Transitional Road Influence Zones for the Southern Appalachian Assessment. Grid Name: transit Version 1 September 1995 ------------------------------------------------------------------------ ------------------------------------------------------------------------ Identification Information Data Layer Name: Transitional Road Influence Zone for the Southern Appalachian Assessment. Description: This data layer is a 90 meter grid of the Southern Appalachian Assessment Region. It identifies portions of the Region which were coded as "Transitional Road Influence Zones" by the team that assessed outdoor recreation settings. Keywords: Roads, Settings, Recreation. Citation: None Native Data Set Environment: Unix; Arc/Info 7.02 (Grid) pathname: //transit.e00.Z Scale: 90 meter cell size File Format: The data layer is a raster in one Arc/Info Grid file that has been exported using Arc/Info's "export" command and compressed using the unix "compress" utility. Use Restrictions: Data users must acknowledge Karl Hermann (University of Tennessee) and Larry Hayden (U.S. Forest Service) in all reports, publications, presentations, etc. Access Restrictions: None ----------------------------------------------------------------------- ----------------------------------------------------------------------- Spatial Reference Information Datum: North American Datum 1983 (NAD83) Precision: single Projection: Albers Equal Area Units: meters Spheroid: WGS-84 1st Std Parallel: 34 00 00 2nd Std Parallel: 38 00 00 Central Meridian: -82 00 00 Origin: 33 00 00 False Northing: 0.0 False Easting: 0.0 Extent: West Bounding Coord.: -420133.156 East Bounding Coord.: 353686.844 North Bounding Coord.: 730770.227 South Bounding Coord.: 16980.227 Distance Resolution: 90 meters Vertical Resolution: n/a ----------------------------------------------------------------------- ----------------------------------------------------------------------- Data Quality Information Thematic Accuracy: unknown Confidence: unknown Accuracy Method: unknown Horizontal Accuracy: unknown Confidence: unknown Accuracy Method: unknown Vertical Accuracy: n/a Logical Consistency: unknown Completeness: unknown --------------------------------------------------------------------------- --------------------------------------------------------------------------- Source Information Source Material: U.S.G.S. Digital Line Graphs from 1:100,000-scale maps Organization: U.S.G.S. Date: unknown Distance Resolution: unknown Contribution: Digital Line Graph coverage of transportation features (roads, utility corridors, railroads). Source Material: U.S.F.S. 1:24,000-scale Forest Service roads Organization: U.S.F.S. Date: unknown Distance Resolution: unknown Contribution: Digital coverage of Forest Service road locations. Source Material: Annual average daily traffic count for selected road segments. Organization: State Departments of Transportation (VA, NC, TN, AL, GA, SC) Date: unknown Distance Resolution: unknown Contribution: Annual average daily traffic counts in both directions for selected road segments in the Southern Appalachian Assessment Region. --------------------------------------------------------------------------- --------------------------------------------------------------------------- Processing History Information Process Description: Introduction We investigated setting descriptors in the Southern Appalachian (SA) region as part of the SA Assessment. A regional map of setting descriptors was made by combining maps of land cover, roads, utility corridors, and protected areas such as Wilderness and National Parks. Briefly, the land cover map was converted to a map of existing land use "themes" such as "Forested" and "Sub-Urban". Information about roads, utility corridors, and protected areas was then used to refine these themes into composite setting descriptors. The areas of different setting descriptors were compiled by ecological "sections" and were compared with existing maps of management designations and recreational features. All of the maps used in this study were part of a standard geographic database which was prepared for the SA Assessment. A commercial geographic information system (ArcInfo [ESRI 1994]) was used for most of the analysis. Metadata for all maps used in this study are contained in the standard SA geographic database. Briefly, the land cover map was derived from Landsat Thematic Mapper (TM) images from the early 1990's by the Pacific Meridian Corporation. In an Albers