U.S. Geological Survey --------------------------------------------------------------------------- Metadata for LANDCOV Table of Contents Identification_Information Abstract Purpose Supplemental Information Data_Quality_Information Spatial_Data_Organization_Information Entity_and_Attribute_Information Detailed Description Overview Distribution_Information Metadata_Reference_Section --------------------------------------------------------------------------- Identification_Information Citation_Information Originator: Georgia Department of Natural Resources, Georgia Natural Heritage Program Publication_Date: 1995, 1996 Title: Landcover Classification of Georgia 1988-1990 Edition: 2.0 2.1 Datum change from 1927 to 1983 (North American Datum of 1983) Data Re-projected using North American Datum (NAD83) - September 2000 Series_Information Series_Name: Digital Data Series Publication_Information Publication_Place: Atlanta, GA Publisher: U.S. Geological Survey Online_Linkage: http://csat.er.usgs.gov/statewide/layers/landcov.html Scale_Denominator: 100000 Description Abstract In 1992, a database of landcover and wetlands for the State of Georgia at a scale of 1:24,000 (and 100 ft resolution) to be used for regional and local environmental investigation and planning was developed by ERDAS, Inc. under contract for the Georgia Department of Natural Resources (GA DNR). In 1995, this database was partially revised by the GA DNR, Natural Heritage Program to reflect alterations to the agricultural landcover information primarily in Southwest Georgia and to provide the data at a 200 ft. resolution. Landcover classification was performed by ERDAS, Inc., Atlanta, Georgia. Landsat Thematic Mapper bands one through five were used in the classification. The landcover classification was derived from a limited number of field samples. Initial testing of concept was accomplished through the performance of a pilot project. Mapping of wetlands within the area of three U.S.G.S. 7.5 minute quadrangle maps was the focus of this effort. A quarter scene of Landsat thematic mapper satellite data was purchased for use in computer analysis by ERDAS. The methods used to classify this satellite imagery for wetlands are much the same as those used for the classification of the entire state, which are discussed later in this report. Purpose Create a landcover and wetlands database for the State of Georgia to be used in the protection of wetlands for the State Planning Act of 1989. Supplemental_Information Procedures_Used For use in mapping of the state, eleven full scenes and two quarter scenes of Landsat Thematic Mapper (TM) satellite data were purchased (see Notes II). Microfiche of available images for appropriate acquisition dates were checked for cloud cover and image quality at the EOSAT field office in Orlando, Florida before the order was made. Winter scenes were specified to obtain maximum penetration of vegetation canopies for wetland delineation. The best available scenes ranged in acquisition dates from December 1988 through January 1990 (see Notes II). The project was completed in two major phases. The northern most portion of the State was completed in Phase I of the project effort and stopped along the physiographic fall line. Phase II of this study included the central and southern part of the State and included additional coastal land cover classes. (see Notes IV) As part of the initial computer processing, the satellite imagery was divided into separate work regions before further analysis. These work regions were a composite of major physiographic regions within the state, the Landsat satellite path (see Notes III), and the state plane coordinate system zones. Division according to physiographic region was done to simplify the classification by isolating some of the edaphic, geomorphic and vegetative differences which occur between regions. These regions include : 1) Cumberland Plateau Section, 2) Southern Valley and Ridge Section, 3) Southern Blue Ridge Section, 4) Upland Georgia Subsection Of the Southern Piedmont Section, 5) Midland Georgia Subsection of the Southern Piedmont Section, 6) Sea lsland Section - Barrier lslands Sequence District, 7) Sea lsland Section - Vidalia Upland District, 8) East Gulf Coastal Plain Section - Fall line Hills District and