On the Interpretation of L- and P-Band PolSAR Signatures of Polythermal Glaciers

• Main Authors: Parrella, Giuseppe, Hajnsek, Irena, Papathanassiou, Konstantinos
• Other Authors: Lacoste-Francis, H.
• Format: Conference Object
• Language: English
• Published: ESA Communications 2013
• Subjects: Radarkonzepte; Austfonna; Ice cap
• Online Access: https://elib.dlr.de/78902/; https://elib.dlr.de/78902/1/s7_parre.pdf

Long-wavelength (e.g. P- and L- band) SAR systems can penetrate tens of meters deep into ice bodies. Hence, they are sensitive to the ice surface as well as to sub-surface (volume) ice structures. Both contributions are present in the SAR signature that has to be interpreted accordingly. For this, significant attention has been given to model-based decomposition techniques of polarimetric SAR (PolSAR) data. In this sense, this paper extends and investigates in detail the modeling of potential scattering contributions with the attempt to explain long-wavelength PolSAR signatures of subpolar glacier ice. The main effort is the development of an extended volume scattering component; different kinds of inclusions typically present in glacier ice (e.g. air bubbles, oriented crystals fabrics, etc.) are modeled by particle clouds with variable particle shape and orientation in a three dimensional space. For the case of oriented particles, the volume anisotropy induces differential propagation effects. The associated differential propagation velocities (phase differences) and losses are accounted for the different polarimetric channels. Second order mechanisms generated from the interaction between adjacent particles or internal ice layers might play a relevant role. For instance, double reflections can cause significant co-polarization phase difference as well as increase in entropy (in the far range), despite their very low scattering amplitude. Finally, a surface scattering contribution from the air/ice interface at the glacier surface must be considered. Covariance matrices of the above mentioned contributions are modeled and incoherently combined. Different possible scenarios are then simulated and analyzed (particle shape and orientation, inclusions nature and volume fraction, distribution of power contributions, etc.). A performance assessment is conducted by comparing modeled PolSAR signature to fully polarimetric SAR data at L- and P-band acquired by DLR’s E-SAR system, over the Austfonna ice-cap, in the ...