Metadata: Identification_Information: Citation: Citation_Information: Originator: Kurt Riitters, Southern Appalachian Field Laboratory, Biological Resources Division, U.S. Geological Survey Publication_Date: 1998 Title: "femf81" Edition: 1 Geospatial_Data_Presentation_Form: Map Publication_Information: Publication_Place: Norris, TN Publisher: Tennessee Valley Authority Description: Abstract: This dataset is a grid map that portrays an index of forest fragmentation calculated at 590.49 ha scale for the United States Standard Federal Region III (SFR 3) at 30 meter resolution. The dataset was created by applying a spatial filter to a published map of land cover known as " U.S. Federal Region III Land Cover Data Set", which was obtained from the Multi- Resolution Land Characteristics (MRLC) Consortium. The spatial filter computed an index of forest fragmentation within 590.49 ha (81x81 pixel) windows from the attribute adjacency table of forest and non-forest land cover types within a window. The [0,1] continuous index was discretized, and a byte representation [1,255] of the index value was stored at the center location in each window. The resulting surface map was then median-filtered to find the median index value in 7.29 ha windows (9x9 pixels), and the result was stored at 270 meter resolution (7.29 ha grain size). Purpose: These data were originally created as part of an assessment of landscape conditions in the mid-Atlantic region (SFR 3) based on digital land cover maps derived from remote sensing (TM) and produced by the Multi-Resolution Land Characteristics Consortium. The forest fragmentation map portrays relative fragmentation, considering only forest versus non-forest cover types, and for a single spatial scale. The map was presented as a surface map for visualization purposes, and was also summarized by watershed for assessment purposes. Time_Period_of_Content: Time_Period_Information: Range_of_Dates/Times: Beginning_Date: 1991 Ending_Date: 1993 Currentness_Reference: Dates of satellite imagery. Status: Progress: Complete Maintenance_and_Update_Frequency: None planned. Spatial_Domain: Bounding_Coordinates: West_Bounding_Coordinate: -84 degrees West East_Bounding_Coordinate: -74 degrees West North_Bounding_Coordinate: 43 degrees North South_Bounding_Coordinate: 36 degrees North Keywords: Theme: Theme_Keyword_Thesaurus: None Theme_Keyword: pattern Theme_Keyword: landcover Theme_Keyword: United States Theme_Keyword: forest Theme_Keyword: fragmentation Place: Place_Keyword_Thesaurus: None Place_Keyword: United States Place_Keyword: Standard Federal Region III Access_Constraints: None Use_Constraints: None Point_of_Contact: Contact_Information: Contact_Person_Primary: Contact_Person: Bruce Jones Contact_Organization: EPA Landscape Ecology Division Contact_Address: Address_Type: mailing and physical address Address: U.S. Environmental Protection Agency, Box 93478 City: Las Vegas State_or_Province: Nevada Postal_Code: 89193-3478 Contact_Electronic_Mail_Address: jones.bruce@epamail.epa.gov Data_Set_Credit: The basic data from which this map was derived were provided by the Multi-Resolution Land Characteristics Consortium Native_Data_Set_Environment: These data are presently stored in ARC/INFO GRID format. Data_Quality_Information: Attribute_Accuracy: Attribute_Accuracy_Report: The fragmentation index is based on a proportion derived from the attribute adjacency frequencies of forest and non-forest pixels (see Process_Step sections; Riitters et al. 1996, 1997). A statistical rule of thumb is that 50 observations are needed in order to accurately estimate a proportion. The analysis window was 81x81 pixels in size, which corresponds to a potential maximum of 12960 observed adjacencies. However, only those adjacencies which involve at least one 'forest' pixel are included in the calculation of the proportion for a given window; therefore, the actual number of observations can range from zero to 12960 and will vary with the number of 'forest' pixels in the window, and also with the number of 'missing' pixels in the window (pixels outside the map boundary). This index is one of many which could have been used to quantify 'fragmentation', for a discussion of relative strengths and weaknesses please consult a standard textbook on landscape ecology or image processing. Like other indices, this index is best interpreted as a relative value, in comparison to other pixels on the same map. Comparisons with other maps prepared from different data sources