Guest post by David Archibald
Frank Hill’s summoning up of sunspots from the vasty deep of the Sun’s convection zone reminds me of some Shakespeare (Henry the Fourth):
Glendower:
I can call spirits from the vasty deep.
Hotspur:
Why, so can I, or so can any man;
But will they come when you do call for them?
Frank Hill says that his sunspots will be with us in three to six months. The Ap Index suggests otherwise. There is a correlation between the geomagnetic indices (aa Index and Ap Index) at minimum and the amplitude of the following solar cycle. Earlier this year I produced this graph of the Ap Index plotted against solar cycle maxima when I thought that the Ap Index would bottom out at three, giving a maximum amplitude of 25:
This is June and the monthly average of the Ap Index is 3.1. What is interesting from that graph is that there will be no sunspots if the monthly Ap Index goes below 2. The heliospheric current sheet is telling us that the month of minimum is possibly a year off and the Ap Index is showing no signs of pulling out of its glide slope of 0.28 per month, as shown in this graph:
The Ap Index enters the no sunspots zone in October at its current glide slope. Will it pull out in time? The Sun is bleeding magnetic flux (for a very good reason), so I don’t think so.
Frank Hill has shouted (his words) that there is no correlation between solar activity and climate. Back in a time when those who studied the Sun were armed with not much more than an enquiring mind, William Herschel in 1801 noted the relationship between the number of sunspots and wheat prices. When there were fewer spots, wheat prices were higher. To bring that up to the current day, when there are no spots at all, wheat prices will be the highest ever.
Back to Shakespeare: Hotspur has some good advice for those who study the Sun and draw implications for public policy:
Hotspur:
And I can teach thee, coz, to shame the devil—
By telling the truth. Tell truth and shame the devil.
Leif Svalgaard writes in with some graphs of his own:
Using Aa which goes much further back than Ap, the relationship between Aa and the size of the next cycle has been used by many people to forecast the next cycle. [one of the standard methods]. The data point with the red dot is the predicted Rmax for SC/24 using the polar fields and is plotted at the 2009 yearly average so far of Aa [9.1]
No spots for Aa = 2.4
Aa = 0.2318 (B Vo2)0.9478
Where B is the interplanetary magnetic field in nT and Vo is the solar wind speed in 100 km/s units. E.g. for B = 4 nT and V = 350 km/s, thus Vo = 3.5, we have Aa = 8.55 and Rmax = 57.
Ap is about half of Aa, but the relationship is not quite linear: Ap = 0.2925 Aa 1.204
So Aa = 8.55 corresponds to Ap = 3.87.
If we plot Rmax for the previous cycle (purple plusses) there is no correlation
Re: Leif Svalgaard (18:13:16)
Thank you for these useful notes.
Leif Svalgaard (16:17:14) :
We do not know how much of that asymmetry is real because we do not know the zero-level error and the plot only uses the single pole-most ‘pixel’ [constant aperture]…
Here is a graph with first and last, not 183 days but 4 years truncated, so no garbage here.
http://www.geocities.com/vukcevicu/PFdif.gif
The point is:
N – S ASYMMETRY IN INTENSITY OF THE POLAR MAGNETIC FIELDS.
If this difference is an artificial result of measurement application, than that should be made clear or date appropriately corrected.
Since there is N/S asymmetry in the SSN count and Area, I do not see a reason why one should not exist in polar field strength as well, and if it does than it must have some meaning to it.
WSO wisely publishes delayed data (for 30 day averaging), filtering should be also delayed for appropriate period. Simple mathematical recalculation could correct for the geometric projection effect (I initialy assumed), no need for numerical LF filtering, with all its drawbacks, of which I am extremely well aware of.
vukcevic (01:13:34) :
Here is a graph with first and last, not 183 days but 4 years truncated, so no garbage here.
The data is not garbage, just your notion that
vukcevic (12:27:03) :
Polar fields are still moving in the opposite directions;
since it was based on misuse of filtered data.
Since there is N/S asymmetry in the SSN count and Area, I do not see a reason why one should not exist in polar field strength as well, and if it does than it must have some meaning to it.
Of course there is, as there is some meaning to the pattern of tea leaves. We do not know how much of the asymmetry is real and how much is instrumental or, more importantly, introduced by my decision to label the measured value in a single pixel as ‘the polar field’.
