| SHATTERED CONSENSUS The True State of Global Warming Patrick J. Michaels (ed.) Lanham, MD: Rowan & Littlefield, 2005 |
Rating: 3.5 Fair |
|||
| ISBN-13 978-0-7425-4922-7 | ||||
| ISBN-10 0-7425-4922-4 | 291pp. | HC/GSI | $? | |
![[Rant Warning]](../../../../RantWarn.gif)
Much as the dynamiting of the Buddha statues at Bamiyan by Taliban fighters represented the shattering of artifacts of one religion by adherents of another, the title of this book intimates the demolition of an edifice of science that has stood intact for a century and a half. And while most of the world, even most Muslims, mourned the loss of those ancient cultural treasures in Afghanistan, who would lament the loss of a climate phenomenon that portends extremely bad times for human civilization in decades to come?
There's one sticking point, though. The mainstream view of climate science is not some religion carried down to us through the centuries in holy scriptures. It is not the opinion of some tribe, widely adopted because it serves the tribe's interests. It is a scientific consensus. That means it is based on evidence obtained by observing and measuring and recording and analyzing natural phenomena. The only way to destroy a scientific consensus is to come up with new evidence that refutes the evidence it was founded on, or a theory that better explains the existing evidence.
Now I gladly acknowledge that this volume is, in the main, a well-intentioned and respectable effort. But it does have its defects. I discuss them here, beginning by examining the Foreword to the book paragraph by paragraph, offering my comments after each. Quotations are in the boxes with brown borders; my comments are in the blue-bordered boxes.
With few exceptions, scientific controversies rarely intrude into the consciousness of the general public, yet the vocal debates, in America and worldwide, over climate change are widely reported. The reasons for this are not surprising: weather affects everyone, every day, and the social and economic impacts of a radically changed climate are hard to comprehend.
In fact, people have little understanding of the exact nature and causes of climate change, in spite of—or perhaps because of—the vast amount of sensational literature available. There are many reasons for this. Specialists often have difficulty explaining their research to the public, in part because the basic principles of scientific methodology are not widely known. Even fundamental principles of climatology—such as that climate always changes, sometimes very abruptly in as little as a few decades—are unfamiliar to most people. The mechanisms and pace of climate change remain the subject of considerable uncertainty and scientific inquiry. Given that, it is not surprising that the public has a difficult time making sense of the debate.
It's possible to interpret that second sentence as a subtle dig at the scientist: "He was serious?" But leaving that aside, this paragraph implies that the National Science Board statement equates to the entire mainstream consensus on climate science. Further, it omits any time scale. In fact, by natural influences alone, Earth should be gradually cooling as the Milankovich pattern edges into a cooling phase. But Milankovich changes take thousands of years to make any appreciable difference. Logically, then, the worldwide warming that scientists currently measure might not be natural.
And let's look at the language: an "apocalyptic bandwagon" gets in motion, propelled by "presumptions of knowledge." If the evidence for the mainstream view is that flimsy, it ought to be easy to knock it down, to stop the "bandwagon" in its tracks. Yet the "skeptics" haven't managed it, even given another ten years of opportunity.
And, as the paragraphs in this Foreword reveal, there is an ulterior motive behind the exaggeration of that uncertainty. It is the desire to block effective action against climate change for as long as possible. The reason is that effective action against climate change means reducing the amount of carbon dioxide entering the atmosphere — and that means cutting back, and ultimately ending, the use of fossil fuels to drive transportation and generate power. There is a large and highly profitable industry built on sales of those fossil fuels. Its leaders want to keep those profits flowing. Indeed, they wish no diminution of those enormous profits, no matter how slight. This is why they (with a few admirable exceptions) have always opposed the relatively minor costs of pollution control measures — costs which have always turned out to be lower than their alarmist predictions.
As you read through my comments on the chapters, remember this quote from Posmentier and Soon:
"We present these criticism with the aim of improving climate model physics and the use of GCMs for climate science research. Furthermore, we are biased in favor of results deduced from observations." – Page 271 |
Here, Michaels introduces the other chapters and provides his list of allegedly deficient IPCC Statements. Let's look at the middle one on page 11. "Incomplete Statement. 'The rate of increase of atmospheric CO2 concentration has been about 1.5 parts per million (0.4 percent) per year over the past two decades' (IPCC 2001, 7)."
