![Information Unit for Conventions (IUC)](iuc1.gif)
Climate Change Information Sheet 6
Has climate change already begun ?
- The earth's climate is already adjusting to past greenhouse gas emissions. The climate system must adjust to changing greenhouse gas concentrations in order to keep the global energy budget balanced. This means that the climate is changing and will continue to change as long as greenhouse gas levels keep rising. But this is not very useful. The real question is how large the change is likely to be relative to the natural climate fluctuations that human societies and natural ecosystems have learned to adapt to.
- Measurement records indicate a warming of 0.3o-0.6oC in global average temperature since 1860. This is in line with model projections of the size of warming to date, particularly when the cooling effect of sulphur emissions is included. But observations are sparse before 1900 and much of the warming occurred between 1910 and 1940, before the largest rise in greenhouse gases. There is clearly more going on than a simple, direct response to emissions. This is to be expected as the climate is a complicated and chaotic system.
- Mean sea level has risen by 10 to 25 cm and mountain glaciers have retreated. As the upper layers of the oceans warm, water expands and sea level rises. Models suggest that a 0.3o-0.6oC warming should indeed result in a 10 to 25 cm sea-level rise. But other, harder-to-predict, changes also affect the real and apparent sea level, notably snowfall and ice-melt in Greenland and Antarctica and the slow "rebound" of northern continents freed from the weight of ice age glaciers. Almost all recorded mountain glaciers show a retreat over the past century but, as with sea level, this is unlikely to be only a response to changes in greenhouse gases.
- The observed global warming trend is larger than the trends that models indicate could be due to natural variability. A key problem in climate change research is that scientists have no direct way of observing what would have happened if humanity had left the climate alone. There is no direct way of comparing the greenhouse "signal" with the background "noise" of natural climate variability. Instead, this background variability can be estimated by running climate change computer models with constant greenhouse gas levels. The results indicate that the warming trend of 0.3o0.6oC per century is unlikely to be a chance fluctuation. However, indirect evidence from past climates suggests that these models underestimate the size of natural climate variability, so they may be overestimating the significance of the signal.
- Climate models omit many sources of variability that could also cause apparent long-term trends. Current model-based estimates of natural variability do not include the effects of volcanic eruptions, which can cool the global climate temporarily by several tenths of a degree. They are also only beginning to include the effects of longterm changes in the power output of the sun. The sun may have been responsible for relatively cool periods during the 16th, 17th, and 19th centuries (the so-called "Little Ice Age") when the northern hemisphere may have been about 0.5oC colder than it is today. Some of the warming over the past century (about 2030% of it, according to some recent model results) may still be a recovery from that time.
- Models can also be used to predict the overall pattern of climate change. Because so many unknown factors may affect the global average temperature, scientists are reluctant to conclude that greenhouse warming has arrived on the evidence of that one number alone. Instead, they look for similarities between the pattern of change emerging in the observations and the pattern projected by climate models.
- Several studies have reported increasingly close agreement between model-predicted and observed patterns of temperature change. Studies of surface temperature records show some evidence that the land is warming faster than the oceans. They also show reduced warming in areas affected by sulphate aerosols and in those ocean regions where surface water mixes down, distributing any warming to the ocean depths - all features of the modelpredicted pattern. But coverage is incomplete, and observations in different regions (e.g. land vs. sea) are made in different ways. A more consistent, but much shorter, record is provided by air temperatures from meteorological stations. These show a pattern of cooling in the stratosphere (above about 10 km) and warming in the troposphere (lower atmosphere), which is also predicted by climate models (see figure).
- The satellite record is still too short to reveal significant trends. The climate has to be observed over several decades before any climate change signal can be distinguished from natural variability. The longest satellite records are still well under 20 years. Models predict that it should not be possible to detect anything in such a short period, so all that can be said about the satellite data for the moment is that they are consistent with climate model projections and with evidence from conventional observations. Satellite data do provide global coverage, which helps to validate models and reduce uncertainties.
- The evidence suggests that recent changes are unlikely to be entirely due to known sources of natural variability. The pattern of change seems to point to some human influence on climate similar to that projected by climate models and larger than expected from natural fluctuations. This point is not yet settled, however, mainly because of uncertainty over the ability of current models to simulate natural variability realistically. Nevertheless, it is reassuring for many modelers because it suggests that the models are pointing in roughly the right direction.
- Uncertainty about the ability of models to simulate natural climate variability remains a significant problem. As with trends in global mean temperature, scientists must use climate model simulations to assess the probability of getting a certain level of agreement purely by chance between the model and the observed patterns of change. There are many sources of natural variability that these models simulate poorly or not at all, and one of these might be associated with a pattern similar to the greenhouse warming pattern. Thus there is still a wide range of uncertainty about the size and origin of the present signal and about the size of future changes.
The Convention - Info for Participants - Info for Media - Official Documents
Daily Programme - Special Events - Exhibits - List of Participants - Special Features
Kyoto Information - COP3 Links - COP3 Home Page - UNFCCC Home Page - Feedback - Sitemap