 |
        |
           |
|
|
|
|
Identification_Information:
Citation:
Citation_Information:
Originator: Kansas Applied Remote Sensing
Publication_Date: 1993
Title: Land Cover
Online_Linkage: http://gisdasc.kgs.ukans.edu
Description:
Abstract: The Land Cover database depicts 10 general land cover classes for the State of Kansas. The database was compiled from a digital classification of Landsat Thematic Mapper (TM) imagery.
Purpose: This database was developed as part of the Core Database for the State of Kansas. It is suited for county-level and watershed-level analyzes that involve land use and land cover.
Time_Period_of_Content:
Time_Period_Information:
Single_Date/Time:
Calendar_Date: publication date of county
Currentness_Reference: publication date of county
Status:
Progress: Complete
Maintenance_and_Update_Frequency: None
Spatial_Domain:
Bounding_Coordinates:
West_Bounding_Coordinate: -102.5
East_Bounding_Coordinate: -94.6
North_Bounding_Coordinate: 40.0
South_Bounding_Coordinate: 37.0
Keywords:
Theme:
Theme_Keyword_Thesaurus: None
Theme_Keyword: Land Cover, Land Use, Cropland, Grassland, Woodland, Urban, Water
Access_Constraints: None
Use_Constraints: This database is not suited for site-specific analyzes. Interpretations derived from its use are intended for planning purposes only.
Point_of_Contact:
Contact_Information:
Contact_Organization_Primary:
Contact_Organization: Kansas Applied Remote Sensing
Contact_Person: Jerry Whistler
Contact_Address:
Address_Type: mailing address
Address: 2291 Irving Hill Road
City: Lawrence
State_or_Province: KS
Postal_Code: 66045-2969
Country: USA
Contact_Voice_Telephone: (785) 864-3107
Contact_Facsimile_Telephone: (785) 864-7789
Contact_Electronic_Mail_Address: whistler@falcon.cc.ukans.edu
Browse_Graphic:
Browse_Graphic_File_Name: landcvr
Browse_Graphic_File_Description: Land Cover for Chautauqua County
Browse_Graphic_File_Type: GIF
Native_Data_Set_Environment: Arc/Info
Data_Quality_Information: Attribute_Accuracy: Attribute_Accuracy_Report: An accuracy report, in the form of an error matrix, accompanies each county dataset. From the error matrix, errors of omission and commission can be obtained. Two summary statistics, percentage correct and Kappa, are provided for the entire county. Accuracy was calculated by comparing classified data with manually interpreted 1985 National High Altitude Photography (NHAP2) and 1986 State of Kansas reappraisal photography. The 1985 NHAP2 photos were color infrared 9" transparencies with a nominal scale of 1:58,200. The reappraisal photos were black-and-white 9" transparencies with a nominal scale of 1:24,000. Due to differences in dates between the satellite imagery and the aerial photos (anywhere from 2 to 5 years), reported accuracies more properly represent percent co-occurrence between data. These aerial photos were used for accuracy assessment because they were the most recent photography available for the entire state. A systematic sampling procedure for each county was adopted to ensure that at least 5% of the total county area was covered by sample sites (county area x 0.05). Individual sample sites within a county were defined as four square miles (2 x 2 sections). Consequently, the number of sites per county ranged between three (Wyandotte County, 149 miles2) and eighteen (Butler County, 1443 miles2). Sites were placed within a county to ensure adequate sampling of major water bodies and urban-rural fringe areas. Prior to photo interpretation, base map information (i.e., roads and section lines) was digitized from paper 7.5' USGS topographic quadrangles and then plotted onto mylar. A Kargl reflecting projector or Bausch & Lomb Zoom Transfer Scope (depending on interpreter preference) was then used to bring the aerial photo into registration with the mylar base map. Land cover classes were manually interpreted from the photography directly onto the mylar, digitized using Arc/Info, and converted from vector to raster format. The TM-derived land use/land cover map was then cross tabulated with the photointerpreted sites, generating statistics for errors of omission and commission, accuracy by land cover class, overall accuracy, and a Kappa statistic. These statistics, along with information such as image source, image date, image path/row, image scene-id, and date of coverage completion, were submitted as data set documentation for the county. Quantitative_Attribute_Accuracy_Assessment: Attribute_Accuracy_Value: >=85% Attribute_Accuracy_Explanation: Accuracy values vary by county and an accuracy report that accompanies each county dataset provides overall and class accuracy figures for the county. In general, however, all counties have an overall attribute accuracy of 85% or greater. Logical_Consistency_Report: Datasets delivered to the Data Access and Support Center are quality assurance tested for major documentation and topological errors using macros. The databases are evaluated based on questions similar to the following: Does the database's documentation properly explain the dataset? Does the data import properly into its native dataset environment? Is consistent methodology used throughout the data? Are all of the database's fields and values legitimate compared to the documentation? Is the database's topology free of errors that could impair the functionality of the data? Once the data is tested a report is written describing the evaluation. A copy of the report is