Fort Valley Plateau District, 9) East Gulf Coastal Plain Section - Tifton Upland District, and 10) Sea lsland Section - Bacon Terraces District and Okefenokee Basin District. These regions are from the Physiographic Map of Georgia, 1976, by Wlliam Z. Clark, Jr. and Amold C. Zisa. Further separation by satellite path avoided possible variations in spectral response due to date or atmospheric conditions (scenes in the same path were acquired on the same day). The Landsat scenes were also georeferenced to their respective Georgia State Plane zones, with duplicate scenes of each zone in places of overlap. IMAGE PROCESSING TECHNIQUES In classifying the satellite imagery, both supervised and unsupervised methods were used. The supervised approach required the field identification and plotting of suitable sites of homogeneous landcover, typically numbering 60 - 100 per region. These "training samples" were digitized on the TM imagery to "train" the computer to group pixels (individual cells of which the imagery is comprised ) with similar spectral characteristics into the same landcover class. In this way the satellite image, which is a picture composed of the brightness values of landcover measured at different spectral ranges, was transformed into a thematic map composed of a discrete number of landcover classes. Unsupervised classification, which creates a user-defined number of classes based on spectral response, was used to aid in the identification of potential training sites before making the actual field observations. Unsupervised classification was also used to help separate spectrally similar landcover classes by breaking these classes into spectral subset which could then be reallocated to the proper landcover classes. Visual checks were made on each class individually by using the computer to overlay the area assigned to a particular landcover type on the satellite imagery and toggling the landcover class off and on to check for proper location and extent. Sample areas of the preliminary classification were field-checked to find and correct problems before the accuracy assessment was conducted. RASTER TO VECTOR CONVERSION AND HARDCOPY MAPPING The next stage was conversion from the ERDAS raster landcover database to an ARC/INFO vector database compatible with other geographic databases kept at the Georgia Department of Community Affairs. The landcover for each physiographic region was mosaicked so that all of the individual U.S.G.S. 7.5 minute quadrangle maps could be cut out. Each quad was then converted to an ARC/INFO coverage with a minimum polygon size of 2.5 acres. For quadrangles which fell on the Georgia State Plane zone boundary, two separate coverages were made, one for each state plane zone. See Notes Section I for copy of SURVEY FORM USED Color paper plots of the generalized database were produced to correspond with the U.S.G.S. 7.5 minute quadrangles at 1 :24000 scale. Individual copies in both state plane zones were produced for quadrangles which fell on the state plane boundary. These plots were prepared with a legend, scale bars, titles, and other text. Three copies of each quad were produced. Revisions 1995 - data were re-projected from original state plane system to an Albers projection (see Spatial Reference Information) 1995 - version 2.0 update of classified data in the Southwest Coastal Plain and aggregation of pixels from 100 ft (30 meters) to 200 ft (60 meters). 1992 - version 1.0 original version 100 ft pixels. Coverage was renamed on October 22, 1997 from LANDCOVER to LANDCOV in preparation for publication to CD-ROM. 2000 - Datum change from 1927 to 1983 (North American Datum of 1983) Reviews_Applied Accuracy Assessment Limited ground-truthing was done during accuracy assessment of this landcover database. Class determinations met a minimum accuracy level of 85 percent, although sample sizes may have been too small for some classes and for some regions. No confusion matrices or higher level accuracy assessments were produced. The ground-truthing of the landcover database as well as training sample collection was performed by Georgia DNR personnel. Training sample collection, which was aided by paper plots of the satellite imagery and an unsupervised classification for the region being mapped, was conducted by