Surface impedance tomography for Antarctic sea ice

• Published in: Deep-sea research. Part II, Topical studies in oceanography Vol. 58; no. 9; pp. 1149 - 1157
• Main Authors: Sampson, C., Golden, K.M., Gully, A., Worby, A.P.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.05.2011
• Subjects: Antarctica; Arrays; Electrical conductivity; Impedance; Inversions; Marine; Mathematical models; Reconstruction; Resistivity; Sea ice; Surface impedance tomography; Wenner array; Wenner array; Surface impedance tomography; Electrical conductivity; Sea ice
• ISSN: 0967-0645; 1879-0100
• DOI: 10.1016/j.dsr2.2010.12.003

During the 2007 SIPEX expedition in pack ice off the coast of East Antarctica, we measured the electrical conductivity of sea ice via surface impedance tomography. Resistance data from classical four-probe Wenner arrays on the surfaces of ice floes were used to indirectly reconstruct the conductivity profiles with depth, involving both the horizontal and vertical components. A common problem with these reconstructions is the lack of uniqueness of the inversions, which worsens as the number of layers in the model increases. In the past, three layer inversions have been used to help avoid non-uniqueness. However, this approach assumes that the conductivity profile of sea ice does not change very much with depth. In order to investigate the conductivity profiles one needs to use more layers in the reconstruction. A reasonable starting model is a useful tool that can be used to regularize the inverse problem, allowing a reconstruction that not only matches the Wenner impedance data but the actual profile. Using measurements of brine volume fraction for 10 cm sections of ice cores taken at the Wenner array site, and various models relating brine volume fraction to conductivity, we compare the predicted conductivity profiles based on the models to the reconstructions from the tomographic measurements. We note the close agreement with the actual data for some models and the inadequacy of others. Such models could be useful in finding a reasonable starting point for regularizing inversions, and using n-layer models to reconstruct accurate conductivity profiles. Our results help to provide a rigorous basis for electromagnetic methods of obtaining sea ice thickness data, a key gauge of the impact of climate change in the polar regions.