Polar sea ice plays an important role in the global climate and other geophysical processes. Although spaceborne scatterometers such as NSCAT have low inherent spatial resolution, resolution enhancement techniques can be utilized to make NSCAT data useful for monitoring sea ice extent in the Antarctic. Dual polarization radar measurement parameters, A and B, are used to identify sea ice and ocean pixels in composite images where A is (sigma) o normalized to 40 degrees and B is the incidence angle dependence of (sigma) o. In particular, the copol ratio and the vertical polarization B values contain useful information about the presence of sea ice. A first estimate of the sea ice extent is obtained through an automated linear discrimination that assigns the decision boundary based upon the properties of the bivariate distribution. This is used to obtain estimates of the statistics needed to perform a more accurate Mahalanobis distance discrimination. Ice edge detection noise reduction is performed through region growing and erosion/dilation techniques. The algorithm is applied to NSCAT data. The resulting edge closely matches the NSIDC SSM/I derived 30 percent ice concentration edge.