Climate Change: What is it? Who can you trust? Is it that bad?

Who exactly can you trust when it comes to climate science? In her 2010 keynote speech at the Canadian Society for History and Philosophy of Science, Evelyn Fox Keller called upon scientists and philosophers of science to engage the public in critical discourse regarding the issue of climate change. She claimed that scientists and HPS’ers (those that study history and philosophy of science), who are best suited to talk about the myriad of complex issues surrounding the subject, have traditionally been reluctant to take their expertise into the mainstream. This blog post, the first in an ongoing series, is designed to answer Fox Keller’s challenge.

As the introductory blog in the series, this entry will be an overview of the multifaceted scientific and political challenges that climate change poses. My goal here is somewhat modest. Simply put, in this post I want to give the basic background needed to enter into a discussion on climate change. Please use the comments section to ask questions about this information. If you already know this information, skip right down to the comments section and ask a question about climate science that you don’t know the answer to. I’ll respond, and maybe even answer it!

Subsequent posts will give more depth to the issues summarized here. I will also attempt reach out to guest authors with more expertise than I, to chime in along the way. On that note, it should be disclosed that I am not a climate scientist (although I studied with one), nor a full-blown environmental ethicist. My attraction to this topic is generated primarily from an academic interest in epistemic issues in scientific modeling, and secondarily from my love of outdoor sports (backpacking, mountain biking, and skiing). I should also disclose that I believe anthropogenic climate change is occurring, and something needs to be done about it – fast. With that being said, I don’t believe that every bit of science that supports this conclusion is correct; in fact, I am sure many of our assumptions are wrong. So with all the disclaimers out of the way, let’s get through the basics.

How do we know climate change is occurring?

The IPCC (Intergovernmental Panel on Climate Change), a group of scientists and legislators from around the world tasked with reporting on climate change, has concluded that warming of the climate system is unequivocal. For the IPCC, climate change is a change in the state of the climate that persists for an extended period, usually decades or longer, and is the result of natural variability or human activity. This kind of change has been detected in global air and ocean temperature, the widespread melting of snow and ice, as well as a rising sea level. Global temperature data from 1995-2006 indicates that 11 of the last 12 years were the warmest on record when compared to instrumental data. This trend has and is expected to continue. Very few people, even climate skeptics, deny that the climate has gotten warmer.

What is responsible for the warming?

What people tend to fight over is whether or not the warm temperatures are the result of human activity or natural causes. If there is scientific consensus on climate change, it comes at this level: the warming has been caused by human activity, most notably the release of CO2 from the burning of fossil fuels.

Carbon dioxide is one example of what is called a greenhouse gas. When the sun’s light reaches the Earth, some of it is reflected off the atmosphere, and some of it penetrates. The penetrating light hits the Earth’s surface and is turned into thermal radiation (heat). The thermal radiation is then emitted back up into the atmosphere where some of it escapes through the atmosphere into space, and some of it reflects back down. When large amounts of green house gases are present, the amount of thermal radiation that can exit the atmosphere is reduced, and much more is kept within our atmosphere. Simply put, this leads to warming. Of course there are some other factors too, factors that exaggerate or mitigate the effects of greenhouse gases, but no factor is as significant to climate change as the release of CO2. We will get into more detail about these other factors in a follow-up post on climate modeling.

So what of it?

Bangladesh during monsoon season. © The Atlantic Monthly Group

This is where things get interesting. So the Earth is getting warmer, why is that a bad thing? Heating of the Earth’s atmosphere, it is predicted, will lead to some long term negative effects, and a few positive ones. It is predicted that with warmer temperature comes increased snow and glacier melting. The beginnings of this process have already been observed. This melting will lead to higher sea levels, which could put seaside counties like Bangladesh underwater, displacing hundreds of thousands, if not tens of millions, from their homes. As seas rise, many believe there will be a loss of fresh drinking water (the scarcest resource on the planet, way scarcer than oil). Where there is fresh water, the higher temperatures are likely to cause increased breeding of insects, and with that increased rates of disease, most notably in countries where there is no money or infrastructure to handle an outbreak. Some species, particularly those in the arctic, are likely to lose their habitats, while others are likely to spread. On the plus side, countries like Russia and Canada are likely to gain arable land, and of course, there is always that battle over the new arctic passageway.

You keep saying that people predict…but how?

The answer is climate models. Scientists have been working hard to create large scale computer models that will help predict the changes in climate 50 to 100 years from now. The models are immense and run on very powerful computers. It is not like sitting down on your laptop, but, you can help that way. In order to reach any conclusion from the models, they are first tuned to known data, then run hundreds of times, and then compared to other models. If this makes them reliable, no one really knows (after all, the test of their results is 50 years away). This issue however, will be another post on its own.

So ah, it sounds bad, what do we do?

Of course, this question is the most important. Most people believe the drastic negative effects mentioned above can be prevented if drastic changes are made now. Many think that we need to prevent more than a 2.5 degree C rise in global mean temperature from current levels. Of course, it seems that we are slow to action. What do we do? Why haven’t we done it yet? Who is responsible for doing it? And what are the moral implications of our (in)actions? That, my friends, is yet another blog entry.


  • W. Dean Reply

    Hello Greg,

    Let me start by saying that my interests in global warming reflect your own insofar as the epistemic and modeling issues go. I also want to commend you on getting the questions right. Too often this debate goes from “Is global warming occurring?” to the necessity of cap and trade or some other carbon emission policy as if there were not a whole host of important questions in between; for example, “Will climate change be good or bad and for whom?” and “Can we even do anything to stop it?” And this is also where I think philosophers can make a contribution to the debate.

