My previous post was a snapshot of Stern’s uncontroversial presentation of the story so far and the underlying science.
The next step is prediction. For this we need two pieces of information: the sensitivity of the climate (represented by average temperature) to greenhouse gas concentrations; and the path of future emissions of these gases under current policies. So we need two different kinds of model, of the climate system and the socioeconomic one. They have different problems and uncertainties. I’ll look at the latter in my next post.
I’m starting to wonder if I’ll ever finish, as I’m still on Chapter 1! Still, to travel in reality is better than to arrive.
Stern sums up the state of the art on climate sensitivity in his table 1.1 (p.12), using a 95% confidence range.
He summarizes (pp 10,11):
If annual emissions continued at today’s levels, greenhouse gas levels would be close to double pre-industrial levels by the middle of the century. If this concentration were sustained, temperatures are projected to eventually rise by 2 — 5ºC or even higher…. If greenhouse gas levels were to reach 1000 ppm, more than treble pre-industrial levels, the Earth would be committed to around a 3 — 10°C of warming or more, even without considering the risk of positive feedbacks.
I think Stern is not presenting the data in the best way here. First, there should be a summary averaging the studies. Second, a 95% confidence level is appropriate in science for establishing a causal connection, but is far too conservative for policy within an established causal framework. The criterion for policy is surely “a strong balance of probabilities”, not “beyond reasonable doubt”. I would present the data like this; the bands represent bands of one and two standard deviations, the confidence numbers are guesses because the distributions are not normal:
The studies Stern cites have strong modal peaks (chart 1.2, page 9). “Sensitivity ” here means the effect of a doubling of greenhouse gases from pre-industrial levels, say from around 275 to 550 ppm CO2 equivalent.
From this, we have a pretty strong central estimate of around 3°C for a doubled concentration. We’ll hit this concentration in under fifty years, and are already at 450 (chart 1.1).
Stern underplays the time aspect. The lags are large, and the climate takes centuries to reach equilibrium. However, according to this IPCC chart from the 2001 report, most of it comes by 2100, except for the highest emission scenario. I suggest policy reports should use this horizon, except for the doomsday risks. On the other hand, the lags mean that (p 12):
Climate models project that the world is committed to a further warming of 0.5° – 1°C over several decades due to past emissions
and ultimate stabilisation will involve unwinding a huge legacy.
The strong insight of the Stern report is that under uncertainty, rational people sum over the risk distribution, not just take the central value or – worse – say that as it’s uncertain, let’s do nothing. So he draws attention to the long upwards tails of all the models. This makes a big difference to the impact assessment. Stern p:9:
These sensitivities imply that there is up to a one-in-five chance that the world would experience a warming in excess of 3°C above pre-industrial even if greenhouse gas concentrations were stabilised at today’s level of 430 ppm CO2e.
What should we do about the sceptics? Ignore them. I propose to follow John Quiggin here. The findings cited by Stern represent an overwhelming consensus of qualified opinion, based (see previous post) on settled basic physics and chemistry. The professionals among the sceptics can publish their arguments in journals. If you are an expert, write to Nature. If you are an amateur not prepared to take the consensus’ word for it, go to realclimate.org to ask your questions and learn. Either way, a current affairs blog is not an appropriate forum for your doubts. Readers are therefore politely but firmly asked not to post here any comments of the form “But [insert name of sceptic] says that…”; I reserve the right to delete them.
If the science were settled, surely there wouldn’t be any qualified sceptics? This objection reflects a false picture of the sociology of science. Consider being a maverick as a professional strategy. As the odds on being right dwindle, so does the payoff to the bet if it wins. It’s like the answer to the puzzle why, when only a few bull seals get to mate, sex ratios are even at birth: the few male winners mate with many females. So there are always likely to be mavericks in academia (and they’re likely to be men). In fact, as the status of an area of enquiry converges on certainty, and the mainstream scholars move on, the proportion of publication by mavericks and cranks will increase. Who’s writing now about the age of the earth, or the authorship of Hamlet?
Another Baconian “idol of the theatre” is the demand for absolute certainty. This is a mistake in positive science, though increasing the assurance of belief is worthwhile up to a point. It is utter folly in praxis, when we are normally compelled to choose under pressure of time. We face not only statistical uncertainty but a wider ignorance of our environment. Remember that clinical trials of new medical treatments are routinely stopped when there is “overwhelming evidence of the benefit of the treatment”. It is unethical to pursue the experiment beyond that point aiming at absolute certainty.
Going back to Stern, the more recent models tend to show higher sensitivities, as they account better for complex feedbacks. But there are big unknowns, the potential positive feedbacks: weakening of carbon sinks in forests and oceans, and methane releases from bogs, peat and undersea hydrate deposits. (Box 1.3, p.11). As Stern calmly puts it (p.10):
It remains unclear whether warming could initiate a self-perpetuating effect that would lead to a much larger temperature rise or even runaway warming, or if some unknown feedback could reduce the sensitivity substantially.
Runaway warming? And it can’t be ruled out? The earth obviously hasn’t suffered this in 4 billion years, or we wouldn’t be here. But there’s a plausible – and controversial – claim that it suffered runaway cooling 700 million years ago.
This is the stuff of nightmares. Runaway warming reduced Venus into a Gehenna that could not support life as we know it. It would be worse than Jonathan Schell’s eerie (if improbable) vision in his 1982 tract The Fate of the Earth of the aftermath of all-out nuclear war, a vast prairie of grass and insects. Runaway warming would extinguish all life.
The Oxford philosopher Nick Bostrom – my kind of maverick! – has written about such low-probability, mega-catastrophic risks as “existential disasters“. Bostrom’s list includes asteroid impact, genocidal aliens, pneumonic HIV, self-replicating nanobots turning the planet into grey goo, and unwise experiments by alien physicists that may trigger
a breakdown of a metastable vacuum state that our part of the cosmos might be in, converting it into a “true” vacuum of lower energy density. This would result in an expanding bubble of total destruction that would sweep through the galaxy and beyond at the speed of light, tearing all matter apart as it proceeds.
Bostrom includes runaway warming, but is over-sanguine about it:
Hopefully, however, we will have technological means of counteracting such a trend by the time it would start getting truly dangerous.
We have a safe baseline technology already – that of the 17th century: horses, sailing ships and windmills, and our options increase all the time. The fear is that we wouldn’t have the time or the political nous to head off a Diamondian collapse.
There’s no point in losing sleep about the Klingons, but Bostrom is right to say we should pay attention to the risks we can do something about and
maximize the probability of an okay outcome, where an “okay outcome” is any outcome that avoids existential disaster.
The main concern of policy should of course be Stern’s conventional risk spectrum. But we also need hedges against runaway warming, just as we do against asteroid impact; initially, enough monitoring to give us timely warning, and a political framework for acting quickly if we have to.