Solar cycle 24 continues weakly, perhaps weakest of the space-age

Leif Svalgaard says (November 9, 2012 at 9:53 pm) :
The power spectrum of the solar microwave emission http://www.leif.org/research/FFT-Daily-F107.png shows the solar rotation period near 27 days and a smaller peak at twice, thrice, etc that. The 155-day peak does not look too hot.
Ray: Did you look at the references in my article on 155 day period? Here is an example, that shows many harmonics of solar rotation simultaneously. The harmonics are pretty much as you might expect given solar rotation and the structure of the neutral sheet. But sub-harmonics adhere to ratios of 2 and 3 as described by Dewey, and predicted by my harmonics theory based on cascading non-linear effects.
Article: http://cyclesresearchinstitute.wordpress.com/2011/07/04/harmonically-related-solar-cycles-in-uv/
Graph expanded: http://cyclesresearchinstitute.files.wordpress.com/2011/07/solar-mgii-index-cycles-harmonics-subharmonics.png
The entire 250+ years of sunspot records shows the 155 day period with very high significance. Many references to this just mention during solar maximum. However if sunspot numbers are transformed to square root of ssn, then the fluctuations at minimum become about equal to those at maximum. This almost entirely removes the modulations. (Lesson: many times modulations by longer cycles are because our data is in a slightly wrong form.)

Just a tiny comment. In the predictions for solar activity dated 1999, 4th paragraph 1st line states’ “The extra ultraviolet (UV) and X-ray radiation created by magnetic fields around sunspots also cause the Earth’s atmosphere to heat up and expand. ” http://science.nasa.gov/science-news/science-at-nasa/1999/ast14oct99_1/ … so the sun has no effect on our climate??? Really? Or is it just the agenda line since global warming was locked into the public troff

pkatt says
“The extra ultraviolet (UV) and X-ray radiation created by magnetic fields around sunspots also cause the Earth’s atmosphere to heat up and expand”
Henry says
So far, I do not exclude a gravitational or electromagnetic swing/switch that changes the UV coming into earth. In turn this seems to change the chemical reactions of certain chemicals Ox & NxOx & HxOx reacting to the UV lying on top of the atmosphere. This change in concentration of chemicals lying on top of us, in turn causes more back radiation (when there is more), hence we are now cooling whilst ozone (O3) & others are increasing.
http://blogs.24.com/henryp/2012/10/02/best-sine-wave-fit-for-the-drop-in-global-maximum-temperatures/

Leif Svalgaard says:
“No need to stoop to my level.”
A great self deprecating and very humorous comment. I shall try to comply with the intent of your comment. I do enjoy our disagreements and find your solar comments very valuable. Dark matter, not so much. Have a good weekend.
Jim

