Wednesday, October 27, 2004

A phony hockey stick?

Willie Soon of Harvard University, one of the global climate experts among the astrophysicists, informed me about their new article

They describe a new paper by Von Storch et al. that was published in a recent (September 30th) issue of Science. The German journal Der Spiegel (The Mirror) made an interview with Von Storch on October 4th, 2004. Von Storch said something along these lines:

  • The assumptions that have led Mann, Bradley, and Hughes (MBH) to their "hockey stick graph" in 1998 and 1999 are unacceptable. Their method is wrong - in fact, it's rubbish.

That's a blunt formulation and one is naturally cautious about it, but it seems that this formulation is justified. It may be useful to say what the "hockey stick graph" means. Try to draw the graph of the "global temperature" as a function of the date - between the years 1000 and 2004. What graph will you choose if your results are supposed to support the statement that the world economy and our breathing should be slowed down because these processes create too much carbon dioxide which leads to an unacceptable and unprecedented global warming?

Well, of course, you will draw a constant function between the years 1000 and 1900 (long, stable shaft) and a quickly increasing temperature between 1900 and 2000 (it's the blade of the hockey stick). Such a profile will prove that the changes are caused by the humans, and they are undisputable and potentially disastrous. That's roughly what MBH did, and they became famous for their result. Their graph was reproduced by the global climate committee of the United Nations (IPCC, Intergovernmental Panel on Climate Change) and it became one of the justifications of the process that culminated with the Kyoto protocols.

Of course, MBH also had a scientifically impressive method based on the principal component analysis. It was really the paper by Ross McKitrick and Stephen McIntyre - the so-called "audit of MBH" - that helped me to understand the methodology of MBH a bit better. I've already written an article in which I stated my opinion that this method may be a good one to isolate a picture that is covered by dust, but in order to estimate the past temperatures from shaky proxy data, it may be better to simply take the average of all the known data - the same average that the kids learn when they're 8 years old. It seems that this simple opinion is becoming more influential.

Von Storch et al. showed that the method of MBH is suspicious, using a kind of sensitivity test. In fact, I would say that MBH themselves should have done such a sensitivity test before they published their paper. How did Von Storch et al. do it? They simulated an artificial world, and extracted the fictious "proxy data" (the widths of tree rings etc.) from this model, and added a reasonable amount of random noise. In their virtual world, they knew perfectly what the "true temperatures" should be. Afterwards, they inserted the "proxy data" to the formula of MBH, and used that formula to reconstruct the temperatures. The result was very different from the "true temperatures". Quite generally, the method of MBH seems to reduce the temperature fluctuations in the past (before the thermometers were known).

What does it mean? Well, most honestly, it means that we know very little about the historical temperature record. However, if the fluctuations were greater than what MBH argued (by a factor of two or more), then the temperature growth in the 20th century is not so exceptional. People used to think that there was a medieval warm period (MWP), followed by a little ice age, and these old theories may stage a comeback.

The importance of the carbon dioxide

I've never understood the hysteria behind the concentration of CO2 in the atmosphere. CO2 is an important gas - it's produced when we breath, and it's necessary for photosynthesis. Animals produce it, plants consume it. The existence of CO2 proves that there are either animals, people, or machines around, and the growth of the concentrations of CO2 is known to be correlated with the growth of the gross domestic product (GDP). Well, CO2 may be used as one of the indicators of life and working economy.

Is it a poison? Well, I don't think so. CO2 is not quite the same thing as Mercury, for example. ;-) Should we really worry if the human civilization doubles the concentration of CO2 in the atmosphere? Today, the concentration is below 400 ppm ("parts per million"), which means less than 0.04 percent. The concentration of CO2 in your bedroom is probably much higher. Does it really matter whether the concentration in the atmosphere is 0.03 or 0.05 percent? Should we pay trillions of dollars just to avoid this increase, even if we don't know any clear negative consequences of such a growth?

