Using Sea Ice Age as a Proxy for Sea Ice Thickness

Since the beginning of the modern satellite record starting in October 1978, the Arctic sea ice cover has been shrinking, with the largest changes observed at the end of the melt season in September. Through 2013, the September ice extent has declined at a rate of -14.0% dec-1, or -895,300 km2 dec-1. The seven lowest September extents in the satellite record have all occurred in the past seven years. This reduction in ice extent is accompanied by large reductions in winter ice thicknesses that are primarily explained by changes in the ocean's coverage of multiyear ice (MYI). Using the University of Colorado ice age product developed by J. Maslanik and C. Fowler, and currently produced by M. Tschudi we present recent changes in the distribution of ice age from the mid 1980s to present. The CU ice age product is based on (1) the use of ice motion to track areas of sea ice and thus estimate how long the ice survives within the Arctic, and (2) satellite imagery of sea ice concentration to determine when the ice disappears. Age is assigned on a yearly basis, with the age incremented by one year if the ice survives summer melt and stays within the Arctic domain. Age is counted from 1 to 10 years, with all ice older than 10 years assigned to the "10+" age category. The position of the ice is calculated on weekly time steps on NSIDC's 12.5-km EASE-grid. In the mid-1980s, MYI accounted for 70% of total winter ice extent, whereas by the end of 2012 it had dropped to less than 20%. This reflects not only a change in ice type, but also a general thinning of the ice pack, as older ice tends to be thicker ice. Thus, with older ice being replaced by thinner first-year ice, the ice pack is more susceptible to melting out than it was in 1980's. It has been suggested that ice age may be a useful proxy for long-term changes in ice thickness. To assess the relationship between ice age and thickness, and how this may be changing over time, we compare the ice age fields to several observational data sets on ice thickness. These comparisons reveal that while a near-linear relationship between age and thickness for ice up to 3m thick existed in earlier years, this relationship is changing.