The Nature Editorial: 'Shooting the Messenger'
A Specially Commissioned Paper for The Lavoisier Group

Bob Foster

1. INTRODUCTION

The English-speaking world enjoys the benefit of two great weekly science journals: Science and Nature. The former is issued in the US by the American Association for the Advancement of Science, and the latter is a commercial venture from England.

Science is a greenhouse booster of long-standing, and its view on this important topic was counter-balanced by a mildly-sceptical editorial tone at Nature. But Dr Jekyll is no more. Now that Sir John Maddox has departed the editorship, and that Macmillan has been bought by German interests, Mr Hyde has broken loose.

The Nature editorial of 12 July 2001 (volume 412, page 103), ie, immediately prior to the opening of the Bonn continuation of COP VI on 16 July, is terrible proof. It is entitled 'Shooting the messenger' and then sub-headed 'The Intergovernmental Panel on Climate Change has a creditable record of developing a scientific consensus and delivering it to policy-makers. What its critics really object to are the facts'.

I will get back to the facts of IPCC's 'scientific consensus' later. First, Nature is a good journal; although I would say that, wouldn't I, having published in it (on palaeoclimatology). But have you read anything half as Goebbelsian in a reputable scientific publication as the editorial quoted below?

2. SHOOTING THE MESSENGER

It will be an unusually long quote, I am sorry to say; but then, as you will soon see for yourself, this is no ordinary editorial. It begins by introducing at once the Great Satan:

It then makes favourable reference to its Guiding Star:

And it makes disparaging reference to the morality of IPCC's critics:

It then questions (my emboldment added) the qualifications and motives of 'experts' who are not of the Cloth, ie scientists whose perspective of climate-change is different from those who are experts in atmospheric science sensu stricto.

And it then mines a much-more-vituperative vein:

A Paean to the Righteous concludes this amazing editorial:

It is reassuring to know that a great journal is keeping an open mind on the complex and multi-disciplinary scientific issues crucial to the future of humanity. Science is in good hands.

3. RUNNING DOGS OF IMPERIALISM

The editorial starts off on the right track. We are witnessing here a battle between wealth creation---as practiced in the twentieth century, based on the ready availability of affordable energy from fossil fuels---best exemplified by the United States, and anti-capitalist fundamentalism---in the form of the decarbonisation of western economies---as advocated by IPCC.

Overstating things a bit, am I? Yes, perhaps so, but not nearly as much as you may think. I have talked elsewhere about the potential economic impact of the Kyoto Protocol (see my paper 'The Carbon Challenge: should Australia decarbonise?' posted at Warwick Hughes' new website). In addition, the attention now being directed to a potential future greenhouse threat---real or not---is already diverting money and zeal from real-life, here-and-now, environmental needs.

4. SUMMARIES MODERATED

In the editorial, and in this analysis, we are talking about the science of climate change. See if you think, as I do, that Nature has allowed its apparent dislike of President Bush, and of the coal and oil industries, to colour its appraisal of IPCC and its science.

First, Nature makes the implausible claim that access of politicians and NGOs to the summary-writing process is a moderating influence on the conclusions of scientists as embodied in the technical chapters of IPCC's reports.

I give below two examples where Nature is wrong.

4.1 SAR Summary for Policymakers

IPCC's Second Assessment Report 'Climate Change 1995: the science of climate change' was the scientific underpinning of the Protocol developed at Kyoto in 1997.

However, there is little chance that 'policymakers' around the world, be they politicians or bureaucrats, could have read this 572-page scientific report before they began negotiation of the treaty. Instead, they would have relied on its 5-page Summary for Policymakers.

To reach the Summary, policymakers would have passed by way of a brief introductory statement---misleadingly asserting that the report 'presents a comprehensive, objective, and balanced view of the subject matter'. Then, over the page, they would have encountered the Preface. This 1 1/3 pages contained but a single reference to climate-change science:

This attention-getting statement is repeated almost verbatim, and again without elaboration, in the subsequent Summary for Policymakers. However, those policymakers who had the time to seek out more detail on this 'key finding' in the body of the text, would have been disappointed.

