were included as highly or moderately suitable in all three submodels, reflecting the importance of these areas to conservation, economic and cultural priorities. Areas identified as priorities in Figure 6 are key areas in which to maintain agricultural production, remaining areas appropriate for reforestation (and/or wetland restoration) and/or public access for recreation. The submodel runs included in the composite green infrastructure vision (Figure 6) identify conservation areas within growth areas (particularly to the west of the Town of Berlin). This should not be interpreted to mean that development is inappropriate in these areas (indeed it has already been decided upon) but to emphasize that there is a need and an opportunity for environmentally and culturally sensitive development and building design. It also includes areas both inside and outside of the green infrastructure network identified by the State of Maryland (Figure 7). Areas identified within this network help to confirm the relevance of the state network and Worcester County’s contribution to state-level ecological sustainability and to set priorities for land acquisition within the Maryland Green Infrastructure. Areas identified outside of the network can be interpreted as opportunities to extend and/or modify the state network to reflect local conditions and goals more accurately.
Refine the model as needed Many edits were made to the model as it approached its final state presented here. Data layers and GIS operations were explored, included and abandoned. Weights were altered and assumptions re-evaluated (see sensitivity analysis in Appendix 2). This process ought to continue in the future, one of the primary reasons it was created using Model Builder. The process and documentation have been left with the County to edit as new goals are developed and new data become available. In particular, there are opportunities to refine the spatial scale of the models further in order to take a closer look at suitable and priority conservation areas within growth areas and inside municipalities. I tested this approach using the Town of Berlin and its surrounding designated growth areas (Figure 8). The resulting map shows areas suitable for conservation in light green and those less suitable in beige. This run of the conservation submodel used a smaller grid cell for analysis (10 m x 10 m, instead of 100 m x 100 m). Physical variables related to sensitive species and their habitat were weighted strongly, while the policy variables (land use and zoning) were given the least weight. The green infrastructure network identified by the State of Maryland is overlaid on the map in dark green. Notably, this network passes right through the growth area to the west of the town limits, indicating opportunities for environmentally-sensitive design in the new
29
development that will take place there. This approach could be used to identify restoration projects as well. Figure 8. Conservation suitability submodel results for the Town of Berlin Layer landuse zoning hydric landcover roadd mdgia protect critarea senspp fidshab floodpl stream
Assigned Weight 1 1 4 4 4 2 2 2 5 5 3 3
Percent Weight 4 4 11 11 11 5 5 5 14 14 8 8
Water bodies
Unsuitable for conservation
Maryland Green Infrastructure
Existing municipalities and developed areas
Moderate conservation suitability
Planned growth areas
Low conservation suitability
Another important change to the model will be the inclusion of the highly detailed land cover layer based on high resolution 2004 aerial photography in the place of the current land cover classification.
Data layers not included that could be included A number of data layers could have been included based on the review of other plans, but were not due to time, data availability and/or data quality constraints. There are likely additional layers that may be explored in future iterations of the model. • • • • • • • • •
Abandoned rail lines and utility corridors Forest size thresholds Forest conservation easements on private lands Steep slopes Wetlands of special state concern (WSSC) Wetland restoration priorities Forest and wetland mitigation banks and restoration projects (existing and proposed) Proposed or identified potential recreation sites Stream density
30
In addition to this list of data that could be included, another look should also be given to the underserved communities included in the recreation submodel, particularly the number of residents in the areas identified as having two times the average percentage of Hispanic residents. This may not be a significant variable, since the average is only 1.3 percent and there may be very few people in each census block, thereby giving high priority to an area on behalf of only a few people.
GIS operations abandoned As tools were explored in creating the green infrastructure model, a few were abandoned due to difficulties in interpretation and the discovery of simpler methods. Clip Before discovering the model environment settings in Model Builder, all incoming vector layers were clipped to the study area before being processed. Instead, the extent over which the model runs is defined globally and the study area is used as a “mask” generating final outputs only for cells entirely within the bounds of the study area. Removing all of the separate clip operations simplified the model and decreased processing time. Buffer and multiple ring buffer For unknown reasons (possibly layer complexity, for example too many stream segments in the hydrology layer), the multiple ring buffer did not function within Model Builder. I explored using only single buffers, which would function in Model Builder. However, using the Euclidian distance tool did work, and simplified the model considerably. In the first instance, the process would have been: clip input feature, create buffer, convert to raster, then reclassify. In the current model the process is streamlined: apply Euclidian distance to line feature and reclassify. Kernel/line/point density This tool calculates the density of linear (roads and streams) or point (well heads) features. Due to difficulties in interpreting the output of this command, particularly for the roads layer, it was abandoned in favor of a simpler measure of proximity to roads, streams and well head locations. Runs of the conservation submodel using road density and proximity to roads did not vary a great deal.
31
Next steps: from a model to a plan Creating and running this series of conservation, agricultural preservation and recreation suitability models, while resulting in a valuable product, does not constitute a green infrastructure plan. In order for this vision (or an amended vision developed by the County as new data become available and better understood) to become a reality it must be coupled with the County’s existing comprehensive planning efforts and linked with specific goals and objectives, as well as conservation tools and funding sources. Fortunately, many of these tools exist and are being used in Worcester County already. These include restrictive conservation and agricultural zoning, critical areas legislation and development restrictions, and agricultural/conservation easements. Thousands of acres are also under protection through US Department of Agriculture (USDA) programs, including the Conservation Reserve Enhancement Program (CREP), the Wetlands Reserve Program (WRP) and the Northern Bobwhite Quail Habitat Initiative. There are also opportunities to link conservation with a transfer of development rights program as well as forest and wetland mitigation banks currently under consideration and development within the County. The green infrastructure vision must also be accompanied by a set of specific and measurable goals and objectives. Tracking progress towards goals (such as acres of forest and farms protected) enables choosing and refining the most appropriate conservation tools and can help secure funding from state and federal government sources. The steps outlined for creating effective conservation plans in the Conservation Measures Partnership (2004) Open Standards for the Practice of Conservation would be ideal to follow in the development of a comprehensive green infrastructure plan for Worcester County.
Looking to the future Looking at comprehensive and green infrastructure plans from more developed areas may be illustrative of future conditions and challenges, particularly in and around the County’s towns and villages.2 What will a Snow Hill twice its current size look like? What will the area around Ocean City, Ocean Pines and Berlin look like after significant growth over the next 20 to 50 years? Suddenly the picture is not so green. However, designing and adopting a green infrastructure plan early on will help to direct responsible development and protect the most important ecological, agricultural and cultural resources in the County. Coordinating development and conservation across county and state lines will become increasingly important as well. Ecological boundaries do not usually match political boundaries, and development decisions in neighboring jurisdictions will have significant impacts (either positive or negative) on Worcester County’s environment and economy. Effecting planning will demand increased collaboration among Worcester, Somerset and Wicomico Counties in Maryland, as well as with Sussex County, Delaware, and Accomack County, Virginia. Coordinating County activities with state goals and strategies in Maryland, Virginia and Delaware will facilitate the broadest conservation success possible.
2
Examples include Montgomery and Prince George’s Counties in Maryland, both of which have begun developing and implementing green infrastructure plans. Cuyahoga County, Ohio is an example of a heavily urbanized area planning for green infrastructure, and is detailed in the case studies in Appendix 4. 32
References Accomack County Planning Commission. 1997. Comprehensive Plan. (http://www.co.accomack.va.us/Planning/comprehensive_plan.html) Accomack County Planning Commission. 2007. Comprehensive Plan Update. (http://www.co.accomack.va.us/Planning/comprehensive_plan_update.html) Benedict, M.A. and E.T. McMahon. 2006. Green Infrastructure: Linking Landscapes and Communities. Washington DC. Island Press. Berlin Planning Commission. 2005. Comprehensive plan for the Town of Berlin (Draft). Carr, M. and P. Zwick. 2005. “Using GIS suitability analysis to identify potential future land use conflicts in north central Florida”. Journal of Conservation Planning. Volume 1, Issue 1, pp. 89–105. Coastal Bays Program. 2004. Coastal Bays Comprehensive Conservation Management Plan. (http://www.mdcoastalbays.org/programs.php?subaction=showfull&id=1096905237 &archive=&start_from=&ucat=8&) Conservation Measures Partnership. 2004. Open Standards for the Practice of Conservation. (http://fosonline.org/CMP/Products.cfm) Cuyahoga County Planning Commission. Cuyahoga County Greenspace Plan. Cleveland, Ohio. (http://planning.co.cuyahoga.oh.us/green/index.html) Cuyahoga County Planning Commission. Towpath Trail Extension Alignment and Design Study. Cleveland, Ohio. (http://planning.co.cuyahoga.oh.us/towpath) Florida Department of Environmental Protection and the Florida Greenways Coordinating Council. 1998. Phase II Final Report: Statewide Greenways System Planning Project. National Oceanic and Atmospheric Administration. 2002. Isle of Wight Watershed Restoration Action Strategy. Maryland Coastal Zone Management Program, Department of Natural Resources. The Trust for Public Land. 2005. Greenprint for King County: A Model for Conservation. Prepared for King County Natural Resources and Parks Water and Land Resources Division by Jones & Jones, Point Wilson Group and The Trust for Public Land. (http://dnr.metrokc.gov/wlr/greenprint/index.htm) University of North Carolina at Chapel Hill. 2001. Kinston/Lenoir County Green Infrastructure Plan for the Neuse River Floodplain. Graduate student workshop, Department of City and Regional Planning. University of North Carolina at Chapel Hill. 2002. Linking Natural and Historic Assets: Green Infrastructure as Economic Development in Lenoir County, North Carolina. Graduate student workshop, Department of City and Regional Planning.
33
McDonald, L., W. Allen, M. Benedict and K. O'Connor. 2005. “Green Infrastructure Plan Evaluation Frameworks”. Journal of Conservation Planning. Volume 1, Issue 1, pp. 12–43.* Meenar, Md Mahbubur R., A.S.M. Abdul Bari, and Kurt Paulsen. 2006. “Developing a watershed inventory for vulnerability assessment using ArcGIS”. ArcUser. January– March 2006, pp. 20–23. Meenar, Md Mahbubur, James Duffy and A.S.M. Bari. 2006. “Life on the floodplain: remapping watersheds, neighborhoods, and lives”. Journal of the American Planning Association. July 2006, pp. 30–33. Merenlender, A.M., C. Brooks, D. Shabazian, S. Gao and R. Johnston. 2005. “Forecasting exurban development to evaluate the influence of land-use policies on wildland and farmland conservation”. Journal of Conservation Planning. Volume 1 Issue 1, pp. 64– 88. North Carolina Division of Coastal Management. 2005. Land suitability analysis – user guide for ArcView 3.x and ArcGIS 9.x. North Carolina Center for Geographic Information and Analysis. December 2005. Noss, Reed F. 1987. “Protecting natural areas in fragmented landscapes”. Natural Areas Journal. Volume 7, Issue 1, pp. 2–13. Town of Ocean City. 2006. Comprehensive Plan. (http://www.town.oceancity.md.us/Planning%20and%20Zoning/DraftComprehensivePlan/index.html) Ocean Pines Association. 2002. Comprehensive Plan and Comprehensive Plan Update. Penfold, B.M., T.H. Funk and H.R. Hafner. 2005. “Exploring the use of suitability modeling to locate air toxics monitors in Massachusetts”. Sonoma Technology, Inc. Petaluma, California. (http://gis.esri.com/library/userconf/proc03/p0997.pdf) Pocomoke City. 1980. Comprehensive Plan. Randolph, John. 2004. Environmental Land Use Planning and Management. Island Press. Washington, DC. Searns, R. “Happy trails: greenways put their stamp on the Denver area”. Journal of the American Planning Association. January 2003, pp. 34–36. Town of Snow Hill. 2004. Comprehensive Plan. Somerset County. 1998. Land Preservation and Recreation Plan. Sussex County. 2003. Comprehensive Plan. The Conservation Fund. 2007. “GreenInfrastructure.net”. (http://www.greeninfrastructure.net)
*
This reference is a particularly exhaustive and relevant review of green infrastructure planning, and is available online at: http://www.journalconsplanning.org/2005/index.html. 34
University of Florida GeoPlan Center. 1999. “Florida Statewide Greenways Planning Project”. (http://www.geoplan.ufl.edu/projects/greenways/greenwayindex.html) United States Census Bureau. 2000. “Worcester County, MD Fact Sheet”. American Fact Finder. (http://factfinder.census.gov) Weber, T. 2003. “Chapter 10: Desktop Parcel Evaluation”. Maryland’s Green Infrastructure Assessment. Maryland Department of Natural Resources. Wicomico County. 2006. Land Preservation and Recreation Plan. Worcester County Department of Comprehensive Planning and Department of Recreation and Parks. 2006. Worcester County Parks, Recreation and Land Preservation Plan. Worcester County Department of Comprehensive Planning. 2006. The Comprehensive Development Plan. Worcester County Department of Comprehensive Planning. 2006. Worcester County Hazard Mitigation Plan (Draft).
35
Appendix 1: Model criteria and workflow diagrams Conservation Suitability Model Criteria........................................ 37 Agricultural Preservation Suitability Model Criteria ......................... 38 Recreation Suitability Model Criteria .......................................... 39 Conservation Suitability Model Diagram ....................................... 40 Agricultural Preservation Suitability Model Diagram ........................ 41 Recreation Suitability Model Diagram ......................................... 42
36
Conservation Suitability Model Criteria Suitability Criteria Category
Layer
source
Unsuitable
Low Suitability
0
1
landuse
S:\WRGIS\data\land_planning\development\l anduse_plans\landuse3_7_06
Atlantic Ocean; Commercial Center; Existing Developed Centers; Growth Area; Industry; Institutional; Municipality; Village
zoning
S:\ArcGISdata\Zoning\ZoningDist
hydric
S:\GISdata\Soils\Soils_05
landcover
S:\GISdata\LandCover\wolu2002
roadd
S:\GISdata\Transportation\Cline
mdgia
S:\GISdata\LandConserv\Green_Infra\Wor_G I
> 300 m
protect
S:\GISdata\LandConserv
> 200 m
shoreline
S:\ArcGISdata\Layer Files\Hydro_NRCS03.lyr
agbmp
policy
B1; B2; CA; CM; M1; M2; OC; Poc; R2; R3; R4; R5; RO; SNH; TwB; V1
Moderate Suitability 2
3
High Suitability 4
Agriculture
R1; E1
A1
11; 73; 241; 242
18; 21; 22; 23; 25; 44
high
med
Zoning districts currently included in the County code and Comprehensive Plan. Field: WOCOZONES
The County's Zoning Ordinance is an enforcable policy instrument based on the land use plan and used to guide development.
yes
County soil types classified according to their hydric rating. Field: hydricrati
Wetland preservation is a priority for maintaining functioning ecosystems. Hydric soils are indicative of the presence of wetlands or potential restoration sites for filled wetlands. Existing forests, wetlands, water features and working farms are primary targets for conservation according to the comprehensive plan. Low density residential areas, institutional properties, open urban areas and bare ground may be suitable for connecting conservation elements.
Road density reflects the number of road segments per square kilometer.
The presence of roads fragments habitats and impedes the movement of species, making areas with "high" road density unsuitable for conservation.
Green infrastructure of statewide significance identified in the 2000 Maryland Green Infrastructure Assessment.
Proximity to MD-identified green infrastructure means connectivity to a larger network of intact forests and access to state conservation funding.
< 100 m
Existing portfolio of protected lands within the county.
Proximity to existing protected areas reinforces the connectivity of the county and state green infrastructure networks, and increases the likelihood of successful conservation efforts (i.e. enforcement, stewardship).
> 300 m
< 300 m
1000' buffer from shoreline.
Proximity to existing protected areas increases the likelihood of successful conservation efforts (i.e. enforcement, stewardship).
S:\Planning\Dana\bmp06_classified
outside
inside
Agricultural best management practices areas
Sites identified by the county already as suitable for agricultural preservation and/or improved nutrient management programs.
senspp
S:\Planning\Katherine\GIS files\DNR data Jan 2006\Lynn Species
> 200 m
100 m to 200 m
< 100 m
Known locations of threatened and endangered species mapped by DNR. Confidential data (ESA, eagles' nests, Delmarva Fox Squirrel, Colonial Waterbirds).
Local habitats of threatened and endangered species should be strongly protected.
fidshab
S:\Planning\Katherine\GIS files\DNR data Jan 2006\Lynn Species
> 200 m
100 m to 200 m
< 100 m
Forest fragments with the potential to support populations of forest interior dewlling birds (FIDS).
Potential FIDS habitat should be strongly protected.
floodpl
S:\GISdata\Floodplain\100_Yr_fl
outside
500-year
100-year; coastal flooding with velocity hazard
Flood zones identified by FEMA. Field: FLDZONE
Focus conservation efforts in the floodplain in order to protect water quality and discourage hazardous development.
stream
S:\ArcGISdata\Layer Files\Hydro_NRCS03.lyr
> 200 m
100 m to 200 m
< 100 m
County rivers and streams ( perennial and intermittent only).
Protect water quality and habitat for aquatic species.
streamd
S:\ArcGISdata\Layer Files\Hydro_NRCS03.lyr
low
med
high
Stream density reflects the number of stream segments per square kilometer.
Protect water quality and habitat for aquatic species.
wellh
S:\WRGIS\data\land_conservation\water_pla nning\water_quality\wells\wocb_wells3m
> 200 m
100 m to 200 m
< 100 m
gwrch
S:\WRGIS\data\physical\water_resources\gro undwater\cosbay_gw_recharge.shp
outside
inside
Recharge areas identified for Coastal Bays surficial aquifers.
Protect water quality in the Coastal Bays and drinking water supply.
tmdlshed
S:\GISdata\Watershed\Worc_Subshed
outside
inside
Worcester County subwatersheds subject to TMDL regulations.
