Recently in the blogosphere there has been a renewal of posts claiming that anthropogenic global warming is not real. Paul Hudson, climate correspondent with the BBC posted a blog in which he claims that there has been no global warming in the past 11 years. The Wall Street Journal and the San Francisco Chronicle took that blog and added the idea that there is now no consensus in the scientific community and that global warming has stopped.
The BBC, Wall Street Journal, and San Francisco Chronicle are normally reliable sources. But they fucked up royally on this one. If we assume their motives are honorable, then the only explanation for the screw-up I can see is a severe misunderstanding of science ... especially science dealing with complex systems like global climate.
First, what do you think is the most important evidence for anthropogenic global warming? Do you think it is reconstruction of paleoclimates which show a steep uprise in global temperatures ... the so-called "hockey stick" graph?
Do you think it is glacier melt?
Do you think it is Arctic sea ice loss?
Do you think it is the threat to polar bears?
While all of these do show the effects of global warming, they are not the data that has convinced the vast majority of scientists that anthropogenic global warming is a real problem and a serious threat that we are swiftly running out of time to do something about. These data sets are valuable but they are ancillary to central data.
However, climate change denialists have focused their attention on these pieces of data criticizing them in unfair manner.
We only have temperature records for slightly over 150 years. The hockey stick recreations of paleoclimates go back much further than that. Thus, the hockey stick graph depends upon proxy data. By that I mean scientists don't use a direct temperature measurement. Instead they measure things that are affected by temperature and infer the temperature from that. This process is going to introduce errors of measurement. In order to minimize this error a multitude of proxies are used. This does a good job. But if you look at any single proxy you can find problems in it. Some proxies will not exactly coincide with others. These are weaknesses in the data. What scientists count on is that each set of proxies has different weaknesses and individual weaknesses will average out. So even though each proxy is going to have its own idiosyncratic problems, by combining numerous proxies we can come up with a pretty good idea of what paleoclimates were like. However, critics can (and have) used the fact that each set of proxies has weaknesses to imply that none of the data is worthwhile ... ignoring that the averaging process of combining the data sets minimizes the weaknesses.
There are glaciers all over the world. While the vast majority are shrinking, not all are. Some glaciers happen to reside in areas where global warming has caused increased precipitation. Some of that precipitation has caused some glaciers to grow. Critics point to these as though they disproved global warming ... ignoring that the vast majority of glaciers are indeed shrinking.
Arctic sea ice loss is very distressing. It is occurring at a much faster rate than predicted by even the worst case scenarios of global warming projections. Critics have used this to cast doubt on modeling. Also the Arctic Ocean is a fixed area in the far north. Normally during the winter almost all of it freezes in the winter. Global warming would have to have a near sterilizing effect to make winter Arctic temperatures incapable of forming ice. We expect winter ice coverage to be about the same. The problem with winter ice is that almost all of it now is new ice, zero to two old ... not the old ice we have seen in the past. That means that even the winter ice is thinner.
Finally, critics point to the Antarctic. Ice coverage there has actually increased. But the Antarctic is the reverse of the Arctic. The Arctic is an ocean surrounded by land. The Antarctic is land surrounded by ocean. The Antarctic is one of those places where precipitation has increased due to global warming. This has caused a movement of ice from land into the water. That has increased the ice coverage. But the Antarctic is showing signs of global warming as well. The ice sheets that protrude from the land are breaking up forming huge icebergs. So while the ice coverage is larger ... it is less stable.
The polar bear problem is tightly linked to the loss of Arctic sea ice. Polar bears normally swim out to the sea ice in their hunt for seals. In the past they have had to swim about 25 miles to get there. Now they are often having to swim 40 to 60 miles to get there. Many are not able to do that. For the first time in history polar bears have been found in the open Arctic Ocean dead from drowning. Other polar bears are not attempting to swim. They stay on shore. But this leads to food shortages for them. Again for the first time in history polar bears have been seen to stalk, kill and eat other polar bears.
