Four Degrees And Beyond - Royal Society Summary Of Latest Pre-Cancun Climate Outlook

EDIT

That there will be large variations in spatial patterns of climate change associated with any global temperature are well established and are well summarized in the IPCC 4th Assessment <18>. Land areas warm more than the oceans, so for almost all areas of human habitation, temperature increases will exceed, frequently by more than one-and-a-half times, the global average. Temperature changes at high latitudes are projected to be especially amplified, largely owing to snow and ice albedo feedbacks; boreal summer temperatures are at least twice the global average warming, and Arctic Ocean winter temperatures warm three times faster than average. While global average precipitation is projected to increase <19>, most areas that are currently arid and semi-arid are projected to dry, while the moist tropics and mid-latitudes are projected to become wetter, a signal that appears to be emerging in recent precipitation trends <20>.

An important question is whether this spatial pattern of change is similar in 2°C and 4°C worlds. Sanderson et al. <21> explore this by comparing global climate models that warm by at least 4°C by 2100 with those that warm less rapidly under the IPCC SRES A2 emissions scenario. They show that the pattern of warming relative to global mean temperature change is very similar between the two classes of climate model, apart from during boreal summers where warming is amplified in models that warm faster. In areas where precipitation decreases, temperature increases tend to be amplified, probably owing to reduced evaporative cooling of the land surface. The broadly constant ratio of local climate change to global temperature change implies that these local changes are amplified in a 4°C world; for example, a local change of 3°C in a +2°C world (1°C greater than the global average) becomes 7.5°C in a +4°C world (3.5°C above the global average).

One of the most certain outcomes of a warmer world is an increase in global sea level, although the actual amount of sea-level rise (SLR) is rather less certain. Nicholls et al. <22> review recent literature on SLR, and propose that in a world that warms by 4°C by 2100, global sea level will increase between 0.5 and 2 m by the end of the century, but with rises greater than 1 m being much less likely. A warming of 4°C will also commit the world to larger SLRs beyond 2100, as the ocean equilibrates thermally to atmospheric warming; these post-2100 increases could be large should irreversible melting of the Greenland ice sheet be triggered and some level of break-up of the West Antarctic ice sheet occur <22>.

There are a range of other potential thresholds in the climate system and large ecosystems that might be crossed as the world warms from 2°C to 4°C and beyond <23>. These include permanent absence of summer sea ice in the Arctic <24>, loss of the large proportion of reef-building tropical corals <25>, melting of permafrost at rates that result in positive feedbacks to greenhouse gas warming through CH4 and CO2 releases <26,27> and die-back of the Amazon forest <28>. While the locations of these thresholds are not precisely defined, it is clear that the risk of these transitions occurring is much larger at 4°C—and so the nature of the changes in climate we experience may well start shifting from incremental to transformative.

EDIT

http://rsta.royalsocietypublishing.org/content/369/1934/6.full