Abstract
Abstract. The diminishing Arctic sea ice pack has been widely studied, but previous research has mostly focused on time-mean changes in sea ice rather than on short-term variations that also have important physical and societal consequences. In this study we test the hypothesis that future interannual Arctic sea ice area variability will increase by utilizing 40 independent simulations from the Community Earth System Model's Large Ensemble (CESM-LE) for the 1920–2100 period and augment this with simulations from 12 models participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5). Both CESM-LE and CMIP5 models project that ice area variability will indeed grow substantially but not monotonically in every month. There is also a strong seasonal dependence in the magnitude and timing of future variability increases that is robust among CESM ensemble members. The variability generally correlates with the average ice retreat rate, before there is an eventual disappearance in both terms as the ice pack becomes seasonal in summer and autumn by late century. The peak in variability correlates best with the total area of ice between 0.2 and 0.6 m monthly thickness, indicating that substantial future thinning of the ice pack is required before variability maximizes. Within this range, the most favorable thickness for high areal variability depends on the season, especially whether ice growth or ice retreat processes dominate. Our findings suggest that thermodynamic melting (top, bottom, lateral) and growth (frazil, congelation) processes are more important than dynamical mechanisms, namely ice export and ridging, in controlling ice area variability.
Highlights
Arctic sea ice extent has declined by more than 40 % since 1979 during summer (e.g., Stroeve et al, 2012; Serreze and Stroeve, 2015; Comiso et al, 2017), primarily as a consequence of greenhouse gas forcing (Notz and Marotzke, 2012) and internal variability (Ding et al, 2017)
Sea ice area in the Community Earth System Model’s Large Ensemble (CESM-LE) is projected to decline in all months in the 21st century, proceeding in three phases: a fairly stable regime of extensive coverage in the 20th century, a decline, followed by virtually no ice remaining in summer and autumn months (Fig. 1)
The same relationship between ice area and its variability is maintained across Coupled Model Intercomparison Project Phase 5 (CMIP5) models, though with more noise resulting from the aggregation of many different models rather than ensemble members from a single model (Fig. 3)
Summary
Arctic sea ice extent has declined by more than 40 % since 1979 during summer (e.g., Stroeve et al, 2012; Serreze and Stroeve, 2015; Comiso et al, 2017), primarily as a consequence of greenhouse gas forcing (Notz and Marotzke, 2012) and internal variability (Ding et al, 2017). Changes in the interannual variability in sea ice coverage have been studied only in a limited capacity, likely because they are only beginning to become visible in September in the present day Both Goosse et al (2009) and Swart et al (2015; their Fig. S6 in the Supplement) reported that maximum ice area variability during September occurs once the mean ice extent declines to 3–4 million km. We analyze a large 40-member ensemble from a single global climate model (GCM), which allows us to isolate internal variability, which is otherwise muddled with inter-model variability in multi-model comparisons This allows us to test the hypothesis that interannual Arctic sea ice cover variability will increase throughout the year in the future as the ice pack diminishes
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
