Science Service System

Summary of Proposal CAL0163

TitleGeometric Validation of TerraSAR-X High-Resolution Products
Investigator Meier, Erich - Remote Sensing Laboratories (RSL), Geography
Team Members
Dr. Small, David - Remote Sensing Laboratories, Dept. of Geography, University of Zurich-Irchel
Dr. Schubert, Adrian - Remote Sensing Laboratories, Dept. of Geography, University of Zurich-Irchel
Dipl. Ing. Jehle, Michael - Remote Sensing Laboratories, Dept. of Geography, University of Zurich-Irchel
Dipl. Ing. Frey, Othmar - Remote Sensing Laboratories, Dept. of Geography, University of Zurich-Irchel
Dipl. Ing. Rueegg, Maurice - Remote Sensing Laboratories, Dept. of Geography, University of Zurich-Irchel
SummaryThree parallel investigations of the TSX high-resolution data will be conducted:
  • Geometric validation of the delivered products
  • Estimation of the tropospheric path delay
  • Verification of elevation AGP correction using the signals returned from corner reflectors placed at different altitudes but similar ranges.
These aspects are described briefly below.

Geometric validation
Using high-resolution products from the spotlight and stripmap modes in ascending and descending configurations, a statistical analysis of the geometric accuracy will be carried out for each product, based on ground control points (GCPs). These will be selected from various sources: (1) corner reflectors placed within the sites listed above, (2) national survey points, and (3) tiepoints selected from digital topographic and cadastral maps. Known GCP positions will be compared to predicted positions, based on the product annotations. The GCP-based error statistics will be calculated and patterns highlighted.

Comparisons between equivalent products will be made with slant-range and terrain-geocoded data processed from TerraSAR raw data in-house. Terrain geocoding will be possible using a Swiss DTM, the SRTM digital surface model (DSM), and a DSM obtained from LIDAR, available for the Duebendorf test area.

Finally, successive ascending (and descending) acquisitions will be used in the context of an interferometric investigation of product geometric consistency, helping to quantify systematic annotation variability.

Validation of Path Delay and AGP Corrections
Corner reflectors at different altitudes but identical ranges will be deployed in the Alpine test area. The arrangement will serve two purposes: (1) Any differences in their reflected intensities will indicate AGP-correction errors, and (2) The nominal correction scheme for the tropospheric path delay will be cross-checked for validity by comparing the predicted and measured range and azimuth coordinates. Range differences between the high- and low-altitude reflectors could help quantify small variations in the path delay.

Project financing, including the required field work, has been confirmed.

Detailed reportExperiment combining alpine and valley corner reflectors: geolocation accuracy estimates (absolute location error of image products) confirmed high accuracy of TerraSAR-X system as well as our own models of atmospheric path delay, tectonic plate movement and solild Earth tide. In the most important study, we acquired a 16-month time-series of high resolution spotlight images from the TSX sensor over a test site in Switzerland (Torny-le-Grand). Four trihedral CRs were deployed and their positions were surveyed with approximately centimeter accuracy.We compared themeasuredCRpositions with their imaged positions, based on the delivered product annotations and corrections made for atmospheric path delay, plate tectonics and SETs. The resulting error distribution is very compact and allows us to assign the TSX system an absolute geolocation accuracy of approximately 13 cm in range and 7 cm in azimuth, with a standard deviation of 34 cm in both dimensions. The sensor has proven to be sensitive enough to (incoherently) detect solid Earth movements on the order of several centimeters. At the same time, this study provided indirect evidence for the validity of the path delay and solid Earth tidal and tectonic models used during the study. Further work would be necessary to validate the models used in non-temperate latitudes, e.g. high arctic and equatorial regions. TSX has set a new geolocation standard for spaceborne SAR systems. However, this accuracy would be severely compromised were the atmospheric path delay not systematically mitigated. Therefore, we highly recommend the widespread inclusion of an atmospheric model in the product annotations, calibrated for the acquired scene by taking the best available meteorological measurements and topography into account.

Back to list of proposals

DLR 2004-2016