In response, Michaels writes:
"Although that is technically true, it ignores the fact that since the mid-1970s, the increase in concentration has been statistically indistinguishable between a low exponent (which is implied by a percentage) and a simple linear increase. An exponential increase is required to generate a constant rate of warming. If the increase is linear, the logarithmic response of temperature to CO2 changes means that the observed linear warming will become slower as the thermal lag of the oceans catches up with the atmosphere. Assuming much of the lag is taken up by forty years, the observed CO2 increases of the past quarter century, if truly linear, argue for a reduction in the rate of surface warming beginning around 2020." – Page 11 |
Michaels is correct that the temperature response to a linear increase in CO2 concentration is logarithmic. But he merely assumes that the concentration rate is linear. Given an increasing population and a rising standard of living, this seems a shaky assumption, to say the least. He also ignores water-vapor feedback. Additionally, I read the part about the "thermal lag of the oceans" to mean that the oceans no longer pull so much heat out of the atmosphere. If this is what he means, it suggests a faster temperature rise, not a slower one.
He continues:
"Further, a statistical fit of emissions per capita data from the Energy Information Administration (figure 1.5) shows that, globally, emissions per capita reached a maximum in the mid 1980s. The statistical significance of that maximum (as opposed to a continued increase) is at the 1-in-10,000 level." – Page 11 |
As an indicator, "emissions per capita" resembles emissions intensity, or the amount of CO2 per dollar of GDP. When energy is used more efficiently, per-capita emissions will fall even if the absolute quantity of CO2 emitted rises. An economic recession will also reduce this indicator, as it did during the Great Recession of 2008, without reducing the total amount of CO2 in the atmosphere, and once the recession ends the indicator will trend upwards again.
So it seems to me that Michaels here is making statements that are both incomplete and misleading.
This chapter is Ross McKitrick's almost day-by-day account of his effort, with Stephen McIntyre, to figuratively "rap the IPCC up'side the head" and get them to recognize the flaws in what has come to be known as the "Hockey Stick Paper." Here's how he describes his motivation:
"After studying in detail how the hockey-stick graph was done, we found mistakes in the data and methods that went unnoticed for years, even as the graph was used by governments worldwide to drive major policy decisions. The story behind the hockey stick provides a cautionary tale about the need to recognize the limited function of journal peer review and the dangers of preceding with major policy decisions without applying a further level of due diligence equivalent to an audit or an engineering study." – Page 20 |
It's noteworthy that McKitrick is an environmental economist and McIntyre is a businessman working in the "speculative mineral exploration" field. Both are Canadians. The essence of their complaint is that Mann and his colleagues misused statistics and data in deriving their conclusions, and that the IPCC did not bother to check their results as they should have. As things turned out, however, investigations completed well before publication of this book vindicated both the ability and the integrity of Mann's team. Since then, multiple independent studies have performed similar research and have arrived at conclusions quite similar to those of Mann et al.
I found McKitrick's tale hard to follow. It seems to rely exclusively on tree-ring data, though the proxy data sources used by Mann et al. include coral reefs, ice cores, and sediments as well. Descriptions of procedure are sometimes puzzling. At one point he refers to McIntyre using the high-level language "R" to do his calculations when, unable to explain his failure to replicate Mann's result, he looked at the data.
"This led him to examine the data set visually, whereupon he saw that many of the PC series that he used began in rows where the year ending is *99 or *49, while Mann's usual practice was to start series in years *00 or *50. On a hunch, he shifted one of the PC networks up a year, and the explained variance shot upward." – Pages 25-6 |
I find it hard to imagine how this shift, using unaltered data, could cause such a drastic change.
McKitrick even questions the use of data from individual trees (e.g. "a tree-ring width proxy from the site at Twisted Tree Heartrot Hill (TTHH) in northern Canada".) (page 30) He also delves into such minutiae as where Mann stored his data and how the files were named. It's all quite unconvincing on its face; and of course the critique has been found not to hold water.