conical equal-area projection, the resolution (pixel size) size was 0.09 ha (30m by 30 m). The study area was about 151,000 km2 for which the land cover map had about 168 million pixels. The map contained 17 land cover classes which were condensed into just five classes (forest, agriculture, developed, water, other) for this study (Table 1). Most of the study area was classed as forest (69.62%), followed by agriculture (20.25%), other (5.39%), developed (3.27%), and water (1.48%) (Table 1). The "other" category was mostly the herbaceous land cover (4.70%) which was not exclusively forest or agriculture according to the classification system. The road maps are also documented in the SA geographic database. Briefly, two road maps were used. The first was a regional map (1:100000 scale) which identified four road classes ranging from city streets (class four) to interstate highways (class one). This map was augmented by adding traffic count data, for class one and two roads only, provided by State Departments of Transportation, for individual road segments. The traffic data consisted of annual average daily vehicle crossings in both directions. Traffic counts were not available for some road segments, for example, for many city streets and short segments of freeways at interchanges. The second road map was for U.S. Forest Service lands only (1:24000 scale). These roads are referred to here as NFS roads. This map was used because it included many restricted-access roads which did not appear on the regional road map. Roads which appeared on both maps were handled as described later. The map of utility corridors (1:100000 scale) included powerlines, piplelines, and railroads. The maps of protected areas (1:250000 scale) identified Federally- designated Wildernesses, Wild and Scenic Rivers, and National Parks. The ecological section map identified sub-regions of similar ecological conditions within the SA boundary and was adopted by the SA Assessment for standardized reporting by different assessment teams. Land Cover Analysis The objective of the land cover analysis was to identify and map six land use "themes": Urban, Sub-Urban, Transitional, Forested, Partially Forested, and Pastoral/Agricultural. Land use themes are similar to "landscape pattern types" (Wickham and Norton 1994) which are mapped land units comprising single types of land uses. The mapping of themes was accomplished by applying spatial filtering procedures to the five-class land cover map. Spatial filtering is a general term used in image processing (e.g., Gonzalez and Woods 1992). Briefly, a "window" is moved in steps over an input image. At each step, a function is evaluated within the window, and the result is mapped to the current window location in a new (spatially- filtered) image. Depending on the content of the input image, and the specific filter function, a variety of information can be extracted by using spatial filtering. Classical applications of spatial filtering include edge detection, pattern recognition, and image enhancement (Gonzalez and Woods 1992). Spatial filtering appears, for example, as "resampling" functions in some geographic information systems. Spatial filtering has also been applied to land cover maps to analyze patterns of, for example, land cover texture (Plotnick et al. 1993), wildlife habitat (Kepner et al. 1994), and land cover diversity (Wickham et al. 1995). The size of the analysis window and the resolution of the output map can be varied in order to extract or display different scales of pattern (e.g., Jones and Riitters 1995). Land use themes were defined in terms of the proportions of forest, agriculture, and developed land covers which were contained in a 65.61 ha (27 by 27 pixels) analysis window (Table 2). Areas of water and "other" land cover categories were ignored when finding the proportions. The land use theme was then mapped at the location of the pixel in the center of the window on a new map, and the window was moved (a distance of one pixel) to start a new analysis window. For example (see Table 2), if the window contained between 50% and 85% forest cover and less than 5% developed cover, then the theme in that window was evaluated as "Partially Forested", and the center pixel on the new map was assigned this value. Thus, each pixel in the land use theme map represents an area of 0.09 ha, for which the surrounding 65.61 ha was categorized as the land use theme which was stored at that pixel's location. If the land cover