Geometric validation of TerraSAR-X high-resolution products

Abstract

Switzerland's diversity of flat, rolling and Alpine terrain makes it ideal territory for geometric validation studies. Two parallel investigations of TerraSAR-X (TSX) highresolution data were conducted: geometric validation and estimation of the tropospheric path delay using measurements of corner reflectors (CRs) placed at different altitudes but nearly identical ranges. Geometric validation was conducted using products from the spotlight and stripmap modes in ascending and descending configurations. The geometric accuracy was investigated for representative products in two ways: (1) CR image positions were predicted based on the delivered product annotations, and these predictions compared to their measured image positions; (2)Products were geocoded using digital height models (DHMs), and the geocoded images were draped over the height models as well as 1:25’000 digital topographic maps for comparison. Terrain geocoding was performed using a Swiss 25-metre digital terrain model (DTM), as well as a 2-metre digital surface model (DSM) obtained from LIDAR (available for the Zurich test area). For the investigation into atmospheric path delay, six TSX stripmap scenes (30 km x 20 km) containing four identical CRs with the same range distances at an altitude difference of ~3000 m were examined. The CR arrangement made it possible to verify the TSX nominal path delay correction scheme by comparing predicted slant range with the slant range obtained by measuring the reflector image coordinates. Range differences between the high- and low-altitude reflectors helped quantify small variations in the path delay.