Observed and modeled patterns of lower stratospheric seasonal trends in Antarctic ozone and temperature in the late 20th (1979–2000) and the early 21st (2000–2014) centuries are compared. Patterns of pre‐2000 observed Antarctic ozone decreases and stratospheric cooling as a function of month and pressure are followed by opposite‐signed (i.e., “mirrored”) patterns of ozone increases and warming post‐2000. An interactive chemistry‐climate model forced by changes in anthropogenic ozone depleting substances produces broadly similar mirrored features. Statistical analysis of unforced model simulations (from long‐term model control simulations of a few centuries up to 1000 years) suggests that internal and solar natural variability alone is unable to account for the pattern of observed ozone trend mirroring, implying that forcing is the dominant driver of this behavior. Radiative calculations indicate that ozone increases have contributed to Antarctic warming of the lower stratosphere over 2000–2014, but dynamical changes that are likely due to internal variability over this relatively short period also appear to be important. Overall, the results support the recent finding that the healing of the Antarctic ozone hole is underway and that coupling between dynamics, chemistry, and radiation is important for a full understanding of the causes of observed stratospheric temperature and ozone changes.

The sea-ice zone surrounding both the Arctic and Antarctic coasts is vulnerable to melt from underlying warm waters residing at depth. In the Southern Ocean, upper circumpolar deep waters (UCDWs) flow below the surface near the continental shelf, where their heat influences both marine-terminating glaciers and sea-ice (Ackley et al., 2015; McKee et al., 2019; Orsi et al., 1995). The Antarctic upper ocean is relatively weakly stratified, such that heat stored at depth may readily be ventilated to the surface. In winter, typical ocean heat fluxes to sea-ice may reach 25-35 W m −2 (Martinson & Iannuzzi, 1989). Ackley et al. (2015) measure heat fluxes of about 8 W m −2 under pack ice in the Bellingshausen Sea and 17 W m −2 under fast ice in the Amundsen Sea, consistent with co-located sea-ice melt rates. In the Arctic, the upper ocean is more strongly stratified, such that the direct influence of deep water-masses on sea-ice melt tends to be weaker. Intermediate Pacific Waters often flow in between the surface layer and warm Atlantic waters, shielding sea-ice from their heat (Carmack et al., 2015). The Arctic is also more quiescent, resulting in relatively weak vertical heat exchanges within the ocean interior (Lenn et al., 2009). Nevertheless, remnant summer waters

A hierarchy of models is used to explore the role of the ocean in mediating the response of the climate to a single volcanic eruption and to a series of eruptions by drawing cold temperature anomalies into its interior, as measured by the ocean heat exchange parameter q (W m−2 K−1). The response to a single (Pinatubo-like) eruption comprises two primary time scales: one fast (year) and one slow (decadal). Over the fast time scale, the ocean sequesters cooling anomalies induced by the eruption into its depth, enhancing the damping rate of sea surface temperature (SST) relative to that which would be expected if the atmosphere acted alone. This compromises the ability to constrain atmospheric feedback rates measured by λ (~1 W m−2 K−1) from study of the relaxation of SST back toward equilibrium, but yields information about the transient climate sensitivity proportional to λ + q. Our study suggests that q can significantly exceed λ in the immediate aftermath of an eruption. Shielded from damping to the atmosphere, the effect of the volcanic eruption persists on longer decadal time scales. We contrast the response to an “impulse” from that of a “step” in which the forcing is kept constant in time. Finally, we assess the “accumulation potential” of a succession of volcanic eruptions over time, a process that may in part explain the prolongation of cold surface temperatures experienced during, for example, the Little Ice Age.

Marginal ice zones are composed of discrete sea‐ice floes, whose dynamics are not well captured by the continuum representation of sea ice in most climate models. This study makes use of an ocean large eddy simulation (LES) model, coupled to cylindrical sea‐ice floes, to investigate thermal and mechanical interactions between melt‐induced submesoscale features and sea‐ice floes, during summer conditions. We explore the sensitivity of sea‐ice melt rates and upper‐ocean turbulence properties to floe size, ice‐ocean drag, and surface winds. Under low wind conditions, upper ocean turbulence transports warm cyclonic filaments from the open ocean toward the center of the floes and enhances their basal melt. This heat transport is partially suppressed by trapping of ice within cold anticyclonic features. When winds are stronger, melt rates are enhanced by the decoupling of floes from the cold, melt‐induced lens underneath sea ice. Distinct dynamical regimes emerge in which the influence of warm filaments on sea‐ice melt is mitigated by the strength of ice‐ocean coupling and eddy size relative to floe size. Simple scaling laws, which may help parameterize these processes in coarse continuum‐based sea‐ice models, successfully capture floe melt rates under these limiting regimes.

A coupled climate model with idealized representations of atmosphere, ocean, sea ice, and land is used to investigate transitions between global climate equilibria. The model supports the presence of climates with limited ice cover (Warm), a continuum of climates in which sea ice extends down into the midlatitudes and the tropics (Cold), together with a completely ice-covered earth (Snowball). Transitions between these states are triggered through volcanic eruptions, where the radiative effect of stratospheric sulfur emissions is idealized as an impulse reduction in incoming solar radiation. Snowball transitions starting from the Cold state are more favorable than from the Warm state, because less energy must be extracted from the system. However, even when starting from a Cold climate, Toba-like volcanic events (cooling of order −100 W m−2) must be sustained continuously for several decades to glaciate the entire planet. When the deep ocean is involved, the volcanic response is characterized by relaxation time scales spanning hundreds to thousands of years. If the interval between successive eruptions is significantly shorter (years to decades) than the ocean’s characteristic time scales, the cumulative cooling can build over time and initiate a state transition. The model exhibits a single hysteresis loop that connects all three climate equilibria. When starting from a Snowball, the model cannot access the Cold branch without first transitioning to an ice-free climate and completing the hysteresis loop. By contrast, a Cold state, when warmed, transitions to the Warm equilibrium without any hysteresis.