Documentation for Reduced-Resolution Radar Imagery, Digital Elevation Models, and Related GIS Layers for Barrow, Alaska, USA, Version 1

Back to Top

Collapse All / Open All

Detailed Data Description

This product set contains high-resolution Interferometric Synthetic Aperture Radar (IFSAR) imagery and geospatial data for the Barrow, Alaska, USA region for use in Geographic Information Systems (GIS) and remote sensing software packages. The baseline geospatial data support education, outreach, and multi-disciplinary research of environmental change in Barrow, an area of focused scientific interest. The data are provided on five DVDs, available through licensing only to National Science Foundation (NSF)-funded investigators. An NSF award number must be provided. Reduced-resolution versions of the data sets are available to the general public.

The primary IFSAR datasets were acquired by Intermap Technologies from 27 to 29 July 2002, and consist of Orthorectified Radar Imagery (ORRI), a Digital Surface Model (DSM), and a Digital Terrain Model (DTM). The ORRI depicts the ground surface as illuminated by airborne radar, and is useful for visualization, mapping, georectification, and a variety of analyses. The DSM is a "first-surface" reflectance model of elevation, including vegetation, buildings, and other structures. The DTM is a filtered model to better represent ground elevation, and is useful for visualization, orthorectification, and environmental analyses. Value-added products include data mosaics, derived layers, and accessory layers.

The unmodified IFSAR data comprise 26 data tiles across two UTM zones. The DSM and DTM tiles (5 m resolution) are provided in floating-point binary format with header and projection files. The ORRI tiles (1.25 m resolution) are available in GeoTIFF format. FGDC-compliant metadata created by Intermap Technologies are provided in text, HTML, and XML formats, along with the Intermap License Agreement and product handbook. The DSM, DTM, and ORRI went through rigorous quality reviews before final acceptance of the data from Intermap Technologies.

Value-added processing helped avoid redundant effort or confusion, and produced a suite of complete and consistent data. To make the primary data sets more usable, the 26 data tiles were converted to data mosaics at original 5 m resolution with a common projection (UTM Zone 4) for the spatial extent of the "Barrow Peninsula" (155.39 - 157.48°W, 70.86 - 71.47°N). The DSM and DTM mosaics are available as floating-point binary files and in ArcInfo grid format. An ORRI mosaic, at reduced resolution (5 m), is provided as a contrast-enhanced GeoTIFF file.

Processing included the creation of derived GIS layers: aspect, shaded relief, and slope-angle grids (floating-point binary and ArcInfo grid format), as well as a vector layer of contour lines (ESRI Shapefile format). Also available are accessory layers compiled from other sources: 1:250,000- and 1:63,360-scale USGS Digital Raster Graphic (DRG) mosaic images (GeoTIFF format); 1:250,000- and 1:63,360-scale USGS quadrangle index maps (ESRI Shapefile format); a quarter-quadrangle index map for the 26 IFSAR tiles (ESRI Shapefile format); and a simple polygon layer of the extent of the Barrow Peninsula (ESRI Shapefile format). To facilitate processing for a smaller area of particular interest near Barrow, each of the primary, derived, and accessory layers was clipped to a subsetted extent of the "Barrow Triangle" (156.13 - 157.08°W, 71.14 - 71.42°N).

Finally, detailed documentation was created for all of the data layers. Federal Geographic Data Committee (FGDC)-compliant metadata are provided in text, HTML, and XML formats. Please read the FGDC metadata and view the thumbnail images (accessed from the File List) for a variety of details specific to each layer or image.

Primary Data Layers

Digital Surface Model

Digital Surface Model
The DSM is a grid of elevation measurements derived from the return signals received by two radar antennas on the aircraft. The signals bounce off the first surface they strike, making the DSM a representation of any object large enough to be resolved, including buildings, vegetation and roads, as well as natural terrain features.

