Global Warming Canards: The Reliability of Global Warming Denialists -- Part 2

In my previous blog I looked at the reliability of the first paper that chillbill, our new global warming denialist linked to "refute" the science behind global warming. That paper claimed to be peer-reviewed and on first look seemed to present reasonable data. It turns out however, that it was NOT peer-reviewed. It was published in a journal devoted to pseudoscientific crap. And the data it published was bogus.

As a short aside, in the comments on that blog chillbill takes me to task for "attacking the source". [sarcasm] Imagine me questioning the reliability of a global warming denialist source in a blog entitled GLOBAL WARMING CANARDS: THE RELIABILITY OF GLOBAL WARMING DENIALISTS -- PART 1. How misleading was that?[/sarcasm] I suspect that he will give the same whine for this blog, since it is PART 2. I will look at the other paper he linked to to refute the science behind global warming. It too is fatally flawed. Chillbill wants me to respond to his claim that the present temperature shows global warming isn't happening. As I already stated I will. I am answering chillbill's canards in the order he presented them. But this is science, it is EASY to post a link, put up a picture, and misrepresent the science. It takes time to correct those misrepresentations. Perhaps if chillbill had taken his time and ACTUALLY INVESTIGATED THINGS he would not have suffered the embarrassment of being shown that his source was bogus. But he didn't investigate his source even though the blog he responded to specifically stated the source was bogus. I also find it instructive that instead of LEARNING from his mistake he resorts to his typical tactic of misdirection. He ignores what the blog does say to bring up something the blog never intended to address -- the global warming denialist canard about the recent cool temperatures as disproving global warming. Never fear, that blog will be the one after this one.

In this blog, I will address chillbill's other claim for a peer-reviewed paper refuting global warming. Actually he link that chillbill provided take you to a global warming denialist website that talks about the paper. The actual papers (there are two of them) can be downloaded here. The paper is by Craig Loehle, a statistician who used to work for Stephen Wolfram, the creator of the mathematics software, MATHEMATICA. The Heartland Institute, a well-known center of global warming denialism, includes him in their list of "global warming experts", but as best as I can tell the two papers in the above link are the only things he has published on global warming. He has though published on other

Again like the previous paper, the Loehle papers are published in a journal with a respectable sounding name ... ENERGY AND ENVIRONMENT. But is it a respectable journal? The answer depends upon how you define respectable. It is certainly more respectable than the JOURNAL OF AMERICAN PHYSICIANS AND SURGEONS that the Robinson et al. paper was published in. But only barely. One way to tell the worth of a journal is to see how well other journals cite the work. There is a journal devoted to just that, the JOURNAL CITATION REPORTS. This journal indexes all major scientific journals. ENERGY AND ENVIRONMENT is NOT one of them. What that means is that ENERGY AND ENVIRONMENT is not a journal held in high esteem by scientists. Other people who have published in this journal include Sallie Baliunas and Willie Soon ... both of whom have been co-author along with the wackaloon Robinsons from chillbill's other reference. When it comes to global warming the only people who publish in this journal are global warming denialists.

Again, that does not make what Loehle published wrong. One of the greatest works of science was published in a no-name journal that was not well repected by mainstream scientists. That was Gregor Mendel's EXPERIMENTS ON PLANT HYBRIDIZATION published in the PROCEEDINGS OF THE NATURAL HISTORY SOCIETY OF BRUNN. However, being published in a lower class journal does increase the level of suspicion. Indeed, even Mendel's work lay unnoticed for about 35 years before its worth became apparent. However, I don't think Loehle is another Mendel.

Loehle's first paper, A 2000-YEAR GLOBAL TEMPERATURE RECONSTRUCTION BASED ON NON-TREERING PROXIES, was published in 2007. This graph shows the gist of the paper:

What makes it potentially important is that it is somewhat different than this graph:

This latter graph is the original "hockey stick" graph published by Mann et al. that sticks in the craw of global warming denialists. Before I can meaningfully discuss Loehle's work, I need to discuss the hockey stick graph and its importance.

The hockey stick graph purports to show global temperature over the past 2000. The sharp rise at the end -- the blade of the hockey stick -- is due to anthropogenic global warming. Global warming skeptics have acted as though this is the only evidence for global warming, but it isn't. The main evidence for anthropogenic global warming comes from the well-established physics of CO₂ absorption and re-emission of radiation, the undisputed observed rise in atmospheric CO₂ concentration and the fact that temperatures are observed (since 1850) to be rising as measured from temperature readings. The Loehle paper questions NONE of those pieces of evidence. What it questions is whether or not the MEDIEVAL WARM PERIOD was as warm or warmer than it is now ... and that is all it questions. So even if Loehle is absolutely correct, the principle evidence for anthropogenic global warming is still sound. But ... is Loehle absolutely correct? To determine that requires a us to carefully examine the evidence behind the graphs.