Simple mathematical recalculation could correct for the geometric projection effect (I initialy assumed),
No, we disagree, as the correction can only be done if we assume a model for the distribution of the field, e.g. that it is radial, or has a specific non-uniform distribution, and we do not want our model assumptions to influence the data.
no need for numerical LF filtering, with all its drawbacks, of which I am extremely well aware of.
Perhaps now, after my ‘stern instruction’, but certainly not when you made your first comment on this [see above].
emphasis mine:
Leif Svalgaard (05:49:26) “No […] as the correction can only be done if we assume a model for the distribution of the field, e.g. that it is radial, or has a specific non-uniform distribution, and we do not want our model assumptions to influence the data.“
Here, here!
Such sensible words.
Paul Vaughan (08:11:00) :
“we do not want our model assumptions to influence the data.”
Hear, hear!
Such sensible words.
But am I not always sensible? 🙂
Now, for that other matter ….
Dr. Svalgaard, to settle the matter:
1. I accept that current appearance of the polar field divergence is not real, but product of numeric filtering (although I did qualify my statement that any interpretation of it significance may be a coincidence).
2. My observation was made under wrong assumption that geometric projection errors are calculated and automatically applied to each set of measurements, instead of floating numerical filtering as may be applied by WSO, in which case I think data should be withheld until such time as most of the filtering effect is out. Depending on the formula used, filtering effect could never be completely eliminated, so it would be useful to quote 2 critical points (considering it is intensity; 6 &12 db down). Filtering in analogue electronic communication circuits is an important factor but it has serious side-effects, numerical filtering in digital circuits is more flexible but not fully devoided of problems, and I happen to be familiar with both types.
3. In past a certain amount of ‘apparent asymmetry’ has been recorded but its meaning and origins are not entirely clear, and a future investigation may revel (if it does exist) does it have any consequential or causal relationship to the N/S asymmetry during the previous and subsequent cycles.
vukcevic (10:09:23) :
instead of floating numerical filtering as may be applied by WSO, in which case I think data should be withheld until such time as most of the filtering effect is out.
We make the assumption [always dangerous to do] that people that “are extremely well aware of the drawbacks of filtering” would not misuse our data. SIDC when calculating smoothed sunspot numbers do, indeed, not calculate for the latest six months, so there is a precedent for this. Apparently they know from experience that one cannot count on people not to misuse the data, even if they be ‘extremely well aware etc…’ 🙂
It is always inviting trouble to change the format of a long-used dataset, so we’ll not do that, but rely on common sense on behalf of the users.
Paul Vaughan (08:11:00) :
“Now, for that other matter ….”
still stands and I’m awaiting a serious reply.
vukcevic (10:09:23) :
My observation was made under wrong assumption that geometric projection errors are calculated and automatically applied to each set of measurements, instead of floating numerical filtering as may be applied by WSO
The website clearly states: “20nhz low pass filtered values eliminate yearly geometric projection effects.”. We acknowledge that perhaps a additional warning would be useful to some, but also feel that there are just things that are obvious, namely that one cannot make data out of thin air… [admittedly done by some 🙂 ]
Leif Svalgaard (11:44:38) :
I have on my desk a strong magnet. At some distance from it an iron key can still feel the magnetic field, right? If I move the magnet across the street, there is still a magnetic field in my office [albeit much weaker], right. Are there any currents in my office or at the Earth in the first example? The only answer that experiments give is “no”. So we have in one region of space magnetic fields without any currents in that region, right?
No. Absolutely wrong!!!
As an engineer trained in electronics I was more than surprised (better word would be shocked) by Dr. Svalgaard ignorance of electro-magnetism basics. Here is a simplest explanation of permanent magnetism quoted for the most basic understanding.
Ampère model: where all magnetization is due to the effect of microscopic, or atomic, circular “bound currents”, also called “Ampèrian currents” throughout the material. For a uniformly magnetized cylindrical bar magnet, the net effect of the microscopic bound currents is to make the magnet behave as if there is a macroscopic sheet of electric current flowing around the surface, with local flow direction normal to the cylinder axis. (Since scraping off the outer layer of a magnet will not destroy its magnetic field, it can be seen that this is just a model, and the tiny currents are actually distributed throughout the material). The right-hand rule tells which direction the current flows. The Ampere model gives the exact magnetic field both inside and outside the magnet. It is usually difficult to calculate the Amperian currents on the surface of a magnet, whereas it is often easier to find the effective poles for the same magnet.
http://en.wikipedia.org/wiki/Magnet