A Statistical Approach for Estimating Sea Ice Thickness

Abstract

Abstract Sea ice reporting and ice charts are a necessary simplification of highly variable sea ice conditions over large areas. Ice conditions are typically shown using egg codes. These codes can be used to determine an equivalent ice thickness as a method for calculating a ship’s ice resistance. This approach does not acknowledge the variability within the chosen region, such as the presence of ridges, ice pressure and snow cover. A new statistical method is proposed to use live capture data including the speed and resistance from a medium ice breaker to estimate the changing ice thickness. The thickness is calculated using the ship’s regression model for the ship’s performance based on model scale tests. Measured thrust data from the ship is used in conjunction with the thrust deduction factor and open water resistance to derive the ice resistance and subsequent thickness. The statistical ice thickness is compared against the equivalent ice thickness derived from the ice egg codes to determine the accuracy of current ice reporting and identify the variance within a defined region. This statistical approach may also assist in identifying features such as ridges, including the frequency of encountering these events. This analysis has high potential for sources of error, such as the changes in force required to bend and break ice sheets. Therefore, the data must be adequately analyzed to eliminate such noise while preserving the ability to identify changes to ice thickness. The potential implications of this research are an improved awareness of the variability in sea ice, and improvements to the methods for reporting and characterizing sea ice conditions.