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
Leif Svalgaard (06:18:25) @ur momisugly vukcevic (00:29:56) :
Let me repeat: you can attach no significance to the last six month of the filtered data. And one more time: you can attach no significance to the last six month of the filtered data.
Please report back with your acknowledgement of understanding.
(Salute) Yes Comrade Dr. Svalgaard !
Reporting my acknowledgement of understanding as instructed. I accept fully, as a view of an expert on the matter, which I am not. In my humble defence may I point to my original statement :
“This may be just a coincidence, but if strong divergence (just started) of polar fields before minima is any guide, then this minimum is not done yet.”
This is, among many other predictions flying about, just another speculation so no need for a repeated repudiation, unless of course there is something to it that I do not know, and you do not like.
It would be, for readers of this blog, as well as for the advancement of science, a useful step forward if you could find time to elaborate on the fact that the polar fields should be so perfectly symmetric as in
http://www.leif.org/research/Solar%20Polar%20Fields.png
I think, and I believe that you know, that can not be case. For time being I will refer to this phenomena as Sole-no-id.
This in a reference to nearly perfect symmetry of magnetic polar fields symmetry of a solenoid supplied by electric current; just a view of a man with ‘worst case of cyclomania’.
I am looking forward to be ‘educated’ about this particular puzzle.
The end of report. (Salute).
vukcevic (09:44:31) :
elaborate on the fact that the polar fields should be so perfectly symmetric as in http://www.leif.org/research/Solar%20Polar%20Fields.png
There are two sets of curves, real data [darker colors], and mirror data [symmetric lighter colors]. The latter just put there to make it easier to judge the relative sizes of the cycles [as explained in our paper on this]. So the perfect symmetry is artificial [whenever you see perfect symmetry like that, it is always artificial, e.g. the plots of the torsional oscillation that were shown last week]. What is plotted is the single-valued difference between the North and South polar fields. The differencing does two things: 1) remove the yearly variation due to the varying tilt of the solar axis, and 2) remove any zero level errors, which is especially important for the Mount Wilson data. The WSO website also shows graphs of the two poles separately, and they show interesting differences you might want to speculate on.
Leif Svalgaard (10:10:58) :
So the perfect symmetry is artificial [whenever you see perfect symmetry like that, it is always artificial…….
The WSO website also shows graphs of the two poles separately, and they show interesting differences you might want to speculate on.
Yes, I agree again, symmetry is artificial, so it should not be there. From your graph it is not possible to tell what each pole is doing .
WSO, true, shows the graphs, but with the annual modulation, which is an observation input, not actually there in the real polar fields. Come on you can do better, and have a graph showing what is actually happening, with no ARTIFICIALITY, with no observation modulation, just polar fields, otherwise people would be forever confused by your Sole-no-ID (solenid) (short for no ID of solar North and South field) and my subterfuge graphs.
vukcevic (11:36:14) :
have a graph showing what is actually happening, with no ARTIFICIALITY, with no observation modulation, just polar fields, otherwise people would be forever confused by your Sole-no-ID (solenid) (short for no ID of solar North and South field) and my subterfuge graphs.
My graph shows the relevant quantity [for my purpose] which is the ‘dipole moment’ and that is with no artificallity, just the observed quantity. Nobody has been confused before [to my knowledge] and your misconstrued graphs are on you. Since the zero level error is not known, it cannot be removed from the data [it is small for WSO, but large for MWO], so my graph gives the best picture we can make of the dipole moment which we believe controls the solar cycle.
Geoff Sharp (05:38:47) :
rbateman (20:52:46) :
Thanks for that, saved me a lot of work. I think there might be 2 sunspots missing, 1 where SOHO was down but one other that should be available. I have assumed a min count of 23 pixels with each pixel requiring a min RGB reading where G>70. This standard method of reading sunspots is supposed to emulate what would have been counted 100 years ago.
According to this method (which needs to be cross checked) it shows only 5 sunspot groups made the grade so far for SC24.
http://i42.tinypic.com/otlssg.jpg
This procedure sounds wrong if you are just basing the analysis on counting pixels in the image file. The count is then totally dependent on image resizing etc. You need to define a standard focal length for the optical equipment (the telescope and any additional barlows/reducers) and a standard physical *sensor* pixel size for this to have proper meaning. Said in a different way, you need to state the field of view (typically in arcseconds) for each sensor pixel. Then that could potentially be used as a basis for such measurements.
