Comparison of DEMs derived from SRTM / X- and C-Band

Abstract

The Shuttle Radar Topography Mission in February 2000 had two ‘single pass’ interferometric radar systems on board which are the C-band system of NASA/JPL and the X-band system from DLR. Both systems have been operated simultaneously during the mission. Independent processors have been developed to produce the digital terrain models. Both systems used the same Position and Attitude Determination Record (PADR) generated by the Attitude Orbit Determination Avionics (AODA) system. The PADR file provides the orbit and baseline information. Meanwhile the processing of all data has been finished and the digital height models and radar images are available to the public. Height error maps accompany the digital terrain models. The proposed presentation shall compare the X- and C-band derived elevation models. Starting from the individual verification results the correspondence of the SRTM DEMs with respect to height and location accuracy shall be investigated. Potential global and local discrepancies shall be discussed. Where available both data sets will be compared to reference data like radar altimeter and local high resolution reference data sets. With SAR interferometry we measure the location or height of a representative phase centre of the backscattering resolution cell. C-band radar waves should have a higher penetration into the vegetation than the X-band waves. This should lead to differences in the interferometric height measured in the two systems over forested areas. In this paper the height differences observed in dependence of density and forest type is shown for some selected typical test sites. The resulting height differences are not very significant and well within the height error boundaries of both systems. Volume decorrelation had no significant impact due to the relative small baseline available for SRTM.