Both purport to show what the climate was like over the past 1000 to 2000 years. But how were the data points determined? We only have reliable thermometer readings since 1850. Any attempt to reconstruct the temperatures before we have reliable temperature readings falls in the province of PALEOCLIMATOLOGY. ... And paleoclimatology requires PROXY DATA. Proxy data are indirect measurments. And since this graph deals with temperature the proxy data we are concerned with are those things that are indirectly associated with temperature.

There are several proxies that were used in constructing the hockey stick graph. This site mentions 6 of the most commonly used proxies. Let's look at them and their limitations:

(1) Historical Data --- People wrote things prior to us having reliable temperature readings. They had newspapers and wrote in diaries and journals. Things like the number of ice-free days in a harbor, length of growing seasons, amount and duration of snowfall, the first and last frosts of winter, etc. are of importance in determining the local climate. This a reconstruction of the paleoclimate around Paris from 1350 to modern times extrapolated from the notes on grape growing seasons:

All proxy data have problems, and historical data is no exception. The writing is often anecdotal, which stresses the unusual. So it may be the exception instead of the rule. Data can be haphazard. Some years appropriate data can be found, some years not. How do you deal with the missing data? The data does not come globally. The three centers in which there is a good amount of historical data are Europe, China, and Japan. All of them are in the Northern Hemisphere. Finally, the data are difficult to calibrate into actual temperature readings. Thus, there is a certain amount of uncertainty that will be unavoidable in any reconstruction of climate from this procedure.

(2) Corals --- When a coral polyp dies its calcium carbonate (CaCO₃) skeleton remains. Succeeding generations of polyps grow on top of the skeletons of the old ones. This is how coral reefs build up. The polyps that grow in warmer waters produce denser, more compact CaCO₃. This means that corals produce annual growth rings. Less dense CaCO₃ from the remains of polyps that grew and died during the winter are bounded by denser CaCO₃ from polyps that grew and died during the summer. Core samples can be taken and paleoclimates can be inferred. Paleoclimatologists have used 3 different ways to infer past climates from corals:

[a] In general corals grow faster in warm water than in cold water. Therefore differences in the width of the annual growth rings are correlated with temperatures, the wider the rings the warmer the temperature.

[b] CaCO₃ contains oxygen which has 2 stable isotopes, ¹⁶O with 8 protons and 8 neutrons in its nucleus and ¹⁸O with 8 protons and 10 neutrons in its nucleus. During warmer temperatures there is more energy in the water and molecules containing the heavier ¹⁸O can more easily rise to the surface where the growing corals are. And enhancement of the ¹⁸O/¹⁶O ratio is correlated with increased temperature.

[c] Strontium (Sr) is an atom with very similar properties to Ca. In fact, it can easily serve as the counter-ion to the CO₃. Thus, within the CaCO₃ skeletons of corals there will always be a small amount of SrCO₃. In warmer weather Sr will be incorporated into coral skeletons in greater proportions. Thus, a shift in the Sr/Ca ratio of corals is associated with temperature change as well.

Just as with historical data, all three of these techniques have their limitations. While it may be a general rule that corals grow faster in warm water, if the water gets too warm the coral is subject to die offs. Indeed with this episode of global warming we are already seeing that in coral reefs around the world. But as for paleoclimates it shows that in really warm waters the annual growth rings will be small and could easily be confused with cooler temperatures.

While it is true there will be more water with ¹⁸O available to corals, corals are living things. They lay down the CaCO₃ skeleton by enzymes. Enzymes preferentially use lighter isotopes. So this enzymatic effect can confound the effect due to the increased availability of heavier oxygen atoms.

Why it is that Sr levels increase with temperatures is not fully understood. Originally it was thought that Sr preferentially partitioned out of sea water more readily in warmer temperature. If this were true then it would be an effect solely dependent upon temperature. However, more recent studies suggest the story is a bit more complex. The temperature dependence of Sr probably involves the active transport (another enzymatically controlled reaction) across cell membranes. This can certainly confound the readings.

Perhaps the most important confounding factor in determining paleoclimates from corals however, is that corals are often associated with algae. The algae live with the corals in a symbiotic relationship ... algae photosynthesize and some of their chemical energy helps the corals; and corals produce a nice home for the algae. The problem is that algae and corals handle water and Ca/Sr metabolism differently.

All these confounding factors means that any data derived strictly from coral studies will also be subject to uncertainties.

