Saturday, 4 February 2012

Fun with numbers

One theme of this site over the last month or so has been the misleading use of numbers. This is very much related to the ongoing discussion of whether warming has stopped or not. One side, the Anti-AGW, says it has stopped and therefore global warming is not happening and never will, conveniently forgetting that a lot of it has already occurred. The other side, the Pro-AGW, says warming still happening and introduces new metrics to demonstrate this. 

In an age when every ‘scandal’ unearthed by budding ‘Woodward and Bernsteins’ is given a ‘gate’ suffix it is easy to forget that President Nixon was impeached not for the original crime of breaking into offices of the Democratic Party but for the cover up. I wonder if something similar is happening with the Climate Science Community at the moment. I have, in earlier blogs, identified examples of presentations and sites trying to prove that the models’ predictions were accurate which have left off data from the recent years (though internal evidence shows that they could have included it.) Yes, it would be inconvenient to admit that the current climate stasis was not predicted but playing with numbers to try and disguise it opens the Climate Science Community to charges of intellectual dishonesty.

Below I ‘prove’ that there has been no increase in temperatures since 1998. I did this by assuming that temperatures from 1998 onwards have been flat and calculating the temperature perturbation from the flat line. I then regressed this perturbation, after adjusting for residual seasonality, against sunspots, aerosols and a multivariate ENSO index and got the following graph.

Wow – there’s no warming! Eat your heart out you pro-AGW crowd. But, how did I prove it? I cheated. Firstly the coefficients I derived suggested that higher solar irradiance is associated with cooling and higher aerosols are associated with warming; only the El Nino index behaved sensibly. Then I adjusted the format of the equation so the coefficient was shown to only two places of decimals. As the data are monthly this means I could have had warming of 0.049 degrees/month (0.6 degrees a year) and still shown ‘zero’ warming. What is more, if I’d played around a bit more with lag times, different indices, and different amounts of smoothing  I could have produced an even more convincing, but even more dishonest, proof.

I’ve been completely upfront about what I have done but as long as the Climate Science Community tries to avoid the reality of climate stasis they will face a barrage of similar ‘proofs’ but with any underlying falsification being undisclosed.

No comments: