Adding Precise Wave-Propagation Information and Geodetic Corrections to Standard SAR Products

2015 
The inherent geometric accuracy of Synthetic Aperture Radar (SAR) imagery was in the past not fully exploited. It is generally limited by the knowledge of satellite orbit, SAR instrument timing knowledge and stability, different coordinate systems of sensor and object, atmospheric propagation delays (ionosphere and troposphere), solid earth dynamics and calibration procedures and annotation. Indeed, the SAR imaging process shares many features with GNSS systems and during the recent years we could demonstrate that centimeter-level positioning accuracy can be achieved, if GNSS techniques are applied properly [1,2,3,4]. Using stereo-SAR, even 3D positioning of points with an accuracy of 6 cm could be demonstrated [5]. Encouraged by our results, we have now defined and implemented a next-generation geodetic SAR product layer (currently tailored to TerraSAR-X) that takes care of the aforementioned factors and provides appropriate corrections. The product may serve as well application oriented users as geodesy scientists. Possible applications are - 2D motion detection of earthquakes or glaciers without using reference points - support of phase unwrapping by reducing fringes from tropospheric stratification - absolute 3D precision-positioning - 3D motion measurement of points. The new product architecture allows for geometric corrections down to millimeter level is presented and the different sources of information are described. We demonstrate some test results by comparing our global product with local GNSS measurements. Furthermore, we demonstrate real-world applications such as 3D-positioning of points on the ground by using Stereo-SAR or ascending/descending combinations. Finally, we present first results for Sentinel-1 where the large image size reveals large scale effects not visible so far.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    3
    Citations
    NaN
    KQI
    []