At this point in time the loss of polar bears has not been great. But it doesn't take a huge leap of imagination to see that if the problem progresses, the loss could become significant and in a hurry. That is the threat to polar bears. Critics, however, point out that polar bear populations are at near historic size. That's true. Global warming helped to start with. It led to an increase in productivity of the Arctic Ocean. In other words, warming allowed small organisms to increase their reproduction rate. These organisms form the base of the food web. Increase them and the effect will certainly trickle up to higher order predators. So long as the sea ice was within reach of the polar bears, we would expect their numbers to go up. If the sea ice is out of reach then that increase in productivity does not reach the bears. The drowned and cannibalistic bears is evidence that it isn't anymore.
But, as I said earlier, none of these pieces of data are the data that convinces mainstream scientists that anthropogenic global warming is real. The evidence that has done that is not generally criticized by global warming denialists. It is either ignored or complaints against the ancillary evidence are used to suggest that the evidence doesn't matter.
The evidence that has convinced mainstream scientists that anthropogenic global warming is a real problem is this:
It is simply the atmospheric concentration of CO2. These are direct measurements ... even the ones from a thousand years ago. Those are taken from "fossil air" ... air bubbles found in ice cores. This shows that atmospheric CO2 is spiking. The only thing capable of causing that effect is human CO2 emissions. Close monitoring of our fossil fuel usage shows that the observed increases in atmospheric CO2 concentrations is exactly what we should expect. Other ancillary tests involving a comparison of carbon isotope ratios in the atmosphere with that found in the fossil air confirm that the increased carbon in the atmosphere comes from fossil fuels.
This increase in atmospheric CO2 concentration is real. Nobody denies that. What is more 120 years of well verified physics says that such an increase in atmospheric CO2 concentration IS GOING to reradiate energy back to the earth that would have been dissipated in empty space otherwise. This energy IS GOING to raise global temperatures. It is simple physics that convinces the vast majority of mainstream scientists that global warming is a real and serious problem. The other data is just confirmation of this.
So why are we still arguing about its reality? It is because climate is a complex system. There are a lot of things that affect temperatures. We cannot do controlled experiments in which we hold everything stable and vary only one component of the global climate. So the data we get is going to have noise in it. Critics can use the noise to claim the exact opposite of what is actually happening.
To see that let's look at the BBC blog I opened with. Paul Hudson claims that there has been no global warming in 11 years. Why does he use 11 years? Because 1998 was an El Nino year..
El Ninos are an abnormal warming of the south Pacific Ocean especially during our winter months. This warming affects global temperatures. It causes them to be warmer than normal. In 1998 there was a very strong El Nino. So the temperature peaked very high that year. One might think that having a very warm year would be a bad thing for global warming denialists. But it was the best thing that ever happened to them.
Having that unusually warm year gave them a starting point that was artificially inflated. Normal global warming hasn't caused the temperature to rise to that level since (although by some measurement 2005 was as hot if not hotter). That has allowed global warming denialists to present a plethora of graphs like this one:
And with graphs like that they can make the superficially plausible claim that global warming is over. The problem is that this decade is going to be the warmest decade on record (even if 1998 isn't a part of it). Global warming is not over. It is going as strong as ever.
Let's look at the data in the proper context:
This picture shows the data all the way back to 1880. From this picture it is clear that global warming is going on at in increasing pace. It certainly isn't over. In fact, if you look at almost any 10 year time span the noise in the data is not going to show the overall pattern. So there is no mystery about where global warming has gone. It here stronger than ever.
We REALLY need to quit arguing about whether or not global warming is real. We need to move to a discussion of what we can do about it. I believe that we are beyond the point where we can prevent global warming from affecting us. But I could be wrong. THIS is where the debate should be.
Even if we are beyond the point of preventing global warming affecting us, their may be things we can do to mitigate the effects. THIS is where the debate should be. The longer we wait the higher the price we will have to pay to make the needed changes. We need to move the debate away from whether or not the phenomenon is real to what we can do about it and we need to do that quickly.