    Unlike you, however, I’m a “qualified skeptic,” not because I’m an expert in climate science (I’m not), but because the case for global warming doesn’t pass what I call the “epistemic sniff test.” I give some of my reasons for skepticism regarding global warming, which, given time constraints, have to be quick and dirty. I don’t presume that any of my reasons for skepticism disprove global warming, of course, which is why I say my skepticism is qualified.

    Odor # 1: Lack of Transparency. Non-experts can assess (indirectly at least) the value of the case for climate change by comparing the scientific process behind it with analogous scientific processes. One of the keys to a good process, especially on issues where big money is involved on all sides, is transparency or full disclosure. In the simplest terms, the two minimum requirements for transparency are [a] that no stakeholder is forced to trust any other stakeholder’s data and [b] that all stakeholders know who is doing the evaluating.

    Regarding [a], it’s now public knowledge that certain repositories of the raw data concerning temperatures have been less than forthcoming when asked to provide it to outsiders for analysis. Whatever you make of McKitrick and McIntyre, for example, there was no reason for East Anglia and Michael Mann to withhold their data from them or from anyone else, no matter how “skeptical” they are. In fact, there is every reason to publish this data in the public domain. What’s even more bizarre is that we now know from the recent inquiry that East Anglia’s raw data has somehow been irretrievably lost.

    Now, ask yourself if we’d tolerate a similar state of affairs from the pharmaceutical industry, for example, or the makers of automobiles and children’s toys. I highly doubt that we’d accept (via Health Canada) Pfizer’s claim that a drug was safe and effective without the raw data along with impartial third-party review. Yet we seem content to accept non-disclosure and partial disclosure of data from the handful of scientists and institutions that actually possess temperature records.

    Regarding [b], I won’t bother with the problems that came up in the most recent review of the IPCC’s report. Instead, I want to point to the problem with the structure of the process itself. On the face of it, the IPCC looks like a meta-study that includes “thousands of scientists” analyzing the latest research. But take a closer look. First, meta-studies usually analyze all experiments involving a single correlation to determine whether the correlation exists (e.g., smoking and cancer). Global warming, however, is a complex phenomena involving an enormous number of variables with different models and different assumptions. It’s really impossible to do a meta-study with so many different models and variables.

    Second, the IPCC report is closer to a textbook constructed by small panels with lead authors and editors composing the final draft. That’s not necessarily a problem, but it’s important to note that the panels are not analyzing but summarizing. The real problem is that the mandate for the summary is to find the consensus among scientists, not to determine the truth about climate change. Thus, the panel cannot help but reflect dominant opinion (or the worst case their own opinion), which does not necessarily reflect the actual state of knowledge concerning climate. After all, the vast majority of scientists work within the framework of a particular model, and usually only in one domain of that model (e.g., historical concentrations of CO2 in the atmosphere). Far fewer scientists actually work on the models themselves (mostly because of the advanced mathematics involved). If the dominant model sees global warming caused by CO2, then the majority of scientists will obviously say that global warming will occur as the concentration of CO2 increases. But that consensus doesn’t support the conclusion that CO2 causes global warming, it only reflects the fact that most scientists operate under the assumption that it does.

    The real question, in other words, is not whether there’s a consensus but whether the model is any good. And the last I heard, no one has constructed a model that predicts past climate without introducing ad hoc fixes to account for unknown variables.

    Odor # 2. Angels, Demons and Propaganda. Regarding the latter, when I look at a graph like the one you’ve presented above, the skeptical hair on the back of my skeptical neck stands up. I don’t mean to suggest that you’re trying to bamboozle anyone with it, but that graph has no scientific import that would not be conveyed in a simple statement of the temperature increase. What possible value could a graph representing only 140 years of temperature on a 4.5 billion year old planet serve, other than to dazzle us with the rhetorical power that every graph full of colours and dots and numbers and dramatic curves has in our scientific age? Besides, it’s been clipped at convenient time periods (after the Medieval warm period) and before the post-2000 decline in average temperature, which might convey the idea that warming has happened but that it’s now over. Come to that, there’s been a scarcity of graphs lately…

    Regarding the first two above, the penchant for ad hominem attacks from global warming supporters would make anyone suspicious. When people are in a strong position, when the facts support them, they don’t have to make others into baddies whose claims are too ridiculous to answer. For me at least, the attempt to portray anyone who disagrees with the “scientific consensus” as an agent of Big Oil (or in Freeman Dyson’s case, a senile old fool—Freeman Dyson!) sets off alarm bells. Leaving aside the insult to my intelligence, the claim itself betrays an ignorance of the interests involved. Big Oil (e.g., Shell and BP) have been huge financial supporters of green organizations and universities. How come that doesn’t compromise these organizations? Are pro-climate change scientists immune to corruption? Moreover, some energy companies have vested interests in emissions controls, and I don’t just mean green industries. Enron, for example, was lobbying the Clinton administration to support the Kyoto, because it had a huge stake in natural gas, which it knew would be used to replace coal in electricity generation. If money corrupts, it corrupts absolutely. No one gets to be on the side of the angels in virtue of their beliefs, so no one should get a free pass in defending their evidence by impugning the motives of their opponents.