The relevant chapter of the report is much more circumspect in its manner of expression than is either the Preface or the Summary for Policymakers. It says (page 439):

And it continues:

After release of the SAR, the man who became the principal representative of the United States (Clinton/Gore) Administration at Kyoto, Timothy Wirth, the Under-Secretary of State for Global Affairs, is reported in a Nature news item on 25 July 1996 (volume 382, page 267) as follows:

And, as a reminder of the intellectual climate prevailing in science at that time (and now):

The 'key finding' of the Preface and Summary for Policymakers is not justified in the text---but the ploy worked. This much-publicised claim was the principal motivator for Kyoto. Contrary to the assertion of the 'Shooting the messenger' editorial, the scientific findings were not moderated at the time of political review---they were sharpened. This isn't science at all, but it is brilliant politics.

After release of the SAR in 1996, a peer-reviewed ex-post justification of the assertion in IPCC's Preface and Summary was published (Santer et al. 1996). A numerical model was used to calculate the warming of the lower atmosphere which should have resulted from the release of human-caused CO2 from the burning of fossil fuels. The model-calculated warming was hemispherically-symmetrical, as one might expect, because the long-lived CO2 emissions (ie, with an atmospheric residence time of decades to centuries) would be well-mixed around the globe.

The atmospheric cooling effect of sulphate aerosols---also emitted from the combustion of fossil fuels---was also calculated. Because 90% of this cooling aerosol is emitted in the Northern Hemisphere, and because it has a residence time of only a few days, the calculated cooling was largely in that same hemisphere.

Hence when the two calculations were combined, it was found that most of the human-caused net warming was in the Southern Hemisphere.

This model-based result was compared with observed temperatures in the lower atmosphere as recorded by weather balloons. Sure enough, a warming trend was detected in the Southern Hemisphere---much as predicted by the model. It was this successful matching of patterns which became IPCC's 'discernible human influence'.

Surely, this conclusion is spurious! During the 22 years of comprehensive records from satellites which we have now accumulated, there is no warming trend for the lower atmosphere of the Southern Hemisphere---all the observed warming is in the north. What gives?

We now know (Michaels and Knappenberger 1996). IPCC relied on only a short (25-year) run of observed atmospheric temperatures, encompassing the 1963-87 interval. Indeed, the years selected do show a warming trend. But that is not the whole story. Available at the time was a 37-year run of data, including 5 earlier years back to 1958, and 8 later years up to 1995. When the full run is used, the warming trend disappears!

IPCC chose a starting year which was cooled by the Mt Agung eruption, and finished in the warmth of the 1987/88 El Niño. If it had gone further, it would have been caught by the cooling associated with the Mt Pinatubo eruption of 1991.

IPCC's warming trend was an artefact of the years chosen for study; and this dissembling was not well known at the time. It is now.

I would describe the Nature editorial as rather economical with the truth, or else naive. In this instance, the government nominees writing the Summary for Policymakers did not moderate the consensus-based science of IPCC's report. Instead, they poured petrol on the flames.

4.2 TAR Summary for Policymakers

IPCC's Third Assessment Report 'Climate Change 2001: the scientific basis' was the scientific underpinning of the Bonn meeting on 16-23 July.

The principal attention-getter in the TAR is that global average surface temperature will rise by up to 5.8 ºC by 2100. This startling projection was judged likely to be so efficacious that it was leaked to the New York Times in the days preceding the US presidential election. It nearly worked, too!

The Summary for Policymakers of the TAR explains it this way:

A forecast warming of 1.4 ºC is not the stuff which might win the battle of hearts and minds. But 5.8 ºC is a different matter indeed. It is the latter figure which is so often quoted---the highest of IPCC's 35 x 7 = 245 projections. No probability distribution was attributed by IPCC to its 245 alternative outcomes; I will return to this topic.

A second point of explanation follows:

If I understand these words correctly, IPCC is saying that the projected upper bound of warming by 2100 has jumped from 3.5 ºC back in 1996 to a much-more-scary 5.8 ºC now, due primarily to an expectation of lower (cooling) sulphur dioxide emissions than previously assumed. This conclusion is crucial to any judgment of the veracity of IPCC's warming projections.