Digital Terrain Model

Digital Terrain Model
The DTM is a custom "bald-earth" model that references the elevation measurements of the bare terrain. It is better suited for derived layers such as slope angle, aspect, and contours. DTM values are a result of a custom process developed for the low coastal plain of Alaska's North Slope. A repeated median filter removed most structures, as well as high-frequency noise. The DTM is well-suited for most applications of elevation analysis.

Orthorectified Radar Imagery

Orthorectified Radar Imagery
The ORRI is a grayscale image of the first reflective surface illuminated by the radar; it was corrected to remove geometrical distortions. This radar-derived product reveals Earth surface features that are not easily identifiable with aerial photography. Furthermore, the ORRI is well suited as a base layer for georectification of aerial photography and other imagery.

Background color on
Derived Data Layers


Shaded Relief
Shaded Relief

Background color on
Accessory Data Layers
Mosaicked 1:250k
mosaicked 1:250,000 scale USGS
Digital Raster Graphic (DRG)
Mosaicked 1:63k
mosaicked 1:63,360 scale USGS DRG
1:250k USGS quadrangle
1:250,000 scale USGS quadrangle boundaries

1:63k USGS quadrangle
1:63,360 scale USGS quadrangle boundaries
Barrow Peninsula
Barrow Peninsula extent boundary
Background color on

The primary, derived, and accessory data layers are provided in a variety of standardized data formats, including floating-point binary, GeoTIFF, and ESRI Shapefiles. The layers vary in terms of grid cell resolution and dimensions. For example, the DSM mosaic for the Barrow Peninsula is available in floating-point binary format, has grid cell spacing of 20 m, and dimensions of 3,255 rows by 3,615 columns. Please see the FGDC metadata (accessed from the File List) for details specific to each layer. To read the floating-point DSM and DTM into an image processing program, refer to the associated header files (.hdr). To import the floating-point DSM and DTM into ArcGIS, use the FLOATGRID command. Or, in ArcToolbox choose Conversion Tools / Import to Raster / Floating Point Data to Grid.

Background color on
File and Directory Structure

File names for GIS layers of the Barrow Peninsula begin with "bp." Names for most of the raster layers also include a number indicating horizontal resolution. For example, "bp20_dsm" is the reduced-resolution version of the DSM with 20 m grid-cell spacing.

Follow the links in the tables below to view FGDC metadata, which provide details about specific data layers, including data sources, processing, data format, spatial and temporal resolution, extent, projection, and datum.

.html, .xml, .txt: Federal Geographic Data Committee (FGDC) metadata
.hdr: ArcInfo header file
.prj: ArcInfo projection file
.tfw: GeoTIFF world file
.dbf, .prj, .sbn, .sbx, .shx: ESRI Shapefiles
.flt: 32-bit IEEE signed floating-point binary file

File List
File Size
File Names Description Format Accompanying Metadata
3 KB 0README_CD.txt Readme File Text
1 KB readme_nsidc.txt Readme File Text
82 KB EULA_Resampled_Unrestricted.pdf Intermap License Agreement PDF
3 MB gt_product_handbook_v2_3.pdf Intermap Product Handbook PDF
180 MB bp05_orri_mosaic.aux

ORRI mosaic raster

ORRI mosaic raster

GeoTIFF .html, .txt, .xml
45 MB bp20_aspect.flt

Reduced-resolution (20 m) aspect raster

20m Aspect

Floating-point binary .html, .txt, .xml
47 MB bp20_contours.shp

Reduced-Resolution (20 m) contours

Reduced Resolution 20m

ESRI Shapefile .html, .txt, .xml
45 MB bp20_dsm.flt

Reduced-resolution (20 m) DSM raster

20m DSM

Floating-point binary .html, .txt, .xml
45 MB bp20_dtm.flt

Reduced-resolution (20 m) DTM raster

20m DTM raster

Floating-point binary .html, .txt, .xml
45 MB bp20_shaded.flt

Reduced-resolution (20 m) shaded relief raster

20m shaded raster

Floating-point binary .html, .txt, .xml
45 MB bp20_slope.flt

Reduced-resolution (20 m) slope raster

20m slope raster

Floating-point binary

.html, .txt, .xml
21 MB bp_drg_250.tif

1:250,000 scale USGS Digital Raster Graphic (DRG)

