Forget El Nino, the error bars from 1995 and the 2008 overlap which means there has been no statistically significant warming during that time. So says Richard Lindzen. And I think he is absolutely correct.
Forget El Nino, the error bars from 1995 and the 2008 overlap which means there has been no statistically significant warming during that time. So says Richard Lindzen. And I think he is absolutely correct.
That strikes me as erroneous. You can’t just blankly state that because the lower bound of one error bar overlaps with the higher bound of a different error bar (and are these 1-, 2-, or 3-sigma error bars?) that there is no statistically significant trend. Error bar range is not considered when a significance test is done. Even I know that.