driving to each site and assessing it on the ground. A sample of the sheet filled out for each sample is included (see Notes I). Accuracy assessment of the landcover database involved examination of random samples by helicopter. The random test sites were selected by computer and stratified according to class prevalence. In order to overcome problems of accessibility, determining exact location, and the large size of the regions to be covered in a limited amount of time, a helicopter equipped with a Loran system was utilized. The Loran enabled the pilot to enter coordinates for a site and be electronically guided to its location. The DNR personnel involved in the assessment were provided paper print of the satellite imagery for each site with the 3 pixel X 3 pixel site delineated as well as the Loran coordinates for each site. The landcover class types found in the field for the test sites were then compared with the class types for the same pixels in the landcover database. As each region was determined by DNR to meet the 85 percent overall accuracy requirement, it was passed on to the next stage of the project process. 1. Documentation review by Chris Canalos, GA DNR, Natural Heritage Program was incorporated into the documentation narrative June 20, 1996. Mr. Canalos was responsible for initial quality control and assurance procedures in the development and revision of the landcover data layer. 2. Colleague review June 20, 1996 by Keith McFadden and Jonathan Musser Digital data was reviewed with respect to positional accuracy, contextual accuracy, attribute accuracy, topological consistency, and metadata content. 3. Approved by Georgia District Chief, Tim Hale, on June 21, 1996. 4. Additional Comments by Chris Canalos, Georgia DNR, Natural Heritage Program (July 12-16, 1996 comm.) Since the published color color scheme for the landcover dataset was developed specifically to separate and highlight each class appropriately, it was agreed upon to include the color lookup values for each landcover value. See section on Related_Spatial_and_Tabular_Data_Sets for the Red Green Blue values for the color lookup table. Related_Spatial_and_Tabular_Data_Sets A summary of landcover statistics by county including acreages and percentages was produced in conjunction with the database and published as Project Report No. 26, State of Georgia Landcover Statistics by County. It is available through the Georgia Geologic Survey, Georgia Department of Natural Resources Map Sales office (GGS Phone 404-656-3214). (Added July 18, 1996 - Jack Alhadeff) Color Lookup Table for each of 15 landcover categories expressed as Red Green Blue (RGB) values. Class R G B 1 0 230 255 2 153 0 153 3 230 230 77 4 153 89 0 5 230 230 230 6 255 255 255 7 255 0 0 8 230 153 0 9 255 255 0 10 0 153 128 11 64 217 145 12 186 255 217 13 230 0 230 14 204 255 255 15 247 247 237 Nodata should be given a white color when producing hardcopy (255 255 255). Other_References_Cited ERDAS, Inc, and Georgia Department of Natural Resources, Wetlands and Landcover Mapping Project, Final Report, Atlanta, Georgia, February 5, 1992 Clark, William Z., Jr and Zisa, Arnold C., Physiographic Map of Georgia, Georgia Department of Natural Resources, Atlanta, Georgia, 1976. Notes I. SAMPLE SURVEY FORM TAKEN BY: DATE: TRAINING SAMPLE #: FIELD VERIFICATION SAMPLE #: ESTIMATED SAMPLE SIZE: TOPO QUAD NAME: LOCATION MAP COORDINATES: INDICATE LOCATION AND EXTENT ON MAP AND/OR PHOTO AERIAL PHOTO INDEX #: LANDCOVER TYPE 1. OPEN WATER 4. CULTIVATED/EXPOSED EARTH ____Stream or Waterway ____Crop type ____Lake ____Irrigated Non-irrigated ____Reservoir ____Gravel pit/quarry ____Bays or Estuary ____Sand bar/beach ____Other (specify)_________ ____Other (specify)_______ 2. CLEAR CUT/ 5. LOW DENSITY URBAN YOUNG PINE PLANTATION ____Clearcut, not planted Pct Vegetative cover_______ ____Pine plantation Land use_________________ ____Age/height of stand _________________________ ____Understory species _________________________ 3. PASTURE 6. HIGH DENSITY URBAN Pct Tree cover___________ PctVegetative cover_______ Tree species___________ Land use_________________ _______________________ _________________________ Grass species__________ _________________________ _______________________ _________________________ _____7. Emergent Wetland _____10. Coniferous Forest ___13.Salt Marsh _____8. Scrub/Shrub Wetland _____11. Mixed Forest _____14.Brackish Marsh _____9. Forested Wetland _____12. Hardwood Forest _____15.Tidal Flats/ Beaches Pct Tree canopy cover ________Average canopy ht.