That I have got it right, is confirmed by a recent exchange in Nature. A 'concepts' piece entitled 'Room for doubt' (Pielke 2001), accompanied by a box containing the words 'Consensus science can only provide an illusion of certainty', doubts the quality of IPCC's work, at least in a conceptual sense:

IPCC's response was not long delayed. On the next week's correspondence page was a letter (Stocker 2001) which began:

Here we have it. Clear as a bell!

The increased projections of warming by 2100, as contained in TAR, are fatally flawed. We know that 90% of the short-lived cooling aerosol is emitted into the atmosphere of the Northern Hemisphere, and yet it is the atmosphere of the Southern Hemisphere which is doing the cooling! The simplest explanation of this surprising observation is that this aerosol is not presently providing the cooling which IPCC attributes to it.

Why might this be so? The probable reason is because the burning of fossil fuels produces both cooling sulphates and warming soot---and they just about balance. For its extra warming in the century ahead, IPCC relies on the implausible assumption that those burning fossil fuels will take the sulphates selectively out of their stack-gases, while leaving in the soot. How could IPCC's consensus of 'many hundreds of scientists' let this through?

But in response to this sycophantic Nature editorial, I want to make a different point. Out of IPCC's 245 cases, only one forecasts 5.8 ºC warming. Is this forecast very likely, and hence justification for the prominence it has been given by IPCC and the media; or has it such a low probability that for all practical purposes we can ignore it?

The Summary for Policymakers remains silent on this crucial point. We have been let down by our government-nominated summary-writers.

In my opinion, the highest extreme has a very low probability indeed. IPCC's 35 SRES scenarios project CO2 concentration in the AD 2100 atmosphere of 490-1260 ppm. These concentrations are then used in 7 models of varying sensitivity to produce the 245 outcomes.

IPCC's range of CO2 concentrations can be compared with a pre-industrial level of about 280 ppm, rising to 315 ppm when detailed measurements began in 1958. It is now about 370 ppm; and the growth rate has stabilised at about 1.5 ppm per year---a bit less in cool (volcanic eruption) years, and a bit more in warm (El Niño) years. The observed increase was 1.1 ppm in 2000. To get to the high estimate, we would need an average increase from now on approaching 9 ppm per year!

So much for the consensus of 'many hundreds of scientists'.

Even IPCC's friends are embarrassed. For instance, Wigley and Raper (2001) say in the abstract to their paper:

Those who wrote the Summary had an opportunity to introduce much-needed moderation here---and they didn't take it.

4.3 Can IPCC foretell the future?

Wigley and Raper accept IPCC's models at face value. They are just trying to make the output intelligible---and hence useful to planners.

But neither they nor Nature raise the crucial question. Even if IPCC correctly predicts the CO2 content of the atmosphere in 2100, can we rely on its forecast of global climate at that time?

The place to start is with a reconciliation of its models with known climate-change during the century just past. Yes, on a globally-averaged basis, hindcasting gives an agreement reasonable enough to quieten most doubters. But it is a spurious agreement.

Global reconciliation is easy. Over-predicting models can be made to match an under-warming world on a global basis, simply by increasing the assumed cooling effect of sulphate aerosol until they do. But when the hindcasts from IPCC's thus-reconciled models are examined---with no intervening changes to the input assumptions, of course---they fail on a regional basis. They show the Southern, not the Northern, Hemisphere warming the faster. They have the twentieth-century world the wrong way round.

Also, they show the lower atmosphere warming faster than the surface. Yet we know from 22 years of satellite coverage, that the surface is warming on average about three times as fast as the atmosphere (of which more below).

There is something seriously wrong with IPCC's models; and until they can match the past, they won't match the future.

In fact the models can't do the job, because they only work in a uniformitarian world. But the real world has a history of abrupt nonlinear transitions between climate states---which the models are unable to replicate.

Two examples of this natural climatic variability are the abrupt reversal of the previous warming trend in the mid 1940s, and the abrupt resumption of warming in the late 1970s. Is the jump at 1976/77 (see below) the last such event to be experienced before 2100? Who can say? But I doubt it.