    (It’s also worth noting how the tenor of the global warming debate compares to other scientific debates. The sympathetic engagement of scientists and doctors with believers in the supposed vaccine-autism link provides a remarkable contrast. I’ve yet to see the sort of denunciations and accusations of corruption from the experts in this debate that I have in the global warming one. Despite the fact that children’s lives are involved and the possibility of the reemergence of long-eradicated diseases, the doctors and scientists I’ve read have been patient and even a little indulgent in dealing with people who have no expertise whatsoever in the science.)

    As I said at the outset, this is not a scientific argument. It’s just some bits of this philosopher’s epistemic heuristic for judging the persuasiveness of the case for global warming. Another might see all the problems I raised in a different light. He might say that these facts just reflect a poor PR campaign on behalf of the scientific community and that zeal has sometimes overtaken prudence when it comes to making the public case. But such a reading assumes an unphilosophical premise that I’m not willing to concede, namely, that we should always assume perfect knowledge and pure motives on behalf of authorities.

    • Boaz Miller
      Boaz Miller Reply

      W. Dean,
      I like your reasoned and provocative reply. I am also sympathetic to the kind of questions you raise. My worry is that you might be setting the bar too high. In an ideal world, science would have to meet the epistemic criteria you mention. But we are not living in an ideal world. If there is something the sociology of science has taught us is that scientific deliberation is a messy process and scientists, like other normal human beings, sometime play dirty. This is normal and happens in many fronts and fields and in itself does not justify blanket scepticism about scientific knowledge, all the more in climate science, which is politically charged from all sides. I think an epistemic “sniff” may not be enough to justify scepticism, but a more thorough epistemic examination is required.

      • W. Dean Reply

        Boaz Miller,

        I agree that in normal cases my reasons wouldn’t justify skepticism, because, as you point out, scientists are as combative and competitive as anyone else. Building a case against Darwinian evolution, for example, by citing Lewontin and Gould’s punctuated equilibrium would ring a little hollow, given the underlying agreement over the evolutionary model. But global warming is no normal case; its advocates aren’t offering a theory or a model of the climate that predicts global warming (in the same way evolution accounts for biological variation), they’re offering us a consensus in lieu of a model. Advocates are saying, in effect, “We can’t prove that global warming is happening, but we all agree, based on what we do know about climate, that global warming is happening. So we must act now, without complete knowledge.” Exceptions to the consensus are summarily dismissed. In a case like this, I suggest, the reasons I adduced do justify skepticism.

    • Allan Olley Reply

      First a factual point, are you sure any data has been lost. I believe East Anglia denies it, also its not the only holder of the data, the data in question was drawn from various national data that is still held locally at those places. Rather than merely analyzing East Anglia’s data sets researchers with alternative points of view, interested in serious scholarship and not cherry picking mistakes to attack credibility, would create their own data sets from the same records and make their own arguments from that. The absence of this gives me reason to doubt the “skeptics.” But its just a smell test…

      Providing data is not necessarily an effortless process, depending on the data demanded, the number of demands made or expected and in what format the data is demanded and in what format it is stored (seems quite possible the researchers rarely use the raw data) there could be real practical reasons not to act on every request. Now climate scientists would be well advised to make their databases easily accessible for any number of reasons, but this is probably true of most of science. With the right protocols in place sharing data of many kinds would be easy, but putting those protocols in place would not be easy.

      Second, ad hominem attacks against those who argue that climate change is human caused may well be as widespread. Depending on what the relevant contrast class is and so on…

      • W. Dean Reply

        Allan Olley,

        With regard to evidence, that’s just not how science works or could work. The onus lies on the claimant to provide the evidence (in this case raw data) supporting his claim; it’s not up to the rest of us to find it and disprove him. Put another way, raw data is really an extension of one’s published findings. So Mann and East Anglia can’t tell all comers to go construct the data themselves. Consider how odd scientific debate would be under this scenario. Imagine NASA’s JPL, for example, claiming to have found life on Mars, while responding to all requests for data with “Go buy your own Mars Rover.”

        I haven’t tallied up all the ad hominem attacks on all sides, because I’m not judging the truth of the case on the basis of politesse. I’m merely pointing out that it’s an all too common tactic of pro-global warming scientists to defend their claims by impugning the character of their adversaries, and that this approach doesn’t exactly inspire confidence.

        Finally, I don’t know for sure that the data are lost. A newspaper report on the Oxburgh investigation stated that the data were lost.

        • Allan Olley Reply

          By your reconstruction of the scientific process its a legitimate criticism of someone’s science that they did not prove all the theorems of arithmetic used in set theory (or whatever formal system you prefer) and you refuse to grant their arguments any credence until they do.

          In reality perfect and complete inference and so on is non-existent in science. In principle some error in basic (or advanced) math may slip into scientific inference, but would that justify the effort of total certainty that could be achieved by adding ever more proofs?

          Similarly in data handling I am sure that in many fields of high reliability and trustworthy-ness, the raw data from instruments, if directly stored at all, is almost never used, rather reduced and corrected data is what the scientists make use of pretty much exclusively. In which case their will be little, if any, impetus to have the raw data in some simple retrieval system. Even though in some rare cases it might have turned out to be useful.

          What is meant by raw data is somewhat arbitrary, you could choice to question the choice of thermometers at each weather station and so on and say that until they do give a full account of that they are hiding something or do not have the “raw data.”

          If it is a valid inference from the fact ad hominem attacks are made by some who defend the claim (A) that human caused climate change is occurring to a conclusion of decreased confidence in the arguments for A, then it must be by substitution valid to infer from the ad hominem used by some of those making the claim (B) that the science supporting (A) is unreliable to a conclusion of reduced confidence in the arguments for claim (B). I really advise against this sort of argument, but if you chose to go down it, I suggest you only do so once you figure out whether the total weight hurts or helps your case…

  • Greg Lusk
    Greg Lusk Reply

    W. Dean –

    Thanks for taking the time to write a thoughtful and interesting response.

