UPDATED: Post note below, with a couple of extra caveats…
Lüdecke, Hempelmann, and Weiss found that the temperature variation can be explained with six superimposed natural cycles. With only six cycles they can closely recreate the 240 year central European thermometer record. There is little “non-cyclical” signal left, suggesting that CO2 might have a minor or insignificant effect.
The three German scientists used Fourier analysis to pick out the dominant cycles of one of the longest temperature records we have. The Central European temperature is an average of records from Prague, Vienna, Hohenpeissenberg, Kremsmünster, Paris, and Munich.
The dominant cycle appears to be about 250 years. There is also a cycle of about 60 years, corresponding to the Atlantic/Pacific decadal oscillations.
Data is of course, always the biggest problem. If we had 10,000 years of high quality global records, we could solve “the climate” within months. Instead, we have short records, and Lüdecke et al, make the most of what we have. The European records are only 240 years long, or (darn) one dominant cycle, and only one region, so to check that the results are valid over longer periods they also analyze a the 2000 year Spannagel Cave [...]