__________Average dbh_____ Dominant tree canopy (>2Oft.) species:____________________________________ Pct Unnderstory or scrub/shrub cover:_________ Dominant under-to- or scrub/shrub (6-2Oft.) species:____________________________________ Pct Groundlayer species cover:________________ Dominant groundlayer (<6f.) species:____________________________________ Pct Bare ground:______________Slope:______________________Aspect_____ Other notes (vegetation phenology, evidence of disturbance, soil characteristics, unusal land forms, surrounding land use, etc.) ____________________________________________________________ ____________________________________________________________ ____________________________________________________________ Altered Wetland: Current Land Cover/Use ____________________ ____________________________________________________________ II. LANDSAT scenes acquired for use in classification of wetlands and landcover Path/Row Scene lD# Date Scene Size 16/38 Y5175415245X0 12/19/88 NW 1/4 17/37 Y5I76115304X0 12/26/88 Full 17/38 Y5176115310X0 12/26/88 Full 17/39 YSI76115313X0 12/26/88 Full 18/36 Y5178415361X0 1/18/89 Full 18/37 Y7178415363X0 1/18/89 Full 18/38 Y5178415370X0 1/18/89 Full 18/39 Y5I78415372X0 1/18/89 Full 19/36 Y52l4315345X0 1/12/90 Full 19/37 Y5214315351X0 1/12/90 Full 19/38 Y5214315354X0 1/12/90 Full 19/39 Y5212715363X0 12/27/89 NE 1/4 20/36 Y5173415490X0 11/29/88 Full III.Sub-regional Analysis Based On Physiographic Regions and Spectral Variation *Physiographic regions have been modified to allow for separation by Landsat Satellite path. 1.) Cumberland Plateau 2.) Southern Valley and Ridge 3.) Southern Blue Ridge 4.) Upland Georgia Subsection of the Southern Piedmont 5.) Midland Georgia Subsection of the Southern Piedmont 6.) Sea lsland Section - Barrier lslands Sequence 7.) Sea lsland Section - Vidalia Upland District (Portion of D) 8.) East Gulf Coastal Piain - Fall line Hills District and Fort Valley Plateau Distrist (Portions A, B, C and D) 9.) East Gulf Coastal Plain Section - Tifton Upland District 10.) Sea lsland Section - Bacon Terraces District and Okefenokee Basin District IV. Additional Classes For the Barrier lsland Sequence District, the following three classes were added: (13) Salt Marsh. Extensive areas dominated by smooth cordgrass are representative of this landcover class. It is spectrally distinguished from brackish marsh and emergent wetlands fairly accurately. This class, the brackish marsh class, and the tidal flat class were used only in the lower coastal region. (14) Brackish Marsh. Brackish marsh is typified by low-salinity emergent wetlands dominated by needlerush or giant cordgrass. These areas are usually found upstream or upslope from the salt marsh, and often adjacent to freshwater marsh and forested wetlands. Where brackish marsh occurs in close proximity with deciduous forested wetland, some spectral similarity may exist. (15) Tidal Flats. This class represents areas with little or no vegetation in the coastal region and includes mud flats, spoil areas, and beaches. It is similar in concept to the cultivated/exposed earth class, but spectrally different enough such that for the lower coastal regions these areas are generally distinguishable from other landcover types. Spectral similarity with urban classes may result in misrepresentation of some urban areas as tidal flats. Time_Period_of_Content Range_of_Dates/Times Beginning_Date: 1988 Ending_Date: 1990 Currentness_Reference Classification complete on data acquired from 1988-1990. Status Progress: Published by GA DNR. Approved for release by Tim Hale, Georgia District Chief Maintenance_and_Update_Frequency Mininal, although re-classification may occur when time and funding are available. Spatial_Domain Bounding_Coordinates West_Bounding_Coordinate: -85.68399187 East_Bounding_Coordinate: -80.58282271 North_Bounding_Coordinate: 35.01832876 South_Bounding_Coordinate: 30.28513896 Keywords Theme Theme_Keyword_Thesaurus: None Theme_Keyword: Landcover Wetlands Classification Imagery Georgia Place