Carlin Economics and Science

(One source added January 28, 2010)

One of the basic problems in reaching rational conclusions with regard to global climate change problems is that AGW proponents and skeptics largely use different data sources and very different analyses of the global temperature data to support their cases.


Global Temperatures According to NOAA (click to enlarge)
This first chart shows a typical global temperature chart similar to those often used by AGW supporters. Like most such charts, it represents global temperatures based on ground-based readings, and is apparently constructed by smoothing annual data. It shows the familiar “hockey stick” emphasizing the rise in temperatures since the mid-1970s that has become so familiar in AGW presentations.

The Satellite Data Shows What Appears to Be a Very Different Picture

And Using Satellite Data (click to enlarge)
AGW supporters almost never use satellite data on near-ground temperatures despite its availability since 1978. This second chart shows what appears to be a very different picture using it. One would think that the two sources of data would yield quite similar trends and have important similarities since they attempt to measure air temperatures so close to each other. But this appears initially not to be the case.

What the satellite data shows is a roughly flat trend with 3-5 year cycles from 1978 through about 1996. This cycle appears to be closely related to the El Nino Southern Oscillation (ENSO) which shows a very similar pattern. But the flat trend shows no obvious effects from the increasing atmospheric CO2 levels during this crucial period for the AGW’s point of view.

Then in 1998 there was a large spike in temperatures (which appears to be largely unexplained by anyone and has no obvious relation to changes in CO2 levels) that occurred just when the 3-5 year cycle might have been expected to bottom.

In the early 21st Century the same essentially flat pattern of 3-5 year cycles appears to reestablish itself in the satellite data but at a slightly higher temperature until a breakdown occurred in late 2007 back down towards what may turn out to be the original levels in the 1978-96 period.

So it appears that the the two charts suggest very different things. The AGW presentation obscures the 3-5 ENSO cycle and emphasizes what appears to be a fairly steady increase in temperatures from the 1970s on that makes the presumed influence of rising CO2 more plausible. The satellite chart, on the other hand, highlights the ENSO cycle and the abrupt changes in 1998, and makes the presumed influence of CO2 appear implausible.


Ground (blue) and Satellite (red) Temps (click to enlarge)
If both monthly ground-based and satellite data are combined on one chart (the third one), there appear to be more similarities between the two charts. In the 1978-96 period the ground data (highlighted in blue) appears to increasingly diverge from the satellite data and increases during the 1978-96 period rather than showing the flat trend exhibited by the satellite data. But there appear to be substantial similarities between the two data sets thereafter.

So in the ground-based case there appears to be a gradual rise in temperatures during 1978-2005 which appears to be at least somewhat consistent with the increase in CO2 levels. And in the satellite case there is no visual support for a gradual increase or any apparent role for CO2 during the whole period.

Increasing Reason to Question the Ground-based Data

There is increasing reason to question the basic ground-based data as well as the rationale for the AGW interpretation of it. There have been a number of studies suggesting that ground-based data is severely compromised by urban heat island effects, inappropriate placement of monitors that increase recorded temperatures over what they would have been if the instruments had been properly cited, and the drop-out of a large number of rural stations in the 1970s. The urban heat island effects arise because urban areas retain heat much more than non-urban areas and many urban areas in the US and elsewhere expanded rapidly in the late 20th Century. This may have been accentuated by the decision to drop many rural sites. So prior to 1997 the upward bias in the ground-based temperatures may represent little more than the results of these three effects.

In addition, there are many reports that the ground data has been subjected to substantial and often repeated adjustments for various factors. In the case of the crucial HadCRUT ground-based data developed in Great Britain, and the source most often used by the United Nations, the custodian has refused to release the data and their manipulations of it and recently said that they have not even retained the original observations. So it now appears to be impossible to reconstruct exactly what the custodian may have done with the data despite the fact that this is the data most often relied upon by AGW supporters to make their case. Finally, the heavily smoothed ground-based annual data appears to hide rather than illuminate what has actually happened.

For more information on the problems with the ground-based data see here and more recently here.