then forwarded to the data originator for his/her review. If errors are found, the data originator is expected to make any necessary corrections and then provide those corrected coverages. After these steps have been taken data is considered archivable and made available for distribution. Data that is archived by DASC has been tested and corrected. Completeness_Report: Lineage: Source_Information: Source_Citation: Citation_Information: Originator: Earth Observation Satellite (EOSAT), Inc. Publication_Date: 198806 to 199009 Title: Landsat Thematic Mapper raster imagery Type_of_Source_Media: 9-track, 1600bpi magnetic tape Source_Time_Period_of_Content: Time_Period_Information: Single_Date/Time: Calendar_Date (YYYYMMDD): 1991 Source_Currentness_Reference: image acquisition date Source_Citation_Abbreviation: EOSAT, Inc. Source_Contribution: month, day, and year image was acquired by the satellite Process_Step: Process_Description: Sixteen Landsat Thematic Mapper scenes were required to completely cover the state of Kansas. Several factors were taken into consideration for data selection. Timeliness of the data was a primary concern, limiting the data to scenes acquired within 1-2 years of the project start date (1/90). For optimal distinction of agricultural land use and forest cover, the data were acquired for the growing season, between June 1 and September 30. Scenes were required to be completely cloud-free, except in the small areas of overlap between scenes. As a result of these constraints, data from 1988, 1989, and 1990 were used. Whenever possible, all scenes for a particular path were ordered with the same acquisition date. Data tapes were inspected for cloud cover, line dropout, and scene noise. Data from each full TM scene were downloaded to disk and areas slightly larger than the individual counties were then subset. In cases where county boundaries extended across two scenes within a row (e.g., Path 29, Rows 33 and 34), common control points and a cubic convolution resampling algorithm were used to combine scene fragments to create a complete county. If radiometric differences between scenes from adjacent rows were relatively small, a histogram matching process (ERDAS programs HSTMATCH and STRETCH) was used to adjust the brightness values for one scene to match the other. Typically, the smaller of the two fragments was matched to the larger fragment. In several cases, data for a county were composed of scene fragments from adjacent paths (e.g., Paths 29 and 30) and the radiometric characteristics of the two scenes differed radically due to differences in acquisition date or data processing. Data for these special counties were processed separately and then joined after analysis was nearly complete. Data were not geometrically rectified before processing, maintaining original radiometric values whenever possible. It was found during the prototype phase that rectification before classification "smeared" the spectral values for stock ponds such that they were no longer classified as water. This was unacceptable because, although they are small, stock ponds represent an important, and sometimes substantial, water resource in many Kansas counties. The data were processed county by county for two reasons. First, this approach served to stratify the data into a areas small enough that a degree of homogeneity within the spectral classes could be maintained. Second, counties were natural units with which to tile the data and also release to users. Initial processing of the digital data consisted of clustering and classification of rural and urban classes and the manual digitization and overlay of specified additional cover types. Six of the TM bands (all except band 6, thermal) were used for clustering. An unsupervised classification technique was used, with the data being initially clustered into a large number of spectral classes and then consolidated by an analyst into the five rural and five urban information classes. Certain features easily visible on the imagery but not easily classified using spectral classifiers were manually digitized and then merged with the classified raster data. For the five rural classes, an ISODATA clustering algorithm was used to create 100 spectral clusters which were then classed using a minimum-distance classifier. The number of clusters was set at 100 based on previous broad-area land cover mapping experience at KARS. For a few counties, different numbers of clusters were tried, but with limited success. A smaller number of clusters (for example, 75) resulted in more confusion classes, while a larger number of clusters (e.g., 125 or 150) failed to resolve the most difficult confusion classes. Tests between the maximum-likelihood and the minimum-distance classifiers during the pilot phase revealed no significant difference in classification accuracy or class area. The minimum-distance classifier, therefore, was chosen because of its speed and relative reliability. After the initial rural classification was completed, the analyst assigned the 100 spectral classes to the five rural information classes: cropland, grassland, woodland, water, and other. This was accomplished using an iterative two-step process. In the first step, the analyst displayed an area of the county for which NHAP2 air photo coverage was available. Normally, the image was displayed in bands 4, 5, and 7 (RGB), which gave best visual separability of the