March 8, 2006
Global Warming "Evidence" Questioned
By Charles Martin
The scientific argument that humans have caused global warming — a major underpinning of the 'Kyoto Protocols' — suffered a major blow last week, with the publication of a new study. The implications have not yet spread very far beyond the rarified circles of specialists, but the gospel of 'anthropogenic' — human—caused — global warming has lost one of its intellectual foundations.
At the root of the argument for Kyoto are a series of reconstructions that claim to model what earth's temperature probably was before human activity caused the burning of massive quantities of fossil fuels. But these reconstructions of 'paleoclimate' by "Mann, Bradley and Hughes":ftp://holocene.evsc.virginia.edu/pub/mann/mbh98.pdf [MBH98], and various confirmatory studies, such as "Briffa et al":http://www.ncdc.noaa.gov/paleo/pubs/briffa2001/briffa2001.html [Briffa01]. all depend on 'proxies' — various observable things believed to correlate with temperature, like the width of tree rings. These proxies are needed because our ancestors foolishly didn't invent the thermometer until around 1600, and didn't start keeping good temperature records until much later.
These reconstructions then depend on mathematical or statistical models to convert the real data, say tree ring widths, into proxy temperature data. Combined with the much smaller collection of real temperature data, the assumption is that we can reconstruct the ancient climate and give an approximate graph of the Earth's temperature.
Briffa et al constructed one example, using six sets of tree ring data, and using the MBH methodology. They got a resulting graph like this:
What's most important to look at is the black line at the right hand end. It shows the characteristic "hockey stick" shape: a sharp increase during the 20th century, one that is much larger than in previous centuries. This is the supposed evidence of global warming, and one reasonable hypothesis of the cause is that it's anthropogenic, caused by the accumulation of greenhouse gasses in the atmosphere.
While I'm not a climate scientist, I ??am?? a computer scientist, and did my graduate work in simulation of complex and non—linear systems — which describes both the physiological systems in which I worked, and also weather and climate systems. One of the things that every modeler learns, and one caution that everyone in simulation needs to keep in mind, is that it is very easy to bias your simulation to give the expected results.
The problem is doubly difficult when dealing with statistical data.
Most non—specialists assume you just plug a set of numbers into a computer and get a nice graph. But real statistical data is 'dirty' and needs to be cleaned or 'conditioned' in order to analyze it. Outlying data, measurement errors, and equipment failures, all have to be dealt with.
The problem is that this process of conditioning can have the effect of throwing away any data that doesn't fit the original hypothesis. The study which casts doubt on the hockey stick graphs has tested the conditioning practices of one of the major studies
"Recently":http://landscape.sdsc.edu/~davids/enm/?p=34#more—34 David Stockwell used a large collection of proxy data, and applied the methods of MBH88 and Briffa, and got the following results:
Once again, we see the dramatic "hockey stick" in the 20th century.
There's only one problem. The 'proxy' data Stockwell used were a collection of random numbers generated to have similar bulk properties (technically, 'red noise') to those of the tree ring data. In other words, purely random numbers, when conditioned by the methods of MBH98, still result in a hockey stick.
The inevitable conclusion is that the hockey stick of Mann and others may be an artifact of their statistical methods.
Now, notice that I don't say 'is inevitably an artifact.' We can't say that. There still may be dramatic global warming. Absence of proof is not the same as disproof. But some of the best evidence that this global warming is caused by the actions of humanity is revealed to be highly questionable. Stockwell's demonstration calls the whole structure of anthropogenic global warming into question.
Charlie Martin (who blogs at "YARGB":http://yargb.blogspot.com under the cognomen "Seneca the Younger") is a writer and system architect in Superior, Colorado.