Internal climate system Feedbacks in Glacial cycles

The link between the periodicity and phasing of orbital variations and the Quaternary glacial cycles is convincing, so there is little doubt that orbital variations play a central role. A great deal is unexplained, however. Much of the original skepticism over the orbital theory stems from the fact that the perturbations to insolation are relatively modest and are hard to reconcile with the magnitude of the climate system response. Sophisticated climate models have been brought to bear on the problem, with little success in simulating either the glacial inception or the deglaciation. This may be a matter of resolution and timescales; it is difficult to conduct millennial-scale integrations with general circulation models (GCMs), the highland locations of ice inception centers are not well resolved, and glacier dynamics (absent in most GCMs) are needed to include the effects of ice advection from inception centers to lower altitudes and latitudes. Long integrations including ice-albedo feedbacks and other climate system feedbacks are needed to describe this process.

The current view is that orbital forcing acts as the trigger for glaciation and deglaciation, but most of the climate forcing that drives glacial-interglacial cycles in fact comes from feedback mechanisms and accompanying shifts in global climate dynamics. Many of the relevant processes are cryospheric, such as the collective albedo and regional cooling impact of expanded sea ice and snow cover, which adds to the area of ice sheets. In addition, increased midlatitude freshwater delivery to the oceans, via iceberg calving and summer melting on the ice sheet margins, affected ocean circulation in the North Atlantic region, reducing deep-water formation and poleward heat transport. The orographic disturbance of the ice sheets also altered patterns of atmospheric circulation.

As discussed in chapter 8, greenhouse gas levels also dropped dramatically during the glaciations. Atmospheric water vapor decreased as the planet cooled, and ice cores indicate that carbon dioxide and methane levels were much lower than preindustrial values. The correlation between Antarctic temperatures, global ice volume, and concentrations of C02 and CH4 hold true over the past several glacial cycles (900 kyr), as far back as the ice cores reach. While the causal link is still enigmatic, greenhouse gas reductions played a crucial role in helping to cool the planet in glacial times, and increases in co2 and cH4 also helped to drive the ice sheet demise during each deglaciation. Changes in oceanic carbon uptake associated with colder oceans, altered ocean circulation, and changes in ocean alkalinity during glaciations are the most likely explanations, although shifts in terrestrial carbon uptake (e.g., methane storage in tropical wetlands, midlatitude permafrost, and subglacial environments) may have played a role.

Overall, glacial cycles provide the textbook example of nonlinear feedback mechanisms in Earth's climate dynamics, which are able to take a relatively small orbital forcing—a seasonal and geographic redistribution of insolation—and amplify it into a global climate shift that transforms the landscape. Cryospheric feedbacks and ice-atmosphere-ocean processes were central to this.

Another aspect of the Pleistocene glacial cycles is important to note in the context of future climate change. There is palynological, sedimentary, ice-core, and sea level evidence that some previous interglacial periods were warmer than the present, in particular isotope stages 11 (ca. 400 ka) and 5e (the Eemian, ca. 125 ka). Arctic sea ice declined at these times, and the southern sector of the Greenland ice sheet retreated dramatically in both of these periods, contributing to sea-level high stands that were 5 to 6 m higher than present. Proxies and modeling indicate that the Arctic was up to 5°C warmer than present in the Eemian, driven by high spring and summer insolation. These periods may be good analogues for the Arctic environment in a warmer world.

0 0

Post a comment