major information classes. The ERDAS CLASOVR function was used to overlay the 100-class raster file over the image, using either the blue or red color gun, depending on analyst preference. Spectral classes were then highlighted one at a time while the analyst made notes of the tentative assignment of the classes to information classes, referring when necessary to the NHAP2 reference photo. The analyst's notes included a judgment of whether classes were relatively "pure" or confused. Following this, the analyst then displayed the 100-class raster file alone and used the ERDAS COLORMOD function to assign information class colors to the individual classes. The analyst first colored up the relatively pure classes, then went back and made a best fit of the remaining, confused, classes using the context of the pure classes as a visual guide. Once this was complete and the analyst was satisfied with the class assignments, the RECODE function was used to create a five-class raster map of rural classes. A similar process, with minor modifications, was used to derive the five urban classes: residential, commercial/industrial, open land, woodland, and water. Whereas in the rural classification, an entire county image area was used for creating the ISODATA clusters, for the urban classification an image subset consisting primarily of urban cover types was used for creating the ISODATA clusters. Fifty clusters were created for the urban classes, a number that was arrived at by trial and error. As with rural classification, a minimum-distance classifier was used to create a 50-class raster file, after which the two-step CLASOVER and COLORMOD process was used to assign spectral classes to urban information classes. Again, RECODE was used to create a five-class raster map of urban land cover. Certain landscape features were easily visible in the imagery but were not easily classified using a spectral classifier. In many cases, a feature such as a highway would be classified with partial success, but would not be recognizable after cartographic generalization. In other cases, certain classes of features would consistently be misclassified (for example, sandbars in rivers and streams were frequently confused with the bare soil of plowed cropland). It was decided that some of these classes of features were desirable in the final land cover map, including four-lane highways, major dams, sandbars, and large quarries and gravel pits. These were screen digitized from images displayed on the computer screen and then gridded into raster GIS files for later merging with the rural and urban raster GIS maps. Since separate urban and rural classifications were performed, it was necessary to create urban masks in order to be able to merge the urban classes into the rural base map. This was accomplished by screen digitizing the urban outlines directly from the imagery displayed on the screen. During this process, reference was made to aerial photography of urban areas in Kansas available in the KARS library. These photos included Agricultural Stabilization and Conservation Service (ASCS) crop monitoring slides, State of Kansas Reappraisal photography, Forest Service photography, and color infrared National High Altitude Photography (NHAP2). As with almost any automated classification technique, the initial raw land cover maps were characterized by "speckling" produced by misclassification of single pixels or small groups of pixels. To eliminate much of the visual noise caused by these misclassified pixels, the raw maps were generalized using a series of both automated and manual procedures. To remove single isolated pixels Remove1, a program written at KARS, was employed. Remove1 removes single isolated pixels for specified classes, for example, crop, grass, and woodland. The water classes, on the other hand, were not included in Remove1 because single water pixels often represented actual water bodies, such as small stock ponds. Following a first run of Remove1, a second program, Smoother, was used to remove short rows or columns of pixels jutting out anomalously from large homogeneous areas. Finally, Remove1 was run a second time to remove any isolated pixels created by running Smoother. This initial generalization procedure not only had the effect of removing noise but also of greatly reducing the number of polygons in the map. This reduction of polygons was important in the raster-to-vector conversion process, because the number of polygons in a county map frequently approached or exceeded the then Arc/Info limit of 10,000 polygons in a coverage. After raster generalization, the county maps were converted from ERDAS raster format to Arc/Info coverages and further generalized. The Arc/Info ELIMINATE function was run in order to remove polygons not meeting the minimum mapping unit requirements of the project. Thresholds for polygon elimination varied according to land cover type: 1 acre for urban woodland and woodland (classes 4 and 8), 3 acres for Urban Residential, Urban Commercial/Industrial, Urban Openland, and Other (classes 1, 2, 3, and 10), and 5 acres for Cropland and Grassland (classes 6 and 7). The Urban Water and Water classes (classes 5 and 9) were not subjected to elimination, the rationale being that in the majority of cases small water polygons represented actual water bodies, as described above, rather than misclassified pixels. As the project progressed generally from