Abstract

The knowledge of glacier zones’ extent and their temporal variations is fundamental for the retrieval of surface mass balance of glaciers and ice sheets. In this context, a key parameter is the firn line (FL), the lower boundary of the percolation zone, whose location is an indicator of time-integrated mass balance changes. Several approaches have been developed in the last decades to map the FL by means of Synthetic Aperture Radar (SAR) imagery, mainly exploiting backscatter intensities and their seasonal variation. In this paper, an alternative approach is proposed, based on co-polarisation phase differences (CPDs). In particular, CPDs are interpreted as the result of propagation through anisotropic firn layers and are, therefore, proposed as an indicator of the presence of firn. A model is employed to demonstrate the link between CPDs and firn depth, indicating the potential of polarimetric SAR to improve firn characterization beyond spatial extent and FL detection. The proposed approach is demonstrated on L-band airborne data, acquired on 21 May 2015 by the F-SAR sensor of DLR in West Greenland during the ARCTIC15 campaign, and validated with in-situ information available from other studies.

Highlights

  • The knowledge of glacier zones’ extent and their temporal variations is fundamental for the retrieval of surface mass balance of glaciers and ice sheets

  • According to Equation (4), for a firn layer with l = 1 m and ρ f irn = 0.6 g/cm3 the L-band co-polarisation phase differences (CPDs) ranges, at an incidence angle θ = 30◦, from slightly less than 1◦ to 5◦ for S p values between 1.05 and 1.4 associated, according to Equation (2), to a dielectric anisotropy (∆ε = ε z − ε x,y ) ranging from 0.015 to 0.085, in agreement with the values reported for Greenland [14] and Antarctica [22]

  • A model-based inversion of CPDs has been performed using airborne Synthetic Aperture Radar (SAR) data acquired along a 200-km long transect in West Greenland covering the transition from the ablation to the percolation zone of the ice sheet

Read more

Summary

CPD Model for Firn Mapping

The CPD model proposed in [18] allows to describe the propagation of electromagnetic waves through birefringent media with slightly different velocities at different polarisations, introducing phase differences between the polarimetric channels. According to Equation (4), the CPD induced by a layer of anisotropic firn depends on two sensor parameters—the wavelength λ and incidence angle θ—and on the three firn properties: density ρ f irn , thickness l and grain shape factor S p. According to Equation (4), for a firn layer with l = 1 m and ρ f irn = 0.6 g/cm the L-band CPD ranges, at an incidence angle θ = 30◦ , from slightly less than 1◦ to 5◦ for S p values between 1.05 and 1.4 associated, according to Equation (2), to a dielectric anisotropy (∆ε = ε z − ε x,y ) ranging from 0.015 to 0.085, in agreement with the values reported for Greenland [14] and Antarctica [22]. A 15% inaccuracy in the knowledge of ρ f irn leads to a 14% uncertainty in the L-band CPD, indicating a smaller impact than for the particle shape

Test Site and Experimental SAR Data
Firn Thickness Inversion
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call