An Inconvenient Scientific Method

Having watched An Inconvenient Truth, I now know that Al Gore himself is not the Pied Piper writing the lyrics, he is merely the instrument singing the tune. As a Degree holder in Mathematics and Statistics with a focus on forecast models, I find the data analysis presented by Gore to be appalling. Mathematicians are taught to obsess over the probability of being wrong and of the potential biases in certain data sets. The variables that go into world climate are great in number and incredibly complex in scope. The idea that you could take one tiny isolated variable and use it to extrapolate a detailed record of world climate going back thousands of years is not only flawed, the rate of error would have to be astronomical. Granted, you can always fit a parameter estimate to the specific data to create a set of guesses, then you can super impose the guessed data back on top of the data it came from and call that correlation in a cleverly produced slide show.

Science by its nature can never be completely settled, and the scientific method exists primarily to prove things right or wrong. It is morally wrong to state that debate must be closed because your flawed analysis has been agreed upon by peer reviewed journals where the peers advocating the cause select who gets to be reviewed. It frustrates me to see people who don't understand multivariate regression proclaim their intellectual superiority due to their belief in a tragically flawed forecast model. I can't begin to tell you how many times I rolled my eyes at Gore's "revelations", because he was drawing impossible conclusions from data that cannot be used to explain what they are trying to make it explain. If they manipulated the math in their equations, the first place I'd look are the margins of error and confidence intervals. I strongly suspect that those have been deliberately deflated to minimize the appearance of making a mistake.

Is this the new future of science? Never having to say you were wrong?