kadaka (KD Knoebel) says (November 10, 2012 at 9:41 am):
Kadaka: If you cannot see how the deep Canadian winters would be a time of hunkering down and waiting for warmer spring weather, in the ancient times before snow mobiles and roads being regularly plowed for easy automobile traffic when travel overland and also by water would be difficult if even possible, thus trade would be suppressed, thus you would not have the maximums dictated by the rigid timing occurring in the depths of those winters, then you are accepting numbers without considering their implications.
Ray: I don’t argue that Canada has tough winters. But they happen roughly every year. That has no effect on a 9.6 year cycle. It it was a 9.6 month cycle then you might have a point. I can see no point in what you say.
Kadaka: Here it says it was mathematically calculated as 8 years. …
Here, in an application of chaos theory, an apparent 10-yr cycle emerges. …
Several “about 10 years” found. …
Ah, finally found 9.6 here:
Figure 1. Canada lynx fur returns from the Northern Department of the Hudson’s Bay Company from 1821 to 1910. The Northern Department occupied most of western Canada. The cycle for these data averages 9.6 years. Data are from Elton and Nicholson (1942).
Ray: The 2nd link appears to have the full data that Dewey used and I have available. It is a pity that they do not know enough about data treatment to take logs, so they over emphasize the later period where numbers are much higher. You can see this in the graph of Fig 1 A which effectively throws away over 50% of the data.
Kadaka: Further searching reveals Elton and Nicholson 1942 to be a seminal work, often cited as a source of the trapping numbers. Finally found a copy here.
First thing to notice is the quality of the data. Dating issues, dropouts, the researchers grouped and adjusted figures as seemed logical to them. Also peaks were not synchronous across the different regions. Several pages of discussing the data later, we find this:
Between the peak years of 1752 and 1935 there were 19 complete cycles, giving an average period of 9.63 years. The frequency of variation around this average cannot be stated reliably from the total figures for Canada, because there is doubt as to the exact year of some of the peaks, e.g. 1809 might be 1808, 1913 might be 1914. It can be partly determined in another way, by counting all complete periods between peaks in Table 6, for the separate regions. Owing to the gaps in records for 1892-6 and 1914, the later series cannot be used except for the last cycle in James Bay, Lakes and Gulf. The result of this is to give a picture of the periodicity mainly for 1821-85. The frequency is: 1 cycle of 7 years, 6 of 8 years, 16 of 9 years, 20 of 10 years, 3 of 11 years, and 1 of 12 years. Of 47 cycles that can be measured, 36 or 78% are 9 or 10 years.
The average of the whole lot is very near that given by the total curve for a longer period, but is not directly comparable.
You are insisting on the rigidity of the timing, calculating that 9.6 yrs fits the data near perfectly. But as seen, that is merely the average of this distribution.
Repeatedly I find a ten year periodicity specified, but the modern use of 9.6 only once, referring to Elton and Nicholson.
You will have to supply the references for the data that yields 9.6 with the excellent agreement you specified, as Dewey goes further than Elton and Nicholson, out to 1962-63. Given the data massaging by Elton and Nicholson to get a usable database, a comparison with the massaging that yields the excellent agreement is indicated.
Repeatedly I find acceptance of a 10 year period. You appear alone in specifying a rigid 9.6 year period. Dewey noted it as “the average time interval between crests or troughs”. What is widely accepted is 10 years, not 9.6. Hopefully someday you will realize the difference.
Ray: I have loaded a graph of rate of change of log of lynx numbers with a 9.6 year cycle fitted. See http://ray.tomes.biz/images/lynx-delta-9.6-year-cycle.jpg The data that I have runs from 1735 to 1948 (and comes from FSC publication by Dewey) and so allows an accurate determination of cycle period. What are you trying to prove? If you are arguing that there are several data points that are suspect, then I agree with you. But it is quite clear that this is a 9.6 year cycle and not a 9.5 or 9.7 or any other thing outside that range. I don’t dispute that there are regional differences but that does not alter the fact that the cycle is 9.6 years.

Ray Tomes says:
November 10, 2012 at 3:29 pm
Actually there are two peaks at 150 and 158 days in that graph which are the highest peaks between 30 and 200 days in the spectrum. Aren’t they?
Simply because they are riding on a rising background. There are 16 peaks that are higher still up to 1200 days, aren’t there? Also all significant cycles?

u.k.(us) says:
November 10, 2012 at 4:41 pm
“Convinced?” I have not even begun to pick your brain.
I think your brain is the important one for this process…

Ray Tomes says:
November 10, 2012 at 5:50 pm
allowing far more accurate periods to be determined than with FFT.
Granted that FFT is not the best, but if a true cycle is there, FFT will find it. FFT may miss a lot of the spurious cycles that other methods can deliver. With your superior method do you find any other peaks between 155 and 2000 days?

Ray Tomes says
So that is why the 88 year cycle doesn’t get the timing right for peak at 1998 while both the 208 and ~60 year cycles do
henry says
with maximum temps we are looking at energy-in. Earth stores energy in, in water, in vegetation, in chemicals, etc. There is also its own core which moves around a bit and there is volcanic action. So energy-in can never be at the same time as energy out (means). There must be a lag, of at least a couple of years. However, if you do my plot from the relevant data (excepting that they are correct,i.e. r2=0.998 on the binomial)
data in degrees C or K per annum are:
0.036 from 1974 (38 yrs)
0.029 from 1980 (32 yrs),
0.014 from 1990 (22 years) and
-0.016 from 2000 (12 years)
what (cycle) fit would you make of it ?