Place_Keyword_Thesaurus: None Place_Keyword: State of Georgia Stratum Stratum_Keyword_Thesaurus: None Stratum_Keyword: Temporal Temporal_Keyword_Thesaurus: None Temporal_Keyword: Access_Constraints Database is in the public domain. Guidance for appropriate use can be obtained from point of contact. Recognition of Georgia Department of Natural Resources would be appreciated. Use_Constraints Complete statewide coverage. To be used for regional analysis and planning, some countywide statistics, but not for use on site-specific land parcels or small watersheds. Limited ground-truthing was done on these landcover map files. Class determinations met a minimum accuracy level of 85 pct. These map files do not represent an attempt to define jurisdictional limits of any federal, state, or local government with respect to wetlands or any other landcover type. The Georgia Department of Natural Resources, the Georgia Department of Community Affairs and ERDAS, Inc. assume no liability in the use of these map files. Point_of_Contact: See Data_Set_Credit Dr. Jon Ambrose, Program Manager Georgia Department of Natural Resources Wildlife Resources Division, Natural Heritage Program 2117 Highway 278 SE Social Circle, Georgia 30025 (706) 557-3032 Christopher G. Canalos, G.I.S. Specialist Georgia Department of Natural Resources Wildlife Resources Division, Natural Heritage Program 2117 Highway 278 SE Social Circle, Georgia 30025 (706) 557-3032 ERDAS, Inc 2801 Buford Highway Suite 300 Atlanta, Georgia 30329 (404) 248-9000 Security_Information Security_Classification_System: None Security_Classification: UNCLASSIFIED Security_Handling_Description: None Native_Data_Set_Environment: dgux UNIX, ARC/INFO version 7.0.4 Cross_Reference Originator: Georgia Department of Natural Resources Publication_Date: 1995 Publication_Time: Title: Landcover Classification of Georgia 1988-1990 Edition: 2.0 Geospatial_Data_Presentation_Form: Series_Information Series_Name: Digital Data Series Issue_Identification: Publication_Information Publication_Place: Atlanta, GA Publisher: GA Dept. of Natural Resources Other_Citation_Details: Online_Linkage: Larger_Work_Citation: --------------------------------------------------------------------------- Data_Quality_Information Attribute_Accuracy Attribute_Accuracy_Report: See Logical_Consistency_Report: Not applicable for raster data. Completeness_Report All areas within the State of Georgia acquired and classified, 1988-1990. Positional_Accuracy Horizontal_Positional_Accuracy Horizontal_Positional_Accuracy_Report: Limited ground-truthing was done on these landcover map files. Class determinations met a minimum accuracy level of 85 percent. These map files do not represent an attempt to define jurisdictional limits of any federal, state, or local government with respect to wetlands or any other landcover type. The Georgia Department of Natural Resources, the Georgia Department of Community Affairs and ERDAS, Inc. assume no liability in the use of these map files. see Reviews_Applied_to_Date section Quantitative_Horizontal_Positional_Accuracy_Assessment: Horizontal_Positional_Accuracy_Value: 60 meters Horizontal_Positional_Accuracy_Explanation: Resolution as reported Vertical_Positional_Accuracy Vertical_Positional_Accuracy_Report: n/a Lineage: See Cloud_Cover datasets chosen to be "Cloud Free" --------------------------------------------------------------------------- Spatial_Data_Organization_Information Direct_Spatial_Reference_Method: Raster Raster_Object_Information: Raster_Object_Type: Grid Cell Row_Count: 8771 Column_Count: 7694 --------------------------------------------------------------------------- Spatial_Reference_Information Horizontal_Coordinate_System_Definition Planar Map_Projection: Map_Projection_Name: ALBERS Longitude_of_Central_Meridian: -83.5 Latitude_of_Projection_Origin: 23 Latitude_of_First_Standard_Parallel: 29.5 Latitude_of_Second_Standard_Parallel: 45.5 False_Easting: 0.00000 False_Northing: 0.00000 Geodetic Model Horizontal_Datum_Name: North American Datum of 1983 Ellipsoid_Name: GRS 80 Semi-major_Axis: 6,378,206.4 Denominator_of_Flattening: 294.98 --------------------------------------------------------------------------- Entity_and_Attribute_Information Detailed_Description Entity_Type Entity_Type_Label: LANDCOV.VAT