Quote:
Originally Posted by Robert Tulip
An interesting paper from NASA - Correlation between sunspot cycles and planets Jupiter and Saturn, published by Slovak Central Observatory at http://www.suh.sk/obs/slnsem/mikula.pdf - searched for a correlation between sunspot cycles and Jupiter’s orbital period using the Fast Fourier Transform method. |
First, it is not a paper from NASA, but just something written down my Mikula, Kletetschka and Adachi. It might sound nice, saying it comes from NASA, but only the second affiliation seems to be from NASA.
Secondly, this has never been published as a real paper. V. Mikula has zero 1st author papers, but co-authored several papers on e.g. Frequency Dependent Susceptibility Analysis of Magnetic Carriers: Application to Fe-Oxides on Mars surface. Mainly, he seems to be in the solid state magnetization stuff.
Thirdly, there seems to be a nice mess, which is almost not visible in Fig. 2 and 3 for the separate intervals. Now, I do not know how they handled their FFT, so I cannot comment on that.
Fourth, why should there be a 4.19% shifting of the results? What is the scientific basis that this is allowed?
Then they say:
Quote:
|
Originally Posted by Mikula et al
We will repeat this calculation using the daily sunspot number. Our prediction is that we will not have to shift the data in this case.
|
So they have the average sunspot number per year, and then have to average again. Now, this could be okay, but I have my doubt about the procedure. Also, why they expect the shift not to be needed then is alluded to, but not explained.
Basically, this "paper" gets a rejection from this referee, if I would get it on my desk, first of all it is too short, secondly the FFT that is used is not explained (do they average over spectral estimates and if so, over how many), thirdly the discussion is totally laughable.
It looks like they tried to find something (for some reason) found a correlation between Jupiters period and sun spot number, but for the rest ... As the solar cycle is on average almost the same length as Jupiter's orbit it is but natural to find a correlation, but a correlation does not mean a causal relationship.
It looks more like a short discussion paper, that in the end went nowhere and they discarded it.