Protect water quality and assist the county in meeting TMDL regulations.
< 300 m
< 200 m
100 m to 200 m
low
< 100 m
inside
protected
habitat
water resources
The County Comprehensive Plan reflects a vision for the County's future land use composition based on goals developed by staff, elected officals and input from residents. It is a general policy guide and does not necessarily reflect actual land cover.
C1; WAT
41; 42; 43; 50; 60; Land cover categories identified from aerial photography 71 current as of 2002. Field: LU_CODE
landcover
Reasoning
5 Green Landuses currently included in the Comprehensive Plan. Infrastructure; Field: LEGEND Major rivers; Waterbody (bays, ponds); Waterway (rivers, streams, creeks)
no, unranked
12; 13; 14; 15; 16; 17
Description
Agricultural Preservation Suitability Model Criteria Suitability Criteria Category
Layer
source
Unsuitable
Low Suitability
0
1
Moderate Suitability 2
3
High Suitability 4
Description
5
landuse
S:\WRGIS\data\land_planning\development\l anduse_plans\landuse3_7_06
Atlantic Ocean; Commercial Center; Existing Developed Centers; Growth Area; Industry; Institutional; Municipality; Village; Green Infrastructure; Major rivers; Waterbody (bays, ponds); Waterway (rivers, streams, creeks)
Agriculture
Landuses currently included in the Comprehensive Plan. Field: LEGEND
The County Comprehensive Plan reflects a vision for the County's future land use composition based on goals developed by staff, elected officals and input from residents. It is a general policy guide and does not necessarily reflect actual land cover.
zoning
S:\ArcGISdata\Zoning\ZoningDist
B1; B2; CA; CM; E1; M1; M2; OC; Poc; R1; R2; R3; R4; R5; RO; SNH; TwB; V1; WAT
A1; C1
Zoning districts currently included in the County code and Comprehensive Plan. Field: WOCOZONES
The County's Zoning Ordinance is an enforcable policy instrument based on the land use plan and used to guide development.
landcover
S:\GISdata\LandCover\wolu2002
12; 60; 71; 11; 18; 73; 41; 42; 43; 50; 44
21; 22; 23; 25; 241; 242
Land cover categories identified from aerial photography current as of 2002. Field: LU_CODE
Existing forests, wetlands, water features and working farms are primary targets for conservation according to the comprehensive plan. Low density residential areas, institutional properties, open urban areas and bare ground may be suitable for connecting conservation elements.
primeag
S:\GISdata\Soils\Soils_05
County soil types classified according to farming suitabilty. Field: frmdcl
Prime agricultural soils are also prime development soils and where feasible should be retained for productive farm use, existing or potential.
protect
S:\GISdata\LandConserv
Existing portfolio of protected lands within the county.
Proximity to existing protected areas reinforces the connectivity of the county and state green infrastructure networks, and increases the likelihood of successful conservation efforts (i.e. enforcement, stewardship).
policy
landcover
protected
Reasoning
Not prime farmland
Prime farmland if drained > 200 m
Prime farmland if irrigated 100 m to 200 m
All areas are prime farmland
Farmland of statewide importance < 100 m
Recreation Suitability Model Criteria Suitability Criteria Category
underserved areas
Layer
source
Unsuitable
Low Suitability
0
1
Moderate Suitability 2
3
4
Description
Reasoning
5
hispanic
S:\GISdata\Demographics\census\2000
> 2 miles < average
> 2 miles up to 2x average
> 2 miles > 2x average
Hispanic communities underserved by recreation. Worcester County average: 1.3%.
Ensure access to conservation and recreation areas for underserved minority communities.
highden
S:\GISdata\Demographics\census\2000
> 2 miles < average
> 2 miles up to 2x average
> 2 miles > 2x average
High density neighborhoods underserved by recreation. Average population density in Worcester County: .15 ppl per acre (38/sq km).
Locating recreation in/near higher density areas provides services to more people at less cost.
black
S:\GISdata\Demographics\census\2000
> 2 miles < average
> 2 miles up to 2x average
> 2 miles > 2x average
African American communities underserved by recreation. Worcester County average: 16.7%.
Ensure access to conservation and recreation areas for underserved minority communities.
landcover
S:\GISdata\LandCover\wolu2002
12; 13; 14; 15; 17
11; 16
241; 242
floodpl
S:\GISdata\Floodplain\100_Yr_fl
outside
coastal flooding with velocity hazard
500-year
100-year
Flood zones identified by FEMA. Field: FLDZONE
The 100-year floodplain is more suitable for recreation than for permanent development. Areas susceptible to severe flooding and velocity hazard may be more appropriate for conservation than recreation.
stream
S:\GISdata\Hydrology\Wor_Streams
> 30 m > 200 m
10 m to 30 m 100 m to 200 m
< 10 m < 100 m
County rivers and streams ( perennial and intermittent only).
Rivers, streams and waterbodies provide opportunities for water-based recreation.
historic
S:\GISdata\LandConserv\Private_Lands\hist_ register\nhrsites00
> 1/4 mile
1/8 to 1/4 mile
< 1/8 mile
National register of historic places.
All trails should lead somewhere, and all historic assets should be considered part of a connected system.
recsites
S:\Planning\Katherine\Land Preservation and Recreation Plan\MEIRS\GIS_projectfiles\meirs2005
< 1600m
> 3200 m
< 100'
Existing recreational facilities.
All trails should lead somewhere, and all parks should be considered part of a connected system.
access
S:\GISdata\Transportation\Cline
> 1/2 mile
1/4 to 1/2 mile
< 1/4 mile
Proximity to roads (access)
Access is important to supporting the use of recreational areas.
viewtr
S:\GISdata\Recreation
> 500'
100 to 500
< 100'
View Trail 100 bike trail.
Future trail development should seek to connect to existing bike and pedestrian trails.
S:\GISdata\Recreation
> 1/4 mile
1/8 to 1/4 mile
< 1/8 mile
Indian Trail scenic beach to bay drive
Car access to scenic areas should be maintained.
> 500'
100 to 500
< 100'
Maryland state greenways.
Support existing and proposed MD greenways (trails) to maximize connectivity of the county trail system and to secure state funding for conservation and recreation.
landcover
water resources
destinations
High Suitability
indiantr access mdgrnway and connections
MD DNR (swgrnway.shp available online)
1600 to 3200 m
21; 22; 23; 25; 73
18; 41; 42; 43; 44; Land cover categories identified from aerial photography 50; 60; 71 current as of 2002. Field: LU_CODE
Recreational areas should be located in natural and agricultural environments, because of their aesthetic qualities and lack of development.
watertr
S:\GISdata\Recreation
> 1/4 mile
1/8 to 1/4 mile
< 1/8 mile
County water trails and paddle routes.
Locate recreational areas and access points near existing water trails.
alltrails
S:\GISdata\Recreation
> 1/4 mile
1/8 to 1/4 mile
< 1/8 mile
All county trail systems, car, bike, pedestrian and watercraft.
Future trail development should seek to connect to existing trails.
Figure 9. Conservation suitability submodel
Figure 10. Agricultural preservation suitability submodel
41
Figure 11. Recreation suitability submodel
Appendix 2: Summary of model runs General Discussion ................................................................ 44 Conservation submodel .......................................................... 48 Agriculture submodel............................................................. 51 Recreation submodel ............................................................. 55
43
General Discussion The following images and descriptions characterize the first runs of the conservation, agricultural preservation and recreation submodels. Several runs are included to display the sensitivity of the models to different policy scenarios (varying land use and zoning regimes) and weighting assumptions. The policy scenarios, restrictive and moderate, are detailed in Appendix 1, but warrant a short description here as well. Conservation submodel The restrictive model runs included only the most appropriate land uses and zoning categories for conservation and agricultural preservation. With respect to the conservation submodel, this meant that the land use categories “green infrastructure”, “major rivers”, “waterbody” and “waterway” received a score of 5; the land use category “agriculture” received a score of 3; all other categories were given a score of 0, unsuitable for conservation (Table 1). Under the moderate policy scenario the conservation submodel included the same land use categories (minus “agriculture”) as the most highly suitable (score of 5), but also included the land use categories “growth area” and “institutional” as highly suitable for conservation (4), “agriculture” as moderately suitable (3), and “existing developed centers”, “municipality” and “village” as less suitable (1). The remaining land use categories, “Atlantic Ocean” and “commercial center” remained unsuitable for conservation and were excluded. At first glance it may seem odd that growth areas were ranked so highly for their conservation suitability, but this was done in order to point out environmentally valuable and sensitive areas slated for development. If these areas are known before development occurs, they can be taken into account when reviewing and adopting subdivision and site plans. Institutional (government property, schools) lands were also considered of high conservation suitability because these places provide the opportunity for experimental, model and educational conservation projects. Table 1. Land use classifications under restrictive and moderate policy scenarios Conservation Suitability
Agriculture Suitability
Land use category
Score Restrictive
Score Moderate
Score Restrictive
Score Moderate
Agriculture
3
3
5
5
Atlantic Ocean
0
0
0
0
Commercial Center
0
0
0
0
Existing Developed Centers
0
1
0
0
Green Infrastructure
5
5
0
3
Growth Area
0
4
0
3
Industry
0
0
0
0
Institutional
0
4
0
0
Major rivers
5
5
0
0
Municipality
0
1
0
0
Village
0
1
0
0
Waterbody (bays, ponds)
5
5
0
0
Waterway (rivers, streams, creeks)
5
5
0
0
44
With respect to zoning, the restrictive policy scenarios of the conservation submodel included only the categories “conservation (C1)”, and “water (WAT)” as most highly suitable for conservation, assigning them a score of 5. The “agriculture (A1)” zoning category was coded moderately suitable for conservation (3), and “estate (E1)” and “rural residential (R1)” were coded as least suitable for conservation (1). All other zoning categories were excluded from the restrictive runs of the conservation submodel. The moderate policy scenario included the same highly suitable and moderately suitable zoning categories (C1 and WAT, E1 and R1), but also included commercial airports and marinas (CA and CM), all towns and villages (OC, Poc, SNH, TwB, V1), as well as suburban, multifamily, general, mobile home and office residential (R2, R3, R4, R5, RO) as least suitable for conservation (1). Only industrial zones were excluded as unsuitable for conservation. As with the land use policy scenarios described above, these zoning categories were included to highlight smaller scale conservation opportunities in more urbanized areas that may facilitate significant connections between larger habitat areas. Conservation in urban areas, which could end up functioning as a greenways system, also provides an immediate connection to nature for residents in more densely settled areas. Table 2. Zoning classifications under restrictive and moderate policy scenarios Conservation Suitability
Agriculture Suitability
Zoning category
Score Restrictive
Score Moderate
Score Restrictive
Score Moderate
Agriculture (A1)
3
3
5
5
Neighborhood Business (B1)
0
0
0
0
General Business (B2)
0
0
0
0
Conservation (C1)
5
5
5
5
Commercial Airport (CA)
0
1
0
0
Commercial Marina (CM)
0
1
0
0
Estate (E1)
1
3
0
3
Light Industrial (M1)
0
0
0
0
Heavy Industrial (M2)
0
0
0
0
Ocean City (OC)
0
1
0
0
Pocomoke (Poc)
0
1
0
0
Rural Residential (R1)
1
3
0
3
Suburban Residential (R2)
0
1
0
0
Multifamily Residential (R3)
0
1
0
0
General Residential (R4)
0
1
0
0
Mobile Home Residential (R5)
0
1
0
0
Office Residential (RO)
0
1
0
0
Snow Hill (SNH)
0
1
0
0
Berlin (TwB)
0
1
0
0
Village (V1)
0
1
0
0
Water (WAT)
5
5
0
0
45
Agricultural submodel For the agricultural preservation submodel, the restrictive policy scenario considered only the land use category “agriculture” suitable for preservation, assigning it a score of 5. The moderate policy scenario included “agriculture” again as most highly suitable for preservation (5) but also included “green infrastructure” and “growth area” as moderately suitable for preservation (3). These latter two categories were included, as in the conservation submodel, to point out opportunities within the growth areas for preservation and sensitive development, as well as opportunities within designated conservation areas where working farms may still be valuable and appropriate uses of the land primarily because of soil characteristics. For zoning, the restrictive policy scenario for the agriculture submodel included only A1 and C1 zones as highly suitable for preservation. Though the Maryland Green Infrastructure Assessment categorically excluded agricultural lands, conservation zones in Worcester County do include some important agricultural areas. The moderate zoning policy scenario included the same two zones as most highly suitable (A1 and C1), but also included E1 and R1 as moderately suitable for preservation, assigning them a score of 3. Estate and rural residential areas were included because they represent large lot development that may be suitable for preservation as working landscapes or rural open space in keeping with the county’s rural character. Understanding the model runs In naming the model runs, the name begins with the submodel (cons, agr or rec), followed by “100” (indicating the 100 m x 100 m cell size at which the model was run), a number (17) indicating the scenario run, and an “m” or an “r” indicating whether more moderate or restrictive model assumptions were used, as described above. The same color range is used to identify conservation, agriculture and recreation suitability in each of the model runs presented below (a legend is included for each image). For the change detection images, red means that the value at a particular location in the second scenario was higher than at the same location in the first scenario, black means that the value in the second scenario was lower than in the first. For consistency, in all cases, the smaller numbered model was subtracted from the larger numbered model to detect the difference between the two. For example: (cons100_7m) – (cons100_1m) = difference between the two grids In addition to varying the policy scenarios in the model runs, different variable weights were tested. In each case the ranking of the variables appears in a table beside the change detection image. The rank translates into a percentage that was applied to the variable in the model. For example, in run agr_1m of the agriculture submodel the policy variables (land use and zoning) were ranked 1, protected areas 2, and land cover characteristics (land cover and prime agricultural soils) 3. These rankings translated into 30%, 20% and 10% weights in the model runs. Similarly, in model run agr_2m equal ranking of all the model variables yielded an equal 20% weighting of variables in the model.
46
Table 3. Rank and percent weight for two runs of the agriculture submodel agr_1m
agr_2m
Variable
Rank
% Weight
Rank
% Weight
landuse
1
10%
1
20%
zoning
1
10%
1
20%
landcover
3
30%
1
20%
primeag
3
30%
1
20%
protect
2
20%
1
20%
Total
10
100%
5
100%
In all cases it was found that ranking policy related variables higher resulted in identifying more suitable areas for conservation, agricultural preservation and recreation. This is a predictable result, since in most places land use and zoning cover large areas.
