Above: step by step animation of solar cycle revisions since 2004
Michael Roynane writes:
On March 4, 2009 Dr. David Hathaway issued a new sunspot prediction for March 2009 which includes sunspot data through the end of February 2009. After no changes in the February 2009 prediction, solar maximum for Solar Cycle 24 was pushed back an additional three (3) months from 2012/10-2012/11 to 2013/01-2013/02. The predicted sunspot number at solar maximum was reduced from 104.9 to 104.0.
Cycle 24 Sunspot Number Prediction (February 2009)
Year Mon 95% 50% 5%
2012 07 128.0 104.0 80.0
2012 08 128.5 104.5 80.5
2012 09 128.8 104.8 80.8
2012 10 128.9 104.9 80.9
2012 11 128.9 104.9 80.9
2012 12 128.8 104.8 80.8
2013 01 128.5 104.5 80.5
2013 02 128.1 104.1 80.1
Cycle 24 Sunspot Number Prediction (March 2009)
Year Mon 95% 50% 5%
2012 10 126.9 102.9 78.9
2012 11 127.4 103.4 79.4
2012 12 127.8 103.8 79.8
2013 01 128.0 104.0 80.0
2013 02 128.0 104.0 80.0
2013 03 127.9 103.9 79.9
2013 04 127.7 103.7 79.7
2013 05 127.3 103.3 79.3
What is very strange about the revised March 2009 prediction is that the smoothed value for Solar Cycle 23 was also pushed forward by one (1) month with no change in the sunspot number at solar maximum.
Cycle 24 Sunspot Number Prediction (February 2009)
Year Mon 95% 50% 5%
2000 08 141.6 117.6 93.6
2000 09 142.0 118.0 94.0
2000 10 142.3 118.3 94.3
2000 11 142.4 118.4 94.4
2000 12 142.4 118.4 94.4
2001 01 142.2 118.2 94.2
2001 02 141.9 117.9 93.9
2001 03 141.5 117.5 93.5
Cycle 24 Sunspot Number Prediction (March 2009)
Year Mon 95% 50% 5%
2000 07 141.6 117.6 93.6
2000 08 142.1 118.1 94.1
2000 09 142.3 118.3 94.3
2000 10 142.4 118.4 94.4
2000 11 142.4 118.4 94.4
2000 12 142.2 118.2 94.2
2001 01 141.9 117.9 93.9
2001 02 141.5 117.5 93.5
I have no idea why this change was made but welcome input from the members. The new animation, with viewing instructions, can be found here.
http://commons.wikimedia.org/wiki/File:SSN_Predict_NASA.gif
http://upload.wikimedia.org/wikipedia/commons/a/af/SSN_Predict_NASA.gif
With these changes by NASA, the variance with the high SWPC prediction remains significant. As the new SWPC numbers are now quite impossible, I expect to see more changes from both NASA and SWPC over the coming months. With each NASA revision the predictions more closely resemble those of Dr. Svalgaard who is on the low-end of the SWPC low prediction faction.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Leif Svalgaard (06:52:44) :
to
vukcevic (04:39:05) :
I never claimed excellent correlation with sunspot number. What
I claim is the excellent correlation with recorded magnetic polar fields
as in:
http://www.geocities.com/vukcevicu/PolarFields_Vf.gif
and I stand firm on it.
This is turning into kind of a Christmas pantomime.
Leif;
Thank you very much for your explanations you are excellent at responding and explaining to us novices. Thanks again.
vukcevic (08:50:04) :
I never claimed excellent correlation with sunspot number. What I claim is the excellent correlation with recorded magnetic polar fields
Yet you keep showing the sunspot graph back a 1000 years.
The polar field correlation has too few degrees of freedom to be significant and fails for 1964-1965.
Carsten Arnholm, Norway (15:58:16) :
to
vukcevic (05:01:58) :
You find the software here, if you need it
http://arnholm.org/astro/sun/sc24/sim2/
Thanks for your link. Found it very useful in pinpointing number of anomalies.
http://www.geocities.com/vukcevicu/CycleAnomalies.gif
Leif Svalgaard (09:50:20) :
to
vukcevic (08:50:04) :
I never claimed excellent correlation with sunspot number. What I claim is the excellent correlation with recorded magnetic polar fields
Yet you keep showing the sunspot graph back a 1000 years.