    There is a lot going on in your comment! Many of the themes that you touch on will be talked about in greater length in subsequent posts. There will also be questions raised under the ‘debatable’ section of the blog where one or two of these topics can be put to the test in the comments.

    However, because the comments section of this post is supposed to be a place where those less familiar with the subject can feel safe asking questions, I will save possibly intimidating nuanced discussions for later posts.


    • Curtis Forbes
      Curtis Forbes Reply

      The debate is uncontainable, it would seem! But, I like the idea of just starting with some questions.

      One thing I’ve always wondered about climate models is how they can be considered good predictors long term, despite not being good predictors short term. I mean, why should I think that a prediction made about the weather in 50 years from now (it will be hot) is trustworthy, given that weather reports about the coming days are notoriously unreliable and uncertain (60% chance of rain)? I think I understand how that makes sense, but I’m curious how you’d explain that feature of climate and weather modeling.

      • Allan Olley Reply

        Curtis that line of argument seem to me almost fallacious. We can predict the climate in July in Toronto next year (and for many years subsequently) relatively well despite not being able to predict the weather at all two weeks hence, because the tilt of the Earth, its orbit and subsequent change in solar radiation on various bits of the Earth creates a trend that is far stronger than those variations.

        Note that in the yearly climate case incident solar radiation alone is not the only factor. If it were climate in Vancouver and in Timmins, Ontario (both around 49 degrees North) would be much the same. In fact they are different.

        So the inability to predict weather tells us nothing about our ability or inability to predict climate and indeed humans have been able to predict yearly change in climate for tens of thousands of years with some accuracy based purely on year to year correlations, despite weather prediction failing after a few days at best.

        As I understand it the basic mechanism of increased capture due to more green house gasses is fairly straightforward and well documented in experimental settings (fill air tight terrarium with different mix of gases and measure results for example). So as in the yearly climate case we have a well understood driver of warming. If the ceteris were parabus when applying these effects to the whole atmosphere warming would be guaranteed to occur given the increase of the level of CO2 and the like in the atmosphere, although if I articles in the public press I’ve read are right such a model is the low end of the trends suggested by the complex models touted by the likes of the IPCC.

        Now there are global cooling effects (anything that increases the Earth’s albido will cause some cooling for example), but is there good evidence of such an effect?

        Apparently many model builders find evidence for further effects that increase warming above the baseline of the green house effect (positive feedback). Given that some apparently models fail to find these positive feedback effects they are clearly in question among climatologists even at the IPCC. Make of that what you will.

        Also, why believe they would ignore potential negative feedback effects that would counter the green house effect?

      • Greg Lusk
        Greg Lusk Reply

        @ Curtis

        That is a good question, and there is a lot to say in response. I will give a basic answer here that presupposes some familiarity with climate and weather models. A future post on climate models will fill in some of the gaps for those that are less experienced with the issue.

        Before I typed this response I checked with a friend of mine who has worked in meteorology, as I knew little about how their models different from GCMs (global climate models). It turns out they don’t differ drastically – both rely on the same basic equations – but the problems each faces is unique.

        Before I mention these problems, I do want to speak more generally about uncertainty. When you say unreliable and uncertain, we should ask “compared to what?” If we measure the reliability of weather models compared to random guess, they are much more reliable. If we measure them compared to a prediction by an expert with meteorological training, the model can beat expert predictions up to 28 days out (according to my friend). So an expert with a good model is the best predictor we have.

        Also on a side note, the 60% chance of rain you mention is not to be taken as a classical probability – it means that 60% of the forecast area will have rain.

        So getting back to the differences…the primary difference between a GCM and a weather model is scale. The GCM covers the entirety of the earth, and the most accurate use grid cells of 100 km on a side. Weather models are made for specific areas and use smaller grid cells (30 km a size). The problem with either model is that the day to day weather events that most people are interested in (rain, snow, clouds) are subgrid entities at 30 km. So forecasting for them is quite difficult. As an example, if your small city is next to a set of hills, their effects would need to be captured with a grid .25 km on a size. But they definitely have effects.

        The small size of the weather models also results in two more difficulties: border edges and initial conditions. In a weather model, what is going on outside of the forecast region isn’t represented, but these things can be very important to local weather. One proposed solution is to nest fine grid models into coarser grid models.

        Because weather models are non-linear, and their forecast time so short, the initial conditions that are plugged into the model make a big difference to predictions. Strange attractors, and other non-linear phenomena, can result in strange outcomes.

        GCMs on the other hand do not suffer as much from these problems. Since the models are truly global, there are no boundary edges to worry about. Non-linearity is still present in the CGM case, but since the predictions are on a longer time scale, strange local occurrences can “work themselves out.”

        Most importantly however, the GCMs are really made to predict one thing: global mean temperature (GMT). The day to day weather (clouds, rain, etc) are very significant in predicting GMT. What is significant is heat transfer items: reflection and absorption of solar radiation, ocean warming/cooling, etc. These processes are not sub grid, and are more well understood than clouds and other weather items.

        So should you trust them? Maybe this is a question for the debatables, but I hope to have shown how they are different from weather models.