47
Conservation submodel The conservation submodel is sensitive to the relative importance of policy variables (land use and zoning). In these two model runs, 1m and 7m, a great deal of the county is identified as at least moderately suitable for conservation. When land use and zoning variables are given greater weight (assigned a weight of 4 in run 7m compared to a weight of 1 in 1m), more areas are identified as having moderate or high suitability (a single cell is given a one point higher score). In each run, the variables are weighted by type (policy, land cover, habitat, protected and water resources). Figure 1. Model runs 1m and 7m (policy variables low vs. high)
Figure 2. Change detection between model runs 1m and 7m Layer landuse zoning hydric landcover roads mdgia protect critarea agbmp senspp fidshab floodpl stream streamd wellh gwrch tmdlshed
Weight 1m 1 1 4 4 4 2 2 2 0 5 5 3 3 0 0 3 0
Weight 7m 4 4 5 5 5 2 2 2 2 3 3 3 3 0 1 1 1
48
The conservation submodel is also sensitive (though to a slightly lesser extent) to the overall weighting of all variables. When all variables are given an equal weight, less land area is identified as suitable for conservation. Though this may be a valid approach in the face of uncertainty over how to weight individual criteria, this result could be interpreted to mean that assigning equal weights to variables is misleading, because it discounts the most pertinent factors. Figure 3. Model runs 2m and 7m (equal weight vs. agreed-upon weight)
Figure 4. Change detection for model runs 2m and 7m Layer landuse zoning hydric landcover roads mdgia protect critarea agbmp senspp fidshab floodpl stream streamd wellh gwrch tmdlshed
Weight 2m 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 0
Weight 7m 4 4 5 5 5 2 2 2 2 3 3 3 3 0 1 1 1
49
Including protected areas in the model actually decreases the number of areas ranking as conservation priorities. This may be an indication that some protected areas do not coincide with the highest suitability lands with respect to at least one of the physical, habitat, or water resource variables. It also means that, in the model run including protected areas, that for a grid cell to appear “highest suitability” it must score highly on all model variables including being currently protected. This will exclude valuable areas that are not currently under protection, which is seemingly against the goals of the model and a green infrastructure plan. While proximity to existing protected areas is a legitimate factor in determining the feasibility of conservation projects, perhaps it is more appropriate to run the model without these areas, adding them in later as items to always be included in the green infrastructure network. In this way more of the underlying natural characteristics can be expressed. Figure 5. Model runs 6m and 7m (no protected areas vs. protected areas)
Figure 6. Change detection for model runs 6m and 7m Layer Landuse Zoning Hydric Landcover Roads Mdgia Protect Critarea agbmp senspp fidshab floodpl stream streamd wellh gwrch tmdlshed
Weight 6m 4 4 5 5 5 2 0 0 0 3 3 3 3 0 1 1 1
Weight 7m 4 4 5 5 5 2 2 2 2 3 3 3 3 0 1 1 1
50
Agriculture submodel These runs of the agriculture suitability submodel show the important differences between the moderate and restrictive assumptions under which the models were run. In the more restrictive scenarios (r), more areas were flagged as highest suitability for agricultural preservation and almost no areas appeared to be moderately suitable. The more moderate scenarios (criteria detailed in Appendix 1) gives a fuller picture of where the agricultural land is and where the priorities are in context, and thus may be more informative for policy making. Figure 7. Model runs 1r, 1m, 5r and 5m (moderate and restrictive scenarios)
51
These runs also show the agriculture submodel’s sensitivity to policy variables (land use and zoning). Assigning the policy variables greater weight (a higher rank, as described above and in Table 3) in the agriculture submodel results in more areas being identified as more suitable for preservation. This is a predictable result since so much of the county is defined as agriculture in both the current land use and zoning plans, and may not provide enough detail to function as a prioritization tool. Some areas ended up two categories higher (shown in green) when scenario 1m and scenario 5m were compared. Figure 8. Model runs 1m and 5m (policy variables low vs. high)
Figure 9. Change detection for model runs 1m and 5m layer landuse zoning landcover primeag protect
Weight 1m 1 1 3 3 2
Weight 5m 3 3 2 2 1
52
Giving all variables equal weight in the overlay analysis identifies more areas as more highly suitable for preservation. This identifies, in particular, areas in the northwestern most part of the county. Areas identified as being the highest preservation priorities remain relatively constant between the two scenarios. Figure 10. Model runs 1m and 2m (agreed-upon weight vs. equal weight)
Figure 11. Change detection for model runs 1m and 2m layer landuse zoning landcover primeag protect
Weight 1m 1 1 3 3 2
Weight 2m 1 1 1 1 1
53
As with the conservation submodel, not including the protected areas in the model run identifies more areas as more suitable for agricultural preservation. In contrast to the conservation submodel, however, so many areas appear as the highest priorities that this model run would preclude strategic decision making. Figure 12. Model runs 1m and 6m (protected areas vs. no protected areas)
Figure 13. Change detection for model runs 1m and 6m layer landuse zoning landcover primeag protect
Weight 1m 1 1 3 3 2
Weight 6m 1 1 2 2 0
54
Recreation submodel In the recreation submodel, there was little difference between the more restrictive and the moderate policy scenario runs. Variable weighting did make a difference, however. When demographic variables (percent Hispanic, percent black, high density) were weighted higher (meaning that greater priority was being given to underserved minority and higher density communities), more areas were identified as more suitable for conservation. However, several of the high priority areas were lost, and the bulk of the county became moderately suitable for recreation. It is also important to keep in mind that in nearly all areas the percentage of the population belonging to a minority group is small, so the difference between an identified minority area and a non-minority area are somewhat insignificant. This scenario may not present the most robust decision making tool. Figure 14. Model runs 1m and 3m (demographic variables low vs. high)
Figure 15. Change detection for model runs 1m and 3m layer hispanic highden black landcover floodpl historic recsites access viewtr indiantr mdgrnway watertr alltrails stream
Weight 1m 1 1 1 2 2 4 4 3 0 0 0 0 3 3
Weight 3m 3 3 3 2 2 2 2 1 0 0 0 0 1 1
55
Similarly, equal weighting of the recreation submodel variables meant that more areas were identified as more highly suitable for recreation. Figure 16. Model runs 1m and 2m (agreed-upon weight vs. equal weight)
Figure 17. Change detection for model runs 1m and 2m layer hispanic highden black landcover floodpl historic recsites access viewtr indiantr mdgrnway watertr alltrails stream
Weight 1m 1 1 1 2 2 4 4 3 0 0 0 0 3 3
Weight 2m 1 1 1 1 1 1 1 1 0 0 0 0 1 1
56
Appendix 3: Editing the model Model Builder Tutorial: Everything you ever wanted to know about Model Builder… almost 1. What is Model builder? a. Cool tool to organize and save commonly executed multi-step actions b. Use to build models and document the process, easy to change c. Creates workflow diagram 2. How does it work a. Add the data you need to your project (vector and/or raster) i. Hint: To help me think about how I was going to reclassify the data I would go ahead and do it here, adds a “sanity check” element against which you can compare the raster you create. b. Open toolboxes c. Right click – add new – toolbox d. Right click – add new – model e. Open file – model properties i. Name your model ii. Set environment (cell size, extents) f. Drag in tools i. Clip ii. Feature to raster iii. Euclidian distance iv. Line density v. Map algebra vi. Weighted overlay g. Drag in files and draw connections or double-click on tool to add files i. Hint: The model does function a little differently if you drag in the file from within your project vs. point directly to the file. I have found that the latter is a little safer, more transparent. The former is great for setting this up, or adding something quickly. You can always change it later by double-clicking on the blue data oval and re-setting the source. h. File – validate entire model (this will let you know if there are any pieces improperly defined that will cause your model not to run right) i. File – save!! j. File – Run!! k. Add the data layers you just created to the project, since they won’t show up automatically (you can probably automate this step, but I haven’t bothered to figure out yet) 3. The “cool green infrastructure model” a. Is actually three submodels – what’s in ‘em? i. Model100agr – agriculture ii. Model100cons – conservation iii. Model100rec – recreation iv. Pieces: I just used this to test sections of the model before adding them to the whole thing and potentially messing stuff up
57
Appendix 4: Case studies Cuyahoga County Greenspace Plan............................................. 59 Florida Statewide Greenways Planning Project .............................. 65 Greenprint for King County, Washington ...................................... 93 City of Kinston—Lenoir County, North Carolina ............................. 119
58
Cuyahoga County Greenspace Plan
1948 Land Use
2002 Land Use
The first county in Ohio to approach build-out, Cuyahoga County’s focus is shifting from greenfield development to redevelopment (and re-greening) of existing urban lands and the protection and enhancement of remaining open space. The greenspace plan aims to create a future where: natural places are an integral part of daily life; natural processes are visible and instructional; and waterfronts are cared for and accessible. Though far more rural in character than Cuyahoga County, planning for ecologically functional and recreationally valuable lands within and surrounding Worcester County’s urban and urbanizing areas is important. Principles • builds off of the County's unique geography and natural history, • emphasizes the environmental, community, and economic importance of greenspace, • intends to inspire decision makers to make greenspace a priority in the community, • promotes connecting neighborhoods in the county to greenspace and the county's natural resources, and • encourages the "regreening" of the more urban portions of the county to make them more desirable places to live. Process and Plan Elements The county mapped natural and man-made features, focusing on geography and major river valleys, and collected ideas from municipal master plans, recreation plans, environmental plans and at public meetings to create open space plan including the following elements: • • • •
System of natural corridors: “broad brush approach” based on geography, corridors planned around major river valleys, the Rocky, Cuyahoga and Chagrin Rivers Countywide trail system: bring trails to where the people live instead of requiring residents to drive out to nature; complete “emerald necklace” around county; facilitate connections to activity centers Preservation of scenic views: identification of major ridges and viewsheds Protection and restoration of critical natural areas: lack data on sensitive areas outside of areas already protected
59
•
•
Greening neighborhoods and property stewardship: increase tree-cover and improve landscaping to mitigate the urban heat island effect (and therefore energy use, particularly during the summer), improve aesthetics and property values, and create an overall healthier environment Public awareness and education: focus on education at all levels and for elected officials, the creation of public-private partnerships and ongoing outreach programs on environmental topics
Greenprint The final map for Cuyahoga County’s greenspace, The Greenprint, identifies the following features and opportunities: • Existing parks (larger than 2 acres) • Greenspace corridors within which to focus management and restoration for environmental health and recreation • Rivers, streams and creeks within culverts • Boulevards and major streets that connect community centers and should be targeted for re-greening • Community assets: shopping centers, institutional uses, colleges and universities, employment centers • Existing and potential trails (including utility easements, greenspace corridors, abandoned rail corridors, institutional properties and rights-of-ways) Towpath Trail Extension Though lacking detailed information on planning methods (mostly likely because studies were conducted by private consultants), planned extensions of the historic Towpath Trail provide an example of uniting environmental and recreational goals with industrial land uses. The proposed plan includes “environmental regeneration” sites for projects to stabilize hillsides, restore wetland pockets, and improve surface and ground water quality along the old industrial corridor. Greenspace Funding Sources The county developed a detailed funding sources chart, which is included on the following pages. Development of a similar chart would greatly benefit implementation of a green infrastructure plan in Worcester County. References Cuyahoga County Planning Commission. Cuyahoga County Greenspace Plan. Cleveland, Ohio. (http://planning.co.cuyahoga.oh.us/green/index.html) Cuyahoga County Planning Commission. Towpath Trail Extension Alignment and Design Study. Cleveland, Ohio. (http://planning.co.cuyahoga.oh.us/towpath)
60
Program FEDERAL GOVERNMENT Environmental Protection Agency Brownfield Assessment Program
X
X
Brownfield Cleanup Revolving Loan Funds
X
X
Brownfield Direct Cleanup Grants
X
X
Environmental Education Grants 66.951
X
Environmental Job Training and Development Pilots
X
X
Environmental Justice Grants 66.604
X
Evironmental Protection Consolidated Research 66.500
X
Wetlands Protection: Develop. Grants 66.461
X
X
X
X
Environmental Research Grants
X
Great Lakes Program 66.469
X
Source Reduction Assistance 66.717 Superfund Innovative Technology Evaluation Program 66.807
X
X
X
X
X
X
X
X
X
X
X
Superfund Technical Assistance Grant 66.806
X
Targeted Watershed Initiative 66.439
X
X
X
X
X X
X
X
Department of the Interior North American Wetlands Conservation Fund 15.623
X
Rivers, Trails and Conservation Assistance 15.921
X
National Coastal Wetland Conservation Grant
X
X
X
Land & Water Consrv. Funds (apply through ODNR)
X
X
X
X
X
X
X
Dept. of Transportation - FHWA Transportation & Community & System Preservation
X
National Scenic Byways Program
X
Public Lands Highways Discretionary Program
X
X X
X
X
X
X
X
X X
Department of Agriculture Conservation Reserve Program 10.069
X
Watershed Protection and Flood Prevention 10.904
X
Watershed Surveys and Planning 10.906
X
X
X
X X
Wetlands Reserve Program 10.072
X
Wildlife Habitat Incentives Program
X
Challenge Cost-Share Grant Program
X
X
X
X
X
X
X
X
Department of Commerce - NOAA Habitat Conservation (PDF) 11.463
X
X
Dept. of Housing and Urban Development Community Outreach Partnership Center 14.511
X
Corp. for National and Community Service
61
Americorps
X
Retired & Senior Volunteer Program
X
X
X X
STATE GOVERNMENT Ohio Department of Natural Resources Greenworks
X
Ohio Bicentennial Legacy Tree Planting Program
X
Clean Ohio Trails Fund
X
Coastal Management Assistance Grants
X
Land & Water Conservation Funds
X
X X
X
X
Natureworks
X
Recreational Trails Program
X
Watershed Coordinators
X
X
X
X
X
X X
X
X
X
X
X
X
X
X X
Boating Infrastructure Grant Program
X
X
Boating Safety Education
X
Clean Vessel Act
X
Cooperative Public Boating Facility Projects
X
Grassland Restoration: Pastures-to-Prairies
X
Wetland Restoration
X
X
X
Ohio Department of Transportation Bicycle/Pedstrian Program (STP Funds)
X
Transportation Enhancements (PDF) (STP Funds)
X
X
X
X
X
Ohio Deptartment of Development Brownfields/Clean Ohio Fund
X
Urban and Rural Initiative Program
X
X
X
X
Ohio EPA Section 319 - Nonpoint Source Pollution
X
Water Pollution Control Loan Fund Water Resource Restoration Sponsor Program
X
X
X
X
X
X
X
X
X
Voluntary Action Program (Brownfields Cleanup)
X
Pollution Prevention Loan Program
X
X
X
X X
X X
Ohio Environmental Education Fund
X
X
List of Other Grant, Loan and Tax Incentive Progs. Ohio Lake Erie Commission Lake Erie Protection Fund
X
X
X
X
X
LOCAL AREA GOVERNMENTS NOACA Transportation Enhancements (STP Funds)
X
X
X
X
X
Metroparks Agency budget
X
X
X
X
X
X
X
Cuyahoga Soil and Water Conservation District Non-point Source Pollution Education Grant
X
Urban Streams Programs
X
X
X
X
Cuyahoga County Clean Ohio Conservation Program - NRAC
X
X
X
62
Community Development Block Grants Brownfields Redevelopment Fund Urban Parks & Recreation Recovery Program
X
X
X
X
X
X
X
X X
PRIVATE & NON-PROFIT Cleveland Foundation Grants
X
X
X
X
X
X
X
X
Gund Foundation Environment Grants
X
X
X
X
Bikes Belong Grant Proposal (TEA-21 funds)
X
X
Captain Planet Foundation Environmental projects for youth and children
X
Chevron Conservation Awards
X
X
X
X
Conservation Fund Kodak American Greenways Award
X
Great Lakes Revolving Loan Fund
X
X
X
X
X
X
X
Environmental Support Center Environmental Loan Fund
X
X
eSchool News Online Links to various environmental education grants
X
Ford Foundation Community and Resource Development
X
X
X
The Foundation Center Search engine for locating grants Great Lakes Commission Soil Erosion & Sediment Control
X
X
X
X
X
X
X
X
X
X
X
X
Great Lakes Protection Fund Environmental Endowment Ittleson Foundation The Environment Land Trust Alliance Midwest Program
X
X
X
X
National Endowment for the Arts Challenge America: Access to the Arts
X
National Gardening Association Youth Garden Grants Program
X
National Fish and Wildlife Foundation Bring Back the Natives Grant Five Star Restoration Challenge Grants
X
X
X
X
FMC Corporation Bird and Habitat Conserv Fund
X
X
Migratory Bird Conservancy
X
X
Pulling Together Initiative (Weed Management) Wildlife Links (Golf Courses)
X
X X
X
X
X
X
X
X X
X
63
National Tree Trust Partnership Enhancement Program
X
X
X
X
The Nature Conservancy Education & technical assistance
X
X
X
North American Association for Environmental Educ. Links to various environmental education grants
X
Rails to Trails Conservancy Education & technical assistance
X
X
Richard King Mellon Foundation American Land Conservation Program
X
X
River Network Watershed Assistance Grant
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Trust for Public Land Education & technical assistance
X
The Pew Charitable Trusts Grant program DuPont Corporate Contributions Program Grant program
X
X
Charles Stewert Mott Foundation Grant program
X
X
X
Surdna Foundation Grant program
X
X
Great Lakes Aquatic Habitat Network and Fund Grant program
X
X
X
X
X
X
64
Florida Statewide Greenways Planning Project Beginning in 1995, the Florida Department of Environmental Protection (DEP) worked with the University of Florida GeoPlan Center, the Florida Greenways Commission and the Florida Greenways Coordinating Council to plan Florida’s Statewide Greenways System, consisting of a network of ecological hubs and corridors as well as a recreational and cultural network. While the statewide green infrastructure in Maryland has already been identified by the MD Department of Natural Resources, and this network has been shown to identify areas of statewide importance in Worcester County, refining the methodology in light of Florida’s ecological and cultural criteria will help develop local green infrastructure plans, such as for the northern portion of the County, which faces significant development pressure. Goals • Respect landscape dynamics, spatial and temporal changes. • Hazard mitigation. • Creation of a statewide recreational network. Network elements • Landscape linkages. Large scale elements connecting large parks, refuges, or forests. Often contain historical or archaeologically important sites, and can be used for recreation. • Conservation corridors. Smaller scale corridors for flora and fauna that may be used as sites for resource-based recreation. • Greenbelts. Large natural or agricultural areas surrounding urban areas that act to preserve agriculture and control urban growth. • Recreational corridors. Sites appropriate for passive or active recreational use by residents and tourists. • Scenic corridors. Corridors, often roads, protected for their aesthetic qualities. • Utilitarian corridors. Corridors established or used for utilities infrastructure (power lines, pipelines, and rights-of-way), but which may also function as conservation or recreation corridors. • Trails. Designated routes connecting natural, historic or recreational sites. Analysis and Classification—GIS decision support model* • First step in deciding on physical plan, and is subject to public review and amendment. • Uses an integrated landscape approach to: select linked reserves and preserve ecological function; identify trailheads, corridors and cultural-historic sites for public access. • Cellular analysis using multiple thematic data layers (180m grid, used GRID program, part of ArcGIS). o Rule-based suitability analysis (i.e. the trail will not be located within 100m of major urban areas, the trail will be allowed to cross wetlands) o Single utility assessment (SUA): need to define “utility” units to compare across layers (i.e. can’t “add” presence of wetland and forest type to achieve a usable number) o Weighted multiple utility analysis (i.e. Bike Trail Suitability (MUA) = (Landuse (SUA) x .25 + View (SUA) x .5 + Slope (SUA) x .25) *
Detailed methodology is included on the pages following this summary. 65
o
•
•
Calculated least cost path: cost surface is the opposite of the suitability surface (i.e. the more “suitable” a grid cell is for the activity, the less it “costs”)
Step 1: Establish modeling parameters o Ecological network (including working lands) and cultural-historic network Ecological Network Goal. To design an ecologically functional and dynamic system that conserves native ecosystems; restores connectivity; and allows biota and natural processes to adapt to environmental changes. Trails/Cultural-Historic Network Goal. To provide public access to and promote appreciation, support and conservation of the region’s natural, cultural and historic features, and to provide diverse opportunities for passive and active outdoor recreation. o Three broad landscape types: riverine and wetland; coastal; upland o Five trail types: hiking, offArcGISoad biking, equestrian, multi-use, paddling Step 2: Model ecological network o Significance and compatibility: priority, significant, other areas based on previous state designation for conservation, presence of native vegetation, and presence (absence) of development o Identify hubs: First: removed unsuitable parcels (high road density, developed, negative ecological value) Second: applied acre size filter (didn’t include parcels smaller than 5000 acres as a state level hub) Third: filled gaps by adding native habitat and other suitable lands Fourth: categorized hubs from step 3 as: riverine, coastal or upland o Identify linkages: First: created suitability surface, each cell designated as “appropriate” or “inappropriate” for linkage to pass through Second: calculated least cost path
66
Third: optimized path to include contiguous native and non-native lands 3: model trails/cultural-historic network Trailheads: provide access and are located on public land with parking and restroom facilities Trail corridors: characteristics meet expectations of user groups (i.e. mountain bikers) Identify the following: cultural-historic sites; existing and proposed trails; points of interest; other linear features (i.e. roads, rail lines); ecological lands classified for trail suitability
•
Step o o o
Public participation • Phase 1: develop goals, objectives, and collect data: workshops and mapping exercises throughout state • Phase 2: review of preliminary GIS model results (examples of changes made include removing areas modeled as suitable but recently developed, identification of additional linkage opportunities) • Phase 3: implementation Implementation • Developed Five-year Florida Greenways System Implementation Plan, including benchmarks to monitor performance. • Created partnerships between government agencies and private sector to carry out plans. Funding sources • The Preservation 2000 program in Florida provided $3 billion over a 10 year period to acquire land for conservation, resulting in the purchase of over 800,000 acres of land critical to the Florida Greenways System. • Used funding from the Transportation Enhancements Program of the federal Intermodal Surface Transportation Efficiency Act (ISTEA) for greenway development along transportation corridors. References Florida Department of Environmental Protection and the Florida Greenways Coordinating Council. 1998. Phase II Final Report: Statewide Greenways System Planning Project. University of Florida GeoPlan Center. 1999. “Florida Statewide Greenways Planning Project”. (http://www.geoplan.ufl.edu/projects/greenways/greenwayindex.html)
67
Phase II Final Report Statewide Greenways System Planning Project
Section 3
The Florida Greenways GIS Decision Support Model This section describes the Florida Greenways Decision Support Model (hereafter referred to as the Model). It includes a section on GIS principles applied in the Model and the process used in the development and application of the Model, including technical review. A detailed description of the Model itself follows and the section concludes with a description of results. The Model was named a Decision Support Model because it was intended to be the first step in delineating or deciding upon a physical plan for the Statewide Greenways System, the results of which were not to be accepted verbatim. As described in the Sections 4 and 5, the Decision Support Model Results were modified in subsequent steps to produce a recommended physical plan.