This blog is a part of Blog Action Day
For those curious about the physics, Wikipedia 'Greenhouse Gas' (or, I'm sure DB has a blog on this somewhere). Short story: these gases absorb certain wavelengths better than others, reflecting Earth's radiation more effectively than the sun's.
In the grocery store near our house, they will give you an option to purchase their "save the earth" green colored cloth bag and that's where they will place your purchased products. And then every time you buy anything from the store and you bring the "save the earth" bag with you, you get to have discounts. This is their way of sharing their contribution to save the earth from global warming. About three weeks ago, we were hit by two very strong storms. We never had that in decades. They left our country with still so many flooded areas, a lot were homeless and lost so many loved ones. Until now, we are still sweeping the streets from mud and tons of ruined appliances drowned from the flood. The effects of those twin storm were devastating. It's not wet season for us here, but we we were informed that we are still expecting four more storms on this last quarter of the year. And just this morning, local news says, weeks from now, we are expecting a strong earthquake to hit the metro city including the nearby provinces. This is the same Metro city hit by the twin storms. And the news says its all because of global warming/climate change. I'm blithely about the issue before, but when the twin storm hit us plus all the bad news, I thought, I must do something too. To solve the global warming problem, it must be stormed at the national and international levels.But the total success is built upon the action of every individual, regardless of nationality, to conserve energy and live in a greener, cleaner community.
Happy BLoG AcTioN WeEk! :-) Chicago interpreter
This topic actually came up a couple weeks ago in another blog I watch. The author posted a link to this article, which claims the "hockey stick" is broken. Since I don't feel I know enough about AGW to really comment on the matter either way, I only posted a forum topic here, which went unnoticed/unanswered.
I'm curious, though, about what your thoughts are on the article, and this seems like an appropriate place to present it.
I am treated as evil by people who claim that they are being oppressed because they are not allowed to force me to practice what they do. ~D. Dale Gulledge
Your link goes to Anthony Watt's blog. Watt is a former TV weatherman. His blog is the most popular of the global warming denialist blog there is. This particular blog mentions a finding by Steve McIntyre. McIntyre writes a blog called Climate Audit. It is the second most popular global warming denialist blog.
The one thing that McIntyre writes about more than anything else is the "hockey stick" graph. He has written about if for years. Recently he has claimed to have found an error in one of the proxy data used to form the graph ... tree ring data. Specifically tree ring data from Russia.
Years ago, a well-respected climatologist, Keith Briffa, published a paper reconstructing the a paleoclimate from tree-ring data. He used data given to him by Russians, the Yamal RCS chronologies. But the Russians evidently didn't give him all their data. They didn't give him some of the data of core sections taken from living trees. So in effect Briffa used a subset of the data.
Now the entire data set is available. McIntyre got it and looked at the data set. If you exclude the data that Briffa used and use the data that had been excluded then the paleoclimate curve doesn't look the same as Briffa's original one.
Here is a picture:
Briffa's curve (shown in red) shows a recent upspike in temperature, McIntyre's curve (shown in black) showed no change overtime.
Global Warming denialists now claim that the entire "hockey stick" curve is bogus and therefore global warming is bogus.
However ....
(1) The hockey stick curve is not the data that convinces mainstream scientists that global warming is real and it is a problem. That data is the rise in atmospheric CO2 concentrations. That data is absolutely unaffected by McIntyre's finding.
(2) The hockey stick curve is not based on a single proxy. It is based on 9 or 10. Whether or not Briffa's data is wrong is going to have little overall effect on the curve.
(3) The ONLY part of the curve that McIntyre's and Briffa's curve diverge is from the second half of the 20th century. WE DON'T NEED PROXY DATA FOR THAT. We have direct temperature measurements. We know beyond reasonable doubt that in agreement with Briffa's data the global temperature DID rise then.
(4) Sometimes data are not reliable. For instance if the trees were sick from an insect infestation then you might expect the growth rings to be narrower (which would be erroneously interpreted as cooler temperatures). On the other hand, if in the area of the trees a policy of fertilizer use had just been established then you might easily get wider growth rings (which would be erroneously interpreted as warmer temperatures). The Russians maintain that it was valid to exclude the data that was excluded.
