Oil spills in the Arctic are becoming more likely as shipping traffic increases in response to climate-related sea ice loss. To improve oil spill detection capability, we used a controlled mesocosm to analyze the multipolarized C-band backscatter response of oil in newly formed sea ice (NI). Artificial sea ice was grown in two cylindrical tubs at the Sea-ice Environmental Research Facility, University of Manitoba. The sea ice physical characteristics, including surface roughness, thickness, temperature, and salinity, were measured before and after oil injection below the ice sheet. Time-series C-band radar backscatter measurements detected the differences in the sea ice evolution and oil migration to the sea ice surface in the oil-contaminated tub, which was compared to uncontaminated ice in a control tub. Immediately prior to the presence of oil on the ice surface, the copolarized backscatter is increased by 13-dB local maximum, while the cross-polarized backscatter is decreased by 9-dB. Ice physical properties suggest that the local backscatter maximum and minimum, which occurred immediately before oil migrated onto the surface, were related to a combination of brine and oil upward migration. The findings of this work provide a baseline data interpretation for oil detection in the Arctic Ocean using current and future C-band multipolarization radar satellites.

The fusion of Iterative Closest Point (ICP) registrations in existing state estimation frameworks relies on an accurate estimation of their uncertainty. In this paper, we study the estimation of this uncertainty in the form of a covariance. First, we scrutinize the limitations of existing closed-form covariance estimation algorithms over 3D datasets. Then, we set out to estimate the covariance of ICP registrations through a data-driven approach, with over 5100000 registrations on 1020 pairs from real 3D point clouds. We assess our solution upon a wide spectrum of environments, ranging from structured to unstructured and indoor to outdoor. The capacity of our algorithm to predict covariances is accurately assessed, as well as the usefulness of these estimations for uncertainty estimation over trajectories. The proposed method estimates covariances better than existing closed-form solutions, and makes predictions that are consistent with observed trajectories.