Data Set ID:

SMAPVEX08 Land Cover Classification Map, Version 1

This data set consists of land cover classification data derived from satellite imagery and of data obtained in the field as part of the Soil Moisture Active Passive Validation Experiment 2008 (SMAPVEX08).

This is the most recent version of these data.

STANDARD Level of Service

Data: Data integrity and usability verified

Documentation: Key metadata and user guide available

User Support: Assistance with data access and usage; guidance on use of data in tools

See All Level of Service Details

Data Format(s):
  • Binary
Spatial Coverage:
N: 39.1, 
S: 38.8, 
E: -75.6, 
W: -76.3
Spatial Resolution:
  • 10 m x 10 m
Temporal Coverage:
  • 1 June 2008 to 31 October 2008
Temporal ResolutionVariesMetadata XML:View Metadata Record
Data Contributor(s):Cosh, M.

Geographic Coverage

Other Access Options

Other Access Options


As a condition of using these data, you must cite the use of this data set using the following citation. For more information, see our Use and Copyright Web page.

Cosh, M. 2015. SMAPVEX08 Land Cover Classification Map, Version 1. [Indicate subset used]. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center. doi: [Date Accessed].

Back to Top

Collapse All / Open All

Detailed Data Description

This data set consists of land cover classification data derived from satellite imagery and of data obtained in the field as part of the Soil Moisture Active Passive Validation Experiment 2008 (SMAPVEX08). Images from the ResourceSat-1/IRS-P6 AWiFS and SPOT HRV of the Maryland study area were retrieved for the summer of 2008 and classified using in situ data collected by sampling teams in the fall of 2008. The land use classification image provides information about vegetation present in the study area at a resolution of 10 meters.


Data are provided in a binary file and a header file called ENVI FST, which indicates it is an Environment for Visualizing Images "dbFast" data file. An associated Extensible Markup Language (XML) metadata file is also provided.

Number of rows: 7956 
Number of columns: 8885 
Number of bands: 1 
File layout: BSQ 
Upper left corner in east-west direction: 375325 
Upper left corner in north-south direction: 4361460 
Pixel size in east-west direction: 10 m
Pixel size in north-south direction: 10 m
UTM zone: 18 N (WGS84)

Background color on
File and Directory Structure

Data files are available at:

Background color on
File Naming Convention

The binary data file is SV08LC_SMAPVEX08_Class.fst, and the header file is SV08LC_SMAPVEX08_Class.hdr.

Class in the file names indicate that these are land cover classification files.

Background color on
File Size

The data file is approximately 68 MB.

Background color on
Spatial Coverage

Southernmost Latitude: 38.8°N
Northernmost Latitude: 39.1°N
Westernmost Longitude: 76.3°W
Easternmost Longitude: 75.6°W

Background color on
Spatial Resolution

10 m

Background color on

UTM 18 N (WGS84)

Background color on
Grid Description

Data are on a rectangular grid with a cell size of 10 m by 10 m.

Upper left corner in east-west direction: 375325 
Upper left corner in north-south direction: 4361460 
Number of rows: 7956 
Number of columns: 8885

Background color on
Temporal Coverage and Resolution

Satellite images were obtained for June and August 2008; in situ data were collected over several days in October 2008. All data were combined into a single map.

Background color on
Parameter or Variable

Parameter Description

The measured parameter for this data set is land cover classification. Land cover classification distinguishes between crop types, water, roads, and urban areas. The land cover classes are designated as such:

Land Cover Classification Designation: 
0 Unclassified 
1 Corn/Corn Stubble 
2 Forest 
3 Soybeans 
4 Water 
5 Grassland 
6 Urban/Roads

Parameter Range

Valid parameter values are as follows:

Land cover class: 1–6 
Unclassified: 0

Background color on

Software and Tools

Various software packages can be used to read the data, such as the Environment for Visualizing Images (ENVI) and Interactive Data Language (IDL).

Background color on

Data Acquisition and Processing

Sensor or Instrument Description

Two scenes were used to construct the land cover classification for SMAPVEX08 on the Delmarva Peninsula. The following steps outline the procedure used to produce this image:

  1. Ground truth data in and around Ruthsburg, Maryland were collected.
  2. Ground truth data were converted to regions of interest (ROIs) within ENVI and one third were set aside for verification purposes.
  3. ResourceSat-1/IRS-P6 AWiFS data for 19 June 2008 and SPOT imagery for 20 August 2008 were collected and imported to IDL/ENVI and geo-registered.
  4. The amount of clouds present was small enough in the study region so that cloud masking became unnecessary.
  5. Within IDL/ENVI, a supervised Mahalanobis Distance classification was conducted with a variety of maximum distances per land cover type:
    1. Corn/Corn Stubble: 10 DN
    2. Forest: 8 DN
    3. Soybeans: 15 DN
    4. Water: 8 DN
    5. Grassland: 10 DN
  6. A classification image was generated and a confusion matrix calculated.
  7. Accuracies were calculated using a set of ROIs that were set aside for verification. Overall accuracy is 99.88 percent. The Kappa Coefficient is 0.9919. Table 1 lists accuracies by land cover type. 
    Table 1. Accuracies by Land Cover Type
    Corn Forest Soybean Water Grassland Ground Truth
    Unclassified 0.01 0.01
    Corn 97.18 3.76 0.99
    Forest 100 0.25 5.99
    Soybean 2.81 93.41 0.02 0.25 0.83
    Water 0.14 99.96 92.07
    Grassland 0.02 2.69 99.49 0.12
    Total 100 100 100 100 100 100
  8. A road network from USGS was overlaid on the land cover image with a swath-width of approximately 30 m per centerline.
  9. Finally, all testing and training pixels were overlaid on the image, thus providing a final correction to the classification scheme.
  10. Generation of a perfectly accurate land cover map was difficult considering the variable planting dates for the soybean and corn fields. Limited summer data were available and little could be done to close the gap in accuracy. However, for the time of consideration, there is little difference in vegetation water content and greenness between soybean fields and corn fields in October.
Background color on
Error Sources

The usage of two scenes, both acquired before the campaign, causes additional error to the classification.

Background color on
Quality Assessment

The quality of the classification is not high due to the effect of the error sources as specified below.

Background color on

Contacts and Acknowledgments

Michael H. Cosh
United States Department of Agriculture - Agricultural Research Service (USDA ARS) 
Hydrology and Remote Sensing Laboratory
10300 Baltimore Avenue
Bldg. 007, Rm. 104
Beltsville, MD 20705

Document Information


June 2015



How To

Programmatic Data Access Guide
Data from the NASA National Snow and Ice Data Center Distributed Active Archive Center (NSIDC DAAC) can be accessed directly from our HTTPS file system or through our Application Programming Interface (API). Our API offers you the ability to order data using specific temporal and spatial filters... read more
Filter and order from a data set web page
Many NSIDC data set web pages provide the ability to search and filter data with spatial and temporal contstraints using a map-based interface. This article outlines how to order NSIDC DAAC data using advanced searching and filtering.  Step 1: Go to a data set web page This article will use the... read more