ASTER Global Digital Elevation Model Version 2 - summary of validation results

On June 29, 2009, NASA and the Ministry of Economy, Trade and Industry (METI) of Japan released a Global Digital Elevation Model (GDEM) to users worldwide at no charge as a contribution to the Global Earth Observing System of Systems (GEOSS). This “version 1” ASTER GDEM (GDEM1) was compiled from over 1.2 million scenebased DEMs covering land surfaces between 83°N and 83°S latitudes. A joint U.S.-Japan validation team assessed the accuracy of the GDEM1, augmented by a team of 20 cooperators. The GDEM1 was found to have an overall accuracy of around 20 meters at the 95% confidence level. The team also noted several artifacts associated with poor stereo coverage at high latitudes, cloud contamination, water masking issues and the stacking process used to produce the GDEM1 from individual scene-based DEMs (ASTER GDEM Validation Team, 2009). Two independent horizontal resolution studies estimated the effective spatial resolution of the GDEM1 to be on the order of 120 meters.

A second version of the ASTER GDEM (GDEM2) is scheduled for release by NASA and METI in mid-October, 2011. Improvements in the GDEM2 result from acquiring 260,000 additional scenes to improve coverage, a smaller correlation kernel to yield higher spatial resolution, and improved water masking. As with the GDEM1, the GDEM2 validation was performed by the U.S. and Japanese partners. Vertical accuracy assessments included a comparison of the GDEM2 against absolute geodetic references over the Conterminous US (CONUS), against national elevation grids over the US and Japan, against the Shuttle Radar Topography Mission (SRTM) 1 arc-second elevation grids over the US and 20 sites around the globe, and against space borne laser altimeter data globally. Horizontal accuracy assessments were conducted as part of the Japan and the global SRTM studies, and horizontal resolution studies were conducted in both Japan and the US. Each group documented changes in artifacts in GDEM2 due to processing improvements.

The absolute vertical accuracy study found the GDEM2 to be within -0.20 meters on average when compared against 18,000 geodetic control points over the CONUS, with an accuracy of 17 meters at the 95% confidence level. The Japan study noted the GDEM2 differed from the 10-meter national elevation grid by -0.7 meters over bare areas, and by 7.4 meters over forested areas. Similarly, the CONUS study noted the GDEM2 to be about 8 meters above the 1 arc-second NED over most forested areas, and more than a meter below NED over bare areas. The global altimeter study found the GDEM2 to be on average within 3 meters of altimeter-derived control, and also documented sensitivity to tree canopy height. The Japan study noted that the horizontal displacement in GDEM1 of 0.95 pixels was reduced to 0.23 pixels in GDEM2. Both teams noted improvements in horizontal resolution, between 71 and 82 meters, comparable to the SRTM 1 arc second elevation model, but at the cost of some increased noise. The number of voids and artifacts noted in GDEM1 were substantially reduced in GDEM2, and in some areas virtually eliminated.

Based on these findings, the GDEM validation team recommends the release of the GDEM2 to the public, acknowledging that, while vastly improved, some artifacts still exist which could affect its utility in certain applications.