PrintAbrupt changes have punctuated climate history. An abrupt climate change is one that occurs faster than its cause, or comes so rapidly that ecosystems or economies have trouble adapting. Abrupt climate changes can involve sudden onset of droughts, collapse of ice sheets, or other features of the climate. Studies have especially focused on the north Atlantic events that punctuated the last ice-age cycle (and, probably, earlier cycles).
In the modern world, the relatively salty Atlantic waters become dense enough in the winter to sink in the far northern Atlantic, and then flow south, while warmer surface waters flow north in replacement. Because of this, the North Atlantic Ocean does not freeze in the wintertime even at high latitudes, so the surroundings remain relatively warm all winter. While the “frozen tundra” of Lambeau Field in Green Bay becomes almost too cold for even American football at 45 N latitude in a Wisconsin winter, the Manchester United football/soccer team runs around in shorts at 53 N latitude in England through the winter. The differences in climate between England and Wisconsin arise from several processes, but it is a safe bet that if the North Atlantic Ocean froze in the winter, Manchester United would not be playing a wintertime season.
There is widespread agreement across a range of climate models, from the simplest to the most complex, that a sufficiently large freshening of the north Atlantic under modern or lower carbon-dioxide concentrations would allow wintertime freezing, changing the oceanic and atmospheric circulation. Furthermore, many models find that the climate undergoes jumps — a gradual freshening can lead to a sudden onset of freezing, which will persist through many winters and then terminate suddenly (in as little as a single year, to a few decades). (In many models, the onset of wintertime freezing occurs with loss of the conveyor-belt circulation, but the continuing Trade Winds across Panama increase saltiness in the Atlantic until the conveyor-belt turns on again.)
The data agree with the models. In the past, large floods from ice-dammed lakes, or surges of the ice sheet in Canada, or slower melting of Canadian ice, have delivered extra fresh water to the north Atlantic and led to loss of the conveyor-belt circulation, allowing wintertime freezing in the north Atlantic, and bringing widespread climate changes. These include very strong cooling in wintertime around the north Atlantic, slight cooling around most of the northern hemisphere, slight warming in the southern hemisphere (the conveyor-belt takes sun-warmed water from the south Atlantic to cool in the north- Atlantic winter, so shutting down the flow gives cooling in the north but warming in the south), a southward shift of the tropical circulation pattern, hence strong drying in the places left behind by the intertropical convergence zone (the ITCZ) and strong wetting in the places to which it moves, and general loss of rain in the monsoonal regions of Africa and Asia. Small northern glacier readvance was observed in such events during the termination of the last ice age, but with no ability to return to the ice age (glaciers mostly care about summertime temperatures, but loss of the conveyor primarily cools northern wintertime). Global-average effects of a conveyor shutdown were small-a bit more cooling in north than warming in south, with ice-albedo feedbacks important.
There has been much discussion of whether such an event could occur in the future. A shutdown would affect ocean currents, fisheries, etc., no matter when it occurred. If a shutdown waited too long into the future, the carbon-dioxide warming would largely block wintertime freezing, and with it the big amplifier of climate change. Model results generally show that a shutdown is more likely in a colder climate, and is more likely when a big ice sheet sits on Canada, steering winds towards Spain rather than Norway. Most models of the future agree that melting of Greenland’s ice and other processes will weaken but not shut down the conveyor-belt circulation, and the Intergovernmental Panel on Climate Change (IPCC) in 2007 assessed a <10% chance (but not zero) of an abrupt change over the next century. The movie The Day after Tomorrow surely was not accurate. (But, if your heroes are larger than life, maybe your problems must be larger than life to make an entertaining movie.)
The Holocene shows stability when carbon dioxide was not changing. After the last ice age ended (with the warming beginning about 24,000 years ago and most of the warming completed by 11,500 years ago), we entered what is called the Holocene. Temperature- wise, fluctuations have been small, except for one brief blip about 8200 years ago corresponding to the last of the outburst floods from a lake dammed by the dying ice sheet in Canada.
Not much happened to greenhouse-gas concentrations during most of the Holocene. The Holocene temperature record is well-explained through the influence of changing orbits (more midsummer sunshine in the north a few thousand years ago than more recently), volcanic eruptions (a degree or so cooling for a couple of years from a big eruption that loads the stratosphere with sun-blocking particles; a few eruptions close together can make enough cooling to matter) and solar fluctuations (reconstructed from sunspot observations, using the recent correlation between satellite-measured solar output and sunspot numbers, or reconstructed from beryllium-10 or carbon-14 using the relation between sunspots, the solar wind, and the penetration of cosmic rays that form those isotopes). Some evidence points to a role for changes in the conveyor-belt circulation, which may act to amplify the other causes by slowing slightly in colder times. Searches for influences from changes in the Earth’s magnetic field, from cosmic dust or cosmic rays, or other causes have come up empty; the paleoclimatic record continues to point to a sensible, understandable climate system. (Farther back, about 40,000 years ago, the magnetic field dropped to near zero for a millennium or so, cosmic rays streamed in to create a large spike in beryllium-10, but the climate ignored it, which argues against any serious role for cosmic rays or the magnetic field.)