The best shot?

UPDATE: David MacKay’s letter is now up in a separate post here

Some weeks ago I wrote an article for The Times about why I no longer find persuasive the IPCC’s arguments that today’s climate change is unprecedented, fast and dangerous.

I was delighted to receive a long and courteous letter from David MacKay, the chief scientific advisor to Britain’s Department of Energy and Climate Change. With his permission I am publishing my reply to that letter. I would put his letter here too (again he agrees), but I only have a hard copy of it, so that will have to follow when he has time to send me a soft version.Now done.

The remarkable thing about this exchange is that far from weakening my doubts about the IPCC case, it has strengthened them. The letter explains why. Essentially, I have realised that almost the only weapons left in the alarm locker are the retreat of the Arctic sea ice and an event that happened 55m years ago and was probably not caused by CO2 at all. Everything else — the CO2-temperature correlation in the Antarctic ice core, the hockey stick, storm frequency, phenology, etc etc — no longer supports the argument that something unprecedented in magnitude or rate is happening. Remarkable.

Here is my letter:

Dear David

I am honoured that you liked my book and I liked yours very much indeed: a brilliant and necessary contribution to the debate. Though it arrived late in my writing process, I managed to squeeze in several references to it in the penultimate chapter of mine.

Thank you for taking the trouble to give such a detailed reply to my Times article – much longer than the constraints of the Times op-ed page allowed for me! I shall now indulge in a longer reply. It is certainly nice that the political `climate’ (sic) now allows articles like mine to receive serious replies, rather than accusations of heresy or sin or threats of prosecution as a criminal against humanity. I appreciate that very much. I surmise from your covering note that perhaps your letter is circulated more widely among DECC colleagues and I would be glad for you to circulate this reply, not least to the secretary of state who showed you my article. I shall post this letter on my blog.

I am surprised to find that I agree with much of your letter, but it changes almost none of my conclusions. How can this be? The gap between the science and how it has been presented is huge. This is as much the fault of bodies like the Royal Society, which should have been a brake on politically inspired extreme statements but was not, as it is of the media. You say scientists know how big the uncertainties are and that the failure to ensure that uncertainties are reported has contributed to the problem. I agree and I wish that the science establishment had paid this issue more attention. They allowed and encouraged their spokesmen to peddle the very opposite impression.

Consider this statement for example: `Earth’s climate can only be stabilized by bringing carbon dioxide emissions under control in the twenty-first century.’ That is the opening sentence of a paper in Nature Geoscience last month.  It is shocking that it got past the editors and reviewers. After 4 billion years of climatic volatility, much of it not caused by CO2 but by orbital variations, solar cycles and so on, how on earth are we to `stabilise’ earth’s climate by adjusting just one forcing factor? I refuse to accept that the climate could ever be stabilised, let alone by adjusting one factor. That sentence has no place in a scientific journal.

Taking your points in turn, then:

You say most climate scientists are nicer than their caricature on the web. I agree, but so are most sceptics. The image of the politicised, right-wing, anti-science zealot fits some, of course, just as the reverse fits Jim Hansen, Bob Ward and Joe Romm, but the ones whose work I have got to know, such as Andrew Montford and Steve McIntyre are quite different.  The polarisation of this issue is a real problem. I learned from writing about the nature-nurture debate that arguments get polarised because people only read their friends’ caricatures of their opponents’ works; it is vital that we all read all sides of the argument.

Next you criticise my argument that current warming is not `unprecedented’ by reference to the Arctic sea ice graph. But this only goes back to 1979! Blackpool’s Football League table position is unprecedented since 1979. In a brief period of warming, of course the warming is unprecedented. You will know the ample anecdotal evidence that Arctic sea ice retreated just as much in the 1920s and 1930s: remember `Warming island’ for example. There is also good evidence from wave-made beaches and driftwood in Northern Greenland of probably ice-free summer months in the Arctic 7,000 years ago. A study published in the journal Quaternary Research of sea sediment cores in the Chukchi Sea shelf in the Arctic Ocean concluded that `during the middle Holocene the August sea surface temperature fluctuated by 5°C and was 3-7°C warmer than it is today’. (Incidentally, I am keen to see a proper test of the hypothesis that black carbon is the main cause of the Arctic sea ice summer retreat of recent years and that cleaning up Chinese coal power stations will reverse the trend. The argument seems quite plausible – and it might explain why Antarctic sea ice has been expanding during the same period —  but it needs a test.)

To be honest, whenever that sea-ice graph is used as an argument, I become a little bit more sceptical. If that is the best evidence of something unprecedented, then the case must be weaker than I thought. It is a change that is not even likely to threaten human or animal livelihoods: even with a total late-summer melt (I presume you do not belong to the school of thought that the ice could fail to reform in winter), there is no great albedo feedback at such latitudes because of the angle of the sun in August, and polar bears will expand their range further north or will survive ice-free summer months onshore as they do already in Hudson’s Bay, on Wrangel island and parts of Svalbard (where one once walked round my tent while I slept).

