Department of Energy Argonne National Laboratory Office of Science NEWTON's Homepage NEWTON's Homepage
NEWTON, Ask A Scientist!
NEWTON Home Page NEWTON Teachers Visit Our Archives Ask A Question How To Ask A Question Question of the Week Our Expert Scientists Volunteer at NEWTON! Frequently Asked Questions Referencing NEWTON About NEWTON About Ask A Scientist Education At Argonne Correlation and Causation
Name: Adrian
Status: student	
Age:  N/A
Location: N/A
Country: N/A
Date: N/A


Question:
Hi from Australia, I have a question about global warming. (It seems to be a very popular topic with our government here, who see it as good justification to introduce a carbon tax). I was looking at climate data on Wikipedia, and there was a definite correlation between atmospheric CO2 concentrations and global temperatures, however the hysteresis between temperature rise and CO2 levels was about 800 years, (with temperature rises occurring first). It seems unlikely that mother earth could preemptively elevate temperatures because CO2 levels were going to rise in the coming millennia. I also noted a very high correlation between CO2 levels and forest fires. According to Wikipedia, a global increase in temperature of around 10 degrees C correlated to an increase in CO2 levels of around 80%. In Australia, we a prone to bush fires. And everybody I have asked agrees that if summer was ten degrees hotter, bush fires would increase by at least 80 percent. My questions are : 1) Is is reasonable to suppose that causation and correlation are being confused? (i.e. temperature increases cause increased CO2 concentrations, not the other way around) 2) What mathematical tools would you use to clarify the relationship between correlation and causation? (other than common sense).



Replies:
You raise a number of questions for which at the present there in on definitive answer.

First, in any position about global warming, the distinction of "cause/effect" and "correlation" is frequently confused or ignored. Associated with this problem is the characterization of the global temperature with a single number.

Second, the models used to extrapolate current conditions are often self-predicting. The "answer" you want is intentionally or unintentionally fed into the model, so the outcome is predictable, no matter how complex the intermediate "number crunching".

Third, beware of the idea that a cause, effect, correlation is "common sense". "Common sense" has led to more errors than any other mind set I can think of. I challenge you to define "Common sense" in a scientifically valid way. Terms like "common sense", "clearly", "everyone knows..." are strategies to avoid rational analysis of a problem.

I hope this leads you to "the answer" -- there are no simple answers to the question you raise.

Vince Calder


Global warming is a fact! There are no more mile thick ice sheets in Canada. Also, it used to snow far more when I was a child in New York. I made a good income shoveling snow every winter, but lately - no big snows.

The big debate is how much of the warming is "man-made" and how much is natural. The existing scientific consensus is that a major part of the present day warming, that is to say the apparent increase in the rate of warming popularly called the "hockey stick curve", is due to human activity The test will be seeing if we can return the warming to its prior rate or less by controlling carbon dioxide emissions.

Although carbon dioxide gets the most press, methane is far more effective as a green house gas, molecule for molecule.

Robert Avakian


I am not going to try to explain the evidence pro or con anthropogenic global warming here. That is not my area of specialization and the topic is quite contentious politically. What I will do is address the general idea of how to scientifically address issues of correlation and causation and to figure out what causes what.

If all you have is two data sets that are correlated in some way, those two data sets in themselves cannot tell you what causes what, or even if the relationship between them is not accidental. To resolve the question, you need more data, and even more importantly, an explanatory theory that can be tested.

The data you have described do not suggest that previous global warming episodes were initiated by increased carbon dioxide levels in the atmosphere. That in itself is not particularly alarming to the hypothesis that the current anthropogenic high levels of atmospheric carbon dioxide will lead to warming. Nobody is claiming that previous warm periods were initiated by humans releasing large amounts of carbon dioxide into the atmosphere. THE RELEASE OF LARGE AMOUNTS OF CARBON DIOXIDE INTO THE ATMOSPHERE BY HUMANS IS AN EVENT UNPRECEDENTED IN THE HISTORY OF THE EARTH. Nothing like this has ever happened before.

The link between carbon dioxide and temperature is more than just observational. Well-established principles of physics explain why increased levels of atmospheric carbon dioxide raise the equilibrium temperature of the surface of the earth.

So, if all we had to go on was the fossil record of atmospheric carbon dioxide and of surface temperatures, then yes, we would be confusing correlation and causation. However, it is not all we have to go on.

I am not aware of a hard-and-fast mathematical method to establish causation. Such conclusions must be based on a review of all available evidence, theory, and the results of investigations suggested by theory.

Richard Barrans, ph.D., M.Ed.
Assistant lecturer, Department of Physics and Astronomy
University of Wyoming



Click here to return to the Mathematics Archives

NEWTON is an electronic community for Science, Math, and Computer Science K-12 Educators, sponsored and operated by Argonne National Laboratory's Educational Programs, Andrew Skipor, Ph.D., Head of Educational Programs.

For assistance with NEWTON contact a System Operator (help@newton.dep.anl.gov), or at Argonne's Educational Programs

NEWTON AND ASK A SCIENTIST
Educational Programs
Building 360
9700 S. Cass Ave.
Argonne, Illinois
60439-4845, USA
Update: June 2012
Weclome To Newton

Argonne National Laboratory