C42B-08:
Spatio-temporal Variability of Arctic Sea Ice from Days to Decades.

Thursday, 18 December 2014: 12:05 PM
John S. Wettlaufer1,2 and Sahil Agarwal2, (1)Yale University, New Haven, CT, United States, (2)University of Oxford, Mathematical Institute, Oxford, United Kingdom
Abstract:
We examine the daily satellite retrievals of Arctic sea-ice extent over three decades. The Arctic basin is divided into 14 different regions and we study how the Arctic wide observations emerge from these regional trends. In order to capture all possible time scales, the temporal dynamics within each spatial region, and in the basin as a whole, are treated as a multifractal. Our approach is Multi-Fractal Temporally Weighted Detrended Fluctuation Analysis (MF-TWDFA), which captures long time scales that are not accessible by standard DFA or MF-DFA and are by definition not part of a procedure that assumes the dynamics are described by an order one auto-regressive process. We examine how the eastern, western, central and the marginal seas control the longest time scales and what processes are responsible for the responses found in the data. Of particular interest is how regions with substantial seasonality contribute to the basin-wide dynamics of the system. The importance of this is intertwined with the general problem of how and when the Arctic may pass on to a state in which the ice cover is seasonal. The transition from a perennial state to a seasonal ice cover is the persistent state of affairs in the marginal regions of the ice cover, which constitute 10 of the 14 subregions, whereas the remaining 4 are in various stages of partial ice cover at various times of the year. In this manner, we use sub-basins as laboratories for the entire basin.