The National Geospatial Intelligence Agency (NGA), formerly NIMA, has a need for a complete global database of topography at DTED-2 accuracy. The Shuttle Radar Topography Mission (SRTM) produced the first consistent, nearly global digital elevation model with DTED-2 accuracy. Because of space shuttle orbital limitations, SRTM was unable to map latitudes north of 60 degrees and south of –59 degrees. The NGA was a major sponsor of SRTM and has an interest in completing the global map north of 60 degrees latitude.

NGA brought together three groups to determine whether it is feasible to use the existing ERS-1/2 tandem database, supplemented with RADARSAT data, to extend the SRTM DTED-2 database north of +60 degrees with an accuracy as good as (or at least approaching) that of SRTM. This current effort is led by the Jet Propulsion Laboratory (JPL), which managed the SRTM project for NIMA (now NGA) and NASA. The team also includes the Alaska Satellite Facility (ASF) at the University of Alaska Fairbanks and Vexcel Corporation, Boulder Colorado. ASF has a large archive of repeat-pass SAR data that can be used to interferometrically and stereogrammetrically produce digital topography. ASF also provides expertise in ERS-1/2 tandem SAR interferometry, use of ICESAT as geodetic control and related areas. The Vexcel Corporation has extensive experience in the area of ERS 1/2 tandem data analysis, the analysis of RADARSAT data, and producing commercial grade software to support this kind of analysis.

The study will provide NGA with a complete understanding of expected results, optimal production strategy to maximize accuracy of results, data availability and anticipated costs for production. It will also provide estimates of the expected accuracy based on the amount of data available, baseline characteristics, coverage geometry, etc.

The NGA is interested in high quality geodetic control from space for use in topographic mapping applications. In the fall of 2003, NGA funded a two-year NIMA University Research Initiative (NURI) project “DEM Control in Arctic Alaska with ICESAT Laser Altimetry” at the Geophysical Institute. This project will evaluate the accuracy and repeatability of NASA’s recently launched laser altimetry mission ICESAT.

Although, there have been technical problems with the lasers on-board the satellite, there have been three 36 day periods of observations; one every six months. Laser pulses at 40 times per second illuminating spots (footprint) 70 meters in diameter, spaced along-track at 170-meter intervals along the Earth’s surface. With its polar orbiting geometry, ICESAT provides tracks spaced cross-track at roughly 12 kilometer (at 70 ° latitude). After extensive instrument calibration work, the data accuracy is expected to have 10 meters (horizontal) and 10-15 cm (vertical) accuracy over ice sheets. The availability of high quality geodetic control that is well distributed is an essential requirement for mapping remote regions such as Alaska.

Barrow Alaska was identified as a study site for phase-1, due to its low topography relief, abundance of lakes in the area, and availability of high quality scientific data for the region. Preliminary results using Differential Global Positioning Systems (DGPS) of frozen lake heights near Barrow show accuracies that are nearly comparable to expected results. Uniform heights of the frozen lake surfaces are easily identified in the data and used to verify geographic location accuracy of the data. The 7 differential GPS elevations on the 7 lakes are within 0.031 to 0.337 m of the corresponding ICESAT/GLAS-derived elevations.  The mean difference is 0.206 m.  All of the differences are of the same sign; that is, the GLAS-derived elevations are slightly higher, by 20.6 centimeters on average. Statistical assessments are being made between high-resolution topography derived from commercial airborne SAR (STAR3I) for the study area. The project goals include evaluation of the use of ICESAT data as a geodetic control source for topographic mapping with SAR interferometry. The effects of terrain relief and vegetation cover on accuracies of the data will also be investigated.

This research will provide an independent assessment of GLAS-derived elevations as a vertical reference to augment the SRTM data set.