Entity_Type_Definition: Arc/Info GRID VAT Based on Original Classification of Data Entity_Type_Definition_Source: Georgia Department of Natural Resources Attribute: Attribute_Label: - Attribute_Definition: Arc/Info GRID VAT Based on Original Classification of Data Attribute_Definition_Source: Georgia Department of Natural Resources Attribute_Domain_Values Enumerated_Domain Enumerated_Domain_Value: - Enumerated_Domain_Value_Definition Enumerated_Domain_Value_Definition_Source: Attribute: Attribute_Label: VALUE Attribute_Definition: Internal feature number for GRIDs Attribute_Definition_Source: Computed Attribute_Domain_Values Enumerated_Domain Enumerated_Domain_Value: Integer Enumerated_Domain_Value_Definition Enumerated_Domain_Value_Definition_Source: Attribute: Attribute_Label: COUNT Attribute_Definition: Number of GRID cells of a VALUE Attribute_Definition_Source: Computed Attribute_Domain_Values Enumerated_Domain Enumerated_Domain_Value: Integer Enumerated_Domain_Value_Definition Enumerated_Domain_Value_Definition_Source: Entity_Type Entity_Type_Label: LANDCOV.STA Entity_Type_Definition: Statistics table Entity_Type_Definition_Source: GRID Attribute: Attribute_Label: - Attribute_Definition: Statistics table Attribute_Definition_Source: GRID Attribute_Domain_Values Enumerated_Domain Enumerated_Domain_Value: - Enumerated_Domain_Value_Definition Enumerated_Domain_Value_Definition_Source: Attribute: Attribute_Label: MIN Attribute_Definition: Minimum GRID cell VALUE Attribute_Definition_Source: Computed Attribute_Domain_Values Enumerated_Domain Enumerated_Domain_Value: Real number Enumerated_Domain_Value_Definition Enumerated_Domain_Value_Definition_Source: Attribute: Attribute_Label: MAX Attribute_Definition: Maximum GRID cell VALUE Attribute_Definition_Source: Computed Attribute_Domain_Values Enumerated_Domain Enumerated_Domain_Value: Real number Enumerated_Domain_Value_Definition Enumerated_Domain_Value_Definition_Source: Attribute: Attribute_Label: MEAN Attribute_Definition: Mean GRID cell VALUE Attribute_Definition_Source: Computed Attribute_Domain_Values Enumerated_Domain Enumerated_Domain_Value: Real number Enumerated_Domain_Value_Definition Enumerated_Domain_Value_Definition_Source: Attribute: Attribute_Label: STDV Attribute_Definition: Standard Deviation of GRID cell VALUEs Attribute_Definition_Source: Computed Attribute_Domain_Values Enumerated_Domain Enumerated_Domain_Value: Positive real number Enumerated_Domain_Value_Definition Enumerated_Domain_Value_Definition_Source: Overview_Description Entity_and_Attribute_Overview All of the data from the original raster imagery were classified into 15 landcover classes. Titles for the classes represent the majority of the landcover making up the class. Some of the data points represent landcover that does not fall clearly into any of the classes. These were nevertheless included in the class that was spectrally most similar. Some of the original data points were spectrally similar to more than one class, in which case the maximum likelihood classifier algorithm selected the most likely class to contain the data. With the possibilities of classes containing data that does not clearly fall into any one class and data that could fall into more than one class, a description of the intended and possible contents of the individual classes is given below. (1-12, below; 13-15 see Notes Section IV) (1) Open Water. Lakes, reservoirs, coastal waters, ponds and wide stream channels with little or no emergent vegetation are included in this class. On the unclassified imagery, open water appears dark, similar to shadows behind northwest-facing slopes; therefore, some shadow areas are included. (2) Clearcut/Young Pine. The spectral characteristics recently cleared in timber harvest operations and planted to pine or left unplanted are usually quite different from those of other landcover types. The clearcuts are often large in area and regularly shaped. The typical clearcut/young pine stand has widely-spaced woody vegetation with a ground cover of herbs and grasses. This vegetation type can be seen as transitional to closed-canopy coniferous forest. Any cleared land can be spectrally similar to timber clearcuts, including some agricultural land such as abandoned pasture and fallow cropland This class