or at different scale are tenuous. Some applications may require adjustments for the amount of forest present in a given analysis window. Logical_Consistency_Report: All cell values are in the range [1,255]. Missing values were assigned to cells for which the analysis window contained no forest cover. Completeness_Report: Every cell in the original land cover map has a non-zero byte value representing the fragmentation index, unless the analysis window for that cell contained no forest land cover, in which case the cell was assigned a 'missing' data value. Positional_Accuracy: Horizontal_Positional_Accuracy: Horizontal_Positional_Accuracy_Report: The positional accuracy of the forest fragmentation map is the same as the positional accuracy of the land cover map from which the fragmentation map was derived. Lineage: Source_Information: Source_Citation: Citation_Information: Originator: Publication_Date: 1996 Title: U.S. Federal Region III Land Cover Data Set Edition: 1 Geospatial_Data_Presentation_Form: remote-sensing image Series_Information: Series_Name: U.S. Federal Region III Publication_Information: Publication_Place: Publisher: Type_of_Source_Media: online Source_Time_Period_of_Content: Time_Period_Information: Range_of_Dates/Times: Beginning_Date: 1991 Ending_Date: 1993 Source_Currentness_Reference: Dates of satellite imagery. Source_Citation_Abbreviation: MRLC-R3V1 Source_Contribution: Land cover map Source_Information: Source_Citation: Citation_Information: Other_Citation_Details: Riitters, K.H., O'Neill, R.V., and Jones, K. B. 1997. Assessing habitat suitability at multiple scales: a landscape-level approach. Biological Conservation 81:191-202. Source_Information: Source_Citation: Citation_Information: Other_Citation_Details: Riitters, K.H., O'Neill, R.V., Wickham, J.D., and Jones, K.B. 1996. A note on contagion indices for landscape analysis. Landscape Ecology 11:197-202 Process_Step: Process_Description: Derivation of forest fragmentation from land cover. Land cover data (MRLC-R3V1) obtained from EROS Data Center in Arc Grid format. Individual pixels were re-classified into forest and non-forest classes by using the following scheme: Original MRLC class Forest/non-forest class ---------------------------- ------------------------------- 1 - water non-forest 2 - low intensity developed non-forest 3 - high intensity developed non-forest 4 - hay/pasture/grass non-forest 5 - row crops non-forest 6 - probable row crops non-forest 7 - conifer forest forest 8 - mixed forest forest 9 - deciduous forest forest 10 - woody wetland forest 11 - emergent wetland non-forest 12 - bare; quarries non-forest 13 - bare; coal mines non-forest 14 - bare; beaches non-forest 15 - bare; transitional non-forest Land cover data exported to ascii format using the ARC/INFO command GRIDASCII, and format converted to an in-house format using an in-house software tool named ASC2TVA.C. The map was then subdivided into sixteen (16) overlapping rectangles for further processing using an in-house software tool named SPLITTER.C. The rectangles overlapped in order to avoid artifacts near image boundaries during the spatial filtering operations. Each of the 16 sub-rectangles were then processed via spatial filtering to estimate the forest fragmentation index as described below. The spatial filtering program is an in-house software tool named SPATCONV.C. Each of the resulting 16 sub-rectangles were then processed via spatial filtering to estimate the median fragmentation index and reduce the spatial resolution as described below. The spatial filtering program is an in-house software tool named SPATCONV.C. After spatial filters, the map of fragmentation index values was reconstructed by a series of steps, including reassembling the 16 sub-rectangles into a single image (via an in-house software tool named LUMPER.C, which removed the overlapping parts of rectangles), converting the format to ascii (using the in-house software tool named TVA2ASC.C), and finally bringing it back into ARC/INFO using their command ASCIIGRID. Spatial filtering proceeded as follows. A 590.49 ha (81x81 pixels) quadrat was centered on each pixel of the original land cover map. An attribute adjacency table was then tabulated for the non-missing pixels in the window, considering only pixel pairs in cardinal directions (i.e., four neighbors per pixel), and counting each pixel pair once only. The fragmentation index was then computed from the attribute adjacency table as the frequency of forest-forest