First sentence: "The overwhelming majority of global warming concern comes from the output of theory-based numerical climate simulation models developed and run throughout the world." No. Just no.
Climate-science Denialists are fond of this claim. It lets them pretend that the inputs to the models are fudged to give the desired result: a scary prediction. This author throws in an extra fillip: the word "theory", which is supposed to suggest that climate science is just somebody's opinion — that's the popular meaning of "theory," after all. But concern about our changing climate would exist without models. It springs from knowing how the Arctic icecap is shrinking; from how most glaciers in the world are melting; from how springtime comes earlier each year, and autumn later; from longer wildland fire seasons and pine-beetle expansion and many other phenomena — not to mention the way measured surface temperatures are climbing, even far away from cities.
Pages 51-2: "Over the entire 1900-2002 period, the near-surface air temperatures rose linearly by 0.069°C decade-1; warming spurts occurred from the late 1910s to 1945 and from 1970 to the present." Calling the overall trend linear is incorrect, but this is merely a semantic objection; the author obviously means the average of the slopes during the three periods equates to a linear trend of the value given. And the discussion of possible biases in the surface temperature (pp. 50-59) appears cogent. I am dubious about the claims (p. 57) regarding warm eddies in summer (would there not be cold eddies in winter?) and that these warm eddies only affected modern thermistor measurements. But at least this comes with a citation; none is provided for the claim (also p. 57) that aging of the paint on the shelters introduces a gradual upward bias.
The major problems arise, however, in the following discussion of alternative explanations for the forcing in the latter period. Page 59: "Given the massive swings in global temperature approaching 10°C from coldest periods to warmest periods, the approximate 0.70°C temperature rise since 1900 is nothing out of the ordinary when viewed over long periods of the earth's history." Yes. Just as 10°C or even 100°C is insignificant when viewed over Earth's history — since Earth was once molten throughout. But prehistorical temperature changes, of whatever magnitude, are not germane. Human civilization was not around when those larger temperature swings took place. The relevant point is what the projected rise of 3°C by 2100 will do to Earth and its denizens as they now exist.
Page 61: "Indeed, solar output has increased by approximately 2.0 Wm-2 over the period 1900 to 2002 (figure 3.5)." I make it 1366.25-1365.5, or 0.75W/m2. But even granting the full 2Wm-2 means the change is only 0.15%.1 It falls to the author to explain how this causes the total forcing observed — especially when solar output has been flat from 1979 on.
Page 62: "Basically, during times of high sunspot numbers, the earth's magnetic field weakens and cosmic rays from the sun and from outside the solar system penetrate into the troposphere." This explanation is garbled. Earth's magnetic field my weaken at times, but sunspots do not weaken it. Rather, when the solar wind lessens, more cosmic rays from outside penetrate the solar system, including Earth's troposphere. But the salient point is that no definitive link between cosmic rays and cloud formation has yet been established.
John Christy's contribution amounts to a first course on measuring temperatures in the bulk of the atmosphere, whether by balloon-borne thermometers or by microwave remote sensing. There is a great deal of highly technical material here, which — to the extent I understand it — shows Christy to be a very astute practitioner of this arcane science.
"Sampling error places the data in the context of time and space and might answer a question such as 'How representative is the trend of this sample (1979-2002) relative to previous and future twenty-four-year periods?' Though we have fairly specific tests to estimate measurement error, it is more difficult to assess sampling error due to the apparent lack of representativeness of statistics over periods of only twenty-four years (Santer et al. 1999, 2001). For example, two significant volcanic events and two of the most intense El Niño-Southern Oscillation (ENSO) events have occurred only since 1979. Events of that magnitude did not occur in the previous forty years, thus the representativeness of the post-1979 period can be quite difficult to quantify (Christy et al. 2000; NRC 2000a; Folland et al. 2001)." – Page 88 |
I may be missing a subtle statistical point here. But since volcanic eruptions are, AIUI, random events, they could simply be subtracted from the record to yield a normal temperature trend for the period. This trend could then be compared to those of earlier periods. (I can't speak to the representativeness of future trends — and neither can anyone else.) Dealing with extreme examples of cyclic events such as ENSO is more difficult. However, I suspect there is enough paleoclimate data to permit estimation of some trend for them.
