Reversibility of Sea Ice and Climate under Global Change

We present work on several topics in physical climatology. Motivated by recent satellite observations of Arctic sea ice age, we investigate how the properties of first-year and multiyear ice regulate sea ice area and volume. Using an autoregressive model of sea ice and numerical simulation with a global dynamic-thermodynamic sea ice model that traces first-year and multiyear, we show that the sea ice mean state, variability, and sensitivity to forcing may be related to sea ice age. Ice age observations may thus offer insights into Arctic sea ice changes under warming.;We next investigate the plausibility of sea ice threshold behavior and irreversible loss under global warming, a subject of much speculation following rapid sea ice declines in recent years. Within a state-of-the-art global climate model (GCM), carbon dioxide is increased until the oceans become ice-free and subsequently decreased until the ice cover returns. We find no evidence of irreversibility over the range of simulated sea ice conditions in the summer or winter ice cover in either hemisphere.;We then shift our focus to quantification of the climate commitment—the warming that would still occur given no further emissions of greenhouse gases or aerosols. We show that elimination of emissions would cause global temperature to increase rapidly as aerosols are quickly washed from the atmosphere, and that large uncertainty in the present aerosol forcing implies large uncertainty in the climate commitment. The multi-millennial atmospheric lifetime of anthropogenic carbon dioxide causes global temperature to remain elevated for the indefinite future, making climate change from greenhouse gas emissions irreversible on human timescales.;Finally, we propose a mechanism for the time-variation of global climate feedback as diagnosed within GCMs. Using a GCM and a low-order energy balance model, we show that the global climate feedback is fundamental related to the geographic patterns of regional feedbacks and surface warming. Its time-variation arises naturally when the pattern of surface warming evolves, activating regional feedbacks of different strengths. These findings have implications for our ability to constrain future climate changes from past and present observations, and for the quantification of climate sensitivity within models and observations.