east to west across Kansas, several problems with the automated classification procedures arose that made it necessary to make modifications in the methodology. The first group of problems involved misclassifications of crop and grassland. With regard to grassland, areas of exposed soil on sparse or over-grazed rangeland often were misclassified as cropland, being confused with the bare soil of plowed fields. A second problem with grassland was that lush riparian areas of mixed grasses and shrubs were often confused with cropland due to their high infrared reflectance. With cropland, some areas of fallow fields were confused with grassland, possibly because of weedy areas that had grown up in them. Because these areas of confusion were often quite extensive and did not lend themselves to refinement with automated classification methods, it became necessary to employ visual techniques to refine them. First, the raster generalization procedure was modified. The purpose was to remove larger clumps of misclassified pixels from crop and grassland before converting to vector format. To accomplish this, the ERDAS CLUMP and SIEVE procedures were used to generalize the two major Kansas land cover classes: cropland and grassland (classes 6 and 7). The clump size thresholds selected for these two classes varied somewhat according to the overall land cover patterns of the county in question. In general, in the eastern part of the state, where parcel sizes tend to be smaller and more interspersed, clump sizes of approximately 10 pixels were used, while in the west, where parcel sizes are larger and land cover is more homogeneous, clump sizes of up to 20 pixels were used. With much of the noise eliminated, it was easier to use visual techniques to remove remaining clusters of misclassified pixels. Our procedure was to display the raster image for the county, usually in bands 4, 5, and 7 (RGB) for best visual discrimination, and then overlay the classified raster GIS map on the computer monitor using the ERDAS GISOVR command. By toggling the display plane for the GIS map on and off, it was possible to quickly see most of the areas of gross misclassification. Then, with the image and GIS map still displayed, the ERDAS DIGSCRN function was used to outline areas of pixels to be eliminated. For difficult areas, the analyst referred to photographs or color slides of the area in question on a small light table next to the workstation. The screen digitized areas were gridded into the GIS map periodically using the ERDAS GRDPOL command. This procedure was repeated until a satisfactory map was created. Water and woodland classification also became problematic in western Kansas. Whereas in the eastern part of the state it had been possible to break out these classes with reliable regularity via the unsupervised classification, in western Kansas it became difficult to separate water bodies from saturated soils (one image in southwest Kansas had been acquired soon after heavy rains had fallen, creating numerous temporary ponds and pockets of saturated soils) and woods from shrub/grass areas along streams and rivers. Our solution was to separately classify woods and water using a supervised technique and then merge them into the classified raster GIS map. Image to map rectification was performed on the vector version of the classified data on a county by county basis. Image column-row positions were transformed to state plane feet coordinates using the Arc/Info TRANSFORM command. Initially, 15 to 20 ground control points (GCPs) were established for rectification of each county depending on county size. GCPs were chosen based on two criteria: they were easily identified on the imagery and they were evenly distributed across the county. Most often the points chosen were at section corners, especially in rural counties. Major road intersections often served as GCPs in urbanized areas. GCPs were interactively selected from the imagery to sub-pixel accuracy. The corresponding map GCPs were digitized from 1:24,000 scale USGS 7.5' topographic quadrangles. The root mean square (RMS) error for digitizer setup of the map from which the GCPs were taken was .003" (at scale, + 6 feet). Trial calculations of the affine transformation (scale, rotation and translation) were performed. They were assessed for how well the derived transformation performed vis-a-vis the reported RMS error for the transform. Transformations were allowed a total RMS error of 1/3 pixel (approximately 33 feet). Individual pixel RMS errors were allowed up to ½ pixel (49 feet). GCPs exceeding the allowable individual error or contributing too much to the overall error were eliminated. In general, between 10 and 15 GCPs were left for performing the transformation. In no case were fewer than 10 GCPs used for a county transformation. Because the counties were classified individually, edge matching problems between adjacent counties arose. Two strategies were employed for edge matching between counties. The first strategy was used with counties that were subset from the same image scene and took place while the data were still in raster form. The procedure simply "borrowed" a narrow strip, 9 pixels wide, from an adjacent, previously completed county. Input pixels from the previously competed county were allowed to overwrite pixels in the current county. The second strategy was applied to counties that were subset from two (or more) image