pairs, divided by the sum of frequencies of forest-forest plus forest-nonforest pairs. Note that this expression yields a value of 0 when no forest pixels are adjacent to another forest pixel, and a value of 1 when all forest pixels are adjacent in all cardinal directions to other forest pixels. In other words, the index ranges from zero (high forest fragmentation) to one (low forest fragmentation). The calculated values were discretized to the range [1,255] and stored at 30-meter spatial resolution. Thus, a pixel value in the new map represents the forest fragmentation of the surrounding 590.49 ha in the original land cover map. A median filter was then applied, such that a 7.29 ha (9x9 pixels) quadrat was centered on each pixel of the intermediate map. The median index value was found, and stored at 270 meter resolution (7.29 ha grain size) in the final map. A pixel value in the final map represents the median fragmentation index value for the neighborhoods surrounding all pixels in the 9x9 quadrat. The transformation used to discretize the index values was: D = ( C * 254) + 1 where D = discretized value in range [1,255] C = calculated value of fragmentation index in range [0,1] If needed, the original calculated values can be approximated by applying the backtransformation: C = (D - 1) / 254 Note that the backtransformed values will be in steps of size approximately 0.00394 as a result of the discretization process (precision has been lost). Source_Used_Citation_Abbreviation: MRLC-R3V1 Source_Produced_Citation_Abbreviation: femf81 Process_Date: 1996 Spatial_Data_Organization_Information: Direct_Spatial_Reference_Method: Raster Raster_Object_Information: Raster_Object_Type: Grid Cell Row_Count: Column_Count: Spatial_Reference_Information: Horizontal_Coordinate_System_Definition: Planar: Map_Projection: Map_Projection_Name: Lambert_Azimuthal Albers_Conical_Equal_Area: Radius of Sphere of Reference: 6370997 Longitude_of_Central_Meridian: -100 Latitude_of_Projection_Origin: 45 0 0 False_Easting: 0.0 False_Northing: 0.0 Planar_Coordinate_Information: Planar_Coordinate_Encoding_Method: Row and column Coordinate_Representation: Abscissa_Resolution: 270 Ordinate_Resolution: 270 Planar_Distance_Units: meters Entity_and_Attribute_Information: Detailed_Description: Entity_Type: Entity_Type_Label: femf81.VAT Entity_Type_Definition: Value attribute table of femf81 grid cells Entity_Type_Definition_Source: ARC/INFO Attribute: Attribute_Label: VALUE Attribute_Definition: Byte value representation of forest fragmentation index (zero represents missing values). Attribute_Definition_Source: software computed Attribute_Domain_Values: Range_Domain: Range_Domain_Maximum: 255 (integer) Range_Domain_Minimum: 0 (integer) Attribute_Measurement_Frequency: None planned Attribute: Attribute_Label: COUNT Attribute_Definition: Number of grid cells with indicated VALUE Attribute_Definition_Source: software computed Attribute_Domain_Values: Range_Domain: Range_Domain_Minimum: positive integer Range_Domain_Maximum: positive integer Attribute_Measurement_Frequency: None planned Overview_Description: Entity_and_Attribute_Overview: The value attribute table associated with this raster: DATAFILE NAME: femf81.VAT 2 ITEMS: STARTING IN POSITION 1 COL ITEM NAME WIDTH OUTPUT TYPE N.DEC ALTERNATE NAME INDEXED 01 VALUE 4 10 B - - Indexed 05 COUNT 4 10 B - - - Entity_and_Attribute_Detail_Citation: Not available Distribution_Information: Distributor: Contact_Information: Contact_Person_Primary: Contact_Person: Bruce Jones Contact_Organization: U.S. Environmental Protection Agency Contact_Address: Address_Type: mailing and physical address Address: U.S. Environmental Protection Agency City: Las Vegas State_or_Province: Nevada Postal_Code: 89193-3478 Contact_Electronic_Mail_Address: jones.bruce@epamail.epa.gov Metadata_Reference_Information: Metadata_Date: 19981014 Metadata_Contact: Contact_Information: Contact_Person_Primary: Contact_Person: Kurt Riitters Contact_Organization: U.S. Geological Survey- Biological Resources Division Contact_Address: Address_Type: mailing and physical address Address:NCSU Forestry Dept. Address: 3024 Biltmore Hall, NCSU City: Raleigh State_or_Province: NC Postal_Code: 27695-8002 Metadata_Standard_Name: FGDC Content Standards for Digital Geospatial Metadata Metadata_Standard_Version: 19940608 Metadata_Security_Information: Metadata_Security_Classification_System: None Metadata_Security_Classification: Unclassified Metadata_Security_Handling_Description: None