of the pixel in the center of the window was "water" or "other", then the land use theme was considered to be "water" or "unknown", respectively (Table 2). The map resolution was reduced from 0.09 ha to 0.81 ha in order to highlight regional patterns and to simplify further analyses. While knowledge of fine-scale variation is important for sub-regional investigations, our interest was on broad regional patterns for which such fine-scale information was not really necessary. We experimented with resolutions of 0.09 ha to about 100 ha before deciding to use a 0.81-ha resolution for subsequent analyses. The resolution of the land use theme map was changed by using a majority-rule spatial filter, a common technique for image enhancement (e.g., Gonzalez and Woods 1992). The procedure is effective for removing small isolated regions, for example, the image "noise" arising from scattered pixels of "unknown" land use themes. The dominant land use theme (this time considering both "water" and "unknown" categories) in a 0.81-ha (3 by 3 pixels) analysis window was found, and this theme was put into a new map with a pixel size of 0.81 ha. Since each pixel in the new map represented nine times more area, the total number of pixels was reduced by a factor of nine. Analysis of Utility Corridors, Roads, and Traffic Counts The influences of utility corridors, roads, and traffic volume were incorporated into the analysis by using the maps and traffic counts to create five separate maps of influence zones -- "Urban", "Sub-Urban", "Transitional", "Rural", and "Roaded - Natural Appearing". First, a buffer region was defined on both sides of each road or utility corridor segment. The buffer distance was either 805 m or 402 m, depending on the class of road and traffic count (Table 3). The total width of the buffer region, on both sides of a road segment, was twice the buffer distance. Once defined, each buffer region was assigned to one of the five influence zones, again depending on road class and traffic count (Table 3). For practical reasons, the buffering operations were done separately for different combinations of road class and traffic count. The union of these separate operations, for a given type of influence zone, yielded the final map for that influence zone. Buffers around class one and two road segments lacking traffic count data were assigned to the "Urban" influence zone if they were in a city, or to the "Rural" zone if they were on the Blue Ridge Parkway or the within the Great Smoky Mountains National Park. Otherwise, class one and two road segments lacking traffic count data were assigned to the "Transitional" and "Rural" influence zones, respectively. The final step of the influence zone analysis was to convert the five maps to a raster (or grid) format with a 0.81-ha resolution to make them comparable on a pixel-by-pixel basis with the map of land use themes. Mapping Setting Descriptors The map of land use themes was intersected with the five maps of influence zones to create a preliminary map of setting descriptors. The rules used to define setting descriptors are shown in Table 4. For example, if a given location was classified as a "Forested" land use theme, and was also contained within a "Rural" influence zone, then the setting descriptor for that location was "Rural - Forested". A given location could have been contained in more than one influence zone because the buffer regions around different road segments sometimes overlapped. Thus it was possible for those locations to be classified into more than one setting descriptor class. The rules were established such that the "more urbanized" setting descriptor had precedence over the "more pristine" setting descriptor (Table 4). If a given location was not in any influence zone, then the setting descriptor was decided on the basis of the land use theme alone (Table 4). As before, the identities of "water" and "missing" pixels were retained in the setting descriptor map. A minimum size contraint was then applied to contiguous regions (patches) of "Semi-Primitive - Natural Appearing" and "Roaded - Natural Appearing" settings. Patches of "Semi-Primitive - Natural Appearing" settings which were less than 1011.72 ha were reclassified as "Roaded - Natural Appearing" settings. Patches of "Roaded - Natural Appearing" settings which were less than 202.34 ha were reclassified as "Rural - Forested" settings. The size constraints were based on existing administrative rules for defining