GIS Principles Applied in Modeling the Statewide Greenways System The University used a geographic information system (GIS) model to define and identify the best locations for greenways within the state. The Model was developed based upon basic raster or cellular geographic information analysis. The principles of geographic analysis are well documented in many geographic information systems texts (Borrough, 1981: Tomlin, 1990: Worral, 1991: Huxhold, 1991: Davis, 1996: etc.), and will be briefly discussed here to provide a basic understanding of technology utilized for the Florida Greenways project. GIS Basics A GIS is defined as a collection of information technology, data, and procedures for collecting, storing, retrieving, manipulating, analyzing, and presenting maps and descriptive information about features that can be represented on maps (Huxhold, 1995: p 3). GIS data representing surface features are organized into individual data layers. Data layers within a GIS are normally organized to represent individual themes, such as topography, habitat or hydrology. Multiple thematic data layers can then be overlaid, one on top of the other, to produce a map or to perform some analysis, such as locating all the residential property adjacent to water and within a mile of a major roadway (Figure 3.1).
Figure 3.1: Multiple data layers are used to represent the surface of the Earth (from Tomlin, 1990: p7).
There are two types of geographic information systems, vector and raster. A vector geographic information system represents surface features using points, lines, or area objects. For example, property parcels could be represented by an area object, called a polygon, while streams would be represented by line objects (Figure 3.2). A point object would be used to represent wells. Each object type in a vector based system also contains attribute information about each surface feature, such as the name of the interstate highway, the type of surface pavement used for the highway, or the classification of the highway by average daily trip capacity.
21
Phase II Final Report Statewide Greenways System Planning Project
Figure 3.2: Vector GIS are constructed of points, lines, and polygon objects. These objects are often referred to as line drawing objects because the GIS is constructed of traditional mapping objects.
Raster or cellular GIS data are constructed by organizing surface features into uniform cells (see Figure 3.3). Each raster cell then represents a specific area on a surface, and contains attributes about the surface. For example, habitat data can be represented by raster cells that cover a specified area and contain the type of habitat found within each cell unit. In the Greenways Model, each cell used to represent spatial features was 180 x 180 meters and contained a specific attribute characteristic about the surface, such as habitat type, landuse category, or roadway location.
Figure 3.3: Raster based GIS uses cells to represent surface features (from Tomlin, 1990: p 9).
Cellular models have the ability to look at cells from many data layers (Figure 3.4) and to use the advantage of common cell location, between layers, to accomplish complex geographic or spatial analysis.
Figure 3.4: Raster models have the ability to access cells, having the same spatial location, within many layers of data. This is perhaps the most powerful advantage for raster models when compared with vector models (from Tomlin, 1990: p 39).
22
Phase II Final Report Statewide Greenways System Planning Project Conversion of GIS Features into a Raster Data Structure for Modeling Modern GIS software allows for the conversion of vector data, i.e., the point, line, and polygon objects, into raster data rapidly and with little difficulty. Figure 3.5 shows the conversion of line features, such as a river, into cells for use within a GIS raster model. Each layer in a raster data model has cells which correspond to cell locations in other raster data layers. Once data have been converted into a raster data structure, modeling activity is possible by utilizing cell attributes for the cells across many data layers. Computer programming can then be applied to the model to determine the suitability of specific cell locations for a particular use, such as the locations of all cells suitable for a backcountry hiking trail.
Figure 3.5: Conversion of vector GIS data to raster GIS data is rapidly accomplished with modern GIS software (from Davis, 1996: p 106).
Advantages of Cellular GIS Models Cellular GIS data have many advantages for surface modeling. Database size can be greatly reduced when data are stored as cells, rather than as polygons, because of raster compression techniques. Modeling with Map Algebra, the ability to write algebraic equations for cells between data layers, is more flexible and complete than overlay analysis techniques in vector GIS. Additionally, the ability to develop a suitability surface with cellular GIS data allows for the more efficient modeling of site locations, such as the location of trail corridors (Figure 3.6).
Figure 3.6: The development of suitability surfaces provides raster GIS models with a powerful tool for identifying locations based upon best suitability (from Tomlin 1990).
23
Phase II Final Report Statewide Greenways System Planning Project Spatial Analysis Basics Cellular based geographic information analysis is accomplished by the utilization of basic functions. These functions can be aggregated into three categories, local, focal, and zonal. Each of these function groups was employed in the Greenways Model. Local functions are functions that act on only one cell per layer (Figure 3.7). Local functions are most commonly used to calculate mathematical equations between layers. The output from a local function is a new raster layer with the result of the query or mathematical analysis from the input layer(s). The simplest local functions reclassify data from one raster layer into a new layer with the data presented in a different classification. An example would be the reclassification of habitat data into a more simplified wetland/upland split. However, more complex reclassifications can be performed with local functions. A local function could be used to query many GIS raster layers in order to determine the best location for commercial activity (Figure 3.8). This type of reclassification is called suitability analysis. The reclassification for suitability can be very complicated, requiring many queries between multiple raster GIS layers.
Figure 3.7: Local cellular analysis functions act one cell per layer and produce a new raster data layer with the results of the analysis from the input layers (from Tomlin, 1990).
Figure 3.8: The darkest regions indicate the areas which are most suitable for the location of new commercial establishments (from Tomlin, 1990: p 81).
24
Phase II Final Report Statewide Greenways System Planning Project Focal functions allow the aggregation or identification of data within a specified area, called a neighborhood (Figure 3.9). The output from a focal function is a raster layer where the individual cells have summary information within the neighborhood of interest from the input layers. For example, the diversity of a selected neighborhood around a specific cell in a raster layer can be calculated by counting the types of habitat in the neighborhood and providing an output layer where each cell has the total number of different habitats inside the neighborhood.
Figure 3.9: Focal cellular analysis acts by summarizing data within a specified neighborhood and produces a new raster data layer with the results of the summarization (from Tomlin, 1990: p 97).
A zone in raster GIS are all cells in the data layer with the same attribute value. Raster regions are zones that have the same attribute data that are isolated in space (Figure 3.10). Zonal functions aggregate or summarize data from many raster layers by using specified zones from a separate raster layer containing the zone data. For example, by using zones from a city raster layer a zonal function could calculate the population density from a raster layer of census population. Additionally, zonal functions could be utilized to calculate the acreage of particular greenways hubs, which could then be utilized to remove hubs with less than a specified acreage minimum. Zones can be reduced to smaller regions by specifying that each zone be comprised of contiguous cells with the same zonal characteristics.
Figure 3.10: Two regions occurring within one zone are spatially distinct areas with the same attribute value (from Tomlin, 1990: p 155).
25
Phase II Final Report Statewide Greenways System Planning Project Map Algebra Map algebra is the methodology for modeling in a raster GIS environment. The ability to analyze data, cell by cell, between raster layers is critical to greenways identification and location. Data analysis between raster layers can be complex (Figure 3.11), producing new raster data layers which are often used to complete additional modeling or analysis.
Figure 3.11: Data analysis between raster layers can be complex, producing new raster data layers that are often used to complete additional analysis. Layers A and E are manipulated to produce a new raster layer G. Layer G is then used with layers C and F to produce new raster layer H (from Tomlin, 1990: p 52).
The analysis between raster layers is accomplished by using map algebra. Additionally, map algebra provides GIS modelers with the ability to perform complex cell based mathematics that may be replicated by peers, given the same input data. Figure 3.12 shows the map algebra relationships developed for an example raster model.
Figure 3.12: Map algebra is used to create raster layer G by dividing the cell values in layer E by the cell values in layer A. the new raster layer G is then added to layers F and C to produce another raster layer H (from Tomlin, 1990: p 53).
26
Phase II Final Report Statewide Greenways System Planning Project The previous model, presented in abstract, could be used to determine the suitability for locations of new commercial establishments in a selected area, or to locate new recreational for Floridas greenways network. The concepts are similar for layer analysis, only the map algebra relationships change. In general modeling, and more specifically raster GIS modeling, is a method for simplifying real world data for analysis and understanding. The GIS modeler must clearly state the data requirements for all layer relationships, identifying which surface features are of importance to the model, how those feature relationships are to be developed, and what priorities exist within the layer relationships. Once these have been identified the model development can proceed. Once processed the model can then be modified based upon the interim results. Additionally, multiple applications of the model and new modifications to the model allow the analyst to develop a greater understanding of the system and to make better planning decisions. Each individual and /or successive applications of the model program provide greater insight into the complexity of the real world process described by the model. Suitability Surface Development Suitability surface development is an application of the raster modeling basics discussed in the previous sections. A suitability surface is constructed by utilizing map algebra relationships between appropriate raster layers. For example, the selection of suitable locations for a recreational trail might be accomplished by organizing a complex map algebra relationship to accomplish the following requirements (Figure 3.13). A recreational trail suitability surface will be utilized to locate a backcountry trail, by assigning suitability measures for the following criteria: 1. 2. 3.
the trail will not be located within 1000 meters of dense urbanized areas, the trail will be allowed to cross wetlands, conservation lands, and large agricultural lands which are primarily used for forestry, and the trail will not be allowed to cross open water without crossing at an existing bridge location, nor be allowed to cross hazardous areas (e.g., bombing ranges).
Figure 3.13: Suitability surface for recreational trail location. Using the suitability surface for the location of a recreational trail, the computer identifies the best location for the backcountry trail. The first identification is for the location of bridges and the distance from roads. The second identification is for wetlands and other land use categories. The darker cell areas have a higher suitability for trail location. The model selects the most suitable trail location.
Suitability is a Mathematical Problem While the development of a suitability model in concept seems simple, in practice there exists a problem. The problem arises from the fact that you can not mathematically combine data that measure different things (i.e. physical characteristics within the environment). This can be commonly stated as you cant combine apples and oranges. For example, suppose you know that landuse, proximity to scenic views, and moderate inclined slopes are highly desired for the development of backcountry biking trails. You cant add raster data classified by land use with raster data representing elevation or slope to define suitability, mathematically the addition of these data is meaningless. The solution to this problem is to re-organize
27
Phase II Final Report Statewide Greenways System Planning Project these individual data layers into utility units of the same measure. Those utility units can then be mathematically manipulated to create a suitability surface for modeling backcountry biking trail locations or other opportunities. In other words, you must ask the same question of all the data layers. Data Measures There are four ways to measure data: nominal, ordinal, interval and ratio. Nominal measurements are not actually measurements. They are a name assignment to represent some characteristic for identification, for example landuse classifications like residential and agriculture. No math is legal for nominal data (e.g. it doesnt make sense to add landuse codes together). Ordinal measurement ranks entities relative to one another. For example the rankings for college football teams. As with nominal data no math is allowed for ordinal data, but testing for relative position is allowed (e.g., show values greater than 4). Therefore, ordinal measurements indicate that one entry in a list is better or more important than other entries in the same list. Interval measurements indicate the magnitude of difference between entities, but not between a value and zero. For example two portions of a travel route are 3 miles and 10 miles in length, respectively. The second route portion is 7 miles longer than the first, however without knowing how far the total route is the difference between route segments could be large or small. It is appropriate to add and subtract interval data, but not to multiply or divided this type of data. Ratio measurements have magnitude and that magnitude is the difference between the value and zero. Suppose the total route traveled from above was 50 miles, the difference between the second and first routes (10 3) / 50 = 0.14 represents 14% of the total miles traveled. All mathematics is valid for ratio data. Figure 3.14: Landuse classification as nominal measurement where identification is by name.
Given that some types of data measurements are mathematically limited, such as the landuse shown in Figure 3.14, the creation of a suitability surface can become complicated. Using GIS query and reclassification functions modelers can reorganize all four data measurement types to produce a surface that represents suitability. Single Utility Assignment (SUA) The assignment of utility provides a means of representing all data values with regard to the same goal, in this case, suitability for use as an backcountry biking trail. The process of transforming values of individual data layers, such as landuse or elevation, into a measure of utility is called a single utility assignment (SUA). Individual landuse categories are transformed by rank order (assigned a numerical value) to represent the single utility of landuse for backcountry biking trails. Elevation data are used to rank order slope and aspect areas with respect to the single utility for backcountry bikers. It is important to understand that the single rank order measures clearly represent utility on an interval scale. Therefore with respect to landuse reclassification for suitability assignment (Table 3.1 and Figure 3.15), a hardwood hammocks SUA of 8 is 7 SUA(s) better than a commercial services SUA of 1.
28
Phase II Final Report Statewide Greenways System Planning Project Table 3.1: Reclassification of Landuse for Single Utility Assignment Landuse Classification Commercial Services Cypress Dome Residential Regional Shopping Malls Backwater Swamp Hardwood Swamp Forested Wetlands Pasture Pinelands Hardwood Hammock Hardwood Forest
1 1 2 2 3 4 5 6 7 8 9
Trail Utility Lowest Utility
Highest Utility
Figure 3.15: Landuse Suitability Assignments (SUA) as described in Table 1.
The reclassification of slope for backcountry biking requires some intermediate processing of elevation data. Figure 3.16 shows the digital elevation model that presents elevation data using a ratio measurement scale. The elevation data are shown in feet above mean-sea-level and are transformed into percent slope (Figure 3.17) using GIS software.
Figure 3.16: Elevation represented by ratio measurement scale in feet above mean-sea-level.
29
Phase II Final Report Statewide Greenways System Planning Project
Figure 3.17: The development of a slope surface from the elevation data in Figure 3.16. Slope is represented in percent slope that is a ratio measurement.
To utilize slope data in the suitability surface the percent slope must be converted to interval measurement that reflects the bikers utility for use of slope when on a backcountry bike ride. Table 3.2, indicates the SUA(s) for slope as indicated by bikers using backcountry trails. By now it has probably become clear that utility assignments for individual layers are a subjective measure. Single utility assignments are supposed to reflect the users preference for riding on steep slopes and therefore are representative of true utility only if they reflect the user community desires. Therefore, utility assignments should be developed with public input. Figure 3.18, represents the backcountry biking communitys preference for riding on inclines as described by percent slope. Table 3.2: Reclassification of Slope Difficulty for Single Utility Assignment 1-5 6-10.1 11-15 16-20 21-25 26-30 31-35 36-40 41-45 50+
Slope percent percent percent percent percent percent percent percent percent percent
Incline low low low low moderate moderate moderate moderate moderate high
Trail View Utility 1 Lowest Utility 2 3 4 5 6 7 8 9 Highest Utility No Data (Not Allowed)
30
Phase II Final Report Statewide Greenways System Planning Project
Figure 3.18: Slope SUA(s) for backcountry bikers.
Figure 3.19, shows the distance in meters from possible views created by topographic relief. The final reclassification in this example, for single utility assignment is for distance from slopes that provide scenic views. Table 3.3 shows the reclassification distances for the SUA(s) for distance from scenic views. Figure 3.20, shows the suitability surface for the SUA(s) that indicates biker preference for adjacency to scenic views.
Figure 3.19: The distance to scenic views in meters. Scenic views are clearly indicated by the areas of variable color.
Table 3.3: Reclassification of Proximity to Scenic Views for Single Utility Assignment Distance 0-75 76-125 126-150 151-200 201-250 251-300 301-400 401-425 426+
Units meters meters meters meters meters meters meters meters meters
Trail View Utility 9 Highest Utility 8 7 6 5 4 3 2 1 Lowest Utility
31
Phase II Final Report Statewide Greenways System Planning Project
Figure 3.20: The SUA(s) for scenic view range from 1-9, with 9 indicating the location of highly prized views.
Multiple Utility Assignment (MUA) Multiple utility assignment (MUA) is used to compute the contribution made by the individual single utility assignments to the final suitability surface. Suppose, in our trail example, bikers believe view is three times more important than landuse. That is, bikers will select a view preference even if they are required (because of their selection for view) to ride across property with dense multi-family landuse that has low SUA for biking. The following equation can account for the MUA of the bike trail suitability surface by assigning importance ratios to the individual SUA(s): Bike Trail Suitability (MUA) = (Landuse (SUA) x .25 + View (SUA) x .5 + Slope (SUA)x.25)
Figure 3.21: Multiple Utility Assignment for backcountry bike trail suitability surface. The darker the color the greater the suitability for backcountry biking.
