See here .
In anticipation of certain comments... (a) we use the analysis as a means of exploring a relationship and as a means to an end, which figure 7 amply demonstrates, (b) we do not use temperature data measured at the surface because it has far too many inconsistencies, and (c) we often use the word "trends" in a general sense rather than a statistical sense (where the exact line will be described to an accuracy of several decimal places).
I've written about the paper in question here, and there are plenty of links in that post to other, more systematic debunkings. But what is important is that, after admitting (along with his co-authors) that the paper in question "does not analyse trends" (meaning it has nothing to offer to the AGW debate), McLeans now wants it to be about trends again--though "trends" in some non-statistical sense. But not, I would argue, about "trends" in the general sense (meaning "general tendency or inclination"), because that was precisely the sense of the term at issue when De Freitas and co. previously admitted their work was not about trends. I admit to being somewhat confused by the 2nd paragraph, and suspect it is straight bullshit. I mean, the statistical sense is the general sense, no?
Sigh. Its hard to explain bullshit.
PS ESNO=El NiƱo