250k USGS raster

GeoTIFF .html, .txt, .xml

325 MB


1:63,360 scale USGS DRG

63k USGU raster

GeoTIFF .html, .txt, .xml
1 MB bp_extent.dbf

Barrow Peninsula extent boundary

Barrow extent boundary

ESRI Shapefile .html, .txt, .xml
508 KB bp_itmquad.dbf

1:63,360 scale USGS quadrangle boundaries

63k quadrangle boundaries

ESRI Shapefile .html, .txt, .xml
625 KB bp_qmquad.dbf

1:250,000 scale USGS quadrangle boundaries

250k USGS quadrangle

ESRI Shapefile .html, .txt, .xml
181 KB bp_quarter_quads.dbf

Quarter quadrangle tile boundary index map

quarter quadrangle

ESRI Shapefile .html, .txt, .xml
390 MB bt_extent.dbf

Barrow Peninsula extent boundary

barrow penisula extent boundary

ESRI Shapefile .html, .txt, .xml
Background color on

The entire data set is approximately 500 MB.

Background color on
Spatial Coverage

Spatial extent of the Barrow Peninsula layers:

West: 157.476330°W
East: 155.394828°W
North: 71.472777°N
South: 70.862456°N

Spatial Coverage Map

The following map displays the spatial extent of the Barrow Peninsula (red lines) and Barrow Triangle (purple lines) data sets. Click the thumbnail for a larger image.

Spatial Coverage Map


The map projection for the Barrow Peninsula data sets is Universal Transverse Mercator (UTM), Zone 4. Horizontal datum is the North American Datum of 1983 (NAD83).

Background color on
Temporal Coverage

Intermap Technologies' STAR-3i airborne radar system collected data from 27 to 29 July 2002.

Background color on

Software and Tools

Software and Tools

NSIDC does not provide any software to read these data, which are easily imported into ArcGIS and image processing software.

Background color on
Quality Assessment

Quality checks were performed throughout all processing steps. Minimum and maximum domain values were checked for each tile, for merged and mosaicked products. Map extents of separate tiles are correct. Overlaps between tiles are seamless and contain identical values. Value-added layers and their ranges were quality checked; these value-added layers were also used to quality check the values in the DSM.

The vertical accuracy of DSMs and DTMs is approximately ± 1.0 m or better root-mean-square-error (RMSE). Horizontal accuracy is ± 2.5 m or better RMSE for slopes less than 20°. Horizontal accuracy of ORRI data is ± 1.25 m or better RMSE for slopes less than 20°. These accuracy statements depend on the sensor's view of the ground, and are applicable to the height of the first surface return. These statements also assume the surface is devoid of vegetation or continuous roof coverage in urban areas (Intermap 2002).

Accuracy statements are based on areas of moderate terrain. Diminished accuracies are to be expected in areas of extreme terrain and dense vegetation. The DSM and DTM contain some small (30 to 50 cm high) ripples or "corn rows." The Intermap data collection systems are very accurately calibrated to control the constant differential phase error between the antennae. The calibration ensures that data are within specification; however, on flat terrain, very low-level artifacts are present from uncompensated differential phase errors. Very small periodic DEM ripples in the range dimension occurred at a consistent location in range and are continuous along the strip. High-frequency variation (speckle) exists in both the imagery and terrain data sets. The high-frequency variation in the DEM is a product of the signal-to-noise ratio that decreases from the near to far range across the track; therefore, the noise increases with the distance from the antenna. The difference in noise between the near and far range of the strip can sometimes be seen at the seam where DEMs from two strips are mosaicked together.

