Climate change brings three challenges, Mark Walport, chief scientific advisor to the UK government, told participants. These are scientific complexity, communication, and policy, and he's not sure which is the hardest. "Some people say that if we deny the science we don't have to deal with the implications," he said. "This is obviously ridiculous." Walport reckons we have three choices - mitigate, adapt or suffer. "We need to optimize the ratio - it's pretty likely we're going to have to do all three." He believes that science isn't finished until it's communicated: "many of the climate sceptics communicate very well indeed. The public can only judge on the quality of the communication."

Thomas Stocker of the University of Bern, chair of the IPCC's Working Group I, said he knows of no other scientific document that has undergone such deep scrutiny but that "doesn't mean there won't be any errors."

According to Stocker, the report team "did not dwell too largely on potentially alarming issues that we covered". For example, there is a large quantity of methane locked up in permafrost , which is going to heat up and melt. But the uncertainties are "really large". "We are only able to say that somewhere between 150 and 200 Gigatonnes of carbon are locked up in permafrost in the high latitudes but we could not be more precise on timescales, critical thresholds, or anything like that," said Stocker. "The science is simply not there."

And how will the IPCC report next time? "The burden is so high for the science community we could never do that again - I think I heard that at AR4," said Stocker, who believes that the IPCC reports are useful in creating a snapshot of knowledge and building up a receptiveness of the public to listen. That said, the process has become "progressively more difficult" as the number of publications to review has grown.

There will now be a short pause

The global warming "pause", or perhaps more accurately the not-so-global-only-surface-temperature-warming hiatus, was discussed by Jochem Marotzke of the Max Planck Institute for Meteorology, who was part of the IPCC team assembling scientific evidence on trends in temperature over the last ten to fifteen years. The team undertook the task in response to feedback from governments on the IPCC's draft reports; Marotzke and his colleagues assembled relevant material from 8 of the IPCC report's 14 chapters.

The linear trend in global mean surface temperature for 1998 to 2012 is 0.04° C per decade compared to 0.11°C per decade for 1951 to 2012, according to HadCRUT4 data.

"Such hiatus periods are common in the record and yet this last one has sparked enormous debate," Marotzke told delegates at the Royal Society. "Does the surface warming hiatus mean global warming has stopped? No. Warming of the climate system continues. Sea ice continues to melt, the ocean continues to take up heat, sea level continues to rise."

Marotzke reckons one of the factors that's added to the attention is that CMIP5 climate models reproduce the long-term global mean surface temperature over the twentieth century but don't reproduce the warming hiatus in the last ten to fifteen years. However, he explained, models have tended to show temperatures higher than observations in the past too.

So what's the cause? Marotzke said the hiatus is a northern-winter phenomenon, found in December, January and February. The spatial footprint is "not entirely clear" but one team reckons there's a cold centre over Eurasia.

A number of explanations have been posed - stratospheric water vapour, stratospheric or volcanic aerosols, the currently downward phase of the solar cycle, an increase in deep-ocean heat uptake, and low sea surface temperatures in the tropical eastern Pacific.

It seems that a decrease in radiative forcing caused by two natural factors over the period in question - an increase in volcanic activity, and hence more reflective particles in the atmosphere, and a downwards trend in solar radiation - account for around half the observed surface temperature trend. And the rest is ascribable to internal variability.

In response to his self-posed question "are models able to reproduce the hiatus?", Marotzke replied "should they be able to?" Climate models are not expected to match the timing of internal variability events, he said. Historical simulations of surface temperature by CMIP5 models start in 1850; predicting internal variability 150 years into the future is like "demanding a weather forecast on a specific day three months in advance".

As a result of the hiatus, explained Marotzke, the IPCC report's chapter 11 revised the assessment of near-term warming downwards from the "raw" CMIP5 model range. It also included an additional 10% reduction because some models have a climate sensitivity that's slightly too high.

"The recent surface warming hiatus poses a fascinating scientific challenge," said Marotzke. "It forces us to think across the entire climate system and possibly to rethink some of our basic concepts, for example how we formulate the energy budget and feedback." That said, "the hiatus is largely irrelevant to the physics (although not the communications) of anthropogenic climate change", Marotzke continued. "The timescale is too short and so dominated by natural - forced and internal - fluctuations."