Well, CO2 is known to protect the Earth from cooling, it's a greenhouse gas. Well, it's not the only greenhouse gas, and on the other hand, there are also other compounds in the atmosphere that tend to reduce the temperature (aerosoles, for example). We may ask a simple question: Is it better to heat the planet up, or to cool it down? Imagine that an observable called "the global happiness" is a function of the "global temperature". An interesting question is what's the derivative of "happiness" evaluated at the current temperature. If it's zero, than up to the first order, it does not matter whether the temperature goes up or down. If it's nonzero, there is 50% probability that it is negative, and 50% probability that it's positive. The expectation value of Delta(happiness) is zero. But of course the expectation value of "Delta(happiness) squared" is positive, and it grows with "Delta(temperature) squared". Do you think it is reasonable to pay 1 percent of your GDP for the chance that Delta(happiness) squared is likely to be a bit smaller, even though you can't predict the sign of Delta(happiness)?

I think that such a decision is an economic stupidity, especially because I tend to think that the derivative of the happiness with respect to the temperature is positive. ;-)

There are many other things whose concentration increases drastically as a result of human activity. The Aliens who observe our Solar system in the microwave range of the electromagnetic spectrum see a pretty bright star - the Earth. Humans have obviously increased the concentration of radio waves and microwaves emitted by the Earth by several orders of magnitude. Should we be worried about the unpredictable consequences of this unprecedented growth of the concentration of something? I don't think so. Experimentally, it seems pretty clear that the microwaves don't imply a disaster. We're not gonna ban radios, TVs, and cell phones because of some uknown hypothetical threats. Is the situation of the carbon dioxide really so different?

Or is the war against the carbon dioxide a simple manifestation of anti-civilization and anti-American emotions?


  1. It certainly does not negate the threat of a long-term global temperature increase. In fact, McIntyre and McKitrick are careful to point out that it is hard to draw conclusions from these data, even with their corrections. Did medieval global warming take place? Last month the consensus was that it did not; now the correct answer is that nobody really knows. Uncovering errors in the Mann analysis doesn?t settle the debate; it just reopens it. We now know less about the history of climate, and its natural fluctuations over century-scale time frames, than we thought we knew.

    If you are concerned about global warming (as I am) and think that human-created carbon dioxide may contribute (as I do), then you still should agree that we are much better off having broken the hockey stick. Misinformation can do real harm, because it distorts predictions. Suppose, for example, that future measurements in the years 2005-2015 show a clear and distinct global cooling trend. (It could happen.) If we mistakenly took the hockey stick seriously--that is, if we believed that natural fluctuations in climate are small--then we might conclude (mistakenly) that the cooling could not be just a random fluctuation on top of a long-term warming trend, since according to the hockey stick, such fluctuations are negligible. And that might lead in turn to the mistaken conclusion that global warming predictions are a lot of hooey. If, on the other hand, we reject the hockey stick, and recognize that natural fluctuations can be large, then we will not be misled by a few years of random cooling.

    A phony hockey stick is more dangerous than a broken one--if we know it is broken. It is our responsibility as scientists to look at the data in an unbiased way, and draw whatever conclusions follow. When we discover a mistake, we admit it, learn from it, and perhaps discover once again the value of caution.

    (This is from the second URL that LM posted.)

  2. Hi Lubos,

    You say:
    Does it really matter whether the concentration in the atmosphere is 0.03 or 0.05 percent?

    No, probably not, and it is likely that on geologic timescales the atmospheric concentration has varied far more than this. But as a physicist you should understand exponential growth! If we can double the concentration in 100 years, then we can raise it from 0.05 percent to 100 percent in a time which is INFINITESIMAL on geologic time scales.

  3. Hi,

    One minor comment to add about my previous "exponential growth" analysis.

    That assumes, of course, that there is no negative feedback mechanism which checks the growth.

    But of course there is such a feedback mechanism which kills the growth of CO2 concentration well before it reaches 100%: the extinction of humanity.