(3) Fossil Pollens --- Different species of plants produce different pollens. Pollens can be seasonally deposited in lake beds. These seasonal deposits form yearly layers called VARVES. By counting the layers we can determine the date. By looking at the type of pollen from a certain level we can get a picture of the plant life that was there at that time. The types of plants that were there indicate the type of climate. That is all straighforward enough, but it too suffers some limitations.

It is not easy to calibrate from the type of plants just exactly what the temperature was. Furthermore, some plants produce tougher pollen than other plants and these plants will be preferentially represented in varve sediments. Thus, pollen does not always give a fully accurate representation of what the plant community was. This will leave some uncertainties in data derived solely from pollen studies.

(4) Tree Rings --- Trees lay down yearly growth rings. By overlapping growth rings in different samples of wood one can readily obtain reasonable growth patterns extending back several thousands of years in a number of sites throughout the world. This picture shows the basics of how it is done:

During warmer years the growth rings in general are wider. Thus, the width of the growth ring is associated with temperature. This technique too has limitations.

Some years can be very hot but with little rain. Under conditions of drought the growth rings will be spuriously thin. Calibration of growth ring width to temperature is tree species specific. In other words an oak and a pine tree differ in how they respond to temperatures. Larger trees may grow at different rates than smaller trees, even of the same species. So again, these limitations impose some uncertainties on any data derived solely from tree rings.

(5) Ice Cores --- These are probably the best known proxies. Ice cores can go back hundreds of thousands of years and have the distinct advantage that embedded within them are air bubbles which harbor "fossil air" from the time of the ice core. This fossil air enables us to get a very accurate picture of the CO₂ concentration in the past. What we can deduce from that is that our present CO₂ concentrations are at the highest levels they have been in the past 650,000 years at least.

Further, ice cores are made of water which contains oxygen. Again we can use the ration of isotopes ¹⁸O/¹⁶O to get an idea of the temperature. This is technique has been used to show correlations between CO₂ concentrations and temperature. However, the temperature measurements do have some limitations. First and foremost, they only tell us the temperature at very selective places on earth. Presently the Vostok ice core is the only one that gives us very long-term information. Other problems include the fact that very old ice is highly compressed. The annual layers become very hard to distinguish so beyond a certain point age must be estimated.

So any data concerning GLOBAL warming that are solely derived from ice cores has its limitations.

(6) Ocean Sediments --- Pollens (mentioned above) are part of this data, although pollens do not deposit as accurately in ocean sediments as they do in lake sediments. However, the sediments contain things other than pollens, fossil animal and plant organisms, decomposed organic material with oxygen isotopes, etc. These are among the least accurate sources and the time resolution is not so good.

So any data concerning temperature readings derived solely from ocean sediments have their limitations as well. ... If you look back at what I wrote you will see that ALL of them have their limitations. That is an unavoidable problem when one is forced to use indirect measurements. Does that mean that we cannot trust past climate reconstructions? Not at all!!

While each proxy has its own set of limitations, those limitations are not necessarily shared by other proxies. By combining multiple proxies the uncertainties inherent in one technique tend to be accommodated by other techniques. When we get a number of different techniques all pointing the same direction then we can legitimately have confidence in the data. That is how the "hockey stick" graph first came into being. It was derived from multiple proxies.

With that information in mind we can return to the original question. Why is Loehle's graph different from the hockey stick graph? According to this website from the Heartland Institute, a noted global warming denialist website, Loehle's climatic reconstruction represents a "major improvement" over the original hockey stick reconstruction. So what did Loehle do make this "major improvement". Did he do a new study that produced new information that showed the hockey stick reconstruction wrong? Nope! What he did was throw out all the tree ring data. So instead of analyzing MORE data, he analyzed less.

This is a dangerous technique in science. One already knows what the data is. If one willy-nilly excludes one part and accepts another then one can get any answer one wants. This is called "selecting your data". In order to legitimately reject data one needs a VERY GOOD reason to do so. How does Loehle justify his rejection of tree ring data? He says that over the 20^th century CO₂ levels have increased. It is that time period that is used to calibrate tree ring growth to climatic temperatures. But since CO₂ in earlier times was less, one can expect trees who need CO₂ to grow those tree rings to have grown skinnier tree rings in temperatures that are just as hot as they are now. This would lead to temperatures readings to being inappropriately skewed to cooler in the past than they actually were. He cites the fact that the "Medieval Warm Period" and the "little ice age" show up more prominently on his reconstruction as evidence that he is right.