32
Phase II Final Report Statewide Greenways System Planning Project The final step in determining the best location for a backcountry bike trail is to use GIS software to determine the best route, using MUA(s), within the suitability surface. The suitability surface is converted to a cost surface by inverting. For example, a MUA of 9 equals a cost of 1 (Table 3.4). The computer calculates the total least cost path utilizing the cost surface values and distance from the start location to the end location. Figure 3.22 shows the best route for a backcountry bike trail using a cost surface based on the suitability surface shown in Figure 3.21. Table 3.4: MUA value transformation to Cost Surface Utility Ranking 1 Lowest Utility 2 3 4 5 6 7 8 9 Highest Utility
Cost Ranking 9 Highest Cost 8 7 6 5 4 3 2 1 Lowest Cost
Figure 3.22: The trail passes through the areas of highest multiple utility assignment, created from the SUAs for slope, landuse and distance from potential views.
GIS Data Sources A discussion and a description of data sources and outputs are significant enough to warrant their own section. Please see Section 6.
33
Phase II Final Report Statewide Greenways System Planning Project GIS Model Development and Application The Model was designed and evaluated in a four county Test Area and a fourteen county Pilot Project. It was then applied using Floridas five water management districts in a stepwise fashion. Maps of the results were plotted using Floridas six DEP districts and were reviewed by the public and private landowners as described in greater detail in sections 4 and 5 respectively. Test Area and Pilot Project. The Model concepts were first developed and tested in 1995 in a four-county Test Area comprised of Orange, Seminole, Volusia and Lake counties. In 1996, the Test Area was expanded to a fourteen-county Pilot Project area comprised of Brevard, Flagler, Highlands, Indian River, Lake, Marion, Okeechobee, Orange, Osceola, Polk, Putnam, Seminole, Sumter and Volusia counties. This Pilot Project area, that represents seven counties within and seven counties contiguous to the East Central Florida Regional Planning Council, was selected because it contained a wide variety of greenways features (i.e. existing and proposed conservation areas and trails) in a broad range of settings (i.e. from intensely urban to dispersed rural land uses). These varying conditions provided an excellent opportunity to test the suitability and sensitivity of the Model. The Model continued to evolve over this two-year period, with useful input from experts during technical review sessions and numerous changes in data treatment and analytical steps that improved the Models effectiveness. Water Management District Model Application Areas. Following the completion of the Pilot Project in late 1996, the University Team applied the Model statewide to identify a preliminary design for the Statewide Greenways System. Floridas five water management districts were selected as the Model s application areas because of their size and the fact that a number of important statewide and regional data layers (e.g. Florida Land Use/Land Cover databases) were already subdivided by these districts. The Model was first applied to the St. Johns River Water Management District. It was then applied sequentially to the South Florida, Southwest Florida, Suwannee River and Northwest Florida water management districts. Sub-state modeling units were used because at the time, equipment to support the Model for the entire state in one run, was prohibitively expensive. This approach, however, had the potential of creating inaccurate results, due to edge effects. For example, when focal functions were used (as described in the first part of this Section), if the neighborhood in which an analysis was conducted contained cells with no values because they fell on the other side of a water management district boundary, then the results of such an analysis would be inaccurate. This issue was addressed following the modeling of St. Johns River Water Management District in the winter of 1997 by adding a 15 km-wide strip within previously modeled water management districts to the current districts modeling area. In this way, the Model was run not only considering the ecological, recreational and cultural/historic features within the modeled water management district but also considering the features present in a 15 km-wide slice of the adjacent, water management district. Following the modeling of each water management district, the results were added to the previous results, thus assembling the results in a step-wise fashion. Model Application. Final changes were made to the Model in early 1997 as it was applied to the St. Johns River Water Management District. Although most parts of the Model were essentially complete by the start of water management district applications, a few final improvements were made in the processing of data and the linkage and trail suitability surfaces. Technical Review of Modeling Assumptions Though the University Team combined with the staff of the Office of Greenways and Trails contained individuals with a broad range of experiences in many fields, it was important to seek the advice of others in the development of the GIS Model. The complexity of the modeling process and the far ranging goals and objectives built into it meant there were numerous decisions to be made on assumptions and specific aspects of the modeling parameters and sequence. To ensure the appropriateness of decisions and to seek input on the use and application of available statewide databases, the University Team sought and received technical input in diverse forums from a variety of experts in the public, private and nonprofit sectors. In 1995, the Florida Greenways Commission and its partners provided most of the technical review. The Commission provided this leadership role through its System Design Committee. Five meetings of the Commission and or the Committee occurred, which strongly influenced the nature of the GIS Model. Details of these meetings can be found in Section 9, Appendix 4.
34
Phase II Final Report Statewide Greenways System Planning Project The technical review of modeling assumptions and process continued in 1996 and spring 1997 through a number of different forums. Participants in the review process included representatives from agencies, nonprofits and others with an ongoing interest in the design of the Statewide Greenways System. The reviews centered on a discussion of modeling techniques and results tied first to the four-county Test Area, next to the Pilot Project and, in 1997, to the St. Johns River Water Management District. Among the individuals important in the review during this phase were representatives from DEP offices and divisions including Ecosystem Management, Recreation and Parks, Environmental Resource Permitting, Mine Reclamation and others including Intergovernmental Programs, the CARL Program and Technical Services. Representatives of federal agencies participated including the USDA Forest Service, the US Fish and Wildlife Service and the National Park Service. Members of the Florida Greenways Coordinating Council and staff from their respective agencies, organizations and institutions became active members in the model development process. And finally, several key individuals with distinguished research credentials in landscape ecology, biological conservation and reserve design reviewed the model. Dr. Larry Harris of the University of Florida, to whom this report is dedicated, is one of the key individuals who were especially important to the effort. Other key reviewers included Reed Noss, Michael Soule, Steve Christman, Rob Jongman, and Lennart Hansson. A detailed description of the technical reviews that occurred during this period can also be found in Appendix 4.
Introduction
Steps in the Florida Greenways GIS Model1
The Model was developed by the University to initially identify areas, corridors and sites appropriate for inclusion in a statewide greenways system. The Model utilized an integrated landscape approach to: (1) select linked reserves and other appropriate lands to protect an ecologically functional system; and (2) identify trailheads, trail corridors and cultural-historic sites that provide public access to and promote the conservation of the systems natural, cultural and historic features. The Model was designed to be consistent with the goals and objectives described in the previous section. The Model used a cellular-based GIS software program called GRID (a module in ESRIs ARC/INFO GIS software package) to compile and analyze existing ecological and geographic data in order to model ecological and trails elements. Following a review of projected processing times with different cell sizes, the University Team chose a 180m cell (approximately 8 acres) for all modeling activities. The Model first selected ecological hubs and landscape linkages for upland, riverine and coastal native landscape units that together form the Ecological Sub-system or Network. The Model then identified a Trails/CulturalHistoric Network consisting of trailheads, trail corridors for hiking, off-road biking, equestrian, paved multi-use urban - urban and urban-rural trails and supporting cultural-historic sites. The results are compiled in the final model step. Figure 3.23. Steps in the GIS Decision Support Model.
Please refer to Section 9, Appendix 5 for detailed tables supporting this narrative description of the GIS Decision Support Model 1
35
Phase II Final Report Statewide Greenways System Planning Project Both the Ecological and Trails/Cultural-Historic Network modeling steps incorporated three fundamental sub-steps. First, ecological, recreational and cultural features were identified that could contribute to the Statewide Greenways System; next, specific criteria were applied to select features to serve as ecological hubs or recreational trailheads; and finally, the most suitable linkages between identified ecological hubs and recreational trailheads were selected. Step 1 - Establish Modeling Parameters Key to the development of the Decision Support Model was an understanding of the Systems fundamental building blocks or elements. Type and significance classify these elements. Types. The two basic types are Ecological and Trails/Cultural-Historic. Ecological elements consist of biological and physical features that occur within the Statewide System as an Ecological Sub-system or Network consisting of hubs or linkages. Working landscapes, such as pine plantations and ranchlands, were incorporated within the Ecological Network where they were complementary to the design objectives (for more discussion of working landscapes, please refer to the Florida Greenways Commissions 1994 Report). Trail/Cultural-Historic elements represent trails and cultural-historic sites that occur within the Trails/Cultural-Historic Network consisting of trailheads, trail corridors and key cultural-historic sites.
Scales of Significance. A premise upon which all of the Universitys work was based is described as follows. A Statewide Greenways System is comprised of elements of statewide, regional and local significance. All are equally important to the formation of the Statewide System, but there is a basic correlation between the management and development of the elements and the scope of responsibility of public and private entities. Elements of statewide significance are generally owned and/or managed by state and national entities, elements of regional significance are generally owned and/or managed by regional entities and local elements are generally owned and/or managed by local entities. The University Teams charge was the identification of the statewide elements that will serve as the spine or backbone of the complete system. For that reason, the Universitys modeling efforts focused on the Systems statewide elements. Although some regional elements have been incorporated within the statewide design, most regional elements and all local elements will be designed and incorporated into the Statewide System by community and regional greenways planning initiatives in later phases. Regardless of their origin, ongoing coordination of the planning and implementation of local, regional and statewide elements will be essential to the ultimate success of the Statewide Greenways System. The development of goals and objectives was also a critical step. These are fully described in Section 2, but are narrowed here to two particularly important goals: •
Ecological Network Goal. To design an ecologically functional Statewide Greenways System that: conserves Floridas native ecosystems and landscapes; restores connectivity among native ecological systems and processes; maintains the ability of native ecosystems and landscapes to function as dynamic systems; and allows the biota of native ecosystems and landscapes to adapt to future environmental changes.
•
Trails/Cultural-Historic Network Goal. To include trails in the Statewide Greenways System to provide public access to and promote appreciation, support and conservation of the Systems natural, cultural and historic features, and to provide opportunities for outdoor recreation and alternative, non-motorized transportation.
The selection of distinct modeling units reflective of different ecological characteristics and different trail user expectations provided a sound rationale for later modeling steps. For the ecological linkage process, the Model divided Floridas native ecosystems into three broad landscape units, each with distinct ecosystem types and associated ecological processes.
36
Phase II Final Report Statewide Greenways System Planning Project Riverine/Large Lake and Wetland Landscapes. Native landscapes including rivers, associated floodplains and large lakes and wetland systems, where the predominant ecological process is freshwater flow and related hydrological processes. Coastal Landscapes. Native landscapes including coastal islands and estuaries, where the predominant ecological processes are the influence of saline water, coastal currents and storms. Upland Landscapes. Native landscapes including pine and hardwood forests, dry prairies and small isolated wetlands, where the predominant ecological process is fire. Similarly, five distinct terrestrial trail types were identified for modeling purposes: Hiking Trails, Off-Road Biking Trails, Equestrian Trails and Paved Multi-Use Urban-Urban Trails and Paved/Unpaved Urban - Rural Trails. (Eventually the two multi-use trail categories were combined). Each of these trail types reflects different user expectations. Hiking Trails. Unpaved, single-use trails that provide hikers with mostly a backcountry experience. Off-Road Biking Trails. Unpaved single-use trails that provide off-road bikers with backcountry and rural experiences. Equestrian Trails. Unpaved single-use trails that provide equestriennes with backcountry and rural experiences. Multi-Use Urban to Urban and Urban to Rural Trails. Paved and unpaved multi-use trails that link urban areas and provide urban residents with access to popular natural or cultural/historic destinations. Paddling trail types can likewise be classified by user expectation. Since their locations are confined to existing water courses, the University Team did not use the Model to identify paddling trail locations; rather, it incorporated existing state designated paddling trails as well as potential future trails identified through public participation (see Section 4). Step 2 - Model Ecological Network To provide a theoretical foundation for the design of the Ecological Network as a linked conservation system, the University Team researched and developed an ecological bibliography that identified most articles, book chapters and other scholarly works on this topic. This was submitted in the Final Report of Phase I. Once the basic modeling parameters were established, the next step in the process was to model the Ecological Network. In this step, automated computer modeling functions were combined with interactive analysis and design considerations to delineate an Ecological Network that addressed the goals and objectives for the conservation of native ecosystems and landscapes. Ecological modeling was accomplished using four sub-steps. Identify Ecological Features. In the first ecological modeling step, available statewide databases were used to identify ecological landscape features that could contribute to meeting the design goal for the ecological sub-system. Once the data were assembled, selection criteria were used to generally categorize both native and non-native landscape features in terms of their significance and compatibility with ecological conservation objectives as follows: •
Priority Ecological Areas consist of ecological features and conservation designations that reflect such national and statewide importance that they are given the highest priority for inclusion and physical linkage as primary building blocks for the Statewide Greenways System. Selection criteria for Priority Ecological Areas include existing conservation lands, the highest ranked natural areas and wildlife sites statewide as identified by the Florida Game and Fresh Water Fish Commission and the Florida Natural Areas Inventory, and areas important to hydrological resources and processes.
•
Significant Ecological Areas consist of other ecological features and conservation designations of statewide or regional significance considered for inclusion as other potential locations for linkages and hubs. Selection criteria for Significant Ecological Areas include large water bodies and moderately ranked Florida Natural Areas Inventory and Florida Game and Fresh Water Fish Commission sites.
•
Other Ecological Landscape Features include lower priority ecological features that may be used to fill-in or expand ecological elements within the Statewide Greenways System.
•
Category I Lands are non-native lands with moderate ecological value such as tree plantations and rangelands.
37
Phase II Final Report Statewide Greenways System Planning Project
•
Category II Lands are non-native lands with low ecological value including orange groves and row crops.
•
Category III Lands, are non-native lands with no to negative ecological value and include urban, residential, commercial and industrial uses.
A review of the relative coverage of Priority Ecological Areas and Significant Ecological Areas demonstrates the inclusive nature of this first ecological modeling step. To ensure that nothing of potential statewide ecological importance was missed, features needed to meet just one of many selection criteria to be classified as Priority or Significant. Identify Hubs. The hubs for the Ecological Network were selected in the second modeling step. Four sub-steps were used to identify the subset of Priority Ecological Areas to serve as hubs: •
First, all unsuitable parcels were removed from consideration. Unsuitable lands included: areas of very high road density, that is areas with greater than or equal to 3 km of roads per square km; areas of inappropriate land use, including Category II and III land use cells; and areas of negative edge effects, which were considered to be areas within 180 meters of Category III land uses.
•
Second, a 5,000 acre size filter was applied. (Results to this point can be viewed in the dataset GWPEAX. Please refer to List of Datasets on Accompanying CD-ROM [page vi] and Appendix 5 for more details).
•
Third, internal gaps within the hubs were filled in and irregular edges were smoothed by adding contiguous native habitat or Category I Lands.
•
Fourth, the resulting optimized hubs were partitioned as riverine, coastal or upland landscapes. (The results of the model through this step are found on the accompanying CD-ROM).
Identify Linkages. Five linkage types were identified in the next step of the Model: riverine to riverine, coastal to coastal, upland to upland and riverine to coastal hubs and cross basin connections between selected ecological hubs. Three substeps were applied to each of these linkage types: •
First, a suitability surface was created for each linkage type. Cell-based GIS was used to divide the entire surface of the study area into a grid with 180m cells. In creating a suitability surface, each cell within a grid is identified as either appropriate or inappropriate for a linkage to pass through. Appropriate or suitable cells are further sub-divided from most suitable, to least suitable using various criteria. Inappropriate cells are classed as not suitable, that is, a linkage cannot past through them. Using riverine linkages as an example, the riverine suitability surface is as follows:
•
Riverine Ecosystem Types that are classified as Priority Ecological Areas are given the Highest Suitability.
•
Riverine Ecosystem Types that are classified as Significant Ecological Areas are given a High Suitability.
•
Riverine Ecosystem Types that are not classified as Priority or Significant Ecological Areas are given a Moderate Suitability.
•
Riverine Ecosystem Types in Areas of High Road Density and/or Negative Edge Effect are given a Low Suitability.
•
All Other Cells including Other Native Ecosystem Types, Category I, Category II and Category III Lands are classified as Not Suitable.
•
Second, the suitability surface was used to identify an optimal path of suitable cells between selected ecological hubs.
•
Third, the linkage was expanded and optimized by adding other appropriate contiguous native and non-native land covers.
•
Create Ecological Network. The final ecological modeling step was the creation of a preliminary Ecological Network by adding together hubs for all native landscape types and the linkages for all ecological linkage types. In this last step, some manual editing was done to correct visible model errors and to reflect some overall design objectives. These
38
Phase II Final Report Statewide Greenways System Planning Project results can be viewed in the dataset GWECO found on the accompanying CD-ROM see page vi and Section 9, Appendix 5 for more details. Step 3 - Model Trails/Cultural-Historic Network The next step in the Model was to design the Trails/Cultural-Historic Network. As noted in the introduction, the trails modeling steps reflected similar procedures as previously described for the modeling of the Ecological Network. The trails modeling objective was to design a recreational trails system, comprised of five trail types, with three defining features: Trailheads, that provide access to the system and are located on publicly-owned land with parking and restrooms; Trail Corridors, with physical characteristics that meet the expectations of each user group; and appropriate Cultural-Historic Sites that support both. • Identify Trail Features. The first trails modeling step was to identify features important for the design of the Trails Network. These features included: •
Existing and Proposed Trails - existing and proposed corridors that can reduce the need for new trail construction and minimize trail redundancy.
•
Points of Interest - ecological, historic, cultural and recreational features or sites that may contribute to the quality of a trail-users experience. People are more likely to use trails that lead them to places where they want to go. Identifying compatible Points of Interest for each modeled trail type will ensure that modeled statewide trails are located in areas where interesting side or spur destinations exist.
