Language selection


Compact Polarimetry: Multi-Thematic Evaluation

PDF Version [PDF, 951.1 kb]

F. Charbonneau and B. Brisco, Canada Centre for Remote Sensing, Natural Resources Canada.

Figure 1: Stoke Representation

Figure 1: Stoke Representation

[larger image]

Text Version - figure 1

Figure 1 - Text version

In fully polarimetric, the product of the Kennaugh matrix multiplied by the Stokes vector at transmission equals the Stokes vector at reception. In compact polarimetry, CL, in right transmission, the Stokes vector is 1,0,0,-1. This Stokes vector, multiplied by the Kennaugh matrix gives the Stokes vector at reception.

This on-going study evaluates the potential of a compact polarimetric SAR mode concept for meeting Earth Observation requirements relevant to Canadian thematic issues. In a hybrid-polarity architecture (CL-pol), SAR transmits circular polarization and receives two orthogonal mutually-coherent linear polarizations. The resulting radar is relatively simple to implement, and has unique self-calibration features and low susceptibility to noise and cross-channel errors. It also permits the maintainence of larger swath coverage as compared to fully polarimetric SAR systems. This configuration is being evaluated by a multi-disciplinary team, for applications relevant to the Government of Canada, for input to future SAR system design and for data selection from the growing array of SAR data products available worldwide.

As preliminary results, a quick comparison based on scattering mechanisms between RADARSAT-2 fully polarimetric data and simulated compact polarimetry is presented as RGB composite images (figure 1), where the Red, Green and Blue channels are representing respectively the double bounce, multiple (vegetation) and single bounce scattering mechanism. Most of the relative intensity between the scattering mechanisms is preserved, nevertheless a less dynamic intensity range is observed

The results will help evaluate the tradeoffs between ground coverage, information content, processing complexity and data volume.

Figure 2: Scattering mechanism decomposition images 2a) Fully polarimetric data (Freeman-Durden decomposition); 2b) m-´ Simulated compact polarimetric mode (CL)

Figure 2: Scattering mechanism decomposition images 2a) Fully polarimetric data (Freeman-Durden decomposition); 2b) m-δ Simulated compact polarimetric mode (CL)

[larger image]

Text Version - figure 2

Figure 2 - Text version

The illustration includes two images identified as a and b, covering approximately 25 kilometres by 25 kilometres in the Beauharnois region west of Montreal. The image pair features scattering mechanism decomposition products. The objective of the image pair is to illustrate the difference between fully polarimetric images (a) and simulated compact polarimetric mode (CL) images.

External collaborators

H. McNairn, Agriculture and Agri-Food Canada
P.W. Vachon, Defence R&D Canada
R. DeAbreu, Canadian Ice Services, Environment Canada
R. K. Raney, Johns Hopkins University, APL

Report a problem or mistake on this page
Please select all that apply:

Thank you for your help!

You will not receive a reply. For enquiries, contact us.

Date modified: