Graham wrote:

> Which data sets are being used here? There are several at http://www.cgd.ucar.edu/cas/catalog/climind/soiAnnual.html#download. The links in the blog post are to darwin.ascii and tahiti.ascii. From just looking at the numbers, these do not seem to be seasonally adjusted and they do look positively correlated. They both go up around July and down around January.

Whoops! That would explain it. Those are the data sets Dara is using!

Sigh. This was supposed to be a quick and easy post, but I guess there's no such thing.

This would also be consistent with Dara getting huge amounts of "power" in his scalogram at a "period" on the order of one year: he's seeing seasons. ("Period" in quotes because I now realize it's only that up to a constant factor of order roughly 1.)