Data Set ID:

SMEX03 Land Use Classification Data: Oklahoma, Version 1

This data set consists of land use classification data derived from Landsat 5 data for the Soil Moisture Experiment 2003 (SMEX03).

This is the most recent version of these data.

BASIC Level of Service

Data: Data integrity verified

Documentation: Key metadata and links to supporting documentation available

User Support: Assistance with data access and basic data usage

See All Level of Service Details

Data Format(s):
  • Microsoft Excel
  • Binary
  • ASCII Text
Spatial Coverage:
N: 37, 
S: 33.6, 
E: -96.6, 
W: -99.3
Spatial Resolution:Not SpecifiedSensor(s):TM
Temporal Coverage:
  • 2 July 2003 to 18 July 2003
  • 10 July 2003
Temporal ResolutionNot specifiedMetadata XML:View Metadata Record
Data Contributor(s):Michael Cosh, Thomas Jackson

Geographic Coverage

Once you have logged in, you will be able to click and download files via a Web browser. There are also options for downloading via a command line or client. For more detailed instructions, please see Options Available for Bulk Downloading Data from HTTPS with Earthdata Login.

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.

Jackson, T. and M. Cosh. 2009. SMEX03 Land Use Classification Data: Oklahoma, 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


Land use classification data derived from Landsat 5 TM are provided in a flat, Band-Interleaved-by-Line binary file called SMEX03_OK_Land_Cover.bil. The Oklahoma study area image dimensions are 12,100 rows by 8201 columns. An Environment for Visualizing Images (ENVI) header file with geolocation information, called SMEX03_OK_Land_Cover.hdr, is also provided.

Additionally, in situ land cover classification data are provided in the Land_Cover.xls file. The file lists the initial surface conditions for each sampling site in the Oklahoma South and Little Washita regional study areas. Sampling sites are labeled by either OS (Oklahoma South) or LW (Little Washita) followed by the site number. For example, OS03 indicates the third sampling site in Oklahoma South study area. OS sampling site values range from 01-52. LW sampling site values include 02, 03, 04, 11, 12, 13, 20, 21, 22, 27, 28, 29, 31, 32, and 33. Also indicated in the Land_Cover.xls file is any noted change in surface condition, such as tillage, planting, or harvesting during the 2 - 18 July 2003 timeframe.

Background color on
File and Directory Structure

All files are available on the HTTPS site in the directory:

Background color on
File Naming Convention

Landsat 5 TM data are provided in a single binary file with a corresponding header file. These files are named according to the following convention and are further described in Table 1:


Table 1. Description of File Name Variables
Variable Description
SMEX03 Soil Moisture Experiment 2003
OK Oklahoma
Land_cover Parameter for this data set
.bil Indicates that this is a Band-Interleaved-by-Line binary data file.
.hdr Indicates that this is a header file.
.xls Indicates that this is an Excel file.

Background color on
File Size

The binary data file is approximately 97 MB, its header file is 1 KB, and the Excel file is 17 KB.

Background color on

The volume of this data set is approximately 97 MB.

Background color on
Spatial Coverage

Southernmost Latitude: 33.6° N
Northernmost Latitude: 37.0° N
Westernmost Longitude: 99.3° W
Easternmost Longitude: 96.6° W

Spatial Resolution

The pixel size is 30 m by 30 m.

Projection and Grid Description

Universal Transverse Mercator (UTM), Zone 14, World Geodetic System 1984 (WGS 84)

Background color on
Temporal Coverage

TM imagery was collected for 10 July 2003. In situ data were collected from 2 through 18 July 2003.

Background color on
Parameter or Variable

The measured parameter for this data set is land use (vegetation) classification. Land use classification distinguishes between crop types, water, roads, and urban areas.

Parameter Range

The parameter range for various Oklahoma land cover types is described in Table 2. Valid parameter range values range from 0 to 8.

Table 2. Parameter Range and Description
Parameter Value Description
0 Unclassified
1 Forest
2 Pasture
3 Winter Wheat/Bare
4 Alfalfa
5 Corn
6 Urban/Quarry
7 Soybean
8 Water

Sample Data Record

Figure 1. Land Use Classification Sample Data Record

Background color on

Software and Tools

Tools appropriate for viewing these data include: ArcView, ENVI, or other similar visualization software packages; any text editor or Web browser; and any spreadsheet program.