•
Other Linear Features - Many other landscape features are needed for inclusion in a trail planning model to make sure that trails are not placed in unsuitable areas but instead placed in locations that maximize the recreational, aesthetic, and conservation objectives (by co-locating where appropriate) for each trail type. Other linear features that could influence the location of trail corridors include busy roads and bridges over open water as well as utility corridors and abandoned rail lines.
•
Ecological Lands Classified for Trail Suitability - Since Floridas native ecosystems and landscapes provide the underlying network of green for the Statewide Greenways System, it is critical to ensure that Trail Elements do not negatively impact sensitive natural areas. To ensure trail corridors are not sited in environmentally sensitive locations all ecological features were divided into three trail suitability classes:
•
Class I Lands - Native lands with the highest ecological value and sensitivity such as wading bird rookeries, and therefore not suitable for statewide trail use.
•
Class II Lands - Native lands with high ecological value and sensitivity such as State Parks and National Seashores and therefore suitable for some restricted statewide trail use.
•
Class III Lands - Native lands with moderate ecological value and lower sensitivity including all other conservation lands and therefore generally suitable for statewide trail use. Select Trailheads. In the second trails modeling step, user groups and existing trails plans were used to identify trailheads for each of the five terrestrial trail types. Where possible, the identified trailheads were located on publiclyowned lands with parking accessible from a paved road. In some cases, managed conservation areas and recreation sites were used for trailheads, in other cases, the trailheads were located at publicly accessible cultural or historic sites.
To assist with the modeling, the Florida Division of Historical Resources developed a recommended list of cultural and historic sites appropriate for use as trailheads and important for incorporation into the Greenways System. A spatial representation of these sites can be found on the GWCHF dataset on the accompanying CD-ROM. Please refer to the List on page vi and Section 9, Appendix 5 for more details. Trailheads for each trail type are also represented on datasets contained on the accompanying CD-ROM. Please refer to the List on page vi and Section 9, Appendix 5 for more details.
39
Phase II Final Report Statewide Greenways System Planning Project
Identify Corridors. Corridors for the terrestrial trail types were identified in the third modeling step. The same three substeps used in the ecological portion of the Model were used to identify corridors for all trail types: •
First, a suitability surface was created for each trail type. For example, to identify hiking trail corridors, the trail suitability surface used the following criteria:
•
Bridges and underpasses/overpasses, uplands proximal to open water and existing trail of the same trail type were classified as Highly Suitable.
•
Areas inside the Ecological Network in Trail Suitability Classes II and III and proposed trail of the same trail type were classified as Moderately Suitable. Areas outside the Ecological Network in Trail Suitability Classes II & III and wetlands were classified as Low Suitability.
• •
Hazardous areas, open water, busy roads, Category III Lands excluding pipelines and utilities, and Trail Use Suitability Class I Lands inside and outside the Ecological Network were classified as Not Suitable.
•
Second, optimal paths of suitable cells were selected between all identified trailheads.
•
Third, all the trail corridors were created by adding 1 kilometer on each side of all selected optimal trail paths. Alignments for each trail type are also represented on datasets contained on the accompanying CD-ROM. Please refer to the List on page vi and Section 9, Appendix 5 for more details.
Create Trails/Cultural-Historic Network. In the last trails modeling step a preliminary Network was created by adding together the trailheads and trail corridors, plus cultural-historic sites for all terrestrial trail types. Paddling trails, identified with the help of the Florida Recreational Trails Council were added in at this point. Step 4 - Identify Recommended Statewide Greenways System Planning Area The final step in the Model was to combine the Ecological Network and the Trails/Cultural-Historic Network to represent an initial physical plan for the Statewide Greenways System.
Results In July 1997, the UF Team completed the application of the model for the entire state. The Results were divided into ecological model results and trails/cultural-historic model results. Model Results as modified by Public Comment and Landowner Comment are described in Sections 4 and 5 respectively. Summary of Ecological Model Results The ecological greenways model results incorporate approximately 57% of the state including coastal waters (Table 3.5). Open freshwater, coastal waters, existing public conservation lands, and private preserves (e.g., Audubon Society and The Nature Conservancy Preserves) compose 53% of the model results (Table 3.6). Another 10% of the model result are composed of proposed public conservation lands (CARL or SOR). Other private lands comprise 37% of the results, with Table 3.5. Area in Ecological Model Results - State of Florida
Land Area Open Water (Fresh and Salt) Total
Acres 18,800,946 4,002,865
% of state 47.0 10.0
22,803,811
57.1
40
Phase II Final Report Statewide Greenways System Planning Project Table 3.6. Composition of Ecological Model Results - State of Florida Land Use Existing Public Ownership Open Water (Fresh and Salt) Proposed Public Ownership (thru CARL, SOR, etc.) Private Ownership in Wetlands* Private Ownership in 100 yr Flood Plain* ^ Private Ownership in Uplands* Totals
Acres % of state 7,958,176 19.9 4,002,865 10.0 2,377,448 5.9 1,731,797 1,623,416 5,110,108 22,803,811
4.3 4.1 12.8 57.1
% eco area 34.9 17.6 10.4 7.6 7.1 22.4 100.0
* Acreage of private ownership in wetlands, 100 yr. flood plain and uplands is calculated as if all proposed public acquisitions are/will be completed. ^ Floodplain data is not available for Bradford, Columbia, Dixie, Gilchrist, Hamilton, Jefferson, Lafayette, Madison, Okeechobee, Taylor, and Union Counties
Narrative Description of Ecological Model Results The ecological model results contain the five largest conservation hubs in the state, which serve as the anchors of a statewide ecological greenways system and therefore are also the major destinations for the landscape linkages and corridors that tie the system together. These hubs are: the Everglades/Big Cypress complex, Ocala National Forest, Osceola National Forest-Okefenokee National Wildlife Refuge, Apalachicola National Forest, and Eglin Air Force BaseBlackwater River State Forest. Other important hubs include the Green Swamp, the Big Bend, and the upper St. Johns River and Kissimmee River basins. Heading north from Everglades National Park, Big Cypress National Preserve, and the Water Conservation Areas, there are two connections to the large ecological hubs in the rest of the state. The first, and most significant of the two, runs north of Big Cypress National Preserve through the Okaloacoochee Slough and then across the Caloosahatchee River to the Fish Eating Creek watershed and Cecil Webb Wildlife Management Area. The second connection runs north from Loxahatchee National Wildlife Refuge to Corbett Wildlife Management Area and the Loxahatchee Regional Greenways Project and then through the ranchland of western Martin and St. Lucie Counties and eastern Okeechobee County to reach the upper St. Johns River basin. Big Cypress National Preserve is also connected to the regionally significant Corkscrew Regional Ecosystem Watershed CARL Project through the Florida Panther National Wildlife Refuge and Camp Keais Strand. From the Cecil Webb Wildlife Management Area and Fisheating Creek, the model results contain a number of connections north: through ranchland to the Lower Kissimmee River; along the western side of the Lake Wales Ridge and across it in two different arms to the Avon Park Bombing Range; up the Peace River all the way to the Green Swamp; and up the Myakka River basin to connections with the Manatee, Little Manatee, and Alafia Rivers. Avon Park Bombing Range, the Kissimmee River Water Management District (WMD) Project, Kissimmee Prairie State Preserve, Three Lakes, Triple N, and Bull Creek Wildlife Management Areas, and the Upper St. Johns River WMD Project are all part of the upper St. Johns River and Kissimmee River ecological hub. The most significant landscape linkage between the north and south halves of Florida heads north from the upper St. Johns-Kissimmee hub. The model output follows the Econlockhatchee and St. Johns River basin to Volusia County. While one connection continues to follow the St. Johns River, the more significant linkage runs through the swamps and pinelands of central Volusia County including Tiger Bay State Forest to the Ocala National Forest. Another linkage follows the Kissimmee River, runs through the Reedy Creek watershed, and then crosses the Lakes Wales Ridge to reach the Green Swamp.
41
Phase II Final Report Statewide Greenways System Planning Project The Ocala National Forest (ONF) hub includes the Wekiva Greenway project, the Ocklawaha River Basin, the middle stretch of the St. Johns River, and pinelands around ONF that have been identified as important additional Florida black bear habitat. Various linkages radiate north and west from the Ocala National Forest hub. The most significant feature, and probably the most important landscape linkage in Florida, runs north from the Ocala National Forest to the Osceola National Forest-Okefenokee National Wildlife Refuge hub through the Cross-Florida Greenway and Etoniah Creek CARL projects, Camp Blanding Military Site, the New River Swamp and Raiford Wildlife Management Area, and the Butler Wildlife Management Area (Type II). This landscape linkage would protect a connected and integrated conservation network running from the Wekiva River basin just north of Orlando to the Okefenokee National Wildlife Refuge in northern Florida and southern Georgia. North of Camp Blanding Military Site the model output also continues north through Jennings State Forest and pine plantations to the Nassau River and St. Marys River watersheds. To the east of the Ocala-Osceola landscape linkage, the model output also continues to follow the St. Johns River to its mouth while incorporating adjacent wetlands and intact uplands along the river wherever possible. East of the Ocala National Forest and the St. Johns river there is another linkage heading north that incorporates swamps and pineland from Tiger Bay State Forest to Durbin Swamp southeast of Jacksonville. The Ocala National Forest is tenuously connected to the Withlacoochee River to the west through the Cross-Florida Greenway. Northwest of the Ocala National Forest, a regionally significant hub includes Paynes Prairie State Preserve, Lochloosa Wildlife Conservation Area, and the Newnans Lake CARL project, which is connected to ONF through Orange Creek. The Paynes Prairie hub is also connected to the Ocala-Osceola landscape linkage to the east through the Ordway Preserve and Sante Fe Swamp. The Green Swamp hub is connected to the regionally significant Hillsborough River Greenway project to the southwest. The most significant ecological linkage included in west-central Florida follows the Withlacoochee River basin north out of the Green Swamp with connections to the Withlacoochee State Forest, Annutteliga Hammock CARL Project, and Chassahowitzka National Wildlife Refuge, Cross-Florida Greenway, Goethe State Forest, and the Gulf Hammock Wildlife Management Area. The Goethe State Forest can be considered to be the beginning of the vast Big Bend ecological hub. The Big Bend includes coastal ecosystems and many coastal public conservation areas, large swamps and shrub wetlands such as Mallory Swamp and San Pedro Bay, and extensive pine plantations. The Paynes Prairie hub is connected to the Big Bend through the Flemington forest and Williston sandhills, the Watermelon Pond CARL Project, and the Sante Fe River. The Osceola-Okefenokee hub is connected to the Big Bend primarily by following the Suwannee River as well as tributaries in the Suwannee River watershed including the New River, Olustee Creek, and the Santa Fe River. The Big Bend hub is connected to the Red Hills landscape north of Tallahassee through the Aucilla and St. Marks River watersheds. A more significant linkage connects the Big Bend hub to the Apalachicola National Forest hub through the St. Marks National Wildlife Refuge and the Wakulla Springs and River watershed. From the Apalachicola hub, the model output follow the Ochlocknee River to the Red Hills and the Apalachicola River basin and the Chipola River to the Georgia and Alabama borders. The last major landscape linkage in the model output connects the Apalachicola hub to the Eglin Air Force BaseBlackwater River State Forest hub in the western panhandle. This linkage starts west of the Apalachicola River at several points and incorporates creeks, forested wetlands, and pine plantations to reach the Ecofina Creek watershed and the Sand Mountain CARL Project. From there the linkage continues west and includes sandhills, pine plantations, and lower Holmes Creek to reach the Choctawhatchee River, and then passes through more pinelands and swamps to the Eglin Air Force Base-Blackwater River State Forest hub through the Shoal and Yellow River watershed and directly to Eglin Air Force Base. The model output also follows the Choctawhatchee River south and includes connections to Pine Log State Forest, Point Washington State Forest and CARL Project, and Topsail Hill State Park. The Eglin-Blackwater hub is then connected to the Escambia River and Perdido River basins through various creek systems and pine plantations, and much of the floodplain of these two rivers are included in the model output.
42
Phase II Final Report Statewide Greenways System Planning Project Throughout the state the ecological greenways model results also include significant coastal water bodies with adjacent wetlands and intact uplands with various degrees of connectivity. Starting in south Florida, the model result includes the water (and keys) within the Key West, Great White Heron, and Key Deer National Wildlife Refuges. The next feature north includes Florida Bay in Everglades National Park, Biscayne Bay National Park, and the Ten Thousand Islands and Cape Romano Aquatic Preserves, and Rookery Bay National Estuarine Research Reserve and Aquatic Preserve. On the east coast the model output incorporates much of the Indian River Lagoon System from the Loxahatchee River to Mosquito Lagoon and then through the Halifax River past Daytona. Some of the larger conservation features included in the Indian River lagoon system include the Indian River and Banana River Aquatic Preserves, Archie Carr Wildlife Refuge, Sebastian Inlet State Recreation Area, Merritt Island Wildlife Refuge, and the Canaveral National Seashore. Further north the model results include the Matanzas, Tolomato, and Guana Rivers with conservation features such as Washington Oaks State Gardens, Pellicer Creek Aquatic Preserve, Faver-Dykes State Park, Fort Matanzas National Monument, Anastasia State Recreation Area, Moses Creek Water Management Conservation Area, and Guana River State Park and National Research Reserve. The last coastal feature on the east coast includes the vast interconnected saltmarshes with some coastal islands associated with the mouths of the St. Johns, Nassau, and St. Marys Rivers. Significant conservation areas and projects include the Timucuan Ecological and Historic Preserve, Big and Little Talbot State Parks, Amelia Island State Recreation Area, and Fort Clinch State Park. On the Gulf Coast north of Rookery Bay and Naples the first significant coastal feature includes Estero Bay Aquatic Preserve and Buffer, San Carlos Bay, Pine Island Sound Aquatic Preserve, Matlacha Pass Aquatic Preserve, and Charlotte Harbor. Other conservation features include Matlacha Pass, Ding Darling, and Pine Island National Wildlife Refuges, Cayo Costa State Park and Preserve, and Charlotte Harbor State Buffer Preserve. The next feature is Tampa Bay and waters to the south and north including parts of Sarasota Bay, Clearwater Harbor, and St. Joseph Sound with conservation features such as the Terra Ceia CARL project, Egmont Key State Park, Pinellas County Aquatic Preserve, Caladesi Island State Park, Honeymoon Island State Recreation Area, and Anclote Key State Preserve. The next coastal feature north includes the saltmarshes, creeks, river mouths, and coastal hammocks all the way from the Chassahowitzka National Wildlife Refuge to the St. Marks National Wildlife Refuge and Apalachee Bay. Numerous other conservation designations are included, some of which are the Crystal River State Buffer Preserve, Cross Florida Greenway, Waccasassa Bay State Preserve, Cedar Key Scrub State Reserve, Lower Suwannee National Wildlife Refuge, Big Bend Seagrasses Aquatic Preserve, various Wildlife Management Areas, and Econfina River State Park. The next feature is St. George Sound and Apalachicola Bay including Dog Island Preserve, St. George Island State Park, Cape St. George State Reserve, St. Vincent Island National Wildlife Refuge, and the Apalachicola National Estuarine Research Reserve. Further west the model output includes St. Joseph Bay including St. Joseph Bay State Park and Buffer Preserve Project; St. Andrew Bay and Sound including Tyndall Air Force Base and St. Andrews State Recreation Area; Choctawhatchee Bay; Santa Rosa Sound including Gulf Islands National Seashore; Pensacola Bay including Gulf Islands National Seashore, Ft. Pickens State Park, Garcon Point WMD Project; and Perdido Bay and Big Lagoon including Big Lagoon State Recreation Area and the Perdido Pitcher Plant CARL Project. Summary of Trails/Cultural-Historic Model Results Routes for four terrestrial trail sub-networks were derived by the Model: hiking, off-road biking, equestrian and multi-use. As previously discussed, paddling trails were not modeled since the options for their location are highly restricted. Separate rather than combined corridors were identified for hiking, off-road biking and equestrian trails, for several reasons. First, trailheads for each trail type are often different, especially for equestrian trails because of the infrastructure needed to support them. Second, while great progress is being made with developing shared use of trail corridors and even treads, there are still conflicts and dissatisfactions that arise from shared use. It is anticipated that through greenways implementation, many corridors if not treads may ultimately become shared, but since the Model was intended to identify the best of all possible worlds, distinct corridors for hiking, off-road biking, and equestrian trails were modeled. These corridors were designed to identify backcountry and rural routes through native and agricultural landscapes.
43
Phase II Final Report Statewide Greenways System Planning Project Table 3.7. Trail Model Results – State of Florida
Trail Type
Multi-Use Hiking Off-Road Biking Equestrian
Existing Corridors (mi.) 8.4 314.3 0.3
0.5 34.3 0.0
Propose d Corridor s (mi.) 43.9 4.6 0.0
4.3
0.4
0.0
%
Corridors of Opportunity (mi.)
%
%
Multi-Use Connector Existing
Multi-Use Connector Proposed
Multi-Use Connector Opportunity
Total (mi.)
2.5 0.5 0.0
1,682.1 598.1 1121.5
97.0 65.2 95.4
53.6 0.0 53.6
N/A N/A N/A
N/A N/A N/A
1,734.4 916.9 1,175.4
0.0
1132.0
99.6
0.0
N/A
N/A
1,136.3
% Represents percentage of trail corridors contained in these results found in existing and proposed corridors and corridors of opportunity. The remaining corridors to total 100% are distributed among the various connectors.