So, a gigantic mountain was made out of a mole hill simply because it could be used to cast doubt on the hockey stick graph. At this time in the world of mainstream climatology, its significance is minimal.
Anthropogenic global warming is pretty much settled science. That means that there is a huge amount of data of such quality that scientists are convinced the phenomenon is real. That doesn't mean that it IS real. We could always be wrong. But the likelihood that we ARE wrong is very, very low. No one piece of data is going to be sufficient to overthrow that consensus. If we are wrong it will take a steady progression of contradictory data to overthrow it. And if we are wrong we should be seeing that data ... but we aren't.
So whenever you see something like:
"Scientist X discovered Y which disproves global warming",
then you can rest assured that it is hype.
Cheers,
DB
===
If a million people say a foolish thing, it is still a foolish thing. - Anatole France
Thanks for the info. I kind of figured it was something along the lines of being another denialist, but I don't really have the extent of knowledge it would take to pick it apart like you have.
I am treated as evil by people who claim that they are being oppressed because they are not allowed to force me to practice what they do. ~D. Dale Gulledge
(1) The hockey stick curve is not the data that convinces mainstream scientists that global warming is real and it is a problem. That data is the rise in atmospheric CO2 concentrations. That data is absolutely unaffected by McIntyre's finding.
I'm not an expert on this but::
My understanding is that the hockey stick is an explicit assumption in all of the climate change models. Essentially they are arguing that the recent temperature rises are unprecedented and therefore can only be caused by forcing caused by human emitted carbon dioxide. They have therefore built this CO2 forcing into their models and as a result they do indeed they do yield the result that increased carbon dioxide result in predictions of higher temperatures. CO2 is practically the only variable that matter in these climate models.
But if this explicit assumption is wrong and the recent recorded temperatures are not unprecedented, then the assumption about carbon dioxide driven forcing is called into question. It is not falsified but it does become questionable. The climate models would be extremely dubious (almost falsified) if it could be shown that in fact there were periods earlier in history (the shaft of the hockey stick) where temperatures were at similar or higher than they are today. For example, the medieval warm period was clearly not caused by CO2 forcing so if it was as warm as today, then the models would be largely falsified because they are based on a false assumption.
Of course we don't know what the tempuratures of the medieval warm period were, we only have the proxies. Some proxies, like Britta's data, actually show that period to have been colder than average which suggests a problem with the proxy because other evidence suggests that it was pretty warm. For example as ice recedes, we keep finding evidence around the world of previous human activity on ground that has been covered by ice for centuries which certainly suggests that these areas were about as warm then as they are now.
The ONLY part of the curve that McIntyre's and Briffa's curve diverge is from the second half of the 20th century. WE DON'T NEED PROXY DATA FOR THAT. We have direct temperature measurements. We know beyond reasonable doubt that in agreement with Briffa's data the global temperature DID rise then.
There is another possible lesson to be learned from this if we start from the assumption that Briffa's methodology was very flawed and selective which appears ti be the case. McIntyre pretty much destroyed it and Briffa's long and very unscientific attempt to prevent the release of his source data suggests that he knew his supposed findings would not withstand scrutiny.
If McIntryre's graph of the data is correct, and it looks far more honest than Briffa's and better matches work done on the same trees by Russian scientists, it suggests that there is very poor correlation between the tree ring data and actual temperatures as measured in the past 50 years. Actual measured tempuratures rose but the proxy tree rings did not show a correlating rise in the proxy. The lesson we should be learning is that if the correlation sucks for the past 50 years, then it probably sucked for the past 2000 years too and this data has no validity as a proxy for historic temperatures. It should be thrown out of whatever compilation of proxies is being used to define the shaft (historical record) of the hockey stick.