settings. Finally, the maps of National Parks, Wilderness Areas, and Wild and Scenic Rivers were used to identify additional settings which are already considered to be Naturally Evolving. The Primitive - Naturally Evolving setting was defined as any land within the boundary of the Great Smoky Mountains National Park that was more than 4.8279 km from any road, subject to a minimum patch size constraint of 2023.44 ha. Semi-Primitive and Roaded - Natural Appearing settings (as defined above) were considered to be Naturally Evolving if they were within designated Wildernesses, Wild and Scenic River Areas, or National Parks. ---------------------------------------------------- Table 1. Land cover statistics and class revisions for the Southern Appalachian Assessment region. See text for description of the land cover map. Original Land Cover Class SAA Area Revised Land Cover Class (%)[a] Northern hardwood forest 0.52 Forest Mixed mesophytic hardwood forest 3.50 Forest Oak forest 45.56 Forest Bottomland hardwood forest 0.09 Forest White pine/hemlock forest 0.25 Forest Montane spruce-fir forest 0.24 Forest Southern yellow pine forest 4.59 Forest White pine/hemlock/hardwood forest 2.21 Forest Mixed pine/hardwood forest 12.66 Forest Herbaceous 4.70 Other Barren 0.45 Other Agriculture - pasture 16.81 Agriculture Agriculture - cropland 3.44 Agriculture Wetlands 0.19 Other Developed 3.27 Developed Water 1.48 Water Indeterminate - clouds, shadows 0.05 Other Background[b] 0.00 Other [a]Percentages sum to 100.01 due to rounding. [b]About 96 ha of the standard SAA Area was labeled as 'background' (i.e., not as a land cover) in the land cover map. ------------------------------------------------- Table 2. Spatial filtering rules used to define land use themes in the Southern Appalachian Assessment region. Because the rules are not mutually exclusive, they were evaluated in the order shown, and the last applicable theme was used. The land cover percentages are based on the five-class land cover map (forest, developed, agriculture, water, other). Land cover composition Land use theme 100% Other Unknown Forest > Agriculture Partially Forested Agriculture > Forest Pastoral/Agricultural 50% - 85% Forest Partially Forested >50% Agriculture Pastoral/Agricultural > 85% Forest Forested 5% - 20% Developed Transitional 20% - 60% Developed Sub-Urban > 60% Developed Urban Indeterminate[a] Unknown Note: The land use theme was set to "water" or "unknown" if the land cover in the center of the analysis window was "water" or "other", respectively. [a]None of the land cover composition criteria were met in the analysis window. ------------------------------------------------- Table 3. Buffer widths and influence zone assignments for buffer regions: (A) by road class and average daily traffic count for class one and class two roads; (B) for other roads and utility corridors. The buffer regions were defined on both sides of each road or utility corridor segment; the total buffer width was twice the distance shown. Traffic counts are annual average daily vehicle counts in both directions. Class One and Class Two roads Road Class Traffic --------------------------------------- count One Two --------------- ----------- ---------- Buffer Distance (m) and Influence Zone > 39,000 805 - Urban 805 - Urban 19,001 - 39,000 805 - Transitional 805 - Sub-Urban 7,901 - 19,000 402 - Transitional 805 - Transitional 1,901 - 7,900 402 - Transitional 402 - Transitional 101 - 1,900 402 - Rural 402 - Rural 0 - 100 402 - Rural 805 - Roaded - Natural Appearing Missing (in city) 805 - Urban 805 - Urban Missing (NPS[a]) 402 - Rural n/a[b] Missing (other) 805 - Transitional 402 - Rural Other roads and utility corridors Feature Type Buffer Distance (m) and Influence Zone ----------- -------------------------------------- Class three roads - not NFS 402 - Rural Class three roads - NFS 805 - Roaded - Natural Appearing Class four roads - all 805 - Roaded - Natural Appearing NFS roads, not class three or four 402 - Roaded - Natural Appearing Railroads 402 - Roaded - Natural Appearing Powerlines 402 - Roaded - Natural Appearing Pipelines 402 - Roaded - Natural Appearing [a]The Blue Ridge Parkway and Highway 441 through Great Smoky Mountains National Park (NPS is an acronym for National Park Service). [b]There were no roads in this category in the study region. ------------------------------------------- Table 4. Assignment of setting descriptors based on land use themes and influence zones. Because individual