Accumulation of dynamic and thermodynamic forces in the Arctic are decreasing the extent of thin annual sea ice which polar bear rely on for survival. It is imperative that we identify the preferred habitat of polar bears to fully understand their future requirements. In this thesis, surveys of polar bear tracks and the surrounding sea ice variables, at various scales, were recorded. Sea ice roughness was measured through surveys of the sample area in situ, with an electromagnetic induction (EM) system (IcePIC) mounted to a helicopter, and analyzed through advanced synthetic aperture radar (ASAR) images of the study area. In situ Polar bear tracks provided a limited association with the EM sea ice roughness and a negative association to ASAR sea ice roughness. Results indicate a significant association between EM mean values and ASAR brightness means. In addition, EM statistics and ASAR texture statistics were correlated through a best fit regression model. These associations display a remote method to identify preferred polar bear habitat and provide a potential linkage between the regional (EM) and remotely sensed (ASAR) assessment of sea ice roughness. / October 2008

Accumulation of dynamic and thermodynamic forces in the Arctic are decreasing the extent of thin annual sea ice which polar bear rely on for survival. It is imperative that we identify the preferred habitat of polar bears to fully understand their future requirements. In this thesis, surveys of polar bear tracks and the surrounding sea ice variables, at various scales, were recorded. Sea ice roughness was measured through surveys of the sample area in situ, with an electromagnetic induction (EM) system (IcePIC) mounted to a helicopter, and analyzed through advanced synthetic aperture radar (ASAR) images of the study area. In situ Polar bear tracks provided a limited association with the EM sea ice roughness and a negative association to ASAR sea ice roughness. Results indicate a significant association between EM mean values and ASAR brightness means. In addition, EM statistics and ASAR texture statistics were correlated through a best fit regression model. These associations display a remote method to identify preferred polar bear habitat and provide a potential linkage between the regional (EM) and remotely sensed (ASAR) assessment of sea ice roughness.

Accumulation of dynamic and thermodynamic forces in the Arctic are decreasing the extent of thin annual sea ice which polar bear rely on for survival. It is imperative that we identify the preferred habitat of polar bears to fully understand their future requirements. In this thesis, surveys of polar bear tracks and the surrounding sea ice variables, at various scales, were recorded. Sea ice roughness was measured through surveys of the sample area in situ, with an electromagnetic induction (EM) system (IcePIC) mounted to a helicopter, and analyzed through advanced synthetic aperture radar (ASAR) images of the study area. In situ Polar bear tracks provided a limited association with the EM sea ice roughness and a negative association to ASAR sea ice roughness. Results indicate a significant association between EM mean values and ASAR brightness means. In addition, EM statistics and ASAR texture statistics were correlated through a best fit regression model. These associations display a remote method to identify preferred polar bear habitat and provide a potential linkage between the regional (EM) and remotely sensed (ASAR) assessment of sea ice roughness.