Incidentally, your account of how how the Russians only gave Briffa part of their data doesn't seem quite accurate. It may be true but it is only part of the story. My understanding is that the Russians published their data shortly after Briffa and that Briffa had many years to correct his findings using the full data set but chose not to and instead continued to publish papers using his flawed and arguably deliberately selectively chosen data-set and even further manipulated it by substituting selected trees from hundred of miles away which re-inforced his previous flawed findings. McIntire explains this selection process in great detail in his blog. The Russian data was largely what McIntyre used to debunk him.,
You keep talking about 9 or 10 proxies. I am no expert on this but my understanding is that NONE of them are considered to be reliable and that Britta's tree rings were considered the BEST of the lot.. Briffa's tree rings almost certainly don't belong in this average. The problem with taking an average of bad data is that you are going to get a bad average that is essentially meaningless. I'm an old computer guy and we used to call this garbage in and garbage out (GIGO).
I guess I need to know more about the 8 or 9 remaining proxies so I can decide if they are garbage like Britta's pathetic and embarrassing excuse for science.
(1) CO2 forcings are not based on the hockey stick curves. They are based upon the 140+ years of confirmed laboratory experiments concerning the absorption and readmittance of radiation by CO2.
(2) There are many different climate models used for a variety of reasons. Some just look at single variables, more complicated ones look at a variety of variables. This picture shows the progression that climate models have made over the years with regard to taking factors into consideration:
(3) If we were to find the medieval warm period were to be as hot as it is today, that would not have any impact on whether or not anthropogenic global warming (AGW) were real or not.
No one doubts that many things affect the climate. The question with AGW is what is affecting it NOW. That question is well answered ... human greenhouse gas emissions. To cause that to come into doubt one would have to come up with a testable alternative theory that fits the facts as we know them now better than does AGW. If we found the medieval warming period to be as warm as it is now, it do nothing to establish that.
However, the present research on that time period does not suggest that it was nearly as warm as it is now. In fact, the best evidence I am aware of suggests that the medieval warm period was a local and not a global phenomenon. Globally the instead of being warmer, the earth might have undergone a slight cooling
(4) We strongly disagree on whose data "looks far more honest". McIntyre used a data set that gave him results he wanted and are contradictory to results obtained from direct measurements. Briffa's data set is in agreement with those obtained from direct measurements. So whose data set is really more valid?
Again, there are times when exclusion of data is warranted. If you have good reason to believe that the data might be spurious then it is good to exclude it. The data that was excluded was core samples taken from living trees in a certain region in Russia. The Russians who collected the data say that it is not good for the purposes of determining climate at that time. Since the data isn't really needed (direct measurements of temperature are WAY WAY better than any proxy data) that assertion has been good enough for me. If it is not good enough for you then try to find out what their exclusion criterion was.
(5) Briffa did not hold up the release of any data. In fact, McIntyre says he had the data all along. In a contentious area like this ignore all pronouncements about people motives and look at the data. What does the data say? What other data supports it? What data tells you something that fits with known facts?
When I do that Briffa's data does very well, McIntyre's doesn't.
(6) All proxy data will have problems. They just wont have the same problems. That is why as many proxies as possible are used. The idea is that problems will factor themselves out. Tree ring data is very good because you have yearly resolutions. Many others don't. But that is the only reason that they are liked. Here is a link to the proxy data sets. If you wanted to you could access them yourself. I counted 306 data sets. I didn't check to see how many are tree rings but I assume quite a few are. The Yamal data set in question is only one of those and it goes back 4000 years. What McIntyre's result question is the last 50 years or so.
(7) With no apparent problem Briffa would have had no reason to review the data. It is my understanding that Briffa did review the data after McIntyre came up with his complaint. After reviewing the data he stood by the conclusion. That seems appropriate to me.
Cheers,
DB
===
If a million people say a foolish thing, it is still a foolish thing. - Anatole France
We strongly disagree on whose data "looks far more honest". McIntyre used a data set that gave him results he wanted and are contradictory to results obtained from direct measurements. Briffa's data set is in agreement with those obtained from direct measurements. So whose data set is really more valid?