Carsten:
Look again at the images:
http://i42.tinypic.com/otlssg.jpg
SIDC official numbers for these spots…
01/04/08 – 10
04/14/08 – 7
05/04/08 – 7
09/22/08 – 8
10/12/08 – 11
11/02/08 – 11
11/13/08 – 9
12/10/08 – 9
Now look at my page: http://www.robertb.darkhorizons.org/DeepSolarMin5.htm
and compare 03/26/08 count of 36.
First things first. We have to come up with a way to relate what is visible 150 yrs ago with SOHO, which we already know what the resolution is. Then we can transfer that to hemispherical area visible.
Patience. Look at the mess the present day counting is in.
It’s not going to be that easy, but we’re working on it.
Carsten Arnholm, Norway (12:33:23) :
Said in a different way, you need to state the field of view (typically in arcseconds) for each sensor pixel. Then that could potentially be used as a basis for such measurements.
Or simpler [as is actually done] divide by the number of pixels of the solar disk. Then multiply by 1,000,000 and perhaps [if you really want to be fancy] adjust for projection.
I agree there is a mess in the counting, no doubt. What I am saying is that if you want to compare with 100 or 150 years ago and get credible results, there must be a proper way to measure that is based on physics as much as possible, and temporary technology as little as possible.
The suggested counting method may work for SOHO images since the optics and sensor is constant. But SOHO was not around 100 or 150 years ago and will not be around 100 years from now.
“We have to come up with a way to relate what is visible 150 yrs ago with SOHO” is indeed the challenge.
This is meant as constructive input. I follow what you are saying and presenting and find it very interesting, so please continue!
Leif Svalgaard (12:23:12) :
My graph shows the relevant quantity [for my purpose] which is the ‘dipole moment’ ……….
Haa! “for my purpose”, Is this some kind of a ‘private science’ or a public forum, but I was sternly instructed: “Please report back with your acknowledgement of understanding.” in regard to what I clearly indicated as “This may be just a coincidence …”.
True scientific endeavour (spelt correctly this time) requires considering everything that one may get data on. I thought any snippet of information on the Sun, in this current minimum might exercise mind of any solar scientist, let alone the top few. Not to mention Babcock-Leighton theory which teaches us that two poles garner their magnetic crop independently.
Now, if you do believe that ‘dipole moment’ controls the solar cycle, than by definition dipole moment is absolutely symmetric, and yet there is asymmetry between two poles. If you take zero as an arithmetic middle value (N-S)/2, it would imply that ‘dipole’ is bobbing up and down in the N-S direction, not to be ignored, just a treat for an enquiring scientific mind. Not one for me. Good night.
Carsten:
The idea of visibility level comes from L&P.
No matter what the imager or resolution, we want to define visibility as a floor of contrast.
Next, by what meets our criteria of visibility, we measure hemispherical area of umbra.
Last, we come up with a hemispherical area that is 10 on the ssn scale.
That is ideally done by finding old drawings/images that relate to the time before the subspots were counted and interfered.
We are helped in this regard by large spots where subspots pale in significance.
We believe it’s not only worth it, it’s something that is way overdue.
vukcevic (15:12:31) :
Haa! “for my purpose”, Is this some kind of a ‘private science’ or a public forum, but I was sternly instructed: “Please report back with your acknowledgement of understanding.” in regard to what I clearly indicated as “This may be just a coincidence …”.
The ‘stern instruction’ was about misuse of filtered data. There cannot even be ‘coincidences’ if you misuse this, only garbage.
by definition dipole moment is absolutely symmetric, and yet there is asymmetry between two poles.
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] while the polar cap changes size with time so that pixel is only a rough indicator of the instantaneous polar field and any asymmetry could well be due to variations of the polar cap size, and the ‘for my purposes’ refer to prediction of the sunspot number [or region count] for the whole Sun for which I made the plot, so the symmetric dipole moment is the right thing to use. You would get very nearly the precise same value if you did a spherical harmonics analysis and extracted the dipole component, as was done here http://wso.stanford.edu/gifs/Dipall.gif [the middle panel]. With your propensity for interpreting noise and garbage it seems prudent to stick to data with real meaning. BTW, the N and S filtered data is available on WSO site, just don’t use the first or last six months of the data, and be careful not to overinterpret the finer details that very well could just be garbage.
Leif Svalgaard (12:46:05) “The difference N-S takes out the annual modulation, so no need for filtering.”
