Comment: Yawn (Score 0) 79
Comment: Re:Then **you're** naive! (Score 1, Informative) 201
Comment: "Combining Microsoft's open, interoperable"... (Score 1) 157
Comment: Re:As somebody who holidays nearby (Score 1) 103
Comment: Re:Funny that people assume this isn't harmful. (Score 1) 103
Comment: Re:When you're out of rational arguments... (Score 0, Troll) 585
Comment: Re:Yes it is! (Score -1, Troll) 585
Comment: Re:moderately skeptical (Score 5, Interesting) 695
Multi-year trends in global temperatures are not.
Prove it.
Please, please show proof that any model with 10+ year response times is demonstrably valid.
The problem is that we only have ~100 years of good data and ~200 years of fair data and very poor data prior to that
The fundamental problem that I (as a person with a good knowledge of statistics) have is that all the data pre-processing (a.k.a. faking) that goes on gives unreasonably low estimates of the variance (or standard deviations or confidence intervals).
A good scientist should define a clear model and use all of the raw data to validate the model. Then and only then can the model be used to extrapolate - say to the future or to the causes of climate change.