"Mathematica" first plots out the data itself (see Atachment 1)
The first part of the instruction set lets "mathematica" do a Fourier Transform on the data, ie. searching out the periodicities, if there are any. The result is shown on Attachment 2.
The transform result shows a sharp spike at the 11 year point (I wonder what is significant about 11 years?). The second part of the instructions now acts upon this observed spike (the Cos 11 bit), to extract it's waveform from the rest of the noise. The result is shown as a waveform in attachment 3, the waves having an 11-year period, with the long-term Sydney warming easily evident.
Attachment 4 shows the original Sydney data overlaid against the 11-year periodicity.
It would appear that the solar cycle does indeed affect temperature.
Link: http://www.anelegantchaos.org/cru/emails.php?page=1&pp=25
Remarkable, it’s the Sun, I reached the same conclusion. We didn’t need to do the Global Warming Nonsense afterall!