Quality Assessment

Table 3 compares the accuracy of the in situ data with the accuracy of the image data obtained via satellite. The far right column lists the accuracy percentages of the in situ data when compared with the satellite image data. For example, of 581 classified forest pixels, only 566 of them are registered in the in situ data as forest, for a 97.4 percent accuracy. Accuracy percentages listed in the bottom row are the satellite image data compared to the in situ data. For example, of the 818 pixels that were deemed to be winter wheat in the field, only 665 of them were classified by Landsat 5 TM to be winter wheat, while 53 of them were classified as pasture and 100 were unclassified, for an accuracy of 81.3 percent.

Note: Urban/Quarry pixels were not subjected to ground truthing.

Table 3. Comparative Accuracy of In Situ Data Versus Image Data
In Situ Data
Landsat 5 TM
Image Data
Forest Pasture Winter Wheat Alfalfa Corn Soybean Water Total Pixels Accuracy (%)
Unclassified 2 - 100
2 1 2 107 N/A
Forest 566
2 581 97.4
Pasture 1 477 53
2 533 89.5
Winter Wheat
665 100
79 73.4
Corn 12
1 565 97.7
38 100
1259 1259 100
Total Pixels 581 498 818 58 567 39 1266 3827 N/A
Accuracy (%) 97.4 95.8 81.3 100 97.3 97.4 99.4 N/A Overall Accuracy = 89 %

Background color on

Data Acquisition and Processing

Data Acquisition Methods

Oklahoma land use classification data were derived from both Landsat 5 TM imagery and in situ data, then classified and cross-verified for accuracy. These data were then compared to SGP97 and SGP99 land use classification data, two experiments which were conducted in the same region as SMEX03. Landsat 5 TM data were collected for 10 July 2003. Data were used from Path 28, Row 34-35 to cover the SMEX03 Oklahoma study area. In situ data were collected from 2 through 18 July 2003.

Background color on
Processing Steps

One scene for 10 July 2003 was used to construct the land use classification for the SMEX03 Oklahoma study area. There was little cloud cover for this particular date and other dates in proximity to the experiment were of low quality.

  1. In situ data were collected in and around Chickasha, El Reno, and the Little Washita Basin in Oklahoma, USA.
  2. In situ data were converted to Regions of Interest (ROIs) within ENVI and one third of these data were set aside for verification purposes.
  3. Landsat TM data for 10 July 2003 were collected and imported to ENVI for Path 28/Row 34-35.
  4. Landsat TM images were masked for clouds. The pixels with values in band 1 (.45 - .52 um) greater than 110 were removed.
  5. Shadowed areas in Band 4 ( .76 -.90um ) with values less than 50 were removed.
  6. Bands 2,3,4,5,7 were extracted from 10 July 2003 and merged to create a 5 band image.
  7. Within ENVI, a supervised maximum likelihood classification was conducted with a threshold of .85.
  8. A classification image was available from SGP99. For areas that were masked due to clouds or shadows in the original classified image, SGP99 classifications were substituted, creating a merged image. Road network information was also superimposed on the image and those pixels selected as Urban/Quarry.
  9. Accuracy percentages were calculated using a set of ROIs that were set aside for verification. Refer to Table 3 in the Quality Assessment section of this guide document for a listing of accuracy percentages.
Background color on

References and Related Publications

Contacts and Acknowledgments

Thomas J. Jackson
Hydrology/Remote Sensing Laboratory
US Department of Agriculture (USDA) - Agricultural Research Service (ARS)
Bldg. 007, Rm. 104, BARC-West
Beltsville, Maryland 20705 USA

Michael H. Cosh
Hydrology/Remote Sensing Laboratory
US Department of Agriculture (USDA) - Agricultural Research Service (ARS)
Bldg. 007, Rm. 104, BARC-West
Beltsville, MD 20705 USA


The investigators would like to acknowledge the USDA ARS Grazinglands Research Laboratory for their support as well as the many graduate students and volunteers who collected the field data.

Document Information


January 2009


No technical references available for this data set.

How To

V0 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. We provide basic command line Wget and curl instructions for HTTPS file system downloads below. For questions about programmatic access to NSIDC... read more