Then you say that if I mean `not unprecedented on 100m year timescales’… But those are not the only two options! I mean not unprecedented in centuries and millennia, ie in human history. It is hugely relevant whether the warming of 1910-40 was as fast as 1980-2010 (it was). It is hugely relevant if the climate was as warm in 1100 AD as now (it probably was) both in attributing cause and in making conclusions about sensitivity.

You will have seen this graph, one of many now making it amply clear that the warmth of the Holocene optimum, peaking about 7,000 years ago, was both global in extent and considerably warmer than today:

And this:

Next you disagree with my characterization that recent warming is not `fast’. Phil Jones himself confirms that the rate of warming in 1975-2009 is statistically indistinguishable in rate from the two other periods of warming in the past 150 years: this is from his interview with the BBC –

I contend that none of these rates are `fast’. Contrast them with the rate of change now known from 12,000 years ago, characterized by `local, regional, and more-widespread climate conditions [which] demonstrate that much of the Earth experienced abrupt climate changes synchronous with Greenland within thirty years or less’ (Alley 2000. Quaternary Science Reviews 213-226), including `a warming of 7 °C in South Greenland [that] was completed in about 50 years’ (Dansgaard, White and Johnsen 1989, Nature 339: 532). That is a change roughly nine times as fast as has happened since 1980 – in Greenland or anywhere else. Another study gives even bigger numbers, saying that the `abrupt warming (10 ± 4 °C)’ at the end of the Younger Dryas and the warming at the end of a short lived cooler interval known as the Preboreal Oscillation `may have occurred within a few years’ (Kobashi et al 2008 Earth and Planetary Sciences 268:397). Nor was this rate of change confined to Greenland. As one article summarises, `temperatures  from the end of the Younger Dryas Period to the beginning of the Holocene some 12,500 years ago rose about 20 degrees Fahrenheit in a 50-year period in Antarctica, much of it in several major leaps lasting less than a decade.’ (Science Daily, Oct 2 1998).

You concede that the rise is running at just 1C per century over the past 50 years, though you do not recognise the degree to which even this is only true of the instrumental record, as adjusted and homogenised by the USHCN and similar bodies. These adjustments have come under question recently since it has become clear that far from correcting for urban warming they seem to be exaggerating it. So the true figure, without adjustments, is probably much closer to that recorded by the SST record and the satellite record, considerably lower than 1C.  Here is the US raw data:

And here it is `adjusted’:

The climate is going to have to get a move on if it is hit 3C this century. One-tenth of the century now over and no significant warming yet. This should have been the fastest bit:  since the curve is logarithmic, the first 100 ppm of CO2 should produce as much warming as the next 200 ppm.

You then say we should not be blasé about 2C in 200 years. I am sorry but I do not find this convincing for four reasons:

If anybody had adopted a policy in 1810 to affect the climate in 2010, they would have made absurd decisions because of uninvented technologies, etc.

There is lots of evidence that climate change is positive in its impacts up to 2C, especially if it takes 200 years to get there.

Remember most of this warming is predicted to be in cold regions, in winter and at night. The daytime temperature changes in temperate regions in summer would be less than 2C.

The thing I think we should not be blasé about is the cost of measures we are taking today. Biofuel policies have caused real hunger. Wind power policies have caused real fuel poverty. Yet these measures would do a statistically insignificant asterisk towards solving the problem even if the warming was happening fast. I refuse to be blasé about the jobs not created, the landscapes spoiled, the deaths caused by indoor air pollution in Africa because people cook over charcoal and above all the distraction and diversion of funds from real problems, including environmental ones.

You then ask me what I think the sensitivity to CO2 doubling is and you guess that I must think it is outside the range 1.5-4.5C. Actually, I think there are lots of sensitivities within that range that are `fairly minor problems’ and so do many of the studies cited by the IPCC. For Malaria, for example, 2C will produce less than 30,000 extra annual deaths on the million we see today. I think the million is a major problem, the 30,000 in a century’s time is a minor problem. Water shortages? The evidence of Arnell 2004 suggests that 2C of warming will reduce the net number of people at risk of water shortage. Etc etc.

So what do I think the sensitivity is? I have no idea. It could be 1C or lower, it could be 3C, but I think it very unlikely from the latest data that it is going to be as high as 4.5C. (Actually, IPCC says that is unlikely, too, if you read the probability right.) I do know this though: the IPCC’s estimates of the sensitivity are utterly worthless because they all – all – assume net positive feedback. You are quite right that we do not know that clouds have negative feedback for sure, but there is good evidence that they probably do, and just 2% change in the albedo of cloudiness could reverse all CO2’s marginal effect.  And you imply that Spencer is a lonely voice in arguing this case. May I refer you to the Nature Geoscience paper quoted above. Despite its catechistic opening sentence, it goes on to say:

It is at present impossible to accurately determine climate sensitivity (defined as the equilibrium warming in response to a doubling of atmospheric carbon dioxide concentrations) from past records, partly because carbon dioxide and short-lived species have increased together over the industrial era. Warming over the past 100 years is consistent with high climate sensitivity to atmospheric carbon dioxide combined with a large cooling effect from short-lived aerosol pollutants, but it could equally be attributed to a low climate sensitivity coupled with a small effect from aerosols. These two possibilities lead to very different projections for future climate change.