was under-estimated in extreme South central Georgia. (3) Pasture. Pasture land is distinguished from other agricultural land by the presence of low-growing herbaceous vegetative cover year round. This class includes actual pastures, as well as lawns, fields, and other open areas within urban areas. Pasture can be spectrally similar to cultivated fields that have vegetative cover during the winter. Pixels of the clearcut/young pine and cultivated/exposed earth classes are often found intermingled. This class was underestimated in Southwest Georgia. (4) Cultivated/Exposed Earth. Agricultural fields with no winter vegetation, and any other areas where vegetation has recently been removed, exposing soil or rock, are represented by this class. Exposed banks around reservoirs with low water levels often are included in this class. Some cultivated fields showing winter vegetation are spectrally similar to pasture. This class may be found within urban areas and in conjunction with the pasture and clearcut/young pine classes in other areas. (5) Low Density Urban. The high reflectivity of man-made structures in urban areas provides for some separation of urban classes from the non-urban classes. The low density urban class represents urban areas with moderate vegetative cover. However any area with high reflectivity, such as isolated industrial sites, may fall into this or the high density urban class. The edges of some bodies of water are spectrally similar to this class. It is typical for residential areas to be shown as a matrix of this class and forest class pixels. Low density urban may be interspersed with high density urban. (6) High Density Urban. This class is distinguished from low density urban by an even higher reflectivity of the landcover. Paved areas with buildings and little vegetation are typical of this landcover class. Roads are often shown as linear features composed of high and low density urban pixels. High density urban pixels found outside of urban areas are indicative of any type of highly reflective structure/ feature such as power substations, grain storage bldgs. (7) Emergent Wetland. Emergent wetlands are spectrally and ecologically transitional between open water and scrub/shrub wetlands. Freshwater marsh vegetation with few woody plants interspersed is typical of the cover type. Where clusters of emergent wetland pixels are found, other wetland types and open water are often in proximity. This class may show up in some non-wetland areas with low-reflectivity cover. (8) Scrub/Shrub Wetland. Intended for wetland vegetation dominated by woody plants less than 20 feet in height, this class contains areas in transition between emergent and forested wetlands. This class is usually found in conjunction with other wetland classes. Where uplands with woody vegetation border open water, pixels from this class may be shown. When found singly within a matrix of low urban density and forest pixels, it is more likely that cover spectrally similar to but not actually scrub/shrub wetland is being shown (i.e., scrubby vegetation over some low-reflective surface). (9) Forested Wetland. Where spectral differences are pronounced, this class may be distinguished from scrub/shrub wetland and upland forest types. Where upland tree canopies overhang river banks or edges of water bodies, pixels from this class may show. These edges may or may not be actual wetlands. Areas of swamp are often shown as mixtures of forested wetland and hardwood forest pixels. Individual or small clumps of pixels in this class when found scattered throughout urban areas may be showing non-wetland areas with spectral similarity to wetlands, such as woody vegetation over low-reflective surfaces. Classification of forested wetlands dominated by deciduous trees is probably more accurate than that in areas with evergreen, closed canopies. In the latter case, the low reflectivity of the wet areas underneath the canopy may not be picked up by the sensor, making them difficult to distinguish from upland evergreen forest canopies. Spectral similarity between this class and shadows behind northwest-facing slopes may account for the presence of forested wetland pixels shown on some slopes. (10) Coniferous Forest. The uniformity of large tracts of planted pines provides for