In the "early days" of global warming, attention was concentrated on the possible consequences of increasing amounts of carbon dioxide, or CO 2 in the global atmosphere.
It soon became apparent, however, that CO 2 was not the only culprit: others gases, although present in smaller quantities, are better than CO 2 at trapping the long-wave radiation from the Earth.
The most obvious suspects were nitrous oxide, methane, ozone and the man-made CFCs, and these ingredients were gradually added to the recipe used by Global Circulation Models to predict the future of the global climate.
By the early 1990s, however, things were still not right. The models were unable to reproduce all the fluctuations in the Earth's temperature that have taken place over the last century. So the modellers rounded up another suspect.
This time they incorporated sulphate aerosols in their calculations - industrial pollutants which reflect some of the sun's radiation back to space. They have the effect of reducing in certain regions the amount of solar radiation available to heat the Earth and there by act as a brake on any tendency for rising temperatures.
Large areas of North America, Europe and south-east Asia are shrouded in a semi-permanent haze of these tiny microscopic particles. When their effect was added, the models were now able to mirror almost exactly the global temperature pattern from 1860 to the present day.
It is these same sulphate aerosols which have hit the headlines in the last few days, linked to the news that the "worse case" prediction for a rise in global temperature over the next century has increased to nearly 6 degrees.
The aerosols facilitate a global cooling; as anti-pollution measures in industrialised countries begin to take effect, with obviously beneficial consequences for the environment in general, there is the disadvantage that they are no longer as efficient a brake on global warming as they were.
Scientists have recently largely removed another nagging doubt they had had about climatic change - the possibility that the observed rise in global temperatures might well have happened without any human input. The question has been dealt with by looking for a measure of how great "natural variability" really is, by examining the patterns which emerge when they allow the models to simulate climate change over a very long period like 1,000 years.
The results suggest that there is only a tiny chance that natural variability could produce the patterns of temperature change now being observed.