Anyway, you agree that climate sensitivity could conceivably be as low as 1C, which is more than the IPCC does, so I should accept this concession with gratitude and I do. It’s a huge change from what was being said by the science establishment two years ago and is still being said by many, namely that 2C is unavoidable.

Then you describe the PETM (it is the Paleocene-Eocene Thermal Maximum – the Pliocene came much later), suggesting that I might not know of it. I not only know it but know the more recent data suggesting that carbon emissions can no longer be reliably interpreted as the main cause of warming then. Gerald Dickens of Rice University last year concluded that CO2 did not even double during the PETM and that something other than carbon dioxide caused much of the heating.

I do think it is revealing how much scientists who are alarmed about climate refer to the PETM. Imagine if the sceptics relied heavily on one episode of uncertain causation and effect, little known and not repeated for 55m years! You would say: is that really the best they can do?

You mention the Toarcian event of 183m years ago, which is new to me, but sounds interesting (by the way I do long to get back to a world where one can discuss paleoclimatic episodes as thrilling stories in their own right without having to draw political lessons from them). Yet the very first abstract I read on the subject after googling it talked about species shifting range in response to `a rapid cooling and their gradual return to former habitat areas in the period of warming’. I will need more evidence that carbon was cause rather than effect here: sounds more like a classic volcanic winter story.

Next you say that sea level is a case where the IPCC has been too conservative. But the graph you show has a trend of 3.1mm per year. This equates to 31cm in a century, comfortably within the IPCC’s estimate of 18-59cm in the present century.

Let me make two final points. I have argued that the two main examples you cite – the Arctic sea ice retreat and the PETM – are weak examples on which to build your case. Five or ten years ago I suspect that you would have cited the Vostok ice core record, showing CO2 and temperature in lockstep, and the Hockey Stick graph, showing recent temperature rises to be unprecedented in a thousand years. These two graphs were very, very important in persuading me to rejoin the consensus view in the mid 2000s, after I had moved towards cautious scepticism in the late 1990s. The fact that both are now discredited as evidence of CO2 attribution has been very, very important in sending me back towards scepticism. When the facts changed, I changed my mind. The Vostok graph now unambiguously shows that CO2 rises follow rather than precede warming. The impact of that discovery is huge. The Hockey Stick graph is largely a statistical artefact caused by the inappropriate use of short-centred principal component analysis and heavily reliant on geographically narrow and methodologically suspect samples of tree rings. If you have not read Andrew Montford’s The Hockey Stick Illusion to understand this, I do beg you to do so.

My last point is this. We always discuss climate change in isolation, as a unique issue. Yet we cannot ignore the history of past environmental alarms, which I catalogue in my book: on population, famine, pesticides and cancer, desertification, sperm counts, acid rain, GM crops, and many other issues, we have been promised catastrophe, often with the backing of peer-reviewed science, and repeatedly these hopes have been dashed. (You may need to remember to switch your sarcasm detector on when reading the last sentence.) My position is heavily influenced by having been science editor of The Economist during the acid rain scare and having been a full-scale alarmist at the time myself.  In 1984 I wrote: `Forests are beginning to die at a catastrophic rate. One year ago, West Germany estimated that 8% of its trees were in trouble. Now 34% are…that forests are in trouble is now indisputable.’ Experts told me all Germany’s conifers would be gone by 1990 and the Federal Ministry of the Interior predicted all forests would be gone by 2002. I was wrong. German forest biomass increased during all these years. Of course, the boy who cries wolf may be right one day. But we are right to grow more sceptical when he keeps being wrong.

Now, if for the past 20 years we had been told that there is a probability of some change in the climate due to CO2, and a very small possibility that it is likely to lead to a drastic lurch, then I could join with you and the consensus. Instead of which I have been repeatedly told that trillions must be spent urgently because there are only a few months to save the world and it is the most urgent problem, more urgent than hunger, malaria and indoor air pollution, likely to lead to the collapse of the entire economy and moreover that the science is settled and to question it is to be equivalent to a criminal. So, apologies if I sound a little exercised on this, but as a huge champion of science I feel very, very let down by the science establishment, especially the laughably poor enquiries on the emails published this year. Ask yourself if these emails had been within a drug company about a drug trial, whether the establishment would have been so determined to excuse them.

Again, I thank you for the courtesy of a proper reply. This is more than I get from most scientists and journalists on this topic. I do not envy the difficult decisions you and your political colleagues face, but I do beg you to review the latest evidence and increase your doubts about the likelihood of catastrophe; also to increase your concern for the costs and damages caused by renewable energy policies.

yours sincerely

Matt