By their silence, those who wrote the Summary for Policymakers allow us to believe that IPCC can foretell the future. It can't. Natural variability, often large and often abrupt, remains unpredictable.

5. THE ATMOSPHERE RULES, OK?

In this multi-disciplinary battle for hearts and minds, Nature is a partisan fighting alongside the atmospheric scientists. But is it right for a respectable journal to be so focussed? Let's look at the facts which, the Nature editorial urges us to believe, are already conclusive enough---but which it doesn't then enumerate.

5.1. A long backward look

The only prolonged periods during the past 300 million years when the atmospheric CO2 concentration remained as low as today were way back in the Permian Ice Age, and over the past few million years leading into the current Ice Age.

About 50 million years before the present, in the warm, wet, and ice-free paradise of the Eocene optimum, it was 2000-4000 ppm; and by the end of the last Glacial at 15,000 yBP it had declined to below 200 ppm. At very long time-scales, CO2 concentration in the atmosphere is an obvious driver of global climate.

But during the current Ice Age, the cycle of 90,000 years cold and 10,000 warm appears to be orbitally driven. The variation in atmospheric CO2 concentration between about 180 ppm in the Glacials and 280 ppm in the Interglacials (such as the present Holocene) is more likely to be effect than cause.

Overprinted on this 100 ky cyclicity is a ca 1500-year warm/cold cycle of which the Roman Empire warm period, Dark Ages, Mediaeval Warm Period and Little Ice Age are the latest manifestations in the North Atlantic Basin. These climatically-disparate periods are responses to changes in oceanic heat transportation which appear to be ice-related and inertially-driven.

Overprinted on the shorter 1500-year cyclicity are prominent climate fluctuations of still-shorter duration. During the Little Ice Age the (four) colder intervals coincide with periods of reduced solar activity. The last such cold snap was the Dalton Minimum at 1800-20. Since then, the world has warmed in three roughly-equal steps in the early/middle 1800s, from the early 1900s to the mid 40s, and from 1976 to the present day.

For the first two periods of post-LIA warming, in the mid-1800s and at 1910-45, facts are scarce; but we do know this warming largely predates the human-caused build-up of CO2 in the atmosphere. There is much circumstantial evidence pointing to twin drivers (which could even be related in ways we don't yet recognise)---increased solar magnetic activity, and increased flow of warm equatorial water into the Nordic seas.

5.2 Newer and stronger evidence

The TAR Summary for Policymakers carries a heading as follows (but emphasised in bold):

A reading of the succeeding fine print confirms that we are talking here of the 'greenhouse effect', and not of urban-heat-island and land-use-change impacts. Global warming during the twentieth century was about 0.6 ºC, in two tranches; and in the last 50 years we have an observed surface warming of about 0.3 ºC. It is this latter warming to which IPCC refers.

Happily, for the last 50 years we have many facts.

The world cooled abruptly in the mid/late 1940s, and temperatures remained depressed until the third period of post-LIA warming began abruptly at 1976/77. The most prominent climatic event in the twentieth century was this step-jump in average global surface temperature; but you wouldn't know from reading the Summary for Policymakers of either IPCC's SAR or TAR.

Is the 1976/77 jump a confirmation of IPCC's 'Greenhouse Effect' hypothesis of global climate change? And is the warming at a lesser rate since, greenhouse warming?

We all know that greenhouse is a phenomenon of the atmosphere. If the lower atmosphere doesn't warm, that warmer atmosphere can't then cause 'greenhouse effect' warming at the surface; greenhouse gas emissions don't warm the surface directly. No problems here, right?

From the start of adequate weather balloon coverage at 1958, right up to the step-jump, both atmosphere and surface cooled slightly---hardly an indicator of contemporaneous greenhouse warming. But then, both atmosphere and surface warmed abruptly at 1976/77. This warming step coincides with a prominent re-ordering of oceanic circulation, as indicated by a sharp reduction in the upwelling of cold, deep, water in the eastern Pacific. This is an inertial event. How do I know? Because of a change in the rate-of-change of length-of-day at about that time.

It is more plausible that the observed atmospheric warming at 1976/77 is a response to the contemporaneous oceanic warming, than vice versa. As corroboration of this line of reasoning, the prominent cooling in the 1940s coincides with a time of increased upwelling.

