Summary
We perform a theoretical analysis of Antarctic sea ice thickness estimates made using a ship-borne electromagnetic instrument. Inversion of theoretical one-dimensional sea ice models assuming a conductive half-space model has shown that the ship-borne system is able to determine thickness to within 10% for level ice up to ~2 m thick. However, the presence of a thin conducting brine layer within the ice, common during the Antarctic summer, results in a large error in thicknesses estimated assuming perfectly-resistive ice. An analysis of the effect of system attitude on the electromagnetic response has shown that pitch and roll of up to 10° can result in significant errors in the interpreted ice thickness in areas of thin sea ice. A preliminary three-dimensional model study of the EM-31 system has shown that it can yield good estimates of maximum sea ice keel thickness. However, sea ice thickness is significantly over- or underestimated near sharp vertical boundaries in the three-dimensional model.