Leif Svalgaard (15:44:50) ” http://www.leif.org/research/WSO%20Polar%20Fields.png
The curve at the bottom is the difference N-S.”
Can you be clear about exactly which ‘data’ you have used to construct this difference-curve? – (which webpage? – & which columns are your N & S?)
Vukcevic, What Dr. Svalgaard is trying to tell you is that the filtered values for recent months are SUPPOSED to depend on FUTURE measurements …which is why they should _not_ be listed at all. The analogy at the beginning of the series is that the first few filtered values are supposed to depend on earlier measurements that were never taken. In other words, the filter-window hangs off the ends of the series [by 6 months for a 1 year window]. It would be helpful if the ‘data’ webpage was unambiguous about exactly what filter was used. (Conventions vary.)
Paul Vaughan (16:54:25) :
Can you be clear about exactly which data you have used to construct this difference-curve? – (which webpage? – & which columns are your N & S?)
I can only give you the Wilcox part of the data [note I removed the quote marks]. The Mount Wilson part is not on a page – you’ll have to contact them to get the gigabytes of data and do the processing yourself.
Here are the WSO data http://wso.stanford.edu/Polar.html
Wilcox Solar Observatory Polar Field Observations
These WSO data are intended to be the most untouched measure of the solar polar field strength, i.e. presented with the least possible amount of processing.
The Sun’s Polar Field strength is measured in the polemost 3′ apertures at WSO each day, north and south.
The polemost aperture measures the line-of-sight field between about 55 and the poles.
As Earth moves above and below the equator each year the solar coordinates of the apertures shift.
Depending on the observatory optics and the date, the square aperture at the pole will be oriented differently on the Sun during each measurement.
Each 10 days the usable daily polar field measurements in a centered 30-day window are averaged.
A 20nhz low pass filtered values eliminate yearly geometric projection effects.
Data are updated about once a month.
Plots are available: [.gif] [.ps]
Latest data
Last updated Thu Jun 25 21:04:09 UTC 2009
1976:05:31_21h:07m:13s 89N -126S 108Avg 20nhz filt: 111Nf -92Sf 101Avgf
1976:06:10_21h:07m:13s 99N -121S 110Avg 20nhz filt: 109Nf -92Sf 101Avgf
1976:06:20_21h:07m:13s 107N -114S 111Avg 20nhz filt: 108Nf -93Sf 100Avgf
….
The columns marked N are for North, S are for South.
—————
And in return you owe me to explain ‘the central thesis’.
rbateman (09:09:23) :
Geoff Sharp (05:38:47) :
Nice. I will check out the missing days and update. How did you come up with G>70 and 23 pixels?
The G value looked like a good value to compare darkness and seemed consistent across many spots. The 23 pixels is a first pass and open to suggestion but the main criteria was to establish a min sunspot size so that sunspecks would not be included like they are currently, in an attempt to level the playing field….we seem to be heading into the realm of the ridiculous.
Here is a pic of the manufactured suggested std sunspot (inside green circle) shown along side sunspot 1007 as a reference.
http://users.beagle.com.au/geoffsharp/sunspotstd.jpg
Geoff Sharp (17:12:20) :
so that sunspecks would not be included like they are currently, in an attempt to level the playing field….we seem to be heading into the realm of the ridiculous.
Wolfer advocated and changed [starting ~1882 as Wolf’s assistant, but in full force after Wolf’s death in 1893] Wolf’s admonition of not counting pores and specks. Since then every sunspot observer and ‘producer’ have followed Wolfer, so to get a comparable count, one MUST count the pores and the specks. I agree that this is silly, but that is the rule. If you change the rule, we get yet another, different series that we don’t know how correlates with the historical data until we have several cycles worth of overlap, so we can cross-compare. One of Wolfer’s sins was that he didn’t run the two methods in parallel long enough [perhaps because he knew that it would be almost useless anyway as he had already ‘contaminated’ the official series with his own counts]. From whatever meager comparisons he made, he suggested a correction factor of 0.6, which we [or I at least] today think is likely not correct.
Re: Leif Svalgaard (17:11:45)
Clarification:
I’ve already plotted the ‘data’ from that page (along with several functions of the ‘data’).