Recently, Hansen et al (2001) have found evidence of a declining (deep) flow of cold dense water out of the Nordic seas, and across the Greenland-Scotland sill into the Atlantic, over the past 50 years. This adds to the (albeit still limited, but growing) evidence that the main climatic driver over the past half-century is much more likely to be changes in oceanic heat transportation than changes in atmospheric composition: Oceans one; Atmosphere nil.

5.3 The past two decades

But what of the continued warming since 1977? There is even more information available for the past two decades. In particular, we have the coverage of satellite-derived atmospheric temperatures since 1979. During this time (now 221/2 years, and counting), the atmosphere has warmed only a third as fast as the surface. The simplest explanation for the discrepancy is that most of the surface warming during this time is not greenhouse warming.

Looking now in more detail, there is no warming of the lower atmosphere in the Southern Hemisphere. This is unsurprising. The dominant greenhouse gas is (naturally-occurring) water vapour, and it will tend to mask the potential warming impact of lesser GHGs such as anthropogenic CO2. Where does water vapour come from? It comes from evaporation; and the surface of this hemisphere is 80% water.

The only observed atmospheric warming trend over the period is in the mid latitudes of the Northern Hemisphere. Here, the crucial clue is that most of the observed warming is over land, and in winter.

It is the intensely-cold, bone-dry, anti-cyclonic cells, which tarry over Siberia and Alaska/Yukon in winter, which enjoy most of the warming---although surface temperatures beneath them still remain far below freezing (and hence, their environmental impact is minimal).

This could well be greenhouse warming: Oceans 11/2; Atmosphere 1/2.

5.4 Looking ahead

Predicting future climate is still beyond us. But there are some things we do know.

We cannot stabilise global climate by 'doing the right thing' about greenhouse gas emissions. But by pretending that the impossible is possible, we are already diverting much-needed effort from environmental threats, some irreversible, on which we can have an impact.

Natural climate variability will always be; and so will extreme climatic events. Greenhouse effect or not, prompt mitigation of weather-related human suffering will remain a need.

We are in the latter days of our 'normal' allowance of 10,000 years of benign climate. Already, summer insolation at higher latitudes of the Northern Hemisphere is on the wane. Will the next 'Little Ice Age' be the one which pushes Earth beyond the point of no return---from whence cooling feedbacks lead inexorably to the impending Glacial? When it comes, Armageddon will be cold, not hot.

Twenty thousand years ago, fortunate Australians walked through the dunes of Bass Strait to Tasmania. In not-so-fortunate America then, a kilometre of ice covered the site of Detroit. How soon will it be before we are welcoming US Boat People to our shores? Just joking.

6. HELLO THERE, NATURE!

• How can you justify limiting the scientific consideration of climate change to atmospheric scientists only?
• Why do non-atmospheric scientists who question IPCC's autistic fixation with the atmosphere make you think of Aids or tobacco?
• Why is reasoned dissent, from IPCC's advocacy of decarbonising the western world, immoral?
• You support the advancement of science, but you object to scientific debate having two sides.
• Finally, you paint scientific consensus as a virtue. But no amount of consensus-building can make IPCC's scientists 'right'. The advancement of science is not a matter of voting; and we know that if it were, science would never advance.

7. REFERENCES

Hansen, Bogi, William R. Turrell and Svein Østerhus 2001, 'Decreasing overflow from the Nordic seas into the Atlantic Ocean through the Faroe Bank channel since 1950', Nature volume 411, pages 927-30.

Michaels, P.J. and P.C. Knappenberger 1996, 'Human effect on global climate?', Nature volume 384, pages 522-3.

Pielke, Roger A. Jr 2001, 'Room for doubt', Nature volume 410, page 151.

Santer, B.D. et al 1996, 'A search for human influences on the thermal structure of the atmosphere', Nature volume 382, pages 39-46.

Stocker, Thomas 2001, 'Climate panel looked at all the evidence', Nature volume 410, page 299.

Wigley, T.M.L. and S.C.B. Raper 2001, 'Interpretation of high projections for global-mean warming', Science volume 293, pages 451-454.