Steven Mosher says (November 9, 2012 at 9:55 am):
Ray: I look at things from a cycles perspective because I find that it exposes new understanding. When very long term data is available, the need for Leprechauns is greatly diminished because statistics give us bounds within which correlation should fall. I would say that as far as 1 goes, there can be natural oscillations of Earth’s climate system including ocean movements etc. Regarding 2, there are heaps of other solar measurements than just TSI. They often have their own peculiar fluctuations, but nearly always cycles are present. Cycles are a great method to work out cause and effect chains because, like fingerprints, they leave their mark as a set of frequencies. Provided of course that you have enough data to measure the frequencies accurately. then the Leprechauns are not needed.
Steven: Ray. The cycle IS the leprechaun. They merely come in different colors. You miss the point of confirmation bias and data snooping entirely. Yes there are many solar measurements. In fact, there are an infinite number of solar measurements. That is, an infinite number of ways of characterizing observations and combining them. It is necessarily the case that if you look hard enough and long enough you will find a cycle. And if you don’t there are any manner of ways you can fool yourself mathematically to create them ex nihilo. And, as you know, there are an infinite number of ways of measuring the climate. Correlation, even long term correlation, is easy to find. In fact you must find it.
Ray: Steven, you entirely misrepresent what I do. I do not manufacture cycles. They are very clearly evident in most time series. Fourier Analysis is a widely accepted means of looking at time series. It is not me that is out of step with scientific practice. Any transformations done are only for the purpose of getting data in the most suitable form and never an attempt to make a cycle appear that was not there before. An example is that in sunspot numbers the fluctuations are much greater at the peaks than at the troughs. I sometimes use square root of sunspot numbers because then shorter term cycles (e.g. 155 day cycle) can be easily followed through both peaks and troughs of the 11 year cycle.
Steven: That is the fundamental problem with data driven approaches. All data confesses. If it doesnt, transform it. Shift it. add a constant, take the reciprocal, you’ll find the cycle. You have to. The test comes when you actually have to put together a physical theory that predicts phenomena not used to construct your system.
Ray: Shifting, adding a constant or taking a reciprocal have absolutely no effect on making a cycle where there was not one. Physical theories are grossly over-rated. Fine for something very simple. But when it comes to solar models, climate models, economic models, there is no such thing as a sensible theory that has a meaningful mechanism. Box and Jenkins showed that all the economic models could be beaten by a technique that relied only on statistics. The same is true of predicting the recent small peak in sunspots. There is no physics theory that you can put forward that gives even a hint of meaning to this. But a number of cycles researchers did predict a low cycle, based on 104 and 208 year cycles that are well established in the Sun. (Yes, Liev, they are modulated cycles, but all cycles are, and we are in the strong phase of 208 year cycle).
Steven: Let’s take Scaffeta as an example. After a modest amount of fiddling we have a ‘theory’ ( a curve fit) that explains global air temperature. Fine. What’s that “theory” say about ENSO. opps. nothing. It says nothing because its not a physical theory about the climate, its an exercise in cyclomania. That disease is addictive because one can always find the drug. Put another way, the philosophy that there are cycles in everything is metaphysics. Pure and simple non falsifiable metaphysics. It’s exactly the kind of nonsense that popper tried to banish from science. But nostradamus lives and he worships the sun. It explains everything and therefore nothing.
Ray: And you have a useful theory that explains solar variations, global temperatures and ENSO? Actually, mathematical models can be built that will include many different types of data and often make useful predictions. But in general long term predictions are not possible of complex systems. Back in 1977 I was using mathematical models to try and make economic forecasts. I was not not looking for cycles, but regression equations that would predict a year or two ahead. Even though I did not put any search for cycles in my method, a number of very clear cycles jumped out at me. Using these cycles I was able to make far more accurate predictions than any economists made. My method was to get many different time series and look at rates of change of them (which is generally what we want to know about most) and use factor analysis to find the underlying dimensions of the processes. These can then be used to predict each other using regression equations. The factors found were highly cyclic (even you would agree) and the regression equations continued the cycles ahead.
Ray: ENSO is not easy to predict very far ahead. I have looked at it and it is a hard one because the cycles present are not of stable period. Cycles analysis methods recognize such things, and all of these (rhythmic cycles vs irregular) were described by Dewey many decades ago. Any improvements in ENSO prediction will come from non-linear system studies as a general technique, not from physical models.