  4. Well, you "calculated" that we will raise the level of CO2 in the atmosphere to 200 percent very soon. That's a great calculation, but we, physicists, not only know the exponential function (that grows) and even more growing functions (such as the generalized factorial), but we can also distinguish craziness from reality. I hope that you were joking. Even if some quasireasonable calculation concluded that the society will develop in such a way that we will multiply CO2 by a factor of 100 during thousands of years, I don't think that such a calculation should affect any policies today. It's just crazy. We don't know what will happen in 15 years, and planning "geological time scales" is insane.

  5. No, of course I don't advocate planning on "geological time scales." That was the point of my first post: that the changes we are making are happening much faster than geologic time scales, by several orders of magnitude.

    The Earth does have various known feedback mechanisms which operate on geologic time scales.

    However, at the rate we are going, its unclear whether the Earth can cope. That's why I pointed out one known fast feedback mechanism: human self-annihilation.

  6. But, if I may be permitted a fourth post :)

    It's hard to say WHY one should worry about these things. It is very unlikely that intelligent life on this planet will survive on geologic timescales, anyway (and it is even more unlikely that intelligent life will evolve in a sustainable way beyond this planet).

    I guess sometimes my romantic half dominates over my rational half, and I hope that maybe intelligent life has some purpose in this universe, which we might discover if we manage to survive long enough :)

  7. Now comes the real shocker. This improper normalization procedure tends to emphasize any data that do have the hockey stick shape, and to suppress all data that do not. To demonstrate this effect, McIntyre and McKitrick created some meaningless test data that had, on average, no trends. This method of generating random data is called “Monte Carlo” analysis, after the famous casino, and it is widely used in statistical analysis to test procedures. When McIntyre and McKitrick fed these random data into the Mann procedure, out popped a hockey stick shape! this method, of the Monte Carlo be suspect, in quantum gravity models?


  8. What we do not know is whether there is a positive feedback mechanism which could cause our small perturbations of the environment to cascade into something big.

    "The evidence also shows that Earth’s climate system has sensitive thresholds. Pushed past a threshold, the system can jump quickly from one stable operating mode to a completely different one—“just as the slowly increasing pressure of a finger eventually flips a switch and turns on a light,” the NAS report said.

    Scientists have so far identified only one viable mechanism to induce large, global, abrupt climate changes: a swift reorganization of the ocean currents circulating around the earth. These currents, collectively known as the Ocean Conveyor, distribute vast quantities of heat around our planet, and thus play a fundamental role in governing Earth’s climate."

  9. GRACE, twin satellites launched in March 2002, are making detailed measurements of Earth's gravity field which will lead to discoveries about gravity and Earth's natural systems. These discoveries could have far-reaching benefits to society and the world's population.

    Any thoughts from the previous poster on Gravity Recovery and Climate Experiment?

  10. I am also romantic about the long future of the humankind that should occupy the galaxy and so forth. But now I believe that the right approach to achieve this goal is to look how to survive in the next 10 or 20 years, and be sure that our children are thinking about the same goal and they teach it their children, including the rule that these new children should continue to teach it. ;-)

  11. "We may ask a simple question: Is it better to heat the planet up, or to cool it down? Imagine that an observable called "the global happiness" is a function of the "global temperature"."

    Actually, thermal expansion of ocean will have a detrimental effect: raise the "sea level" reducing land mass. At least the "global happiness" function is not a monotonically increasing function of "global temperature":)

  12. From

    "Thermal expansion: Most of the expected sea level rise due to climate change will be from physical thermal expansion of Earth’s oceans. When water temperature increases, its density decreases. Density in this case is equal to mass of the ocean divided by the ocean’s volume. Since the mass can’t change, the volume must increase; the ocean expands. Scientifically, it is the best-supported effect of global warming and climate change, with widespread observations already revealing the 4 to 10 inch rise in sea levels during the last century. IPCC projects a rise of 10 to 20 inches during the next century."

