McMurdo Ice Shelf Sounding and Radar Statistical Reconnaissance at 60-MHz: Brine Infiltration Extent and Surface Properties

Tuesday, 16 December 2014
Cyril Grima, Arami Rosales, Donald D Blankenship and Duncan A Young, University of Texas, Institute for Geophysics, Austin, TX, United States
McMurdo Ice Shelf, Antarctica, is characterized by two particular geophysical processes. (1) Marine ice accretion supplies most of the ice shelf material rather than meteoric ice from glacier outflow and snow-falls. (2) A brine layer infiltrates the ice shelf laterally up to 20-km inward. The infiltration mainly initiates at the ice-front from sea water percolation when the firn/snow transition is below sea-level. A better characterization of the McMurdo ice shelf could constrain our knowledges of these mechanisms and assess the stability of the region that hosts numerous human activities from the close McMurdo station (USA) and Scott base (New-Zealand). McMurdo ice shelf is also an analog for the Jovian icy moon Europa where brine pockets are supposed to reside in the ice crust and accretion to occur at the 10-30-km deep ice-ocean interface.

The University of Texas Institute for Geophysics (UTIG) acquired two radar survey grids over the McMurdo Ice Shelf during southern summers 2011-2012 and 2012-2013 with the High Capability Radar Sounder (HiCARS) on-board a Basler DC-3 aircraft. HiCARS transmits a chirped signal at 60-MHz central frequency and 15-MHz bandwidth. The corresponding vertical resolution in ice is 5-10 m. An important design goal of the radar was to maintain sufficient dynamic range to correctly measure echo intensities.

Here we present the brine infiltration extent and bathymetry derived from its dielectric horizon well distinguishable on the HiCARS radargram. We complement the ice-shelf characterization by classifying its surface thanks to the novel Radar Statistical Reconnaissance (RSR) methodology. The RSR observable is the statistical distribution of the surface echo amplitudes from successive areas defined along-track. The distributions are best-fitted with a theoretical stochastic envelop parameterized with the signal reflectance and scattering. Once those two components are deduced from the fit, they are used in a backscattering model to invert surface properties such as roughness, density, and/or impurity load. This combined analysis gives new insights into the superficial processes and exchanges at the McMurdo ice shelf.