    • W. Dean Reply

      Greg Lusk,

      It’s often said that correlation is not causation. The flip side is that causation without correlation needs a lot of explaining. In this case, warming is not always preceded by a rise in CO2. It’s sometimes the reverse. That means the issue comes down to the climate model. And so far as I’m aware, no one has developed a model that predicts past climate without introducing auxiliary hypotheses to keep the model working. Without a model (let alone correlation), therefore, I find it hard to accept a consensus, no matter how widespread, as compelling evidence. That’s why, for me, the case must be judged inconclusive until such time as [a] a model is discovered or [b] we’re overwhelmed by the effects of global warming. Hopefully, you’ll focus on this stuff in the coming installments.

      • Greg Lusk
        Greg Lusk Reply

        I’m not exactly sure what point you are responding to – but regardless – just because the warming of the climate is not always proceeded by a rise in CO2, it does not follow that CO2 does not produce warming.

        Could you please detail what auxiliary hypotheses you know are added?

        You seem like you want certainty from your science, but you just can’t get that from inductive practices until [b] we’re overwhelmed by the effects of global warming. But of course, then it’s too late.

        • W. Dean Reply

          Greg Lusk,

          Sorry, that came off the wrong way. You responded to my post by saying you’d address some of the issues about models later on. I’m fine with that. I only meant to offer some remarks about framing the problem. The key philosophical issue, for me at least, is a model that predicts past climate. I’ve been fairly cautious in my statements in this regard, because it’s been awhile since I looked at the models. But I still haven’t heard that anyone has built a comprehensive one.

          As for those auxiliary hypotheses, it depends on the model. If one takes the simple model, where carbon dioxide is a greenhouse gas, then the amount of CO2 in the atmosphere tells us how much heat is retained by it. The trouble, of course, is that the real world doesn’t behave in a linear fashion. More heat causes more evaporation, causing more water vapor, which increases heat retention; yet clouds also reflect solar radiation. The varying energy from the sun, the rate of uptake of CO2 by the oceans and by plant matter all come into the equation. In this case, one needs knowledge about all these factors, as well as how they will react when one variable changes. Often, this evidence is lacking so auxiliary hypotheses about how they interact are introduced to keep model predicting climate. If the model doesn’t predict the Medieval warm period, for example, an auxiliary hypothesis about one of the other variables is introduced to account for it (e.g., solar activity). So far as I’m aware, climate scientists still have to do this with their models.

          I’m not looking for certainty here. I have no problem, for example, with the (inductive) claim that smoking drastically increases one’s likelihood of developing cancer, even though the explanation for any given case isn’t in yet. But that’s because of the strong correlation between the activity and the disease. I’m just not convinced that the case for global warming is a strong.

          (By the way, it’s only too late to do anything when it’s possible to do something, which is a point I’m also skeptical about.)

          • Jason Hornosty Reply

            Uhh.. that’s not true at all. You’re correct that no models predict past climate, but your understanding is incorrect. Scientists do not introduce auxiliary hypotheses; rather, what they do is tune parameters in the existing model so that the models history matches historic data.

            For instance, the albedo of Arctic ice might be modified between 0.6 – 0.8 based on how the model performs relative to data (these numbers are made up for example; further, usually the parameterization is on more uncertain feedbacks).

          • W. Dean Reply

            Jason H.,

            It’s too bad Quine’s dead, because I think I found a perfect case of interchangeability salva veritate: “auxiliary hypotheses” and “tuning in parameters.” Seriously, auxiliary hypotheses and tuning in parameters are ad hoc adjustments that make a model work. You should have looked it up before proclaiming my remarks were untrue.

            As for your comments on temperature, below, you’ve offered no substantial reply to my initial assertion regarding the graph. The curve drops off after 2000, which spoils the rhetorical effect of that rapidly sloping upward curve. It’s not for nothing that the IPCC included the famous “hockey stick graph” in its report.

  • Chris Hall Reply

    W. Dean,
    I’m failing to smell the same odors here you detect, but perhaps I have an epistemic cold…
    On scent #1, I think your analogy is a bad one. The pharmaceutical industry isn’t voluntarily transparent. Often we find out that there was evidence of harmful side effects that was withheld. More to the point, when a pharmaceutical industry develops a new drug, their chemistry is proprietary information for a period of time, so that they can make use of the market advantage they have earned by conducting the research. This part of the the pharma industry does seem analogous to academic research. Hence there is an incentive for a researcher to be protective of their data (at least for a time). You also seem to overlook the impact that the politicized nature of climate research involves. Many researchers rely on government money to fund their work. This gives them a reason to be cautious about what information they release. Given our current political climate (at least in the U.S.) information is often taken out of context, mis-characterized, etc., all of which might make a researcher gun-shy about throwing open the curtains for all to see.

    On #2, I really don’t get your point about the graph…the focusing on the past century and a half is not an arbitrary decision, or one made simply because it best suites the wishes of climate change scientists. It’s the period that roughly matches the industrial revolution, which marks the beginning of the most significant physical changes human beings have made to the planet. I take it your point is that, if we look on a long enough time-scale our current climate data is but a mere blip. I fail to see how that shows very much…on a long enough time scale, the temperature of the universe is on average just above absolute zero, but that doesn’t mean it’s not warmer in the summer than it is in the winter.

    Okay, off to take some epistemic sinus medicine, I’ll see if that clears things up 😉

    • W. Dean Reply

      Chris Hall,

      On 1. On the contrary, pharmaceutical companies are required to provide the exact chemical structure of their products for approval. The composition a given drug is only proprietary (in the proper sense of the word) in the R&D stages, which is why drug companies need patent protection once the drug is approved. In addition, they’re obliged to pay for independent examination of their findings. All of this adds up to a fairly transparent process, which mirrors the impact of their products on our lives. The response to global warming will affect us as drastically, so a similar transparency should be on offer when it comes to climate science. I don’t think it’s there.