accurate classification of this landcover type in upland areas. These stands may be fringed or bisected by the other forest types. Spectral similarity with evergreen hardwood forest in the Coastal Plain may result in difficulty in distinguishing between these two cover types. Where pine canopies are dense, as is often the case, it may be difficult to determine whether the sites are upland or wetland. (11) Mixed Forest. Typically, this class represents mixed stands of hardwood and coniferous trees, neither type exceeding 60-70 percent of the stand. Pine plantations in transition from early stages to forest may be shown in this class, although few if any hardwood trees may be present. Edges of coniferous stands and areas of transition between coniferous and hardwood forest are often shown with this class. Also included may be abandoned cut-over areas. (12) Hardwood Forest. Stands of deciduous hardwoods are generally distinguished from forested wetlands and other forest classes accurately. Evergreen hardwood forests may be spectrally similar to mixed and coniferous classes, and, due to a closed canopy, may be difficult to distinguish from evergreen forested wetlands. River floodplains are often depicted as a mixture of forested wetland and hardwood forest pixels, with drier areas shown as hardwood forest. Cut-over lands with young, shrubby hardwood growth, although not forest, may make up part of this class. Entity_and_Attribute_Detail_Citation: Not Available --------------------------------------------------------------------------- Distribution_Information Distributor Contact_Person_Primary Contact_Person: Jonathan W. Musser Contact_Organization: U.S. Geological Survey Contact_Position: Hydrologist Contact_Address Address_Type: mailing address Address: U.S. Geological Survey 3039 Amwiler Road Suite 130 City: Atlanta State_or_Province: GA Postal_Code: 30360-2824 Country: USA Contact_Facsimile_Telephone: 770-903-9199 Contact_Electronic_Mail_Address: jwmusser@usgs.gov Standard_Order_Process Digital_Form Digital_Transfer_Information Format_Name: ARCE ARC/INFO Export format File_Decompression_Technique: gzip Digital_Transfer_Option Online_Option Computer_Contact_Information Network_Address Network_Resource_Name: http://csat.er.usgs.gov/statewide/layers/landcov.html Distributor Contact_Person_Primary Contact_Person: Jonathan W. Musser Contact_Organization: U.S. Geological Survey Contact_Position: Hydrologist Contact_Address Address_Type: mailing address Address: U.S. Geological Survey 3039 Amwiler Road Suite 130 City: Atlanta State_or_Province: GA Postal_Code: 30360-2824 Country: USA Contact_Facsimile_Telephone: 770-903-9199 Contact_Electronic_Mail_Address: jwmusser@usgs.gov Standard_Order_Process Digital_Form Digital_Transfer_Information Format_Name: ARC/INFO GRID Format File_Decompression_Technique: gzip & tar Digital_Transfer_Option Online_Option Computer_Contact_Information Network_Address Network_Resource_Name: http://csat.er.usgs.gov/statewide/layers/landcov.html Distributor Contact_Person_Primary Contact_Person: Jonathan W. Musser Contact_Organization: U.S. Geological Survey Contact_Position: Hydrologist Contact_Address Address_Type: mailing address Address: U.S. Geological Survey 3039 Amwiler Road Suite 130 City: Atlanta State_or_Province: GA Postal_Code: 30360-2824 Country: USA Contact_Facsimile_Telephone: 770-903-9199 Contact_Electronic_Mail_Address: jwmusser@usgs.gov Standard_Order_Process Digital_Form Digital_Transfer_Information Format_Name: ERDAS image files (ERDAS Corporation) File_Decompression_Technique: gzip Digital_Transfer_Option Online_Option Computer_Contact_Information Network_Address Network_Resource_Name: http://csat.er.usgs.gov/statewide/layers/landcov.html --------------------------------------------------------------------------- Metadata_Reference_Section Metadata_Date: 19971219 Metadata_Contact: dmleckie Metadata_Standard_Name: FGDC Content Standards for Digital Geospatial Metadata Metadata_Standard_Version: 19940608 Metadata_Time_Convention: Local Time Metadata_Security_Information: Metadata_Security_Classification_System: None Metadata_Security_Classification: UNCLASSIFIED Metadata_Security_Handling_Description: None --------------------------------------------------------------------------- Last modified: 2008-03-20