"Santer et al. neglected to report the many other tropospheric data sets that support the UAH long-term trends. They even cited one of the papers (Lanzante et al. 2003), yet failed to mention what the paper actually reported—there was very high consistency between the LKS trends and those of UAH. As a result, it is not useful to consider that methodology as addressing the issue of precision in long-term trends." – Pages 91-2 |
Is Christy accusing Santer and the other authors of scientific misconduct here, or only of carelessness? In any case, mistakes in use of a technique do not necessarily render that technique irreparably useless.
"Since there have been serious questions raised about the magnitude of the future concentrations of anthropogenically produced greenhouse gases (i.e., the magnitude of future forcing) as well as the climate responses to that forcing, this issue of long term projections continues to be wrapped in considerable uncertainty." – Page 100 |
Based on the tenor of Christy's previous comments in this chapter, I tend toward the view that he overestimates (or overemphasizes) that uncertainty. Why would he do this? I can only speculate.
Randall S. Cerveny concludes that there is no evidence for the contention that climate change has increased either the strength, frequency, or speed of formation of severe weather (hurricanes, thunderstorms, tornadoes.) Since the mainstream view agrees, this is valuable but unoriginal. It does nothing to challenge the consensus.
David R. Legates elucidates the complexity of the hydrologic cycle at regional scales, and convincingly demonstrates that the models used in support of the IPCC's AR3 cannot reproduce phenomena observed at such scales. But he ignores the certainly of improvements in the models, both in computer hardware and in the ability of the programming codes to incorporate physical phenomena. Moore's Law alone guarantees greatly enhanced processing speed and spatial resolution.
Oliver W. Frauenfield similarly demonstrates the fallibility of 2005 modeling of the El Niño-Southern Oscillation and other cyclic phenomena. His conclusion:
"Only after we identify these factors and determine how they affect one another can we begin to produce accurate models. And only then should we rely on those models to shape policy. Until that time, climate variability will remain controversial and uncertain." – Page 176 |
The necessity of making decisions without having full information in hand arises in business every day. Dr. Frauenfeld is a research scientist, not a businessman. Yet I feel he should be aware that climate policy too might require such ad hoc decision-making at times.
There is always the possibility that some of these conclusions are not supported by the sources they cite. That has been the case in claims by certain Denialist politicians — notably James Inhofe, the Senator from Petrofulia. In reading this volume, I got the sense in places that such obfuscation was being attempted. One such place is Chapter 8, and especially the section on diseases in that Chapter.
I point out some examples, opposite. Of course, this is only a supposition on my part. Confirming it would require examining the cited sources as well as other papers on the subject, which might provide contrary data. I can't spare the time for that.
Still, I think the possibility of bafflegab is worth noting, because any findings that would truly shatter the climate consensus would be widely heralded; there would be no need to publish them in a book like this.
Robert E. Davis casts doubt on projections of increased deaths due to higher temperatures — at least in urban areas of the United States. He goes on to question the view that rising temperatures will lead to the spread of tropical diseases. I think some of his arguments may have merit.
At the same time, I have to question the veracity of his overall argument. For example, here is how he leads off his chapter:
Global warming is one of the most important scientific issues of our time. The increasing concentration of greenhouse gases (carbon dioxide, methane, nitrogen oxides, ozone, etc.) has been implicated in the temperature increases observed in the latter part of the twentieth century. Many of the greenhouse gas increases can be directly linked to human activities, such as fossil fuel combustion, certain agricultural practices, land-use change, cattle production, and so on. A substantial number of climatologists believe the warmth of the late twentieth century is at least partially related to anthropogenic factors, although convincing proof remains a matter of debate within the climate community. – Page 183 |
Shorter Robert E. Davis: "It is well established that temperature increases are linked to rising greenhouse-gas concentrations due to human activities, but this is still a matter for debate." I can only conclude that the paragraph I quote above is contradictory. How does it seem to you?