      Second, I don’t see government sponsorship as a justification for keeping one’s data secret. Quite the opposite. The public has a right to see research it’s paid for; and the fact that this particular research has enormous political ramifications only reinforces the imperative of transparency. Besides, all this secrecy betrays the scientific spirit of openness.

      On 2. I appreciate your frankness on this one. I suspected someone might take the bit about the graph the wrong way: if I’m not claiming it’s false, why am I complaining? It’s not what’s in the graph, it’s what’s left out. A graph that represents only a small section of a long history implies that what happened before and after is immaterial. But climate has changed considerably. We’re not shown the warm period that preceded it or the cooling off period that followed 2000. Nor for that matter, are we shown a million year correlation between warming and CO2. Yet the graph implies that it exists in virtue of the fact that it’s omitted. In a nutshell, that’s my beef with these graphs.

      • Jason Hornosty Reply

        Of course, the exact same misrepresentation can be made in text. Take your statement that the mean global temperature has decreased post-2000. Relative to what? It has decreased relative to the year 1998. It has increased relative to the 1990s decadal average.

        • W. Dean Reply

          Jason Hornosty,

          In order to “misrepresent” the facts, I’d actually have to “represent” them. I didn’t. I pointed to omissions in the graph, so obviously there’s no question about what the cooling I mentioned was relative to – it’s relative to the warming depicted in graph, which contains the temperature for 1998.

          • Jason Hornosty Reply

            Except that the graph DOES show the years 1999-2004. Other things you seem to be unaware of:

            * 2005 is the warmest year (in the last 100 years, at least) — warmer than 1998.

            * 2006, 2007, and 2009 all had warmer annual global mean temperatures than the years 1999, 2000, 2001 and 2004.

            In other words, the graph is cut off right before the warmest year on record (in the last hundred years), and before 4 of the top 7 years on record (in the last hundred years). So explain to me in what sense this is a “convenient” cutoff point for the data.

            Quite simply, there is no “post-2000 decline in average temperature” unless you are comparing the 2000-2009 decadal mean to 1998 alone. If this is what you meant, then there IS a need to specify what the cooling you’re referring is relative to, because no scientist would ever make this comparison.

  • Mike Thicke
    Mike Thicke Reply

    When being skeptical I think it’s important to be clear regarding what you are skeptical about. As I understand the situation, there are many reasons to be skeptical of specific claims and specific predictions made by climate scientists. However, there doesn’t appear to be good reason to be skeptical of the overall claim that human-caused warming is occurring, and that it will lead to significant consequences within a pretty short timeframe. On the general claim there was scientific consensus long before global warming became a political issue, so recent political maneuverings on the part of climate scientists doesn’t seem particularly relevant.

    • W. Dean Reply


      I’m skeptical of the claim that [a] global warming is occurring and [b] its anthropogenic cause. The reason I’m skeptical of [a] and [b] is [c] I’m not convinced that there is a scientific consensus and [d] I don’t believe in accepting a consensus without definitive proof. My original post was, as I said, a quick and dirty account of my reasons for [a], [b] and [c].

      • Mike Thicke
        Mike Thicke Reply

        What subjective probability would you assign to claims [a] and [b]?

        • W. Dean Reply

          Mike Thicke,

          As far as climate catastrophes go, I think the smart money’s on an ice age. We’re overdue by about 2000 years.

          • Mike Thicke
            Mike Thicke Reply

            As far as I know, the consensus regarding the ice age theory is that the time scales on ice ages are so large compared to the time scales of human-caused climate change that any cooling due to an impending ice age will be completely swamped by warming due to CO2.

            Anyways, you’re dodging the question!

          • Curtis Forbes
            Curtis Forbes Reply

            I love the idea of unpacking disputes about Climate Science in terms of Bayesianism. I do suspect that it all comes down to a different assessment of a priori probabilities.

          • W. Dean Reply


            Picking one outcome out of a variety of possibilities makes more sense to me than assigning a probability to a particular outcome. For me, the cosmic forces operative for billions of years get more weight than human activity for the last hundred.

      • Bernhard Isopp Reply

        With regards to [c], please see,

        Anderegg, Expert Credibility in Climate Change:

        Anderegg, W.R.L. (2010) Moving Beyond Scientific Agreement. Climatic Change, 101 (3) 331-337.

        Doran PT, Zimmerman MK (2009) Examining the Scientific Consensus on Climate Change. Eos Trans. AGU 90.

        As well as the well-known Oreskes article, published in science, which can be found here:

        (Sorry about the long link).

        Considering these sources, it becomes difficult to maintain that there is no consensus, unless one reworks a general meaning of consensus (what working definition is appropriate? A majority? A substantial majority? An overwhelming majority? The consensus on climate change appears to meet all of these criteria) or takes issue with the above studies. If this is the case, please refer to

        • W. Dean Reply

          Bernhard Isopp,

          Ah yes, the drive-by bibliography…enlightening people everywhere since the birth of the Web…

          If you’d taken the trouble to read the thread, you’d know that the papers you cite are tangential to the point I raised about the consensus. I suggested that it’s wholly unpersuasive because nowhere near 93-7% of climate scientists are in a position to know whether anthropogenic green gases are causing global warming. That’s why it smells less like a “consensus” of experts and more like inner-disciplinary orthodoxy. In other words, it’s not that 93-7% of climate scientists “believe” global warming; it’s that (a likely) 70-80% are in no position (or have no desire) to dispute the far smaller number of their peers who deal with the models and the collective evidence. Hence, the consensus is illusory.

