Day 4

Detailed paper information

Back to list

Paper title Polarimetric and interferometric signatures for change detection in agriculture using SAOCOM-1A Synthetic Aperture Radar images
  1. Anne Orban Centre Spatial de Liège - University of Liege Speaker
  2. Dominique Derauw Centre Spatial de Liège - University of Liege
  3. Christian Barbier Centre Spatial de Liège - University of Liege
  4. Murielle Kirkove Centre Spatial de Liège - University of Liege
  5. Santiago Bustos Revol Comisión Nacional De Actividades Espaciales (CONAE)
  6. Danilo Dadamia CONAE
Form of presentation Poster
  • B6. National missions TPM
    • B6.01 National EO satellite missions
Abstract text SAOCOM-1A and -1B are a pair of satellites developed by the Comisión National de Actividades Espaciales (CONAE) of Argentina launched in October 2018 and August 2020, respectively, orbiting on a Sun-synchronous orbit at 620 km altitude 180 deg. apart, and providing imaging of the Earth's surface with an effective revisit time of 8 days. Their main payload is a full-polarimetric Synthetic Aperture Radar (SAR) operating in L-band with selectable beams and imaging modes.
This paper reports the first result of collaborative work between CONAE and CSL, investigating the use of polarimetric and interferometric SAR signatures to detect changes in agricultural zones in Argentina. This work is the natural continuation of a previous pre-SAOCOM activity [1] that made use of airborne SAR images from the national SARAT sensor and the NASA/JPL UAVSAR, and from the spaceborne JAXA ALOS-PALSAR-2, all operating in L-band.
The test site of interest is referred to as the SAOCOM Core Site, a highly agricultural zone within the Pampas region, located in the surroundings of Monte Buey (a small rural village southeast of the Córdoba province, Argentina (-32º 55', -62° 27')). This site is regularly imaged by the SAOCOM satellites, and regular field measurements are carried out in conjunction with image acquisitions.
We have used full-polarimetric, Stripmap-mode SAOCOM-1A images over the region of interest, acquired between March 2019 and February 2020 in both right-looking ascending and descending orbits, making a temporal series of useable interferometric pairs covering a full year.
Each image was the subject of polarimetric processing, involving generation of backscattering coefficient and Radar Vegetation Index (RVI) maps, Pauli decomposition of the scattering vector, generation and diagonalization of the polarimetric coherence matrix, resulting in derived quantities like the entropy, the anisotropy, and the alpha angle. Interferograms and coherence maps were generated by InSAR processing of the interferometric pairs. Finally, adding full polarization information to the interferometric, i.e., carrying out polarimetric interferometry (PolInSAR) processing, optimized coherence maps were generated, allowing to highlight one or the other backscatter mechanism and follow the evolution along a full season.
This multi-dimensional information was put in relation to terrain events by cross-correlation with field data. All the products were co-registered in order to perform time-series analyses for change detection.
This work was performed under a Belgium-Argentina bilateral collaboration. The CSL contribution was supported by the Belgian Science Policy Office.
[1] Danilo J. Dadamia, Marc Thibeault, Matias Palomeque, Christian Barbier, Murielle Kirkove and Malcolm W.J. Davidson, “Change Detection Using Interferometric and Polarimetric Signatures in Argentina, 8th International Workshop on Science and Applications of SAR Polarimetry and Polarimetric Interferometry”, ESRIN, Frascati, 23-Jan-2017.