What does mainstream science have to say about Loehle's reconstruction. Well for one thing, mainstream paleoclimatologists are well aware of the limitations of tree ring data. They are the ones who discovered those limitations in the first place. To minimize the difficulties, many different species of trees are used. These tree ring data have been calibrated in numerous sites. Some of these sites have been at different altitudes and the concentration of CO₂ varies with altitude. If trees responded significantly differently at different concentrations of CO₂ this should have been easily seen in the calibration process. However, the effect was minimal. Tree ring calibration is among the best of the proxy datas.

Second, ALL PROXY DATA HAS LIMITATIONS. None vary solely with temperature. The case against excluding tree rings is no better -- and often much worse -- than the case for excluding any other proxy. Thus, Loehle's reconstruction amounts to no more than cherry-picking the data you want to make a conclusion.

Third, and even more telling, Loehle didn't even do his reconstruction right. He, being unaware of the climatological conventions, made the same mistake as did chillbill's previous source ... the Robinsons. He mistakenly concluded that "bp" (before present) referred to the year of publication instead of the standard climatological year, 1950. To his credit once his mistake was pointed out he redid his reconstruction. That is his second paper published in the same journal in 2008. This is he new "updated" graph:

He says this graph is about the same as the first one. But look closely. Where Loehle's original graph ended at the year 2000, this one ends at 1950. It was the later part of the 20^th century that experienced the sharpest global warming. Thus, even on Loehle's graph by 2000 we are at as high or even higher temperatures than we were in the Medieval Warming Period.

There are numerous other problems with Loehle's reconstruction as well. He used 18 data sets. His requirements for those data sets were that they contain a record over the past 2000 years and that they not be tree rings. However, the time resolution of those data sets is not particularly good. Only 5 of them have data useful for determining the temperature in the 20^th century. And that is perhaps the most important part of his graph. He claims to show that time is cooler than it was during the Medieval Warm Period.

Loehle claims that the hockey stick graph does not show the Medieval Warm Period. But in actuality it does. The picture of the original hockey stick graph did not cover the time period of the Medieval Warm Period fully. Here is a more recent reconstruction:

Indeed this reconstruction was done prior to Loehle's and the Medieval Warm Period and the Little Ice Age can clearly be seen. Also what can clearly be seen is the sharp rise in temperature after 1950. Thus, Loehle's "major improvement" is not that at all. It uses a highly selected data set, the results it claims is not exactly the results it actually shows, even it did show what it claimed the consequences for anthropogenic global warming would be very minimal.

So again chillbill chose to use a discredited source. Why? Because he obviously didn't believe it was discredited. The reason he didn't believe it was because he preferred to get his information from blogs and people that tell him what he wanted to hear instead of doing the hard work to check it out. And it is hard work to check it out. I have read several thousand pages worth of material now on global warming. While I do not claim a level of scientific expertise in the field, I think I am getting a good foundation in the basics of the field. STILL however, it required several hours of reading up on proxy data, checking to see what the mainstream response to Loehle's paper was, and hard thinking to integrate this data with what I had learned in the past. I don't think it is reasonable to expect everybody to do that. Most people have to chose to believe someone. One would be well advised to believe reliable sources.

Science is an adversarial process. Before a scientist can get a paper published he must show to adversaries (peer reviewers) that his results are (1) important within the field; (2) without apparent flaw with respect to technique of obtaining and analyzing the data; and (3) the conclusions he draws are appropriate from the data presented. After the paper is published it is fair game for other scientists to test and extend. If the paper is truly important, any faults will be discovered by future investigators. The information that is left has survived a barrage of attacks on its credibility. This suggests to me at least that the information is trustworthy ... or at least as trustworthy as we have any right to expect anything to be.

Science denialists, however, have a different standard. All they need is a statement that contradicts mainstream science. Anybody can make that up. It is only a little harder to come up with something that will give it a superficial plausibility. Thus, science denialism is easy especially when the results of science suggest conclusions that people would rather not believe. Unfortunately, reality is not affected by what we want to believe but by what really is.

Science denialism is dangerous. It prevents us from taking appropriate actions to fix what is wrong in a timely manner. There is no credible scientific doubt at the moment that global warming is real and its principal cause is man-made. I issue this challenge to global warming denialists ... find a paper published in a mainstream climatology journal (ie one indexed in Journal Citation Reports) within the past 20 years that produce results that call either of those two conclusions in question. I don't think they are likely to find a single one, but I am not a climatologist. Perhaps they will be successful. Here is what I can do, however ... I can find as many papers as can be reasonably be asked for that produce results consistent with those two claims.

Science denialism is causing us to waste resources to come up with more and more data that tells us what we already know. Perhaps it is time to shift our focus in research to better data on what can be done to alleviate the mess.

Coming up next: GLOBAL WARMING CANARDS: THE RECENT COOL SPELL DISPROVES GLOBAL WARMING