So what should we make of this seemingly perplexing but very important battle of the global temperature charts? A reasonable interpretation is that the AGW charts cleverly hide the ENSO cycle through use of annual data, heavy smoothing, and questionable data and show a gradual increase in temperatures starting in the mid-1970s due to an upward bias in the data. But If monthly data are used, no smoothing is carried out, and the increasing ground-based temperature data during 1978-96 is assumed to be replaced with the flat trend line shown in the satellite data, the two data sources would actually appear to be quite consistent except that the ground-based data is generally higher than the satellite data. This difference may also be due to biased ground-based data. But if this substitution is made, the same major features would appear in both cases.

I believe it is incumbent on AGW supporters to show why the monthly satellite data without smoothing should not be used since it avoids using the suspect ground-based data and provides a more accurate picture of what actually happened. It is they, after all, that are asking the world to spend astronomical amounts on the basis of their hypothesis. If this is done, their “hockey stick” disappears and along with it the familiar AGW arguments for their hypothesis. This is not proof of biased 1978-96 ground data but appears to be the most reasonable explanation of the remaining differences between the two data sources.

The satellite data may have its own problems, of course, but there is reason to believe that it may be far more reliable than the ground-based data. There are also two independent data services which have delivered very similar data. A much more detailed analysis of what the satellite data can tell us can be downloaded from here, which is also the immediate source of the charts above.

(Updated February 28, 2010 to include added sources)

In a previous post I explained why I believe that the United Nations GHG hypothesis that significant global warming will occur as a result of increasing greenhouse gas (such as CO2) levels is implausible.  In this post I will explain why I believe that the best available evidence indicates that the hypothesis is not just implausible but rather should and can be rejected on scientific grounds.

For a broader view of how science progresses see here. Clearly I am in the Popper camp in this regard. Kuhn’s view may more accurately describe how science has unfortunately sometimes been historically conducted, but certainly not how it should be.

Before going further, it is important to explain that the important word in the definition of the UN GHG hypothesis is “significant.” There is little doubt that higher levels of greenhouse gases are likely to lead to slightly higher global temperatures since that is why they are called greenhouse gases. The United Nations, however, claims that increases in the levels of these gases in the atmosphere are the predominant influence on global temperatures. Hence the qualification “significant” in order to include the UN claims while excluding the minor warming that has probably been caused by increasing levels of greenhouse gases in the atmosphere.

As explained in my Comments, the models relied on so heavily by the UN do not prove anything scientifically one way or the other.  They simply show what the model builders believe would happen if the hypothesis and all their other assumptions were correct. The model results are interesting, perhaps even useful, but irrelevant in deciding whether their AGW hypothesis should be accepted or rejected. This is because they do not compare the implications of the hypothesis with real world data other than past temperature data which the models have been modified to emulate. 

The Critical Role Played by the Scientific Method

According to the scientific method, a scientific hypothesis must be tested by comparing real world data with the implications of the hypothesis.  This is how Albert Einstein was able to persuade the world that his ideas on relativity had merit.  Scientists kept proposing real world tests of his hypothesis but each test confirmed its validity. After a number of these tests, the opposition conceded that his hypothesis was valid. (For a description of this extended process see, for example, Jeffrey Crelinsten’s Einstein’s Jury: The Race to Test Relativity). A similar process resulted in the acceptance of Newtonian mechanics and other hypotheses which gradually assumed the status of theories.

If the comparison with real world data does not confirm the hypothesis, the hypothesis should be rejected. There are only two alternatives from a scientific viewpoint when this happens: Discard or at least modify the hypothesis or discover an error in the data used to reject it.  From a scientific viewpoint, it is totally irrelevant how many public officials or scientific organizations–or how prominent they may be–support a particular hypothesis.  A hypothesis has scientific validity only by comparison with real world data. Joanne Nova has expressed this very well in her Handbook downloaded here.

There are numerous inconsistencies between the UN CO2 hypothesis and observed data.  Gregory explicitly compares the explanatory power of the UN hypothesis with the competing Svensmark hypothesis and finds the UN hypothesis wanting.  Idso and Singer provide extensive scientific evidence against the UN hypothesis.