scenes and took place after the data were converted to vector coverages and were transformed. The concept is similar to the first strategy in that a narrow strip, also 9 pixels wide was "borrowed" from an adjacent, previously completed county. In this process, a strip of the previously completed county was spliced onto the current county using the Arc/Info UPDATE command. Minor editing of arcs and polygons was necessary after the update to improve the cartographic appearance of the splice. For example, slight differences in orientation between the two transformed classified maps resulted in arcs with "jags" along the edge of the splice. In addition, small polygons were sometimes created during the update. These were deleted. Because Kansas is split into two state plane zones, North and South, some counties required temporary projection of the coverage to the Universal Transverse Mercator projection for the edge matching procedure. After edge matching, the coverages were projected back to their appropriate state plane zone. Contract specifications required the coverages to be delivered in decimal geographic coordinates. In order to maintain spatial precision for the projected coordinates, the geographic coverage was represented with double precision real numbers. The geographic coverage was then clipped with the appropriate double precision county boundary. The source data for the county boundary coverage were digitized by the Kansas Geological Survey from 7.5' USGS topographic quadrangles as part of their Kansas Cartographic Database (KCD). The KCD is part of the Data Access and Support Center's Core Database. It was discovered that the clipping process sometimes shifted the intersection location of arcs lying along the county boundary by a very small amount (1 x 10-14 degree). This necessitated a final edge matching between counties. The procedure simply snapped intersections occurring along the county boundary to a previously completed coverage. Process_Date: 199002-199302 Spatial_Data_Organization_Information: Direct_Spatial_Reference_Method: Vector Spatial_Reference_Information: Horizontal_Coordinate_System_Definition: Geographic: Latitude_Resolution: 1 x 10-14 Longitude_Resolution: 1 x 10-14 Geographic_Coordinate_Units: Decimal degrees Entity_and_Attribute_Information: Overview_Description: Entity_and_Attribute_Overview: The Land Cover dataset for the State of Kansas includes two descriptive attributes. The xx_landc.pat (xx represents any county abbreviation) contains the items GRID-CODE and COV_CLASS. A list of the attributes with a short definition follows: GRID-CODE - Land cover class type code, acceptable values are 1 through 10. COV_CLASS - Name of the land cover class; acceptable class names include Residential, Commercial/Industrial, Urban-Grassland, Urban-Woodland, Urban-Water, Cropland, Grassland, Woodland, Water, and Other. The following definitions are used for the land cover classes: Residential - residential land cover consists of areas of medium density, with a more or less even distribution of vegetative cover and house/garages, to high density, represented by multiple-unit structures such as apartment complexes. Linear residential developments along transportation routes extending outward from urban areas are included. Rural subdivisions not directly connected to the core of an urbanized area are also included. The main buildings, secondary structures, and immediate surrounding landscape are all included (i.e., house, apartment complexes, streets, garages, driveways, parking areas, lawns, trees, etc.). Commercial/Industrial - commercial/industrial land consists of areas of intensive use with much of the land covered by structures. These areas are used predominantly for the manufacture and sale of products and/or services. This category includes the central business districts of cities, towns, and villages; suburban shopping centers and strip developments; educational, governmental, religious, health, correctional and institutional facilities; industrial and commercial complexes; and communications, power, and transportation facilities. The main buildings, secondary structures, and areas supporting the basic use are all included - office buildings, warehouses, driveways, parking lots, landscaped areas, streets, etc. Urban-Grassland - urban-grassland consists of areas with uses such as golf courses, zoos, urban parks, cemeteries, and undeveloped land within an urban setting. This category also includes tracts of land that have been zoned residential or commercial, but have yet to be developed. Urban-Woodland - urban-woodland consists of wooded tracts within a town or city. These wooded tracts maybe associated with golf courses, zoos, urban parks, and other undeveloped land. Urban-Water - urban-water consists of any open surface water within a town or city. This includes ponds, lakes, sewage settling ponds, etc. Cropland - cropland includes all areas in row crop and small grains, as well as harvested land and large, uniform areas of bare ground. Grassland - this category includes all pasture (hayed land), rangeland, and other grasslands having insufficient trees and/or shrubs to be classified as "Forest". It does NOT include conservation reserve program (CRP) land. Woodland - this class includes any wooded areas having a canopy closure of 50% and greater. Water - all open water bodies, including reservoirs, lakes, ponds, rivers and