You are starting from the unproven assumption that tree core data is a decent proxy for temperature. When all the data is examined, it turns out that it is not. As a layman I don't find that to be overly surprising. I think I learned in about the the 4'th grade that tree rings were a good proxy for the precipitation record. I don't recall any discussion of temperature.
Briffa was able to force-fit tree rings as a temperature proxy by discarding trees from the data-set that did not fit this hypothesis. By the time he got to the pointy tip of his hockey stick there was only something like a statistically meaningless sample of 10 trees left in his data-set that could be forced to support his hypothesis My understanding of the scientific method is that this is not normally how things are done. Normal scientific procedure is to formulate a hypothesis and test it with data and let the chips fall where they may. Either the data supports the hypothesis or it does not.
(5) Briffa did not hold up the release of any data. In fact, McIntyre says he had the data all along. In a contentious area like this ignore all pronouncements about people motives and look at the data. What does the data say? What other data supports it? What data tells you something that fits with known facts?
Simply not true! Briffa was extremely unforthcoming with the data even though he published in several peer reviewed journals that supposedly require data to be published to the public domain. McIntyre had data that he got from the Russians. He had no idea if it was the same data that had been used by Briffa. And Briffa not only refused to publish his data or provide it to McIntyre. He had denied others as well.
Briffa unforthcoming with data
(1) Tree ring data is a good proxy. But since ALL proxy data by there very nature has problems. Yes, one of the problems is that it is possible to have a wet year without it being warm and you would get a spuriously large tree ring. During those years the proxy would be giving you wrong data. Did you think climatologists were unaware of this? But the fact is that during warmer years there will be more water in the air due to evaporation and therefore you are more likely to have high rainfall. So most years in which there is high rainfall will be years in which the weather is warmer. And added factor is that during warmer years the growing season will be longer as well. So while tree rings don't give you perfect data ... no proxy does -- we KNOW that ... it gives you data that has does correlate with the temperature.
Climate change denialists can and do point out problems with each proxy data type and present it as though you cannot trust any of itl. However, since no two proxy data types share the same problems by using several different proxies the problems average out and strong conclusions can be drawn. That is how science works. You can't always get the data you want, you have to make do with the data you can get.
(2) Why is it that you are TOTALLY CREDULOUS of claims made by McIntyre, someone with no credentials in climatology, but dismiss everything contradictory that qualified climatologists say? McIntyre's blog seems to have an inordinate fondness for criticizing the hockey stick graph. The National Academy of Science, the most prestigious group of scientists in the country, did an independent review and found the data valid.
In short, McIntyre does not have a track record that warrants uncritical acceptance of anything ... as we will see in a bit.
(3) Briffa didn't "force fit" anything. He used the data he had. As for the hockey stick graph, for the period in question, Briffa's data is meaningless. Why? Because we have direct data to use to make up that part of the graph. We have actual temperature measurements. Those would trump proxy data no matter what.
Briffa's data is important with respect to how well it corresponds with other proxy data for the period before temperature records. There it does very well no matter what data sets one uses.
The chips ARE falling where it may ... unfortunately for climate change denialists they are falling right in line with anthropogenic global warming.
(4) For the sake of argument, let's throw out all of paleoclimatology. Let's say the hockey stick graph is meaningless. ... Nothing changes. Anthropogenic Global Warming is still shown beyond reasonable doubt. The predictive models are still valid.
(5) Don't rely on Climate Audit and WattsUpWithThat as your sole source on global warming. In fact, I would recommend you not rely on them at all. But if you insist on getting a contrarian point of view then at least read a more mainstream source for counter-claims.
The reason that McIntyre had the data in 2004 in the first place is because Briffa directed him to the Russians. It seems to me a little disingenuous to claim that he didn't realize it was the same data that Briffa had used.
(6) OK, ... I have just found Briffa's formal response to McIntyre. It was evidently posted today. From a quick reading this is what I get:
(a) Briffa's original work was designed to determine the paleoclimate back 2000 years. None of this brouhaha has anything to do with the vast majority of the work.
