That is, he claimed that the 11 - year sunspot cycle plus its secular and millennial
variation, which I was modeling very precisely with my model, could be produced also by this kind of formula f (t) = A *
cos (2p * (t - T1) / p1) + B *
cos (2p * (t - T2) / p2) Some
variation on that formula does a good job, e.g. the one I used in my toy - example: «Sunspot Number» = SQRT (ABS (k *
cos (π / p1 * t) +
cos (π / p2 * t)-RRB--RRB-
I think Gavin &
co are probably relying
on the obscure reasoning behind the anthropogenic «fingerprinting» analysis summarized in Figure 10.5, where Natural + Internal
variation are shown as negligible while ANT fully explains all the observed temperature change.