Trail Percentage 100% 80% 60% 40% 20% 0%
Corridors of Opportunity
Proposed Corridors
Equestrian
Off-Road Biking
Hiking
Multi-Use
Multi-Use Connector Existing
Existing Corridors
Trail Type
Trail Length 2000 1500
Corridors of Opportunity
Miles 1000 Multi-Use Connector Existing Equestrian
Off-Road Biking
Hiking
0
Multi-Use
500
Proposed Corridors Existing Corridors
Trail Type
Table 3.7 presents a summary of the results of the trails modeling. A modeled trail route can be comprised of 6 different types of segments: • Existing single use trail of the same type • Proposed single use trail of the same type • Opportunity single use trail of the same type • Existing multi-use trail (called a connector) • Proposed multi-use connector • Opportunity multi-use connector To qualify as a proposed trail, a segment has to be identified in a governmental program, e.g. State Rails to Trails or as a local government project. The Corridors of Opportunity and Multi-use Connector Opportunities were segments newly identified through the modeling process. Almost 2,000 miles of multi-use trail were identified by the model and approximately 1100 miles of off-road biking and equestrian trails. The model identified a little over 900 miles of hiking trail. More than half of the routes identified for all four
44
Phase II Final Report Statewide Greenways System Planning Project trail types will have to be newly developed. This is less so for the hiking trail route, since it closely follows the corridor of the Florida National Scenic Trail, but gaps in the existing route, and an alternate route east around Orlando still require the addition of approximately 600 miles or more than 65% of new trail. Only off-road biking had any corridors coincident with multi-use trails (3%). Narrative Describing Trails/Cultural-Historic Model Results The hiking corridors identified by the Model are very similar to the existing (and proposed) alignment of the Florida National Scenic Trail (FNST). This is not surprising since the modeling approach encouraged use of existing trails wherever appropriate. The modeled corridor extends from the Gulf Islands National Seashore near Pensacola to the Pinecrest trailhead in Big Cypress National Preserve. It includes two options through the central Florida area, one to the east of Orlando and one to the west. The eastern leg does not follow the existing alignment of the FNST, rather it moves north along the St. Johns River Basin and turns west into the Ocala National Forest north of Deland. Trailheads identified for inclusion in the hiking corridors of opportunity are listed in Section 9, Appendix 5 and represented in the dataset called GWHTH on the accompanying CD-ROM. The hiking trail route identified by the Model is represented in the dataset called GWHTR on the CD-ROM. Off-road bicycling corridors begin in the Panhandle with Blackwater River State Forest and end in Broward County at Markham County Park. Three additional off-road biking trailheads were selected for use in the Model south of Broward, John Pennekamp Coral Reef State Park, Bahia Honda State Park and Fort Zachary Taylor State Historic Site; but because of the way in which the suitability surface for off-road biking was constructed the Model was unable to find a path to link these additional trailheads. As with the hiking corridors, the off-road bicycling corridors split around Orlando suggesting an eastern and a western leg. Trailheads identified for inclusion in the off-road bicycling corridors are listed in Section 9, Appendix 5, and are shown in the dataset called GWBTH on the accompanying CD-ROM. The off-road bicycling route identified by the Model is represented in the dataset called GWBTR on the CD-ROM. The route for the equestrian trails also begins in Blackwater River State Forest but terminates in the Everglades at the Pinelands area. As with the hiking, and off-road bicycling corridors, the equestrian corridor splits in the central Florida area. Trailheads identified for inclusion in the equestrian corridors are listed in Section 9, Appendix 5, and are found in the dataset called GWETH on the accompanying CD-ROM. The equestrian route identified by the Model is represented in the dataset called GWETR on the CD-ROM. The purposes of the multi-use corridors are different than the three types of corridors previously described. They are intended to: 1) connect key urban areas, possibly supporting alternatives to vehicular transportation and 2) connect urbanurban multi-use corridors with key natural destinations. Multi-use trailheads in the urban areas include city parks, culturalhistoric sites and museums. A number of interim trailheads that are also cultural-historic sites with the capability of handling visitors were also identified. The natural destinations that served as trailheads were: • • • • • • • • •
Gulf Islands National Seashore Pine Log State Forest St. Marks National Wildlife Refuge Stephen Foster State Park Three Lakes Wildlife Management Area Highlands Hammock State Park Pahokee Marina and Campground Big Cypress National Preserve - Oasis Ranger Station Everglades National Park - Pinelands
Please refer to Section 9, Appendix 5, for a complete listing of the multi-use trailheads. A spatial representation of these trailheads is called GWMTH and is found on the accompanying CD-ROM. It is appropriate for the multi-use corridors to pass through native landscapes, agricultural landscapes as well as suburban and urban landscapes. Though, multi-use trails, even more so than other trail types, should not be placed within environmentally sensitive areas including strictly protected areas for biodiversity/ecological conservation. The multi-use corridors identified through the Model include an east west corridor through the Panhandle to Jacksonville, with a spur to Panama City, a north central Florida corridor from Lake City to Orlando and then two legs to the south, one following the west coast to Naples and the other following the east coast to Key West. A spatial representation of these corridors is called GWMTR and is found on the accompanying CD-ROM.
45
Greenprint for King County, Washington Completed in 2005, the Greenprint for King County is an award-winning1 comprehensive conservation strategy prepared by the Trust for Public Land (TPL) and the county Department of Natural Resources and Parks (DNRP). The plan is based on the ideas that it is vital to maintain the region’s unique natural areas, “jewels,” and that now is the time to do so, before development pressure decreases land availability and increases land values. The plan consists of a GIS model analysis as well as recommendations and “conceptual corridors” arrived at via public input and consultation with experts. Final recommendations emphasize county-city cooperation and communication. Highly focused on identifying specific conservation priorities and identifying financial tools with which to realize projects, the Greenprint for King County provides concrete examples for creating and implementing a green infrastructure strategy in a rapidly developing region that is also characterized by significant ecological resources and existing protected areas. Attention to flood hazard areas and providing open space and recreational opportunities to underserved populations is relevant to green infrastructure planning in Worcester County. Overall process • Create inventory of existing parks, open space and resources lands (working farms and forests). • Define a shared conservation vision for the community. • Develop strategies to guide future land acquisition for parks and open space. • Provide specific recommendations for plan implementation. Conservation vision • Ecological health: The vision for King County’s future should continue to focus on protecting natural areas that are important to water quality, salmon habitat, healthy forests, and floodplains. A conservation vision that is grounded in current information, strong partnerships, and a commitment to focusing limited resources in high conservation value areas will improve the county’s ability to protect the ecological health of the land for communities today and in the future. • Cultural and economic values: Communities also value land of historic significance and land that supports trails, parks, viewpoints, working farms and forests, fairgrounds, and regional recreation complexes… part of the local and regional culture that is shared between generations. • Connectivity: The value of individual open space and resource land properties is dramatically enhanced when it is part of an interconnected system… Connecting parks and trails enhances their value for recreation; wildlife species benefit when core habitat areas are connected; and connections between working farms and forests enhance their economic value. GIS modeling approach The decision support model built by TPL and DNRP evaluates regional conservation values according to six program areas (ecological lands, forest, farm, flood protection, regional trails and marine shorelines) individually and in combination. Operating on a 100’ x 100’ grid, the model uses ESRI products, Model Builder and ArcView 9 to assign conservation
1
2005 Honor Award in “emerging technology” from the American Planning Association and the Planning Association of Washington. (http://dnr.metrokc.gov/wlr/greenprint/about.htm) 93
values (0-low to 5-high) based on 60 thematic data sets and 50 sub-models. The data included: • • • • • • • •
Hydrology, Topography and Soils (wetlands, streams, slope, soil) Landscape level considerations (land cover, forest fragmentation) Wildlife (birds, fish, connectivity, management, ownership, agricultural and forestry uses) Conversion risk (annexation areas, proximity to planned/existing development) Trail linkages Marine shoreline analysis (slope stability, sensitive vegetation/habitats, impervious surface) Shoreline public access Park equity analysis (population density, income, minorities, children)
Each analysis (scenarios based on varying weights applied within criteria classes) created new countywide raster layers highlighting existing conservation values. Using the “zonal statistics” function in ArcView 9, the model can also analyze polygon layers (i.e. for a parcel-level study). In addition to identifying valuable areas, TPL and DNRP calculated the number of parcels and acres identified and their status as protected or unprotected. The model is flexible to accommodate future analysis.
Ecological lands model run. Cells are evaluated according to their ecological value. Ecological lands (forests, stream and river corridors, wetlands, shorelines) are responsible for maintaining valuable ecosystem services, so highly functioning ecosystems scored “5 (High)”, shown in dark orange and poorly functioning areas scored “1 (Low)” or “0 (Zero).”
Combined model run. Cells are evaluated according to their “program agreement.” For example, a cell scoring 4 or 5 (highest) in three different program areas got a score of “3” in the final program agreement count depicted in Figure 1. Likewise if the cell scored 4 or 5 in all program areas it would receive a final score of “8”, shown in dark orange.
Incorporation of the GIS model into DNRP programs • Acquisition strategy: Focus funds on highest priority acquisitions by identifying the highest value parcels and maintaining a common database to evaluate purchases across all programs.
94
• • • • •
Water Resource Inventory Areas (WRIA) Salmon Recovery: Assist in implementing the plans by identifying specific problem parcels and being flexible enough to incorporate updated criteria as the need arises. Maintenance and operations: Focus limited resources and influence stewardship and management of protected lands. Capital program: Upgrade restoration projects for degrading stream and river basins. Regulations: Improve implementation efficiency by targeting mitigation reserve areas and facilitating federal and state coordination. Growth management act: Direct growth into designated areas, enforce the urban growth boundary and maintain the area’s rural character by targeting acquisitions and land conservation actions on the rural side of the urban growth boundary.
Outreach strategy During planning stages of the Greenprint project TPL and DNRP conducted interviews and discussions with key city personnel and agency directors focused on the following questions: • • • • • • •
Does your city have a clear conservation vision? Is that vision graphically represented? What are your major land acquisition/conservation priorities? Is there political support to achieve these priorities? What should the county’s priorities be? Is there potential for better city and County coordination? What are your stewardship goals and management capacity? What are your principal funding mechanisms?
What they found was that most communities had a conservation vision, but lacked formal stewardship goals (i.e. management plans, environmental education) or dedicated staff. Cities felt they should focus on local priorities and projects, but supported the county focusing on a regional strategy and providing opportunities for involvement. Many responded that parks and open space should be used as community buffers, creating an individual sense of place in each community, an idea that was reflected in the final Greenprint. On the whole, city representatives desired more regular communication with the county, including technical assistance. Improved communication can foster increased trust and future collaboration on open space and other initiatives. County funding for trails, connecting greenway segments and other open space needs can increase good will between the county and municipalities and facilitate greater funding for future projects. Attached • Sample model matrix, Appendix B: pp. 124-146 References The Trust for Public Land. 2005. Greenprint for King County: A Model for Conservation. Prepared for King County Natural Resources and Parks Water and Land Resources Division by Jones & Jones, Point Wilson Group and The Trust for Public Land. (http://dnr.metrokc.gov/wlr/greenprint/index.htm)
95
Appendix B
King County Greenprint Model Criteria Matrix November, 2004 I. Hydrology, Topography, and Soils H1a Proximity to Wetlands H1b Proximity to Rivers and Streams H1c Proximity to Other Water Bodies H2: Springs H3: Alluvial Fans H4: Surface Runoff H5: Slope H6: Impervious Areas H7: Channel Migration Zones H8: Flood Plains H9: Pre-existing Flood Protection Facilities H10: Critical Aquifer Recharge Areas H12: USDA Agriculture Capability H13: Forest Site Soils H All: Hydrology, Topology, Soils Weighted Overlay II. Landscape Level Vegetation Measures L1: Vegetative Cover L2: Forest Fragmentation L All: Landscape Weighted Overlay III: Landscape Level Wildlife Measures W1: Birds W2: Wildlife Network W3: Fish Habitat W_all: Wildlife Weighted Overlay IV. Gap Analysis and Connectivity Measures G1: Adjacency to Public Lands G2: Adjacency to non-profit protected lands (trusts, etc) G3: Forest Land Connectivity G4: Within Ag Production District G5: Forest Stewardship Plans G6: Proximity to FPP Sites G7: Farm Size G8: Forest Size 1
G9: Actively Farmed G10: Ag Current Use Tax G11: Forest Current Use Tax G all: Gap Weighted Overlay V. Conversion Risk D1: Urban Growth Annexation D2: Proximity to Planned High Density Development D3: Proximity to Existing High Density Development D_all Conversion Risk Weighted Overlay VI. DNRP Acquisition and Conservation Targets P1: DNRP Target Acquisition Parcels VII. Trail Linkages T1: Proposed Trails VIII. Marine Shoreline Analysis S1: Slope Stability S2: Eelgrass S3: Salt Marshes S4: Forage Fish Spawning S5: Marine Reserves S6: Important Bird Areas S7: Armoring S9: Marine Riparian Vegetation (Trees, Shrubs, Grasses) S10: Impervious Surface (shows shoreline modification) S11: Large Woody Debris and Drift Logs IX. Shoreline Public Access S13: Shoreline Public Access
2
King County Greenprint Model Criteria Matrix - SAMPLE June 28, 2005 Criteria Class
Criteria
Hydrology, Topography, and Soils (H) H1a. Proximity to wetlands Data Source: H8_Wt1.shp, derived from WETLD
H1b. Proximity to rivers and streams Data Source: H8_St1.shp, Derived from WTCTRS H1c. Proximity to other water bodies Data Source: H8_Wb1.shp, Derived from WTRBDY.
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 12.5%
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
(11.3%) 5 = within 100 ft of wetland 3 = 101-200 ft of wetland 1 = 201 - 300 ft of wetland 0 = greater than 300 ft (11.3%) 5 = within 100 ft of river/stream 3 = 101-200 ft 1 = 201 - 300 ft 0 = greater than 300 ft
(11.4%) 5 = within 100 ft of water body 3 = 101-200 ft 1 = 201 - 300 ft 0 = greater than 300 ft
3
Criteria Class
Criteria
H2: Springs (Not implemented. Data not available)
H3: Alluvial Fans Data Source: Alluvialfans.shp: Derived from SURFGEOL
H4:Surface Runoff (soil permeability and depth of water table) Data Source: SSUR_mod3.shp, Derived from SSURGO dataset entitled Snoqualm iaP e ss A reaaW , shing ton
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5 <= 300 ft (0%) 0 > 300 ft
5 = alluvial fan on parcel 0 = alluvial fan not on parcel
(9%)
[ permeability ranking + depth to water table ranking ] / 2
(7%)
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
(Permeability ranking and depth ranking or on a scale of 1-5 from more permeable/ depth to less permeable/depth)
4
Criteria Class
Criteria
H5: Slope Data Source: Slope3.shp, generated from a mosaic of USGS 7.5 minute DEMs
H6: Impervious Area
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) (9%) 5 = 0-3 degrees 4 = 3-6 3 = 6-12 2 = 12-24 1 = > 24 0 = no data Change to %
5 = Impervious Areas 0 = other
(7%)
5 = severe CMZ 3 = moderate CMZ 1 = potential CMZ 0 = none or no data
(9%)
5 = in 100 year flood plain 0 = other
(9%)
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Data Source: imperv ArcInfo coverage
H7: Channel Migration Zone Data Source: Chnlmigr.shp (poly)
H8: Floodplains Data Source: Fldplain.shp
5
Criteria Class
Criteria
H9: Pre-existing flood protection facilities (levees, revetments, etc) Data Source: Riverfac.shp
H10: Critical aquifer recharge areas
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) (7%) 5 = flood facility is contained in the parcel 3 = parcel is located next to a parcel that contains a flood facility 1 = parcel does not contain a facility
5 = Category I 3 = Category II 1 = Category III 0 = None
(9%)
5 = Presence of USDA class 1, 2, 3, and 4 soils 0 = Soil classes 5 and above
0%
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Data Source: CAO1Only.shp,C AO2Only.shp, CAO3Only.shp H12: USDA Agriculture capability Data Source: Soil_ag_1to4.shp, Soil_ag_5to8.shp
6
Criteria Class
Criteria
H13: Soils: Forest site class Data source: WADNR (on website)
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5 = class 1 or 2 0% 4 = class 3 3 = class 4 0 = class 5
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
12.5%
Landscape level considerations (L) L1: Vegetative cover
5 = 65% or more forest cover
Data Source: 20040109_Scenari os.shp
3 = 25% forest cover and 65% combo of forest, scrub/shrub and grass
(55%)
1 = 25% forest cover and 25-65% combo forest, shrub/scrub or grass 0 =< 25% forest cover
7
Criteria Class
Criteria
L2: Forest fragmentation Data Source: NLCD 83 National Land Cover Data
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5 = Transitional or (45%) perforated 3 = patch 0 = uniform interior/exterior
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
12.5%
Wildlife (W) W1: Birds Data Source: 20040109_Scenari os.shp
5 = 3 or more nests or presence of priority bird species (bald eagle, osprey, goshawk, great blue heron, big eared bat, peregrine falcon, vaux's swift)
(31%)
3 = 1-2 nests, no priority bird presence 1 = 1 nest w/in 1/2 mi 0 = no nest w/in 1/2 mi or no data
8
Criteria Class
Criteria
W2:Wildlife network
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5 = Some wildlife (31%) network present
Data Source: 20040109_Scenari os.shp
0 = 0 or no data
W3: Fish Habitat
5= chinook or bull trout spawning or highly productive salmonid habitat
Data Source: 20040109_Scenari os.shp
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
(38%)
3= no/low chinook or bull trout spawning; moderately productive salmonid habitat 1= low presence of salmonids 0= no salmonids or no data Gap Analysis and Connectivity Measures (G)
12.5%
9
Criteria Class
Criteria
G1 Connectivity to protected, public lands
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5 = immediate (38%) adjacency 0 = no adjacency
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Data Source: Public.shp, public_addl.shp
* Data is currently compiled into one mega public lands dataset. Future possibility to assess adjacency according to individual program lands? I.e. farm, forest, fhr, eco, and parks? G2 Connectivity to protected lands, via trusts, conservancies, etc.