locations could occur in more than one type of influence zone, they were evaluated in the order shown, and the first applicable setting descriptor was used. Land Use Theme Influence Zone Setting Descriptor -------------- --------------- ------------------- Unknown Urban Urban Sub-Urban Sub-Urban Transitional Transitional Any other, or None Unknown Water Any, or None Water Urban Any, or None Urban Sub-Urban Urban Urban Any other, or None Sub-Urban Transitional Urban Urban Sub-Urban Sub-Urban Any other, or None Transitional Pastoral/Agric. Urban Urban Sub-Urban Sub-Urban Transitional Transitional Any other, or None Rural - Pastoral/Agricultural Partially Forested Urban Urban Sub-Urban Sub-Urban Transitional Transitional Any other, or None Rural - Partially Forested Forested Urban Urban Sub-Urban Sub-Urban Transitional Transitional Rural Rural - Forested Roaded - Natural Appearing Roaded - Natural Appearing[b,c] None Semi-Primitive - Natural Appearing[a,c] Note: One additional setting, Primitive - Naturally Evolving, was defined for certain areas within the Great Smoky Mountains National Park as described in the text. [a]Subject to minimum contiguous region size of 1011.72 ha, otherwise Roaded - Natural Appearing. [b]Subject to minimum contiguous region size of 202.34 ha, otherwise classed as Rural Forested. [c]In addition, the setting was considered to be Naturally Evolving if contained within the boundaries of a designated Wilderness, Wild and Scenic River Area, or National Park. ------------------------------------------ Literature Cited Gonzalez, R.C. and R.E. Woods. 1992. Digital Image Processing. Addison-Wesley Publishing, Reading MA. Jones, K.B. and K.H. Riitters. 1995. Evaluating wildlife habitat suitability using a multi-scaled landscape assessment approach. In Proc. of the 27th Symposium on the Interface: Computing Science and Statistics, June 1995, Pittsburg PA. Interface Foundation of North America, Fairfax Station VA. Kepner, W.G., Riitters, K.H., and J.D. Wickham. 1995. A landscape approach to monitoring and assessing ecological condition of rangelands -- San Pedro case study. Environmental Monitoring and Assessment, In Press. Plotnick, R.E., Gardner, R.H., and R.V. O'Neill. 1993. Lacunarity indices as measures of landscape texture. Landscape Ecology 8:201-211. Wickham, J.D., and D.J. Norton. 1994. Mapping and analyzing landscape patterns. Landscape Ecology 9:7-23. Wickham, J.D., Wade, T.G., Jones, K.B., Riitters, K.H., and R.V. O'Neill. 1995. Diversity of ecological communities of the United States. Vegetatio, In Press. ESRI (Environmental Systems Research Institute, Inc.). 1994. ARC/INFO (Version 7). ESRI, Inc., Redlands, CA 92373, USA. ---------------------------------------------------------------------- ---------------------------------------------------------------------- Entity/Attribute Information Entity: transit.vat Definition: transitional road influence zone (yes or no) Attributes: Value Definition ----- ---------- 1 Within a transitional road influence zone Missing Not within a transitional road influence zone ---------------------------------------------------------------------- ---------------------------------------------------------------------- Status Information Data Set Status: Available Release Date: September 1995 ---------------------------------------------------------------------- ---------------------------------------------------------------------- Metadata Reference Information Date: September 5, 1995. Review Date: n/a Contact: Kurt Riitters (address below). ----------------------------------------------------------------------- ----------------------------------------------------------------------- Distribution Information Distribution Contact: Karl A. Hermann National Biological Service Cooperative University of Tennessee 17 Ridgeway Road Norris, TN 37828 (423)-632-1452 samab@utk.edu Distribution Liability: File Compression Technique: UNIX compress, Arc/Info export format Transfer Size: ----------------------------------------------------------------------- ----------------------------------------------------------------------- Contact Information Contact Person: Kurt Riitters Contact Mail Address: Tennessee Valley Authority, Historic Forestry Building, 17 Ridgeway Road, Norris TN 37828 USA Contact Phone: 423-632-1651 Contact Fax: n/a Contact email: n/a -------------------------------------------------------------------- --------------------------------------------------------------------