Here's a statement that strikes me as false:
Globally, the warming experienced over the last half-century has been predominantly concentrated in the cold half of the year (Balling et al. 1998) – Pages 190-191 |
I think the great quantity of recent record highs in absolute temperature measurements gives the lie to this statement.
Here's another questionable passage:
Weather-mortality relationships in winter are significantly more complicated than in summer. In nearly all cities examined globally, winter mortality is much higher than summer mortality, for reasons that are not well understood [multiple citations removed]. Influenza and related respiratory diseases have a pronounced winter peak, and other causes of death also tend to be correlated with influenza epidemics (Simonsen et al. 2001) To date, no research group has been able to link the strength or duration of influenza epidemics or seasonal influenza deaths to weather or climate factors. – Page 191 |
Who does not know that flu is a disease of autumn and winter? The matter is less clear-cut for other respiratory diseases, such as asthma, which Davis discusses later. But here, in a subsection headed "Cold," he should focus exclusively on direct effects of lower temperatures. He does that later in the section, but finds no firm evidence either way.
Next up is the topic of air pollution. Davis defines this to include total suspended particulates (TSP), and more specifically PM10 ("suspended particulate matter greater than 10 microns in diameter"), on page 194. Yet later he declares:
Air pollution does not appear to exacerbate asthma occurrence or severity [four citations removed]. – Page 197 |
"Lie" is such an overworked word. "Howler" comes to mind as a substitute. Snark aside, the impacts of air pollution on asthma sufferers have been well-known since long before 2005. Also I suspect (and will check ASAP) that Davis is confused about whether PM10 means particles larger or smaller than 10 microns.
The last section of this chapter on health effects of climate change is devoted to vector-borne diseases. In discussing malaria, Davis notes:
Although small temperature increases can raise the risk of malaria transmission, very high temperatures can be lethal to the mosquito and the parasite (Bradley, 1993). – Page 198 |
Obviously, a sufficiently high temperature (Fahrenheit 451, say) will destroy both parasite and mosquito. But notice that Davis leaves the magnitudes of his temperatures unspecified. If we are to assume the range involved in climate-change discussions, as is logical, we must allow that both organisms survive tropical temperatures quite well. Global warming of the projected magnitude (at least 3°C) will certainly expand the range of malaria mosquitoes.
Davis mentions the concern that malaria will, with warming, move into high-altitude regions that are presently free of it, and cites (Epstein et al. 1998; McMichael et al. 1998). He then debunks this worry by citing other sources (Mouchet et al. 1998; Reiter 1998) to assert that instances of high-altitude malaria have occurred in Madagascar and Costa Rica. Probably they have; but without checking his sources I would not concede the point. Could a storm have blown a population of malaria mosquitoes into a high valley, where they lived long enough to infect local people? Could someone have inadvertently carried water containing mosquito larvae up there, with the same result? Yes to both questions. By the way, "Epstein" is undoubtedly the late Paul Epstein, co-author of Changing Planet, Changing Health. I tend to trust Dr. Epstein's word on this subject.
A general note: Davis discusses health effects in terms of mortality almost exclusively. Even when discussing asthma, he barely mentions non-fatal cases of illness. This is not the way to conduct a fair and comprehensive analysis of health effects.
Here is how Davis closes his chapter:
In summary, developed nations are now, and in the future will likely remain, largely immune from major effects of climate change on human health and well-being. These societies have the infrastructure and demonstrated ability to adapt to climate variability. Underdeveloped nations will likely be less successful. Yet the extent of human biophysical adaptations, though not well understood, will certainly be a factor in mitigating against major impacts. Furthermore, the climate changes likely to occur are small relative to the sensitivity of the systems they will purportedly influence. Given the inherent nature of living things to adapt, it is likely that the net impact of projected climate changes on human health will be very limited. – Page 203 |
Remember, kids: climate change is only a purported threat — and we'll survive it just fine.
Sallie Baliunas probes possible effects of solar variability on Earth's ecosystems. Dr. Baliunas is an astrophysicist. Is this not a rather odd topic for an astrophysicist to tackle? Perhaps not, in Denyalistan.
