Sea Ice Non-Linearities Act to Rectify and Filter Oceanic and Atmospheric Forcing

Sea ice is undergoing rapid alteration in the context of climate change. Understanding its sensitivity to atmospheric and oceanic forcing, constraining its variability and predicting its trend are critical challenges to improve climate projections. The non-linearities controlling sea ice thermodynamics integrate forcing from the ocean and atmosphere in surprising ways, rendering it difficult to understand the processes affecting sea ice response to climate change. In this study, a simple ice thickness model is forced by realistic stochastic atmospheric and oceanic heat fluxes. Ensemble experiments show that the non-linearities in the system rectify the added zero-mean noise on weather timescales leading to a change in mean sea ice state. Most notably, there is a thinning in summer when sea ice is already at its minimum. Whether the ice system can recover during the following months depends on the mean state of the ice. The sea ice system integrates high frequency forcing to influence longer time scales, thus changing not only the mean state but also the interannual to decadal variability of sea ice. Adding a trend to the forcing variables yields estimates of the dominant drivers of the current and future ice loss in the Arctic, with a prevalent role of ice-ocean heat flux over surface heat fluxes. This study reveals sea ice as a fundamental climate component, absorbing the energy into its mean state and transforming weather fluctuations with time scales of days to weeks into energetic internal variability on time scales of months to decades.

Journal of Climate (in review, preprint available upon request)