(b) Briffa original data (that McIntyre now questions) used trees from 3 different sites. McIntyre's threw out that data and used trees from a single site. The advantage of doing this is that more trees were used, but the disadvantage of doing this is that by confining yourself to a single site you increase the chance of error due to something strange having happened at that site. Evidently the reason the data from that site was excluded in the first place was because the tree ring data from that site was inconsistent. In other words, tree ring thicknesses varied between trees at that site. That is consistent with some of the trees suffering from disease, which can give you spurious thin tree rings which will be interpreted as colder years than it really was.
(c) Briffa made graphs using the complete tree ring data set, and the proper RCS (Regional Curve Standardization). It is not clear to me whether McIntyre, not being a climatologist, knows how to do this.
Here is his new findings:
The top graph shows the entire 2000 year reconstruction (year 0 to 2000). The bottom graph shows reconstruction for the last 200 years (year 1800 to 2000). The red curve is Briffa's data from his paper published in 2000, The blue graph is the reconstruction from updated study published in 2008. The black curve is the graph produced when one includes McIntyre's data.
None of the graphs are significantly different. McIntyre found an incredibly small molehill and is trying to make it out as though it is a mountain that casts all of climatology in doubt. That is bullshit. What McIntyre found makes even less of a difference than I had imagined. That is what makes me think that McIntyre must have improperly standardized his data ... which did as it turns out have valid reasons for being excluded in the first place.
Cheers,
DB
===
If a million people say a foolish thing, it is still a foolish thing. - Anatole France
First you claim
Tree ring data is a good proxy.
And then you say:
Briffa original data (that McIntyre now questions) used trees from 3 different sites. McIntyre's threw out that data and used trees from a single site. The advantage of doing this is that more trees were used, but the disadvantage of doing this is that by confining yourself to a single site you increase the chance of error due to something strange having happened at that site. Evidently the reason the data from that site was excluded in the first place was because the tree ring data from that site was inconsistent. In other words, tree ring thicknesses varied between trees at that site.
My understanding is that he did more than just take data from 3 sites. He also steadily and systematically reduced the size of his data set as he selectively excluded trees in recent years. He has not adequately explained why he eliminated all but 10 trees from his dataset when a great deal more data was available at all three sites. But it certainly appears that he was apparently starting with the assumption that the trees were a valid proxy for temperature and any tree that did not yield data that fit this assumption was assumed to be an abberation. Apparently, in the last 50 years, without being highly selective about which trees are allowed to be included in the dataset, trees are apparently a rotten proxy for temperature.
The problem with this methodology is that how do you know which trees to exclude from the historical database? Obviously a lot of the trees that were sampled in the current 50 years were yielding data that did not agree with actual recorded temperatures. It seems extremely likely that there were a lot of trees in the historical database that were also aberrant. But which ones?
So which is it? Are the trees a good proxy for temperature or are they only a good proxy if you manipulate the dataset?
Here are the final 10 trees from the Briffa Data.
It looks like the blade of the Briffa hockey stick was inordinately driven by a SINGLE tree: YAD061. It looks like an outlier to me but for some reason it was not excluded. Maybe a moose died at its base and gave it an inordinate dose of fertilzer.
To answer your original question, the reason people are questioning global warming is because the scientists involved seem willing to be unethical and unscientific and are in many cases refusing to yield up their source data like Briffa which is very contrary to good science. A similar instance is that all of the source data for the Jones and Wigley temperature record was apparently discarded and all that remains is their highly manipulated "value- added" data set. How is that science? The scientific method is supposed to require things to be verifiable and repeatable. And it is not like their have not been mistakes in temperature records. For example NASA embarrasingly had to revise their "value-added" temperature record for the USA a few years ago resulting in the 1930s being significantly re-stated as much warmer then originally claimed.
The Dog Ate Global Warming
I used to put amazing faith in science but no longer. I'll be a lot more willinging to give credit where credit is due when the scientists start behaving like scientists rather than high priests who we are supposed to accept their dubious pronouncements on faith rather than verifiable and repeatable science.