streams. Other - the "other" class is used to identify land cover land use classes not previously defined. In general, this class is used for exposed bare ground other than cropland. Examples include rock quarries, sand and gravel pits, sandbars, and built-up. Entity_and_Attribute_Detail_Citation: Distribution_Information: Distributor: Contact_Information: Contact_Organization_Primary: Contact_Organization: Data Access and Support Center Contact_Address: Address_Type: mailing and physical address Address: University of Kansas 1930 Constant Avenue, West Campus City: Lawrence State_or_Province: KS Postal_Code: 66047-3726 Country: USA Contact_Voice_Telephone: 785-864-3965 Contact_Facsimile_Telephone: 785-864-5317 Contact_Electronic_Mail_Address: dasc@mongogis.kgs.ukans.edu Hours_of_Service: 0800-1700 Distribution_Liability: The State of Kansas Geographic Information Systems Core Database's digital data have been tested and their documentation carefully reviewed. However, the State of Kansas Geographic Information Systems Policy Board's Data Access and Support Center and its representatives make no warranty or representation, either expressed or implied, with respect to the digital data and their documentation, their quality, performance, merchantability, or fitness for a particular purpose. The digital data are distributed on "as is" basis, and the user assumes all risk to their quality, the results obtained from their use, and the performance of the data. In no event will the State of Kansas Geographic Information Systems Policy Board or its representatives be liable for any direct, indirect, special, incidental or consequential damages resulting from and defect in the State of Kansas Geographic Information Systems Core Database's digital data or in their documentation. This disclaimer of warranty is exclusive and in lieu of all others, oral or written, express or implied. No agent or employee is authorized to make any modification, extension, or addition to this warranty. Standard_Order_Process: Digital_Form: Digital_Transfer_Information: Format_Name: ARCE Transfer_Size: 5-10 MB Digital_Transfer_Option: Online_Option: Computer_Contact_Information: Network_Address: Network_Resource_Name: http://gisdasc.kgs.ukans.edu Access_Instructions: The State of Kansas Landcover databases are stored in ESRI's Arc/Info Interchange Format and can be downloaded from the DASC home page or by connecting directly to the DASC anonymous FTP server at gisdasc.kgs.ukans.edu. To connect to the FTP server use the login name of anonymous and your E-mail address as the password. Offline_Option: Offline_Media: CD-ROM, 3.5 inch floppy disk, 4 mm cartridge tape, or 8 mm cartridge tape Recording_Format: ISO 9660 Fees: Under the Kansas Public Records Law, DASC will attempt only to recover the costs related to the processing and distribution of core database requests. The following is a description of our Basic and Supplemental Conversion Services, as well as costs associated with the distribution of digital data: Basic Conversion Services-DASC will provide coversion services to all Federal/State/municipal tax-support agencies/entities for the cost of media and shipping and handling. Basic conversion services shall include the exportation of the Core Database in their native projection and tiling scheme into DASC supported spatial data exchange formats and technical support for the loading and importation of the data. Basic conversion services are provided to other than tax-supported organizations on a fee-for-service basis. Supplemental Conversion Services-DASC will provide supplemental conversion services to all organizations on a fee-for-service basis. Secondary services shall include the alteration of a Core Database's native projection, tiling scheme, or topological structure. Supplemental services also includes custom map production. Where applicable, recoverable cost include: 1. Labor to process the request 2. Computer processing time to extract/convert database 3. Magnetic media to distribute the data 4. Shipping and handling charges 5. Tax Below are the fees associated with each of the recoverable items: 1. Labor: $35.00/Hour 2. Central Processing Unit (CPU) computer time: $.14/Minute 3. Media: 3.5" HD floppy $.90 DC 2120 Mini data cartridge tape $22.00 8mm-112m Data cartridge tape $19.00 4mm-120m Data cartridge tape $20.00 Compact Disks $15.00 Network Transfer No Charge 4. Shipping and Handling: Varies 5. Tax (State of Kansas): 6.90% Metadata_Reference_Information: Metadata_Date: 19970320 Metadata_Contact: Contact_Information: Contact_Organization_Primary: Contact_Organization: Data Access and Support Center Contact_Address: Address_Type: mailing and physical address Address: University of Kansas 1930 Constant Avenue, West Campus City: Lawrence State_or_Province: KS Postal_Code: 66047-3726 Country: USA Contact_Voice_Telephone: 785-864-3965 Contact_Facsimile_Telephone: 785-864-5317 Contact_Electronic_Mail_Address: dasc@mongogis.kgs.ukans.edu Hours_of_Service: 0800-1700 Metadata_Standard_Name: FGDC Content Standards for Digital Geospatial Metadata Metadata_Standard_Version: 1.0 |
Sponsored by: For information regarding advertising rates Click Here! |
Copyright© 1995-2012 MindSites Group / Privacy Policy
GeoCommunity™, Wireless Developer Network™, GIS Data Depot®, and Spatial News™
including all logos and other service marks
are registered trademarks and trade communities of
MindSites Group