Tuesday, June 30, 2009

Lies, Damn Lies, and Statistics

I'm a statistical process control analyst for my day job. This means that I deal with statistics every day and I know what they can tell you and what you can MAKE them tell you. The trouble with statistics is that it's pretty easy to manipulate your data to tell you what you want to hear instead of the truth. This appears to be what has happened with CO2 and climate change.

I can hear the outrage already from some, but the theory (yes, it's only a theory) that CO2 emissions has caused global warming isn't supported by the unmanipulated data. The main theory around the CO2 levels causing global warming comes from the work of G.S. Callendar and C.D. Keeling who based their work on historic arctic ice core readings. The problem is that they didn't use all of the data. They used only the low readings over the years and discarded the high ones. In essence, they manipulated their data.

I decided to do some research on this and found a wonderful site with lots of excellent, though technical, information in an article written by Dr Tim Ball in the Canada Free Press. Much of what I'm going to share with you comes from that article.

Here is what is called a scatter plot. It's a very simple tool we geeks use to see what's going on. In this scatter plot, the historical CO2 levels from the ice core readings are logged. All of them. The circled dots are the ones that were selected for use by creators of the CO2 emissions theory on global warming. Notice anything a little wonky? Only the low readings were used. By throwing out the high readings they can make the statistics say whatever they want. And the world swallowed it hook, line and sinker. And think about this, if the historical CO2 levels we base our theories on are wrong, then what they put in the climate model is wrong, and all of the dire predictions that model gave us are, you guessed it, wrong.

But let's not stop here. One of my favorite graphs in the analysis is that as the CO2 levels rise so does the temperate. But what they don't tell you while looking at all the lines and dots on the graph, is that the temperate actually rises BEFORE the CO2 levels. It just keeps getting curiouser and curiouser. This looks more like a temperature increase is causing higher CO2 levels than the currently accepted theory of the opposite.

Here's the bottom line, statistics are only as good as the data used to calculate them. And the data is only as good as the people obtaining it. The data collection, and therefore the data, used to determine that CO2 is causing global warming (or climate change as it's now known) has been called into serious question, but are we taking another look at the data as we should? Nope. Instead we're working to pass expensive legislation to impact the CO2 levels that real data shows is not our problem.
Could global warming be accelerated by man? Yes, it is a possibility. But the data, the real data, all the data, shows that CO2 is not what's causing it. By focusing on a theory that has been disproved, we may be missing the real issue. We're acting like bad cops who ignore exculpatory evidence and work to convict the wrong man while letting the real criminal go free. Of course, man may also have little to no impact on global warming at all. Who'd have thunk it.

No comments:

Post a Comment