 Remote sensing has become increasingly popular for mapping land cover in the Arctic due to its ability to capture large areas quickly and accurately. However, there is still much work to be done to improve the accuracy of these maps. This paper examined the use of remote sensing data to create a land cover map around the Canadian forces station alert in Nunavut, Canada. The authors tested various predictive models and found that the most effective was one that incorporated soil-adjusted vegetation indices and hydrological predictors related to water bodies and snow banks. The resulting map had an overall accuracy of 85% and produced less bias than other classifiers. Additionally, the authors identified some challenges such as shadows cast by boulders and snow covered by soil material, which need to be addressed before more accurate maps can be created. This article was authored by Emily Desjardins, Sandra Lai, Loron Howe, and others.