Velocity field in the McMurdo shear zone from annual ground penetrating radar imaging and crevasse matching

The McMurdo shear zone (MSZ) is strip of heavily crevassed ice oriented in the south-north direction and moving northward. Previous airborne surveys revealed a chaotic crevasse structure superimposed on a set of expected crevasse orientations at 45 degrees to the south-north flow (due to shear stress mechanisms). The dynamics that produced this chaotic structure are poorly understood. Our purpose is to present our field methodology and provide field data that will enable validation of models of the MSZ evolution, and here, we present a method for deriving a local velocity field from ground penetrating radar (GPR) data towards that end. Maps of near-surface crevasses were derived from two annual GPR surveys of a 28 km2 region of the MSZ using Eulerian sampling. Our robot-towed and GPS navigated GPR enabled a dense survey grid, with transects of the shear zone at 50 m spacing. Each survey comprised multiple crossings of long (> 1 km) crevasses that appear in echelon on the western and eastern boundaries of the shear zone, as well as two or more crossings of shorter crevasses in the more chaotic zone between the western and eastern boundaries. From these maps, we derived a local velocity field based on the year-to-year movement of the same crevasses. Our velocity field varies significantly from fields previously established using remote sensing and provides more detail than one concurrently derived from a 29-station GPS network. Rather than a simple velocity gradient expected for crevasses oriented approximately 45 degrees to flow direction, we find constant velocity contours oriented diagonally across the shear zone with a wavy fine structure. Although our survey is based on near-surface crevasses, similar crevassing found in marine ice at 160 m depth leads us to conclude that this surface velocity field may hold through the body of meteoric and marine ice. Our success with robot-towed GPR with GPS navigation suggests we may greatly increase our survey areas.