Back in 2007, the IPCC’s fourth assessment report found that "most of the observed increase in global average temperatures since the mid-20th century is very likely (>90% probability) due to the observed increase in anthropogenic greenhouse-gas concentrations”. Now, researchers from the UK, Canada and the US have explained the methods behind the statement in a manner that they hope is suitable for nonspecialists.

"It is not widely appreciated that this conclusion is not based on model agreement over the 20th century but on detailed time and spatial patterns of change that allow [us] to disentangle – with uncertainties, of course – the changes due to a variety of external influences," Gabi Hegerl of the University of Edinburgh, UK, told environmentalresearchweb. "We found there was no succinct paper accessible to a wider nonspecialist readership that explains how such a conclusion was reached, and to what extent alternative explanations were considered."

Together with Francis Zwiers of Canada’s University of Victoria and Claudia Tebaldi of Climate Central, US, Hegerl discusses the information that goes into fingerprint-detection studies. "These use the pattern of change in space and time to estimate what part of the observed change may be due to greenhouse-gas increases, internal climate variability, aerosols, other anthropogenic influences, response to volcanic eruptions and changes in solar radiation," she said.

It is this pattern in time and space that forms the "fingerprint" for climate change caused by a particular influence. For example, higher greenhouse-gas concentrations should warm the troposphere but cool the stratosphere and mesosphere, whereas a brighter sun should warm the upper stratosphere and mesosphere as well. Statistical techniques known as "optimal fingerprinting" can assess which fingerprints are seen in the observed climate change, as well as calculate the amount of the change that they are responsible for.

The researchers stress that the statement that a large part of recent warming is due to man is based on a "much more comprehensive and layered analysis than a comparison at face value of model simulations with observations". They reported their work in Environmental Research Letters (ERL).