          • Bernhard Isopp Reply

            I read the entire thread, and I was merely responding to a single statement you made, namely, that you didn’t believe there was a consensus. While I doubt that you made more than a drive-by glance at my “drive-by bibliography,” it makes it a little more difficult to maintain such a view. Also, my reasons for providing a short bibliography instead of a long response were done in the interests of brevity and the hopes that you would actually read and critically reflect on the points made in these sources – which in fact address in part your scepticism of the very idea of consensus. Much of what I wanted to say and much of what has been said here in this thread has been addressed ad nauseam in other locations. But, lest I be accused of not contributing my fair share to the nauseum, I think the major problem with your scepticism is as follows.

            You being unconvinced that there is no scientific consensus rests on redefintion of the notion of scientific consensus. You are not sceptical that scientists agree on the basic hypothesis of AGW, but you are sceptical about their very capacity to have this agreement. You suggest that they agree with their more senior colleagues out of fear, or a desire to fit in, or some other social motivation – but not because they are experts qualified to adjudicate on such matters. First, I think this is a fairly dismal view of the expert credibility of climate scientists and a fairly unreaslistic view of the eagerness to agree amongst practicing scientists. And keep in mind that the scientists included in these studies are actively publishing in the field, and thus are regularly engaging with the collective evidence. If your scepticial standards leads you to think that 70-80% of them are incapable of adjudicating on this collective evidence, then I urge you to review the climate science literature and demonstrate this hypothesis more substantially. As it stands, I see no reasons based on what you have written anywhere in this thread, why I should a priori doubt 70-80% (why likely? where do you even get this number?) of the pronouncements made by experts actively contributing to the field of climate science (Are we allowed to believe 20-30% of the most renowned and active climate scientists in the world? If so, I think you need to rethink your scepticism).

            You also mention models, but models pertain to only one dimension of climate change science. However, in your original post it does take a central role in justifying your scepticism, but in your attempts to distance yourself from the actual scientific content of the debate and present your scepticism as philosophically justified, you rush over this issue far too quickly and haphazardly. You seem to suggest that the hypothesis that increasing GHG emissions have been contributing to climate change hinges on the efficacy of a limited number of “models” – if these models are ineffective then we can simply kiss the AGW hypothesis goodbye. Part of the problem is your ambiguous use of the word “model.” You state, “The vast majority of scientists work within the framework of a particular model…If the dominant model sees global warming caused by CO2, then the majority of scientists will obviously say that global warming will occur as the concentration of CO2 increases.” You say this as if the theory that an increase in concentration of CO2 in the atmosphere will cause a heightened greenhouse effect is merely the base assumption of the GCMs. You try to impose a circular reasoning to climate scientists (scientists make a model which assumes that increases in CO2 contribute to increases in global mean temperature, then the model shows that temperatures increase if CO2 increases, so CO2 is the cause of global warming) which is unsubstantiated. Upon wading into the scientific literature it is clear that this assumption is not as arbitrary as you imply. What is required from you is to demonstrate why you think it is so, in the face of the vast literature in atmospheric physics which studies the radiative forcing of GHGs, with particular regards to CO2. The consensus (whatever that means) around the AGW hypothesis is the result of a complex mix of theory, observation, and predictive modelling, which cuts across many disciplinary lines. Furthermore, to simultaneously provide a rationale for allowing your scepticism to avoid actually engaging with the content of climate science, while bolstering your scepticism on an overly-simplified of a technical issue (“last I heard, no one has constructed a model that predicts past climate without introducing ad hoc fixes to account for unknown variables”) seems a little fishy. Again, you make it sound like all of climate science in support of the AGW hypothesis hinges on this solitary issue, the effiacy of “the model” (one might ask what is “the” model that you are referring to?). But this grossly oversimplifies the complexity of climate science.

            But this of course, is a techical scientific issue, and we’re not talking about science, we’re talking about philosophy. So here are my “philosophical” objections. Mostly, as others have mentioned above, I think your scepticism is unrealistic in that it is too stringent of a standard. You seem to expect some sort of objective, indepedently verifiable confirmation of climate change. I think this is summed up pretty well in your statement, “The real question, in other words, is not whether there’s a consensus but whether the model is any good.”

            Here you are basically saying “It doesn’t matter if there’s a consensus (because consensuses are worthless because scietists just blindly go along with the assumptions of the dominant paradigm), what matters is if climate science is really happening.” And how do we know if climate science is really happening? If we have an effective model. Leaving aside the issues that I just went over surrounding making a belief in climate change hinge on a solitary model, yes, you’re right, it would be nice if we have this kind of definitive proof. But as someone commented earlier, science is rarely, if ever, this straightforward. Climate change is worthy of serious attention because of the serverity of some of the possibilities. And this is why your longing for a simplified, absolute logic for scientific confirmation simply will not do. In so far as we see scientific knowledge as suggestive of future action, we cannot simply wait for the fulfillment of an abstract criteria of knowledge which approaches 100% certainty. Furthermore, it is highly doubtful that your criteria can or should provide the basis for a meta-applicable scepticism. As you inadvertently point out, different scientific questions, and different scientific practices will have different criteria and means of scientific confirmation.

