loading page

Modelling the Backscattering Properties of Anisotropic Rough Snow Surfaces: Implications for the Evolution of Near Surface Features on the Ice Sheets
  • Shashwat Shukla,
  • Stef Lhermitte,
  • Bert Wouters
Shashwat Shukla
Delft University of Technology

Corresponding Author:sshashwat93@gmail.com

Author Profile
Stef Lhermitte
Delft University of Technology
Author Profile
Bert Wouters
IMAU/TU Delft
Author Profile

Abstract

Understanding the near-surface roughness is paramount for the evolution of surface features and its contribution to the ice sheet dynamics. In regions experiencing the transition from dry snow to melt conditions, the surface becomes rough which along with liquid water content influences the radar backscatter. Such surfaces are driven by wind resulting in the formation of features oriented at different azimuth directions, ranging from ripples to dunes. Different scales of roughness also induce depolarisation in the radar signal, thereby making it difficult to infer the effective contribution of densification process on total backscatter. Here, we present a physical understanding of a wind-induced snow surface having anisotropic roughness from the modelled radar backscatter. As part of the approach, we propose to model the horizontal component of roughness (autocorrelation length) as a function of azimuth direction in the surface scattering model, i.e. Integral Equation Model. We performed our experiments for strongly anisotropic (a < 0.2), weakly anisotropic (a > 0.7), and isotropic (a = 1) surfaces, where a is the anisotropy, tuned for C-band frequency. Also, the minima of backscatter intensity varied over entire range of azimuth angles at a particular incidence angle is derived for retrieving the most dominant wind direction. From our analysis, significant azimuth modulation occurs in the radar backscatter at low incidence angles. Moreover, we found that the most dominant wind direction at 30-deg incidence angle changes from a strongly anisotropic surface (i.e. 0-deg) to a weakly anisotropic surface (i.e. 90-deg), however it remains unaltered for 45-deg and 60-deg incidence angles (i.e. 90-deg). Our modelled results are in strong agreement with wind scatterometer data of ice sheets and further show the potential of 30-deg incidence angle for tracking the changes in wind flow pattern based on directionality of roughness. In this regard, we foresee the importance of azimuth modulation for understanding the properties and orientation of surface features in the ice sheets, with particular application to wind flow. As a future scope, we plan to apply our model on radar images for the retrieval of wind directions and compare the directionality of roughness with the katabatic wind flow patterns and AWS measurements.