5 = immediate adjacency 0 = no adjacency
(38%)
Data Source: TBD *data set does not currently exist for King County
10
Criteria Class
Criteria
G3. Proximity to Forest Production District or Rural Forest Focus Areas Data Source: forpddst.shp, rffa.shp
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 0%
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
5 = parcel is located within a RFFA or 5 = parcel is located within a FPD 0 = parcel not in RFFA or FPD
Data sets: 1. Forest Production Districts 2. Rural Forest Focus Areas
G4. Proximity to Agricultural Production Districts
5 = parcel is within APD 0 = parcel outside of APD
(24%)
Data Source: fpp.shp, agrpddst.shp, zoning_new.shp
Data: Agriculture production district boundaries
11
Criteria Class
Criteria
G5 Proximity to parcels with Forest Stewardship Plans Data Source: forest_tech_asst.s hp
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5 = 2 or more 0% forest stewardship plans within 1 square mile 3 = 1 FSP within 1 square mile 0 = none
G6. Proximity to Farmland Preservation Program properties
5 = parcel is adjacent to FPP parcels (or other conserved farmland)
Data Source: fpp.shp
4 = parcel is within 2 miles of FPP parcel or other…
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
0%
3 = parcels is within 3 miles of FPP site or other… 0 = parcel more than 3 miles from other FPP sites or other…
12
Criteria Class
Criteria
G7. Farm size Assessor’s data – contiguous ownership and parcel size
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5 = > 40 acres 0% 3 = 20 - 40 1 = 10 – 20 0 = < 10
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Data Source: agrpddst.shp
G8. Forest size Assessor’s data – contiguous ownership and parcel size
5 = > 160 4 = 80 – 160 3 = 40 – 80 2 = 20 – 40 0 = < 20
0%
Actively farmed property = 5 No = 0
0%
Data Source: agrpddst.shp
G9. Actively farmed Data set: survey done by R. Reinlasoder; KC GIS Data Source: active_ag.shp
13
Criteria Class
Criteria
G10. Current use taxation agriculture
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) Parcel is in 0% Agriculture CUT =5 Not in CUT = 0
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Data Source: cut_ag.shp
G11. Current use taxation –forestry
Parcel is in forestland, timberland CUT = 5 Not in CUT = 0
Data Source: cut_forest.shp
0%
12.5%
Conversion Risk (D) D1: Urban Growth Annexation (UGA) Data Source: Udb_prop.shp
(46%) 5 = outside of annexation boundaries 0 = within potential annexation area and urban boundary
14
Criteria Class
Criteria
D2: Proximity to planned high density development Data Source: parcels_hdd.shp
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5 = not adjacent to (27%) planned high density development 1 = adjacent to planned, high density development
D3: Proximity to existing high density development
5 = not proximate to high density development (determine distance appropriate) 1 = proximate to high density development
Data Source: TBD
DNRP Conservation Targets, by program
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
(27%)
0%
15
Criteria Class
Criteria
P1: DNRP acquisition and conservation targets, by program
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 100% 5 = 2-3 program targets
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
3 = 1-2 program targets 1 = 0-1 program targets
Data Source: TPL2004forestryt argets.shp, TPL2004fhrstarge ts.shp, TPL2004ecotarget s.shp, TPL2004farmtarg ets.shp
0 = no targets programs = farm, forest, ecolands, flood hazard reduction, and parks 12.5%
Trail Linkages T1: Trail Linkages
100% 5 = within 100 ft of proposed trail 3 = 101-200 ft 1 = 201 - 300 ft 0 = greater than 300 ft
Data Source: proptrail.shp
Marine Shoreline Analysis
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
12.5%
16
Criteria Class
Criteria
S1 – Slope Stability Data Source: kcslopestabilitylin e.shp
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5=unstable 10% 4=mod-unstable 3=mod-intermed 1=modifiedstable,stable
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Washington Department of Ecology (WADOE) via King County Drift cell data S2 – Eelgrass Data Source: ssz_eelline.shp
5=continuous 3=patchy 0=none
10%
Washington Department of Natural Resources (WADNR) ShoreZone database
17
Criteria Class
Criteria
S3 – Salt Marshes
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5=Presence 10% 0=Other
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Data source: marsh.shp Washington Department of Natural Resources (WADNR) ShoreZone database
S4 – Forage Fish Spawning
5=Presence 0=Other
10%
Data source: sph_herrspwn.shp sph_sandlanz.shp sph_smelt.shp Data layers originate from Washington Department of Fish and Wildlife (WDFW). Via Washington Department of Natural Resources (WADNR) ShoreZone database
18
Criteria Class
Criteria
S5 – Marine Reserves
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5=Marine Reserve 10% Polygons 0=Other
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Data Source: MarineReserve.sh p WADNR, Seattle Parks and Recreation, and King County Department of Natural Resources S6 – Important Bird Areas
5=IBA Polygons 0=Other
10%
5=Absence 0=Presence
10%
\iba.shp Audubon Society
S7 - Armoring Armoring.shp Anchor Environmental (Anchor 2004) and King County
19
Criteria Class
Criteria
S9 – Marine Riparian Vegetation (Trees, Shrubs, Grasses) mrv.shp
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 5=Trees (all 10% categories) 3=Shrubs (all categories) 1=Grasses (all categories)
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
Anchor Environmental (Anchor 2004) and King County
S10 – Impervious Surface (shows shoreline modification)
5=Low 3=Medium 0=High
10%
5=DRIFT or LWD 0=Other
10%
Impervious.shp Anchor Environmental (Anchor 2004) and King County S11 – Large Woody Debris and Drift Logs Lwd.shp Anchor Environmental (Anchor 2004) and King County.
20
Criteria Class
Criteria
Shoreline Public Access
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 12.5%
S13 – Shoreline Public Access
½ mile shoreline buffer
KC_szline.shp kingco_blkgrps2.s hp park.shp
5=>1000 pers/sqmi 3=>500 pers/sqmi 0=other
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
100%
King County, US Census
Park Equity Analysis P1 – King County Parks and Open Spaces Geolytic 2000 census data
12.5 %
1/8 mile buffer for parks less than 1 acre, ¼ mile buffer for parks greater than 1 acre
21
Criteria Class
Criteria
Population Density
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 33 % 1 – 1700 = 1 1700 – 10k = 2 10k – 16k = 3 16k – 44k = 4 44k – 87k = 5
% Low Income
33 %
0 – 19 = 0 20 - 30 = 0 31 - 57 = 3 58 - 76 = 4 77 - 95 = 5
% minorities
0%
0 - 24 = 0 25 - 30 = 0 31 - 68 = 3 69 - 90 = 4 91 - 112 = 5
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
22
Criteria Class
Criteria
% kids under 18
Scenario 1 – Ecolands Program Criteria Criteria Class Criteria Ranking Strategy Relative Relative (0-5) Weighting Weighting (sum to 100% (sum to 100% WITHIN Criteria ACROSS Criteria Class) Classes) 33 % 0 - 10 = 0 11 - 21 = 2 22 - 31 = 3 32 - 42 = 4 43 - 53 = 5
Criteria Class Relative Weighting (sums to 100% ACROSS Criteria Classes)
Scenario 2 Criteria Ranking Strategy (0-5)
Criteria Relative Weighting (sums to 100% WITHIN Criteria Class)
23
City of Kinston—Lenoir County, North Carolina With the help of the Conservation Fund and the University of North Carolina, the City of Kinston and Lenoir County, North Carolina drafted the Green Infrastructure Plan for the Neuse River Floodplain in 2001. This plan was supplemented by a plan completed in 2002 addressing the economic benefits of green infrastructure, and later influenced the development of a heritage tourism plan for the area. The Kinston—Lenoir example is interesting because the green infrastructure plan capitalizes not just on the ecological resources of the region, but on cultural assets as well in designating priority areas for preservation and recreation. The plan also focused on the adaptive reuse of flood buyout properties unsuitable for development, but important for flood mitigation and habitat restoration. The Kinston—Lenoir plan is relevant to planning for green infrastructure in northern Worcester County, particularly around Berlin, given the area’s dedication to environmental protection (particularly of streams and wetlands) and heritage tourism. Principles • Use green infrastructure as the framework for conservation, planning for ecological resources before development. • Focus on connectivity of ecological resources, creating links with existing gray infrastructure (transportation and utility networks) and connecting sites (schools, tourist destinations) to enhance both ecological and social networks. • Promote diverse social, ecological and economic benefits. • Finance preservation as public investment. • Engage the public, creating opportunities for sharing ideas and generating a shared vision for the future. Goals • • • •
Identify areas suitable for preservation/restoration, recreation and development. Improve upon and connect existing recreation and green infrastructure assets. Contribute to the area’s economy and quality of life. Enhance existing flood mitigation.
Analysis and Classification • Conducted a cell-based GIS suitability analysis using criteria for conservation and recreation potential (10m grid using map calculator tool in ArcView spatial analyst). o Conservation potential defined by {City of Kinston 2001, figure p. 35}: Soils (hydric, other); Land cover (two forest types, other); Land use (undeveloped, low density residential, other); Degradation (possibility for restoration); Proximity to river/tributaries and pond; Proximity to publicly-owned lands; Proximity to intense development; Proximity to roads (access, conservation risk). Based on physical landscape features (most important for conservation, weighted 70%) and proximity to surrounding features (30%). Equal weights given to each of four landscape categories (25%); higher weight given to proximity to river than tributary. o Recreation potential defined by {City of Kinston 2001 figure p. 36}: In/near buy-out properties; In floodplain; Near existing historical/cultural resources and recreation sites or connections b/w
119
•
sites/facilities; Near areas with few recreational opportunities; Near planned recreational facilities near urban area; Access, proximity to roads. Proximity to surrounding figures weighted more than landscape characteristics. Categorized suitability results according to three activity classifications. o Conservation/preservation: will leave the natural environment essentially undisturbed. Though there will be no permanent structures built, earthen trails may be included where suitable. o Mixed use recreation: includes projects that have moderate impacts on existing habitats and land cover. Paved trails, campgrounds, and educational forest projects. o Active recreation: activities that substantially change the existing land cover but are consistent with green infrastructure because they provide open space, retain elements of nature, and are not covered by a large amount of buildings or impervious surface.
High Conservation suitability
• •
•
Low
Recreation High Mixed uses/passive recreation Active recreation
suitability Low Conservation/ preservation Not suitable for green infrastructure
Compared resulting map to aerial photography to identify specific project areas. Ranked project sites according to the following criteria: o Benefit: identified in urban area plan, potential to operate year round, ability to serve a large population and draw people from a wide area o Cost: priority given to low cost options o Feasibility: ability of land to support a project, FEMA land restrictions, political support, noise pollution o Environmental impact: positive, neutral or negative o Sustainability: prioritize projects not thought to need major recovery after a flood event (all identified projects were located within 100-year floodplain) Identified focus areas for cost-acreage-ownership analysis (to determine finer detail project feasibility and location)
Relation to heritage tourism • “Green infrastructure planning harnesses the value and appeal of these resources to stimulate and diversify economic development, all without compromising long-term viability.” (City of Kinston 2002) • Developed a walking tour of the historic downtown and Civil War battlefield sites to connect with a river walk. • Planned a “museum row” connecting nature center, downtown museums and historic cemeteries. • Heritage/cultural tourists spend more money and stay longer than average tourists, and so provide significant resources with which to develop projects. • Festivals, fairs and cultural activities are popular attractions, so creating activities around local amenities will draw tourists and project resources.
120
Funding resources • Grants, tax revenue, bonds, public purchase/lease-back, public take-over (eminent domain), tax abatement, citizen/corporate sponsorship, charitable contributions, easements, fundraising. • ISTEA federal transportation funds used to revitalize historic roads and rail stations. Attached • Appendix B: Detailed Methodology (City of Kinston 2001) • Figure 3: Combined Recreation and Conservation Suitability • Figure 4: Potential Green Infrastructure Hubs and Corridors References University of North Carolina at Chapel Hill. 2001. Kinston/Lenoir County Green Infrastructure Plan for the Neuse River Floodplain. Graduate student workshop, Department of City and Regional Planning. University of North Carolina at Chapel Hill. 2002. Linking Natural and Historic Assets: Green Infrastructure as Economic Development in Lenoir County, North Carolina. Graduate student workshop, Department of City and Regional Planning.
121
Appendix B. Detailed Methodology In order to plan a green infrastructure, the land in the study area must be analyzed to determine its suitability for recreation and for conservation. Figures B.1. and B.2. below present flow charts of the suitability analysis process. In addition, Table B.1. presents further detail about the weighting process described below. Suitability analysis is based on two primary factors, physical features of the landscape and proximity to surrounding related features. Conservation suitability depends upon the proximity of the land to natural resources and upon the physical attributes of the land, such as land use, land cover, whether or not the land is located in the floodplain, and if it has wetland soils. Physical attributes of the land are the most important factor for conservation. To reflect this fact, physical characteristics are weighted 70% of the overall suitability score, while proximity is weighted 30%. For both of the broad categories described above (i.e., physical characteristics and proximity to natural resources) it is necessary to determine the relative importance of the characteristics they incorporate. Toward this end, each characteristic is assigned a weight that indicates the percentage the given characteristic contributes to the broad suitability category (out of 100%). For example, for conservation suitability, all four physical characteristic categories received equal weights of 25%, meaning none is more important than any of the others for determining conservation suitability. However, proximity to the river is more important for conservation than proximity to a tributary or to Kelly’s Mill Pond, and this fact is reflected in the assigned weights. The next step is to determine how well each plot of land meets the characteristics of land suitable for conservation. The study area was divided into a 10-by-10 meter grid to perform this analysis. Using the map calculator tool in ArcView’s spatial analyst extension, each plot of land was assigned a score from 1 (lowest suitability) to 5 (highest suitability) that reflects the degree to which the given land plot possesses the suitability characteristic in question. For example, landuse is divided into three categories. The land may be vacant, making it highly suitable for conservation (score = 5). It may have low-density residential development, making it moderately suitable for conservation (score = 3). Or the land plot may be more intensely developed, making it unsuitable for conservation (score = 1). After each plot of land is scored for all characteristics, an overall suitability ranking can be generated for each plot. The map resulting from this process is presented in Figure B.3. A similar process was followed for recreation suitability using characteristics important for recreation and weighing them accordingly. The recreation suitability map is presented in Figure B.4. The primary difference between recreation suitability and conservation suitability is that proximity to related features is weighted more than physical landscape features.
31
Figure B.1. Conservation Suitability Flowchart
Suitable for wetland plants (5) Other (1)
Wetland soils (0.25)
Bottomland hardwood (5)
Land cover (0.25)
Oak/gum/cypress m (4) Other (1)
Physical
Inside 100 year (SFHA) (5) Other (1)
Floodplain (0.25) x 0.70
Vacant (5) Low density residential (3) Other (1)
Current land use (0.25)
Within 50 feet (5)
River (0.5)
Between 50-100 feet (3) Between 100-500 feet (3)
Proximity
Tributaries (0.3)
Other (1) Within 500 feet (5)
x 0.30
Kelly’s Mill Pond (0.2)
Other (1)
32
Figure B.2. Recreation Suitability Flowchart
Floodplain (0.55)
Physical
Inside 100 year (SFHA) (5) Inside 500 year (SFHA) (3)
Current land use (0.45)
Other (1) Vacant (5) Low density residential (2) Other (1)
Point of interest (0.20)
W/in 1/8 mile (5); ¼ mi. (4); ½ mi. (3); ¾ mi. (2); 1 mi. (1)
Road buffer (0.05)
Within ½ mile (5)
x 0.15
Buyout property (5)
Proximity
Buyout properties (0.26) Within 50’ of buyout property (4); within 150’ (3)
Underserved areas High density(0.08); low (0.07)
x 0.85
> ¾ mile (5); bt/w ½ and ¾ mile (3); bt/w 1/8 and ½ mile (1) Within 10 feet (5); within 50 feet (4); within 1/8 mile (3)
Proposed recreation (0.15) River (0.08) Adkin Branch (0.06)
Within ¼ mile (5); within ½ mile (3) Within 1/8 mile (5); within ¼ mile (3)
33
Table B.2. Recreational Suitability Weighting Matrix Macro level Category
Micro level
Weight
Category
Name of Grid
Floodplain
Physical
Classification
rec_flood
0.15 Current land use
rec_landuse
Point of interest
rec_interest
Category
Weight
0.55 Within 100-year floodplain
5
Within 500-year floodplain
3
Outside of floodplain
1
0.45 Vacant
5
Low density residential
2
Other land uses
1
0.20 Within 1/8 mile
5
Within 1/4 mile
4
Within 1/2 mile
3
Within 3/4 mile
2
Within 1 mile
1
Road buffer
rec_majroads
0.05 Within 1/2 mile
5
Buyout properties
rec_buyout
0.26 Buyout property
5
Areas underserved by rec.
rec_highdens2
(high density) Proximity
Weight
Within 50 feet of buyout
4
Within 150 feet of buyout
3
0.08 More than 3/4 mile Between 1/2 and 3/4 mile
0.85
Between 1/8 and 1/2 mile Areas underserved by rec.
rec_lowdens2
(low density) Areas proposed for recreation
rec_proposed
0.07 More than 2 miles
5 3 1 5
Between 1 and 2 miles
3
Between 1/2 and 1 mile
1
0.15 Within 10 feet
5
Within 50 feet
4
Within 1/8 mile
3
River
rec_river2
0.08 Within 1/4 mile
5
Within 1/2 mile
3
Adkin Branch
rec_adkin2
0.06 Within 1/8 mile
5
Within 1/4 mile
3
0.05 Within 1/4 mile
5
Kelly's Mill Pond
rec_kelly
34
Figure B.3.
35
Figure B.4.
36
FIGURE 3
23
FIGURE 4
24
