John wrote:

>I’m not trying to get anyone to check the DIFF12 calculation, because as far as I can tell the RealClimate people agree with that. As far as I can tell, that’s not the problem with this paper.

Actually Nathan seems to agree as well:

> I coded this up myself and got an extremely similar result. (I used the Mauna Loa data which I had on hand, which is slightly different from the global data, so it’s not exactly the same.)

As said I find this an unexpected result. It goes quite against my mathematical intuition.
So Nathan you say that Diff12, as defined in the
rcode by the Humlum paper
ist given by the following:
If $(q_0,q_1,....q_n)$ is a list of monthly measurements and

$filter(q_i)= 1/12*\sum_{j=0}^{11} q_{i-5+j}$

then

$diff12(q_i)= filter(q_{i+12})-filter(q_i)$

Is that correct?

If I look at the code then the temperature values in fig 2 of the realclimate post (red line) seem to come from: