On Monday, we published a paper in the scientific journal Nature Geoscience that re-evaluated how much carbon dioxide we can still afford, collectively, to emit into the atmosphere and still retain some hope of achieving the ambitious goals of the Paris climate agreement to “pursue efforts” to keep global temperatures to 1.5C above pre-industrial levels. The carbon budget we found, to yield a two-in-three chance of meeting this goal, was equivalent to starting CO2 emission reductions immediately and continuing in a straight line to zero in less than 40 years: a formidable challenge.
Formidable, but not inconceivable. The distinction matters, because if it were already completely impossible to achieve the Paris ambition, many might argue there was no point in pursuing those efforts in the first place – or that the only option left is immediately starting to cool the planet with artificial volcanoes.
We knew this finding would be controversial, since previous estimates had indicated that to meet the same goal, emissions might have to reach zero in well under a decade, which really is inconceivable. Crucially, the reason for the correction was not that we had a new estimate of the climate response, or warming per tonne of CO2 emitted – we used exactly the current consensus range – but that we took better account of past emissions and where human-induced warming has got to already.
It was a relatively technical paper, so we prepared as best we could, wrote a non-technical blogpost and organised a press briefing with the Science Media Centre. Almost all of the initial coverage on Monday and Tuesday was accurate: both the Times and Telegraph had headlines about ‘wrong’ or ‘faulty’ models, but in the articles beneath them, Ben Webster and Henry Bodkin were careful not to say there was any evidence the models were systematically over-responding to CO2. We took pains at the briefing to stress the discrepancy was likely due to other, more transient, factors. Those who were there evidently understood.
Then the opinion writers piled in. Writing in Breitbart, James Delingpole announced that our paper “concedes that it is now almost impossible that the doomsday predictions made in the last IPCC assessment report of 1.5C warming above pre-industrial levels by 2022 will come true.” Which would be exciting, except that the 2013 IPCC report made no such prediction. In fact, the IPCC specifically assessed that temperatures in the 2020s would be 0.9-1.3C warmer than pre-industrial, the lower end of which is already looking conservative. Anyone who had troubled to read our paper would have found this “IPCC AR5 Ch11 projection” helpfully labelled on two of our figures, and clearly consistent with our new results.
In the Daily Mail, Graham Stringer, a Labour MP, weighed in: “According to these models, temperatures across the world should now be at least 1.3C above the mid-19th century average.” Except, they don’t show that, and nor did our paper. Because CO2 concentrations are much better known than emissions, models calculate both the emissions needed to reach observed concentrations and the temperature response. In many models, total accumulated CO2 emissions do not reach today’s level until well after 2020, by which time the level of human-induced warming in the models can indeed be about 0.3C warmer than it is in the real world today, depending on how you measure it. Which means you need to be careful about using these models to work out the remaining carbon budget for an ambitious goal like 1.5C – precisely the point of our paper. In a rapidly warming world, a few years makes a big difference: the discrepancy between modelled temperatures today and observed temperatures today is much smaller.
Moving on to the Sun, James Delingpole (again) announced that our paper would “scotch the […] myth […] that man-made carbon dioxide (CO2) is causing the planet to warm at such dangerous and unprecedented speeds that only massive government intervention can save us.” Note the careful wording. Delingpole presumably knew that our paper said nothing at all about revising the impact of CO2 on climate. But by adding a swipe at government intervention, perhaps he hoped to pack just enough into that sentence to squeak it past the UK Independent Press Standards Organisation (which presumably has no sway over Breitbart).
So after reasonably accurate initial reporting, suddenly our paper was about a downgrading of the threat of climate change, when it was actually nothing of the kind: our predictions for warming rates over the coming decades are identical to those of the IPCC, and we do not assess the impacts of climate change for any warming level. And then, of course, these ideas were picked up by sympathetic editors all over the world.
Who really loses from all this? While Delingpole and Stringer were making out that our paper was about something it wasn’t, it seems to have prompted much more interesting conversations among scientists around the world about what the true level of human-induced warming really is, and what the Paris goal actually means.
These are important questions. For such a tight target, the actual remaining carbon budget is sensitive to a number of assumptions, including even how we define global average temperature. Significant uncertainties remain, and while we believe our paper improves on previous estimates, it is by no means the last word. But debating the current level of human-induced warming and how it relates to the 1.5C goal feels a bit like discussing how best to steer a spacecraft into orbit around Saturn while Delingpole and Stringer are urging their readers to question whether the Earth goes round the Sun.
Critics of mainstream climate policy frequently complain that they feel excluded. The real problem is that they exclude themselves, and their readers, from the discussion as soon as it starts to get interesting.