However, recent research reveals surprisingly good correlation between measures of solar variability and local terrestrial ecosystem change suggesting sun-climate influences on periods of decades to centuries and longer. If such observed environmental variations were to be explained solely by solar total irradiance changes, they would be much larger over periods of decades to centuries than has been suspected. Alternatively or additionally, terrestrial processes such as ocean-atmosphere interactions may amplify small changes in total solar irradiance or specific aspects of solar change (e.g. variations of the ultraviolet portion of the sun's electromagnetic spectrum or the sun's highly variable fast-moving particle output). Finally or additionally, the flux to the earth's upper atmosphere of energetic particles produced primarily by galactic supernovae, called cosmic rays, changes as the sun's varying magnetism modulates the cosmic ray flux. That has led to speculation that cosmic rays contribute to environmental change, perhaps through the intermediary of terrestrial clouds. Proposed solar and cosmic ray processes that seek to explain the observed, apparent exo-terrestrial influences seen in environmental reservoirs cannot yet be adequately quantified. Hence, it remains difficult to apportion ecosystem change among exo-terrestrial and other natural effects, plus anthropogenic changes such as the enhanced greenhouse effect, black carbon emission, and landscape modification. – Pages 210-211 |
Leaving aside the oddness of the writing (which the editor should have caught), I want to say this about that:
Thus, the SPM statement [quoted in the previous paragraph], emphasizing the impact of total solar irradiance alone, is incomplete, in part because, as Working Group I concludes, the sun's environmental influence remains uncertain. *
* * That lack of knowledge of the causes of the correlations—beyond those of just total irradiance observed over two decades—on time scales of decades to centuries means that their impacts on climate cannot be reliably reproduced by simulations. – Pages 211-212 |
Dr. Baliunas has devoted the previous portion of the paragraph from which this came to quoting the several qualifications made by the IPCC about its then-current knowledge of solar variation. Why does she apparently insist that they factor in those unquantifiable (perhaps unreal) unknowns? Also, it's my understanding that simulations using both natural and anthropogenic forcing factors do reproduce the observed late-twentieth-century warming. As she should know,3 it's not science when you fold conjecture into your theorizing — though it may become science as evidence for the conjecture appears.
Further, TAR4 SPM concludes that "surface warming over the past 100 years" "was unusual and unlikely7 to be entirely natural in origin" (10), where the superscript 7 denotes a footnote defining "unlikely" as "10-33% chance" that a result is true, based on "judgmental estimates of confidence" (2). That mix of quantitative and subjective numbers appears despite poor confidence, meager quantitative knowledge about solar and other exo-terrestrial factors, and the correlations that have been observed5 linking solar magnetic changes to local terrestrial ecosystem changes on decadal to millennial scales. – Page 212 |
So whereas above she advocated for the IPCC adding total unknowns into the climate-science assessment, now she complains that they did not. She goes on, in a part of the paragraph I omitted, to complain that landscape alterations and biological responses to increased CO2 "have only crudely if at all been incorporated into the simulations." For the record, that superscript "7" applies only to the word "unlikely" and is perfectly justified as a way of labeling a confidence interval.
I will grant that Dr. Baliunas knows her solar physics and also the history of astronomy, so that reading this chapter is quite an education. However, when it comes to proposing possible mechanisms for the alternatives to greenhouse-gas forcing, or negative feedbacks opposing it, that she mentions, I find only a series of speculations. While some of those speculations are intriguing — in particular the putitative (speel?) link between cosmic rays and cloud formation, and the idea that solar UV might exert an unsuspected strong forcing effect — they have scant scientific support in 2014, and had less in 2005.Likewise, the contention that phenomena operating on times scales of centuries or millennia are somehow involved seems irrelevant to the current situation. Such phenomena may become relevant, but not until new evidence appears.
I must preface my remarks about this chapter by noting that, while I have some experience with simulation of electronics design (SPICE and related tools), my familiarity with climate modeling is only that of the educated layman. While I'm not able to evaluate specific details of models, I can make some general observations. I can also evaluate the language used.