Rephrasing:
Are you trying to suggest that your N-S at …
http://www.leif.org/research/WSO%20Polar%20Fields.png
… is column 2 minus column 3 from the following?
http://wso.stanford.edu/Polar.html
Paul Vaughan (17:39:18) :
Are you trying to suggest that your N-S at …
http://www.leif.org/research/WSO%20Polar%20Fields.png
… is column 2 minus column 3 from the following?
http://wso.stanford.edu/Polar.html
No, I missed that you were referring to a different plot than I thought you were [which I was discussing with Vuc]. The N-S curve on the plot you did refer to is a 50-day smoothed version of the single N-S differences. As the N-S is used [by me] as a predictor for the amplitude of the next cycle, a wider smooth is used, trying [not always successful] to suppress the noise caused by winter weather [clearly visible for the first 2-3 months of e.g. 2006 and 2006] that is still present even in the 30-day averages. You may also note a decrease of the noise starting in the summer of 2006 [just close in time to where the N in N+S is]. This is due to the replacement of a green-light filter at that time. The filter is necessary because the spectrum is made by a diffraction grating and we observe [in the green] in the fifth order that is overlapped a bit by both the 4th and 6th order. The old filter was getting weaker and weaker letting less and less light through and our noise is photon-limited: less photons = more noise. You can see how poor the agreement between observations of N+S [brown] and my model for the expected behavior [green curve] was getting in the year prior to the replacement of the filter. Also note the effect of the thousands of wildfires in California in 2008 [marked by WF] that caused excessive scattered light due to haze and dirtied mirrors. Scattered light reduces the polar field strength see http://www.leif.org/research/Reduction%20of%20Spatially%20Resolved%20Magnetic%20Field%20by%20Scattered%20Light.pdf
All in all, these measurements are difficult [and we are using 35-year old technology to boot], and care [and some distrust] must be taken in interpreting them.
Leif Svalgaard (18:13:16) :
suppress the noise caused by winter weather [clearly visible for the first 2-3 months of e.g. 2005 and 2006]
Geoff Sharp (17:12:20) :
I can definitely see your prototype spot in my 70mm refractor in the real world.
It’s a good fit.
Leif Svalgaard (17:37:22) :
Yes, you are correct, and I believe I can see why. There’s no telling how many subspots were counted, and as time goes by, it get’s worse as fainter “nebulae” are drawn in like the Hubble Deep Field.
It’s off the deep end.
SWPC is counting in to their area number anything that sits just above the background noise. It’s open-ended.
Leif Svalgaard (17:37:22) :
Would Wolfer have seen some of the specks that have been counted during SC24, the equipment these days is vastly different. Also his changes came after the last grand minimum, so to compare the Dalton to today we need to use a comparable system. During times of grand minima there is a far higher likelihood of sunspecks as L&P are finding out. The periods between grand minima when the Sun is behaving more normal this is not nearly the same kind of issue…its important we dont lose sight of this impending grand minimum and its very different output and make the necessary adjustments to reflect reality.
We certainly can run a parallel system.
Geoff Sharp (19:33:41) :
Would Wolfer have seen some of the specks that have been counted during SC24, the equipment these days is vastly different.
Yes, he would. And the very same telescope still exists and is being used for sunspeck counting as we speak: Figure 1.2 on page 11 of http://www.leif.org/research/Friedli2005.pdf I’m trying to get their counts.
There is no sign of any impending Grand Minimum, just of some low cycles like during the 1800-1820 period and 1900-1920.
Leif Svalgaard (20:27:02) :
There is no sign of any impending Grand Minimum, just of some low cycles like during the 1800-1820 period and 1900-1920.
You continue to deny the Dalton Minimum, most likely because its the most common type of grand minimum that occurs on a regular basis and defies the logic of the Babcock-Leighton theories.
The Leif Svalgaard universe certainly is a strange place.
Geoff Sharp (20:48:47) :
You continue to deny the Dalton Minimum
There was a minimum like in the early 1900s and coming now, but they do not qualify as Grand Minima [where there were an order of magnitude or two less visible spots]. And there are no signs of an impending Grand Minimum, only speculation.
As I have said before, a Minimum can take any form it wishes to. It could be brief, like 1900 or 1913, not as deep or long as the Maunder (like the Dalton),
or it could be as deep as the Maunder but as brief as the Dalton.
Even a long Dalton.
Existing theory says 2 very bum cycles, which says Dalton length.
How deep?
We are along for the ride.
Like the man said, take your best guess, come up with a test for being right, and wait.
In the meantime, I am seeing a ton of work to go through the ssn records and make a system with data that takes us into the age of technology no matter how improved it gets. Somebody has to do it.
Time is money.
Money is tight.
I’m open to suggestions.
(doing nothing is not a solution).