    "A 2.7° F to 8.1° F (1.5° C to 4.5° C) window of temperature rise is expected to raise sea levels from 8 to 34 inches. If the worst-case estimate was realized, major coastal cities in the U.S., such as New York, Miami, and New Orleans, could be inundated. Foreign coastal cities and the world's islands would face a similar fate."

  13. Concerning the sea level. A couple of inches is a completely irrelevant change, and I am sure that virtually everyone understands it. If the sea level increases by a centimeter per year, the owners of houses closely adjacent to the ocean may pay 1 percent of their income to protect themselves against the increasing sea level. But we are talking about the 1 percent slowdown of the *whole* economy, which seems like a bigger price to pay by several orders of magnitude.

    Another example why a rational person cannot be impressed by "20 inches per century" change of the sea level even if it were true: 60 million years ago, the sea level was 100 meters higher than today, for example.

  14. "Another example why a rational person cannot be impressed by "20 inches per century" change of the sea level even if it were true: 60 million years ago, the sea level was 100 meters higher than today, for example."

    Yes, but those are not ideal for human existence, were they? I am sure lots of species will thrive under such conditions, just not humans. And the earth and the universe will go on, whether or not humans survive another 1000 years, just as the extinction of the dinasours did not lead to end of life.

    I would be more comforted if any dramatic change in sea level had happened in the past 10,000 years or so. That is, how much has the sea level has changed in the past 10,000 years, and has there been any noticeable acceleration in the past couple of hundred of years? Maybe you have figures on that.

    Where do you get the 1% figure from? I am sure it is as reliable as global warming forecasts which you have no problems easily dismissing. In any case, a lot of the current fossil fuel industries gets lots subsidies from the government, which also costs taxpayers (not to mention health and other costs). If those are transferred to cleaner industries, the cost should be significantly less. If gradually done, it should be relatively painless. After all, the world economy is still thriving even though cars today are a lot cleaner and more fuel efficient (mandated by law) than they were a few decades ago.

  15. "Or is the war against the carbon dioxide a simple manifestation of anti-civilization and anti-American emotions?"

    If eating greasy burgers is proven to be bad for health, and it turns out Americans eat the most greasy burgers per capita per year, then pointing out the health risks of greasy burgers would be neither anti-American nor anti-civilization.

  16. Some professional economy papers are talking about 1-4 percent shock for the GDP, see e.g.

    You know, the Kyoto protocol is a rather well-defined set of rules and constraints, and it is much easier to predict its impact than the impact of the CO2 concentrations. If you must close a company, the economists simply know what it means for the output etc.

    There are thousands of pages that talk about 1 percent. I have not seen a single page that would indicate that the economic impact would be positive, or much less negative than 1 percent per year.

    Also, if the humans were tossed to the world as it existed 60 million years ago, they would probably survive easily. I don't see anything dangerous about the sea level moving by 100 meters in millions of years. You know, half of Boston stands on landfills that have already been produced centuries ago, and I am sure that people, if they survive, will be even more skillful in millions of years from now. Also, my guess is that the human species is one of biggest "survivors" among the existing "large" species of animals.

  17. Thanks for the link!

    However, as the abstract mentions, that calculation is based on not allowing emissions trading.

    "If the United States is not able to take advantage of "where" flexibility (reducing emissions wherever it is cheapest globally) through international emissions trading to meet the Kyoto target, the cost in terms of lost output will range from about 1 percent to over 4 percent of GDP."

    However, emissions trading is a part of Kyoto protocol.

    "Ironically, the EU is about to launch the world's first international carbon-trading system. America, which pioneered such trading, is left out. Canada, Japan and maybe even cheeky California will soon join Kyoto-land (see article)."

    True, humans may 'survive' hard climate, but that is not the same as living 'well'. Further, as evolutionary biologist Mayr noted "at least judging by biological success: beetles and bacteria, for example, are vastly more successful than humans in terms of survival."