This final chapter starts off well. But then, in its third paragraph, I find:
Given the host of uncertainties and unknowns in the difficult but important task of climate modeling, the unique attribution of observed current climate change to increased atmospheric CO2 concentration, including the relatively well-observed latest twenty years, simply is not possible. We further conclude that the incautious use of GCMs to make future climate projections from incomplete or unknown forcing scenarios is antithetical to the models' inherently instructive value. Unfortunately, such uncritical application of climate models has led to the commonly held but erroneous impression that modeling has "proven" the hypothesis that CO2 added to the air has caused or will cause significant global warming. – Pages 241-2 |
A common contrarian conflation is the claim that models are inherently imperfect, and therefore AGW is flawed if not wholly incorrect. Here we find an example of that conflation. That models have limitations is a truth acknowledged by all. However, "the unique attribution of observed current climate change to increased atmospheric CO2 concentration" does not depend on models. It depends on physical laws elucidated before the twentieth century began, on temperature measurements conducted since that time on an increasingly widespread basis, and on isotopic analysis to attribute the source of the extra CO2. Remember the line I quoted above, about how Posmentier and Soon are biased toward observations? One could wish they had given some evidence of that by recognizing the reality of such observations.
Here's another unconvincing objection to the usefulness of climate modeling:
An additional difficulty concerns the logistics of modeling a system with spatial and temporal scales that account for such factors as cloud microphysics and global circulation. Fortunately, that difficulty can be circumvented because of empirical "loopholes" such as the existence of gaps in the energy spectrum of atmospheric and oceanic motions that allow for the separation of various physical and temporal scales. Say, for example, that climate is viewed as an average over a hypothetical ensemble of atmospheric states that are in equilibrium with some slowly changing external factor. Then, under a regular external forcing factor, it seems possible to anticipate the change (Houghton 1991; Palmer 1999). Essentially all calculations of anthropogenic CO2 climatic impacts make that implicit assumption (Palmer 1999). But for such a calculation to have predictive value, rather than merely representing the sensitivity of a particular model, a GCM must be validated specifically for the purpose of its specific type of prediction. As a case in point, we note that the prediction of climate responses to individual forcings such as the long-lifetime greenhouse gas CO2, the shorter-lifetime greenhouse gas CH4, the inhomogeneously distributed tropospheric O3, and atmospheric aerosols would all require separate and independent validation. A logistically feasible validation for such predictions is essentially inconceivable. – Page 243 |
This can be interpreted as questioning the ability to combine in models the simulations of individual forcings (e.g. sunlight and aerosols.) But I don't interpret it that way, because I doubt that Soon and Posmentier are so uninformed about models. Rather, I think this is a sophisticated form of the claim that a prediction cannot be validated without arriving at the future time for which it was made. Thus, for example, if a hypothetical model were to predict a certain amount of cooling due to a doubling of atmospheric aerosol content by 2035, the authors mean that only in the year 2035 can such a prediction be validated, because the real climate system is too complex to model accurately.
I believe this overstates the difficulty, because as climate models and the hardware that runs them continue to improve, they will more closely simulate the real climate system. Will they ever simulate it closely enough to be used in deciding policy? That of course is a matter of judgment, on which reasonable people can disagree. Based on what I continue to hear from a number of climate contrarians, I predict a very long wait until disagreement becomes negligible.
Are there any problems which will finally preclude sufficiently realistic modeling of climate? I doubt this because of my understanding of modeling in other fields. Numerical simulation is good enough to be used in the design of nuclear weapons, where it obviates actual underground detonations in certain cases. Other complex problems where it is useful include stellar astrophysics, computational fluid dynamics (CFD) in aerospace, and drug design. Also, few people, to my knowledge, refuse to ride in airplanes because they are designed by computer models. Therefore, my personal judgment is that climate models will be good enough to depend on quite soon, if they are not there yet.
Two more things are apparent. First, the large number of citations in this chapter makes it clear that claims of criticism being suppressed are totally unfounded, at least with respect to models.6 Second, as with open-source software, this abundance of contributors guarantees that any problems which can be fixed will be fixed, and probably in short order.
To contact Chris Winter, send email to this address.