            To return the issue of consensus, you state, “The real problem is that the mandate for the [IPCC] summary is to find the consensus among scientists, not to determine the truth about climate change.” This seems to resemble an individualistic, verificationist notion scientific truth. As I already mentioned, you imply that a consensus and truth are not the same thing. But this notion of truth leaves us laypeople in a very difficult position when it comes to science. You and I are simply not going to verify every fact presented to us by science (or any other way of knowing for that matter). Our knowledge systems depend on us trusting the knowledge of others. Does every doctor “know” that smoking causes cancer? Do you “know” that smoking causes cancer? Probably not by the standards you present here. I don’t know about you, but I prefer not to smoke, because I feel justified in the belief that smoking causes cancer. Why? Not because I’ve perused the vast medical and scientific literature establishing tobacco smoke as a carcinogen, but because I trust the expertise of those doctors and scientists who have come to a consensus on the issue (even if we are assuming that 70-80% of all scientists are blind-followers of some received wisdom). Without this trust in consensus, I think you would find yourself justified in very, very, few beliefs. I do not think your distinction between “the truth” about climate change and the scientific consensus on climate change is very tenable, nor does it have the consequences that you seem to imply. What exactly is the distinction you are trying to make? If we are not going to base our beliefs about climate change on the consensus of climate scientists, then who? Certainly not on your own scientific knowledge of climate change. Are you going to pick certain scientists to believe? Would you pick scientists that go against the consensus? By what criteria would you do so? In short, I think you deeply underestimate the importance and need for consensus in science, and your attempts to generally dismiss scientific consensus as unimportant to justifying beliefs is unconvincing. So, if we are not going to dismiss the idea of scientific consensus generally speaking on philosophical grounds, you need to show specifically why the consensus of climate scientists is somehow a uniquely problematic scientific consensus. You try to do this by referring to specific issues with climate science, but this needs to be born out with much more detail to specifics of climate science, and of course, this would take you into the scientific arena, but you state is not even necessary to justify your scepticism. Otherwise, you would need to show why climate scientists in particular are bigger dupes than most other scientists, but this seems fairly unlikely.

            By the way, I don’t think my mere suggestion that you read a few articles warranted a sarcastic and condescending response.

          • W. Dean Reply

            Bernhard Isopp,

            1. Scientific consensus and “straightforward” science

            One could argue that there is a “consensus” among physicists that F=ma (i.e., Newton’s second law of motion). But you rarely find the authors of physics textbooks appealing to a “consensus” among physicists. Instead, you’re provided with an experiment that confirms it. The same can be said of most of the content of chemistry, biology, geology, etc., namely, that what matters is not a consensus of experts, but facts gathered and explanatory or predictive success. So, contrary to your assertion, consensus is irrelevant for most scientific knowledge. For that matter, we laypeople can do much to satisfy ourselves that a given theory is true in the basement or the backyard.

            Theoretical physics, of course, is more nebulous stuff. Multiple theories have been proposed for unifying physics. But suppose, in a break with the conventional practice, that a group of physicists claimed that “string theory” was “probably true,” because a study showed that 90-95% of physicists believed in string theory. How persuasive could this be? We know that only a small subset of physicists are theoretical physicists, so nowhere near that number of physicists are in a position to know. Even if physicists are in a better position than we are to judge the merits of string theory, all we could really conclude from the consensus is that string theory was orthodoxy in physics.

            Nor do we have to explain this kind of consensus as fear or malfeasance. It may reflect the success of string theory advocates in making their case to their peers. The same must go for climate science. As for the numbers I suggested (70-80%), it may be more or less. But it is a fact that climate scientists are specialized in various areas, and, like any other science or academic discipline, few actually deal with all the data and the models.

            2. As the stakes rise, so does the burden of proof

            You mentioned that you believe your doctor when he says that smoking causes cancer, so why not believe climate scientists? Notice that the “opportunity cost” of believing that smoking causes cancer is pretty low for you, assuming you don’t smoke and you don’t sell cigarettes. But suppose your doctor looks at you as you stroll in for a regular checkup and says, “You have cancer and I’m scheduling you for chemotherapy beginning tomorrow.” I have a feeling that you wouldn’t be so quick to accept his authority. You’d want something stronger than his confident assertion, even if, at the end of the day, you don’t have the expertise to judge yourself.

            The same goes for climate science. It’s easy to accept a consensus opinion instead of a prediction based on a comprehensive theory when there’s little at stake (e.g., as in the case of string theory and smoking). But when the economic and political ramifications are as drastic as they are in the case of climate change mitigation (not to mention the profit to be made and lost), it’s incredible than anyone would expect a consensus among a small group of scientists to be enough.

            3. Climate models

            Climate scientists use a number of discrete models covering various aspects of climate phenomena (e.g., CO2 as a GHG), because there’s no unified theory of climate change that predicts past climate and (thus) will predict future climate. This is common knowledge, not a “fishy” supposition of mine that relies on an ambiguous notion of model.

            4. Note on etiquette

            I guess you don’t think it’s presumptuous to direct someone to a paper without argument. Let me suggest an experiment to see who’s right. The next time you overhear two people debating something, interrupt the conversation and tell them both to “Go read books x,y,z.” Let me know how it turns out.

            • Bernhard Isopp Reply

              Yes, we obviously have different conceptions of the rules of etiquette that apply to public discussions that take place in the comment sections of blogs.

  • Jason Hornosty Reply

    If W. Dean returns, I’d like him to tell me which data he wants to access that he can’t. For those interested, you can access pretty much everything imaginable through links at:

Leave a Reply

Your email address will not be published. Required fields are marked *