Yesterday we received a mail about a special seminar organized by The Union of Concerned Scientists (UCS). Many great scientists, including the Nobel prize winners (and also some string theorists), are active members. This organization is a very clear example of a strong anti-Bush sentiment in the significant part of the current scientific community.
The main document produced by UCS was called
"Scientific Integrity in Policymaking: An Investigation into the Bush Administration's Misuse of Science"
I was not too impressed by this document. Well, maybe it's because I did not read it carefully enough. But the findings described in this text - those that I have read - either seemed relatively unimportant - like the complaints that a title of an article was changed from "summer hotter" in the draft to the final "hotter summer", or something along these lines - or they seemed biased.
My feeling was that a group of people who dislike Bush has simply put together criticisms or anecdotes invented or experienced by all of them, and they converted this conglomerate of random stuff into an important document. But I admit that I was not able to read every sentence of the full report carefully because it looked like a waste of time.
I think that a recent article in the New York Times was much more balanced. The main hero of that article is John Marburger, a Democrat, a physicist, and the president's science advisor. I found his comments reasonable. He explains that many scientists simply feel bitter because they think that their voices should be much more important for policymaking than they are.
He also realizes that science can teach you about some causal relations between different events and phenomena, and it can impose limits on what's possible, but it cannot really tell us what we should do. It's a matter of moral and political values how we decide to behave, given the laws of Nature as we know them - and democratic countries are those in which these decisions are finally made by the people, not by self-declared know-it-all wise men. These political decisions therefore should reflect the results of the elections, and this is a fact that many scientists are unable to accept.
Now it's hard to believe that in the modern world and the world of internet, it is possible to hide an important piece of information completely. I don't believe in all these amazing conspiratory theories that the whole government is controlling all of us and prevents everyone from seeing the truth. In fact, the amount of sensational insights flowing to my mailboxes is larger than what should be appropriate.
I agree with the authors of the text that scientific integrity is important for scientists not only when they actually do research, but also when they're hired as policymakers' advisers. However, I see very different threats than they do. The "twisting" that the Bush administration is accused from simply does not seem serious enough to write another paragraph about it. Consequently, let's focus on the real threat as I see it. Although UCS tries to address virtually all questions in the world (including commercials for condoms), one of the most important goals of UCS are associated with the so-called global warming theory, and let's switch to this topic.
A certain powerful, politically organized - and financially and politically motivated - group of scientists (not only UCS) would like to call for "scientific consensus" about the global warming issues. They want all other scientists to confirm that all following points are true:
- their research has rigorously proved that the global temperature grows much more rapidly than in previous thousand(s) of years
- it has been proved that most of this increase is due to human activity
- the research has established that most of the increase is caused by the human-produced carbon dioxide
- science can show that this increase is the cause for the hurricanes and all possible "undesirable" climatic phenomena
- it is known that the future increases of temperature will be drastical and disastrous unless we dramatically change the way how the world economy works
- it has been established that the increase of temperature is even worse than the decrease of temperature, and the animals and plants are unable to accomodate
- the very notion of "global temperature" is problematic. Temperature depends on position, and the best observable we may define is the average temperature. Obviously, we don't want to include the whole mass of the planet into this average because the interior of the Earth has almost nothing to do with the climate. We want some sort of "average temperature of the atmosphere". However, the measurements are usually surface measurements, mostly from populated and industrial areas, and therefore these results are more or less guaranteed to be dominated by the "surface hotspots", and the conclusions about the higher layers of the atmosphere - which are really the relevant ones - and not too justified
- the very idea of global warming is relatively new; in the 1970s, "global cooling" used to be a much more popular catastrophic scenario; other factors (aerosoles) are known to support cooling of the atmosphere
- there are simply too many factors that must be taken into account, and focusing on carbon dioxide only is not enough to get reliable total results
- the historical temperature record of the Earth is a very controversial topic. It is not really known whether the Middle Ages were warmer than the present and I will discuss some stories later
- it is unknown whether the recent increase of temperature is a statistical fluke, or a consequence of human activity
- the mathematical models used to derive the conclusions about the climate have too much arbitrariness (and free parameters) in them, and the models from a similar class can virtually lead to any conclusion we want. The validity of these models is a conjecture without any real justification. We know very well that people are not really able to predict weather for the next week, and multi-century predictions seem like a piece of fantasy
- the link between global warming and a particular hurricane is a speculation. Hurricanes are phenomena affected by a large number of factors, and therefore the link - even if it exists - is statistical at best. It is often the case that the global warming is blamed for increases as well as decreases of temperature; for too big fluctuations as well as too small fluctuations, and so forth. I don't think that this is science
- it is far from clear whether the increase of the temperature, even if we assume that it is true, is bad news. Obviously, some species etc. will benefit from it, and the first obvious intuitive answer is that a higher temperature is better for life
- Nature is able to self-regulate and accomodate to new conditions. The Universe has been doing it since the Big Bang, and life has been doing the same thing for billions of years. It just does not sound good if someone says that such an increase would be a disaster for life, even though the increase per year is smaller than the annual fluctuations by one or several orders of magnitude
- let me not continue with these simple statements; anyone who is able to distinguish reality from "The Day After Tomorrow" (by the way, a horrible piece of propaganda, scientific dishonesty, and stupidity that UCS have no problems with) knows that these environmental emotions have usually little to do with the real laws that govern the climate
Let me start with a not-quite-scientific prejudice. Yes, I believe that e.g. the 13th century was warmer than the 20th century. Finnland is really called this way because it used to be a "wine land", much like Greenland used to be a "green land". There used to be a lot of wine in Prague, too. There are scientific papers that support the idea that this "Medieval Warm Period" was global in character.