Background color on

Data Acquisition and Processing

Theory of Measurements

The DSM, DTM, and ORRI are derived from airborne Interferometric Synthetic Aperture Radar (IFSAR). Interferometry involves detecting the radar return signal using antennas at two different locations. With an interferometric radar, the patterns of electromagnetic radiation emanating from the same point on the ground strike each antenna independently because of their different ranges. These waves are out of phase, meaning they do not always overlap each other. This phase difference result, and the geometry formed by the separation of antennas, provide all the information required to derive the height and corresponding geographic position of targets that interact with the transmitted energy (Intermap 2002).

Background color on
Data Acquisition Methods

Raw radar and navigation information are combined to produce flight strip information referenced in a radar-specific coordinate system, which are then subject to quality control (QC). Once they pass QC, the strips are transformed to the desired map projection, merged into map sheet products for elevation, radiometrically balanced, and mosaicked into map sheets (Tennant, Coyne, DeCol 2003).

Background color on
Derivation Techniques and Algorithms

Processing Steps

IFSAR data for the Barrow region were originally delivered by Intermap Technologies as 26 tiles of data split across UTM Zones 4 and 5. Manley et al. imported the tiles into ArcInfo, merged tiles within each zone, and projected Zone 5 to Zone 4 to produce a mosaic across both zones. Data were then clipped by a Barrow Peninsula extent and a Barrow Triangle extent. Refer to High-Resolution Radar Imagery, Digital Elevation Models, and Related GIS Layers for Barrow, Alaska, USA for data layers subsetted to the extent of the Barrow Triangle. These mosaics were then regridded to 20 m, to make a free, reduced-resolution version available to the general public.

Error Sources

Several potential sources of error exist in the DSM and DTM, as well as other layers. For more information, refer to the FGDC metadata, under File List for more information. 

25 April 2004
NSIDC discovered the file "bp20_dsm.hdr" contained an unnecessary reference to ENVI software. The correct header information should be:

ncols 3615
nrows 3255
xllcorner 555660
yllcorner 7865740
cellsize 20
NODATA_value -9999
byteorder LSBFIRST

This file was replaced and labeled "Version 1.1".

Background color on
Sensor or Instrument Description

IFSAR data were acquired from Intermap Technologies' STAR-3i airborne radar system aboard a Learjet 36A, 31,000 ft above ground level. The system transmits a radar pulse 2,200 times per second at a peak power of 8 W. Bandwidth is 135 MHz, digitization rate of the return signal is 300 million times per second, and information is recorded at 90 MB per second. The radar is mounted on a pedestal that rotates to allow the crew to collect data from either side of the aircraft, using two parallel antennas separated by 1 m. Each terrain height calculation represents the average height within the 1.25 m resolution cell (Tennant, Coyne, DeCol 2003). Acquired data are "interfered" by a digital correlation process to extract terrain height data and geometrically correct radar images. Areas of missing data are interpolated using continuous curvature spline over non-data areas. Most of these instances are due to radar shadow and layover due to steep terrain. The system's accuracy was independently validated by the U.S. Army Topographic Engineering Center, the Institute of Navigation, Stuttgart University, and by NASA.

The STAR-3i system was developed by the Environmental Research Institute of Michigan (ERIM) under contract to the U.S. Defense Advanced Research Projects Agency (DARPA).

Background color on

References and Related Publications

Contacts and Acknowledgments

Investigator(s) Name and Title

William F. Manley
Institute of Arctic and Alpine Research (INSTAAR)
University of Colorado
Boulder, CO, USA

Leanne Lestak
Cooperative Institute for Research in Environmental Sciences (CIRES)
University of Colorado
Boulder, CO, USA

Craig Tweedie
Arctic Ecology Laboratory
Michigan State University
Lansing, MI, USA

James Maslanik
Cooperative Institute for Research in Environmental Sciences (CIRES)
University of Colorado
Boulder, CO, USA


This research was funded by the National Science Foundation Award OPP-0224071, "High-Resolution Imagery and Terrain Model for Collaborative Research of Environmental Change at Barrow, Alaska."

Document Information


February 2005


Febriaru 2005


February 2005