  18. You don't like my post or the part about the children?

    lubos:But now I believe that the right approach to achieve this goal is to look how to survive in the next 10 or 20 years, and be sure that our children are thinking about the same goal and they teach it their children, including the rule that these new children should continue to teach it.Lubos, then we must speak to the public as though we are speaking to one's eight year old grandson?:)

    You can imagine that a child's mind can be so trouble by anomalies in nature, that like Einstein and his compass given as a gift to him at a early age, that such features are more then attractive to the enduring wonders of why.

    That they become the fuel for further forays of the intellectual world. Shadows cast and geometries that develope, arise from fertile grounds of the imagination, faced with those same anomalies?

    When all the foundations of thought about gravity are one day shown to be violated, then indeed what physics has gone on unnoticed? Is that day here soon? Liking the toys that we do, superconductors floating, has made this realizatin a step closer and holds in awe the children for what we as adult parents, had always taken for granted.

    Even as fathers , we are mothers, mothers as fathers. A harmonical balance?

  19. In my case, the "war against carbon dioxide" is motivated by fear. Here's a informal list of putative recent "greenhouse symptoms", drawn from a book, "Tomorrow Now", by Bruce Sterling:

    Eastern USA, July heat wave 1999. 250 dead. Record 119 degree Fahrenheit temperature in Chicago.

    Texas, record downpours, 1998. $1 billion damage and 31 dead.

    Florida, 1998, worst wildfires in 50 years.

    Mediterranean, intense drought and fires, 1990s. 1.2 million acres of Spanish forest burnt in 1994, .5 million in Greece in 1998, .5 million in Italy in 1998.

    Indonesia, 1998. Fires across "up to" 2 million acres.

    Eastern USA, driest growing season on record, 1999. April-July 1999 was driest in a century of recordkeeping in four north-eastern states. Agricultural disaster areas declared in 15 states.

    Wstern USA, worst fire season in 50 years, 2000. 84 large fires spread across 14 states.

    Mexico, worst fire season ever, 1998. 1.25 million acres burned.

    He also has a list of 14 glaciers around the world, all in rapid retreat during recent decades.

    That is just an informal survey. I have started keeping my own notes from the international news, and could easily make a different list just as long.

    At what point does one stop trying to explain it away, and start thinking seriously that this might be anthropogenic climate change in action? How many more meteorological records have to be broken, how many millions of acres of forest have to burn down, how many heatwaves, brownouts, floods, hurricanes does it take? At what point does epistemic caution simply become denial, hiding behind the complexities of atmospheric physics? Maybe it's an observer effect, maybe it's the solar maximum, maybe it's population spillover or bad forestry practices or... Anything but the greenhouse effect, apparently.

  20. The biggest fear is fear itself, or how the well-known quote goes.

    Some of your entries are amusing. "Biggest Florida's wildfires in last 50 years in 1998". You can be sure that there should be a biggest fire in the last 50 years in one state of the USA approximately every year - not just every 6 years - because there are 50 states in the USA.

    Your approach lacks statistical framework, it lacks direct and provable links to the subject we discuss, and it is a random set of anecdotal examples that are not enough to imply any statements about the role of carbon dioxide as long as we stay rational.

  21. I know you need to calibrate against a background before you can say you've got a signal. For me these anecdotal lists are a first step, a checklist of phenomena whose ostensible significance might then be investigated more rigorously. So, American forest fires. Do we have any *aggregated* data on trends here? Here's a start, from the national forest service:

    "Given favorable weather conditions, forest structure, fuel overload, and other factors, the number of wildland fires has been growing, getting larger, and gaining in intensity."

    They tell us we have a trend. What are the figures? They say 4 million acres is the annual average burned, with figures since 2000 being: 8.4m acres in 2000, 3.6m acres in 2001, 7.2m acres in 2002, 'nearly 4m' acres in 2003. So far, the 21st-century average is way above 'the' average, however that was calculated.

    Unfortunately I have yet to find a single authoritative source of such figures on the global scale, for the variety of phenomena that I want to be able to look at. This would be the 'Particle Data Book' of meteorology. It must exist, I just don't know where to find it. For now, to find the analyses I want, I'd have to look through half a dozen different technical literatures.