Professor Howard Zebker

Fall 2001-2002

Date: Wednesday, October 17, 2001
Time: 4:15 PM; Refreshments at 4:00 PM
Location: 460-334

Measuring Topography of the Entire Earth with Radar Interferometry: The SRTM Mission

Dr. Paul Rosen
Jet Propulsion Laboratory, NASA


Digital Elevation Models (DEMs) and global topographic products of heterogeneous origin have grown in availability with the development of digital computing. Yet no global DEM exists of uniformly fine quality. Using the methods of ScanSAR single pass interferometry, the Shuttle Radar Topography Mission will deliver the first fine resolution, uniformlyderived, DEM suite of nearly global extent. Each 30 meter resolution cell on the Earth's surface between the 60th parallels will be imaged at least twice, and will go through the same interferometric processing steps, quality checks, mosaicking with continental scale bundle adjustment, and calibration and verification steps. In this sense, the SRTM DEM products will be the first all-digital topography of the world. The end-to-end design of the topography algorithms takes into account the unique characteristics of the SRTM hardware, as well as the extent of available ground control and verification points. The SRTM mission was flown in February 2000, and acquired data over 99% of its targeted area during its 10 days of mapping. Global data production was scheduled to begin in October of 2001. This talk will describe the mission, the processing algorithms and methodology, and present some examples of processed data and its quality. From the globally distributed data that has been processed, it has been determined that the overall statistical performance of the C-band interferometer typically is excellent, out-performing design specification often by a factor of 2 or 3. Systematic effects are difficult to evaluate without full area processing, however we expect that the system will meet specification globally, and the data will meet the required absolute accuracy of 16 m in the vertical direction. A global systematic error model will be a unique product from SRTM that will characterize these errors.