But perhaps the most fundamental comparisons are between the major physical effects of the UN hypothesis and available real world data.  There are four particularly telling physically-based basic comparisons in this regard.  According to the scientific method an inconsistency even in one of these comparisons means that the hypothesis should be rejected from a scientific viewpoint.  It is important to deal with the uncertainty introduced by the word “significant,” however. This uncertainty increases the likelihood that a few of the comparisons may prove positive. Hence it increases the strength of any negative finding. In fact, if a number of tests should prove negative it makes the tests very powerful evidence against the hypothesis.

Four Critical Comparisons with Real World Data

Acceptance of the hypothesis requires that each of the following four observations are present:

1. There is a hot spot in the upper troposphere in the tropics as predicted by the UN. If greenhouse gases are significantly warming the Earth the first signs of it are supposed to appear about 10 kilometers above the tropics. The lack of such a hotspot is discussed in my Comments in Section 2.9 as well as by Joanne Nova downloaded here. She discusses the major objections that have been raised to this comparison and why she believes they are not credible. For more detailed information see here.
2. There is heating of the oceans.  The added heat generated by increasing greenhouse gas levels in the atmosphere must be stored somewhere. It has not been showing up in the atmosphere in the last decade, so if the hypothesis is valid it must be going into the oceans. But in the last few years this has not been the case. An extensive discussion of the evidence can be found here. The bottom line is that the AGW hypothesis fails this test as well.
3. The observed outgoing radiation fluxes from the Earth decrease with increases in sea surface temperatures. Satellite data, however, shows an increase, which is inconsistent with the high climate sensitivities to increases in CO2 and positive feedback so crucial to the UN’s case. A new peer reviewed paper accepted for publication on this subject can be found here. For a video of Christopher Monckton’s presentation on this study on the Glenn Beck program see here. For a more up-to-date but considerably more technical presentation, see slides 50 through 66 here.
4. The atmospheric response times for volcanic sequences would be longer than they would be without the UN hypothesis. If climate sensitivity is as high as the UN claims, it should show up in the atmosphere’s response time from volcanic eruptions. The reason for this is that climate sensitivity is also a measure of how tightly air and sea temperatures are coupled. High sensitivity is associated with weak coupling, allowing the establishment of significant disequilibration of the sea surface temperature. A discussion of this can be found in a 1997 report from the National Academy of Sciences here.  The discussion may be a little technical, but the conclusion that the data “is consistent with low [climate] sensitivity,” which is inconsistent with one of the UN’s crucial conclusions, is clear.

The conclusions are the same in each of these four cases: The UN hypothesis is not supported or even partially supported by these comparisons with real world data.  As Professor Richard Lindzen of MIT has recently observed with regard to his findings on comparison 3 above, “In a normal field, these results would pretty much wrap things up, but global warming/climate change has developed so much momentum that it has a life of its own – quite removed from science.”

The data are far from perfect, of course, perhaps in part because of a lack of effort to gather it.  But they all tell the same story.  This means that the hypothesis should be rejected scientifically based on current information. Future testing could lead to other conclusions, of course, but for now rejection is the rational course of action.

Implications

Accordingly, using this hypothesis has no scientific basis based on current knowledge concerning these four comparisons.  Attempts to argue that it is anything more than a religious or superstitious belief must show that the data used in each and every one of these tests (as well as others that may be proposed in the future) is wrong.

Accordingly, using the UN hypothesis as a basis for formulating policy is not useful or relevant from a scientific viewpoint.  Attempts to do so are likely to lead to scientifically unsound policy. Given that the current proposed “solution”–radically reducing CO2 emissions–would cost many tens of trillions of dollars, it is particularly incumbent on those advocating this very large expenditure (for which there are many other uses if it should actually become available) to show that their solution should not also be rejected since it is based on a hypothesis that should be rejected.

The UN reports issued to date do not show that the data used in these four important comparisons is incorrect, and therefore the reports should not be used as a basis for policy in my view. Reports substantially based on the UN reports, such as the draft EPA Endangerment Technical Support Document reviewed in my Comments, should also not be used for policy purposes for the same reason in my view.