The paradigm that the 20th century is unusually hot is mainly supported by a paper by Mann et al. from 1998. The main result of this paper is the "hockey stick" - a temperature graph that shows a nearly constant temperature for 600 years and a sharp growth in the 20th century. A year ago or so, I spent tens of hours with reading this paper, and papers that disagreed with it - namely papers by Willie Soon and Sallie Baliunas from Harvard, and by Stephen McIntyre and Ross McKitrick (MM) from Canada.
There have been just too many good contributions in these papers. Although it turned out that there were also some errors in MM, for example, it is clear that MM corrected many problems of Mann et al. - and some of these problems could almost be called "scientific fraud". Don't get me wrong: I would have problems to believe some other papers written by one of the Mc's - for example the paper that proposed that the best formula to calculate the temperature is on the GDP growth and the membership in the Soviet Union, among other factors - but otherwise I was really impressed by their knowledge of math, advanced statistics, and especially by their obviously unbiased attitude to the problem.
I could say similar things about Soon and Baliunas. But of course, I would need to study all of their papers for my statement to be fully justified.
Finally, my impression is that there are some problems with the MM paper, too, and fixing them would lead to a graph not too different from the graph in Mann et al. But MM used the same methodology as Mann et al. and the application of this methodology in this context just seemed rather crazy to me. It is based on the "principal component analysis" which is a method to extract "pattern" from a noise. Imagine a picture covered by dust that makes the picture almost invisible. The principal component analysis can help you.
The application of this sophisticated method to the calculation of the past temperatures seems to assume that there is some nice "pattern" that must be highlighted and isolated. I am not getting it. If we calculate the average temperature, we must treat all reliable data as equally serious, otherwise the results are biased. I don't know why the data should be dividable to "pattern" and "noise". Which part of the data is the "pattern" we're looking for? I think that the temperature graph is obviously a rather chaotic curve, and it is really the noise component of the data that we should extract and calculate because the "pattern" is the unreliable piece.
Moreover, at some points I had a feeling that the method guaranteed that the results for the temperature record would be much more constant than the temperatures in reality - sometimes the method even did not seem to distinguish "warm weather" from "the temperature growth", i.e. a function from its first derivative.
I am sure that some of my statements may sound stupid to the people who have spent much more time with similar statistical analyses. But I still think that most other scientists outside climate science have no idea whatsoever what sort of science supports these statements about the "unusual 20th century", and therefore they cannot be able to support these statements by their authority without crippling their scientific integrity.
OK, so my feeling was that Soon, Baliunas, McIntyre, and McKitrick were doing the type of science that I more or less understand. It's science that can sometimes be in error but whose methods are pretty transparent, they are correctly applied, and we don't assume anything about the result.
I wanted to hear the other side, too - simply because these four people might have manipulated with me, or something like that. So I asked Mann, but also a rather well-known "leader" of the global warming "alarmists". Let me call him SHS from Stanford.
SHS has received an e-mail from a person who is obviously interested in the issue and who has read a couple of papers and who knows something about science and math. Moreover, the person seemed to be open-minded, but inclined to the "sceptics". Would you expect anything else than a reply with a lot of hard data and references to hard work, supporting the "alarmist" viewpoint?
The actual reply from SHS has surprised me, shocked me, terrified me. It was a rather long e-mail, but it contained nothing else than personal insults against the four "sceptics" listed above. The only argument he offered against the MM paper was his opinion that it should not have been published, because it was not refereed by [his friends], but he seemed to have no idea what arguments could be written in such a referee's report. Such a reply from SHS was really surprising because at that moment, I would already have been able to present better arguments against MM than SHS did - at least some minor problems.
Mann's answer was not terribly interesting either, but it was the answer from SHS that I was overwhelmed with. SHS and his friends looked like the Church that used to control everything, and whoever disagreed, had to be destroyed. (SHS is a new-born believer, in a sense, who used to promote the theory of global cooling.) It just seems to me that they have the power to reject any article - a theory or a treatment of data - that they don't like, and the final results are therefore ineviably biased.
My confidence in these global warming scientists and their scientific integrity is tiny at this point. The whole field should either be supplemented by fresh, unbiased, new people, or its funding should be reduced.
Incidentally, the name "Union of Concerned Scientists" sounds ridiculous to me, especially if I try to translate it to Czech. Any translation I've tried so far looks like a cliche from various jokes about the people who try to flatter the Communist Party and to satisfy its best definitions of "moral rules" for a communist.