(1) Tree ring data is proxy data. As proxy data goes it is good data. However, ALL PROXY DATA IS GOING TO HAVE PROBLEMS. That is a given. The question then is can one derive useful information from proxy data at all. The answer is ... yes. But to do so you cannot rely on any single proxy. Also, you need to realize what the limitations of the data is.
(2) As best as I can tell, Briffa acted in full accord with good science. As best as I can tell McIntyre hasn't. Briffa was trying to reconstruct a paleoclimate. That is, the question he was interested in is what was the climate like BEFORE we had temperature records. The important part of his data is the paleodata. There is very little importance to his recent climate data. Abso-fucking-lutely no one has contested his paleoclimate reconstruction. If you use McIntyre's curve (that seems to be poorly standardized) it shows the paleoclimate the same as Briffa's.
(3) You don't throw data out unless there is good reason to do so. Just because a data point seems to be an outlier you don't throw it out. ON THE OTHER HAND, you do throw data out if you have reason to believe the data is spurious. The data that Briffa through out had internal inconsistencies. Tree rings size did not correlate well from one tree to the next. For example: Let's say that you had data from 5 trees from the same region. For one particular year the tree rings showed a thickness like so:
A -- .2 cm
B -- .3 cm
C -- .4 cm
D -- .5 cm
E -- .6 cm
From this data you assume that it gives you a ratio of how fast each tree is growing with respect to the other. Tree E is growing 3 times faster than tree A twice as fast as tree B, etc.
But suppose for the next tree ring for the next year you get thicknesses like so:
A -- .4 cm
B -- .3 cm
C -- .1 cm
D -- .2 cm
E -- .2 cm
Oops, you don't get anywhere near the same ratios. If this process goes on for several years and no clear ratio emerges then that tells you something is wrong with the data. The most likely thing is that during some years some of the trees were infected with pests. Your data will be skewed. Data like that SHOULD be thrown out.
Briffa says that data from that site had the highest level of inconsistency. So perhaps discarding that data was the thing to do. BUT WHO CARES ... that data is not the important data anyway.
McIntyre throws out the data that Briffa used ... the more internally consistent data ... and uses the data that was thrown out, and comes up with a different curve showing cooler temperatures. Of course he would, if the trees were had significant infestations, the wouldn't grow as much as trees would have in adjacent areas that weren't infected. In fact, as Briffa noted the tree rings were generally smaller there.
But there is a way to handle those problems somewhat. That is standardization of the curve. Suppose you have trees from one area in which the average tree rings are .4 cm in width. Suppose from an nearby area the average tree rings are .6 cm in width. Since the areas are close to each other presumably the difference is not due to climate. Perhaps it is due to the soil conditions. You need to standardize the two data sets.
Briffa did that on the data that McIntyre used. He standardized the data set. Guess what ... the original Briffa curve re-emerged. So even the data that was thrown out when properly analyzed supports the original findings.
McIntyre's used the data like somebody who doesn't know what he is doing. That's understandable, he's not a climatologist. He doesn't know what he is doing. Yet you insist that it is Briffa who is being the poor scientist here. Sorry, but you don't know what you're doing either.
But again ... WHO CARES ... that is not the important part of the graph. Throw it out if you want. Don't use it. Fine with me ... you wont hear any objections. The question then becomes what is left over.
Here is the data you get from glacial retreat:
Here is the data you from boreholes:
Here is the data you get on direct measurements of CO2 concentrations
Here is the data you get from Arctic reconstructions without using the Yamal data:
Here is the combined data you get when you exclude the tree ring data:
Here is the data you get when you look at the direct temperature data:
And finally here for the second time is Briffa's curves of 2000, 2008 and with McIntyre's data included:
They all show the same thing. What's more these are only the graphs I could find. I suspect there are plenty more. They too will show the same damned thing. What gives us confidence in the data is not any single proxy, it is the consilience of all the proxies.
Do you REALLY think that criticizing one miniscule piece of evidence (with invalid criticism) cast doubt on the entirity of the evidence? Come on, Jack ... you know better than that.
DB
===
If a million people say a foolish thing, it is still a foolish thing. - Anatole France