The polar field correlation has too few degrees of freedom to be significant and fails for 1964-1965.
I will agree with that, failed for 1 year out of 45 (1964-2009).
I call that excellent correlation.
Geoff Sharp (01:54:50) :
Carsten Arnholm, Norway (15:58:16) :
You find the software here, if you need it
http://arnholm.org/astro/sun/sc24/sim2/
If your still around, is the sim1 version still available?
Reply: “you’re” ~ charles the grammar nazi moderator.
The sim1 is at http://arnholm.org/astro/sun/sc24/sim1/ but currently only as screen shots (sim2 is available as windows executable). As there seems to be renewed interest, I shall seek to make also the sim1 executable available for download. Give me a few days. I have also ported sim2 to Linux now.
Good to know we also have a grammar moderator, I need it sometimes 🙂
Carsten Arnholm, Norway (14:41:40) :
Thanks Carsten. I will look out for it.
vukcevic (14:09:29) :
“fails for 1964-1965.”
I will agree with that, failed for 1 year out of 45 (1964-2009).
I call that excellent correlation.
This is because you do not [apparently] understand the ‘number of degrees of freedom’. There are not 45 independent values, but only about 5. To take a similar example, if I fail in my prediction of the sunspot maximum, I can say that since my prediction of SSN for 2000 was correct, I correctly predicted the values on Jan 1st, 2000, and on Jan 2nd, and 3rd, and … March 1st, and Oct 21st, etc for all days of 2000, so my prediction fails only once in 366 cases. This is not something we ought to spend a lot of time on here, you have some reading to do about time series with autocorrelation.
vukcevic (14:09:29) :
I will agree with that, failed for 1 year out of 45 (1964-2009).
I call that excellent correlation.
Because the polar field data has a very high serial autocorrelation there are not a thousand points or even 50 points, but only about 5 independent data points that determine the curve. This means that one can very easily find a function with about five parameters that has an excellent correlation with the data. Here is one that I just whipped up:
http://www.leif.org/research/Vuk-polar-fields-1.png
As you can see it has an even higher correlation than yours. A high correlation does not mean that the formula expresses any physical or significant relationship.
Most of the point-to-point variation of the PF is just noise and is not significant. You can clearly see the larger noise level of the Mount Wilson instrument before year 10 (years are counted from 1966.67 when we have the first good data point from MWO).
Dr. Svalgaard
I can only say, thank you for your effort. Your latest contribution is just may confirm once for all, validity of my formula.
1. If I reproduce your equation I get something totally different, what you show. Regardless of the numbers involved, anyone familiar with basic maths knowledge can observe that your equation is a product of a parabola (goes to infinity) and a sin wave. Product of this is an oscillating bell shape curve going to infinity, and that is exactly what I get if I put your formula into Excel.
I am sure you will not mind submitting the Excel entry that you used for your chart, so that I, and anyone else interested, can reproduce the curve.
2. In my case, everyone can see how I devised my formula. It is correct that it has only 4 degrees of freedom (2 orbits, phase shift and the amplitude) and that is that.
You have not shown us how you have obtained your formula. I can see two ways:
a) using an algorithm to break my formula into components (in which case is the same formula in different guise), and you have validate my work.
b) Using available data for a simulating algorithm to produce your formula, which again proves I am correct, since two curves for this time-period appear to be almost identical, and your formula is just expressed in different (less meaningful form), selected by programmer of software used. As you are well aware, same mathematical expression can be shown in many different forms, but selecting most meaningful is the preferable one.
R=98% and R=97% are not here or there (if you used polar data for algorithm it is slightly better then my numerical entries, perhaps correcting my Saturn orbit as you previously suggested , or as Mr. Sharp suggests using Uranus effect we could get that 1% more, but that is irrelevant)
Since I have reproduced my formula from 1850-2030 (and can do for any period you whish), I expect you to show us in a graphical form the same.
To conclude I expect to see following:
– Excel entry for your formula (mine is readily available)
– method you used to produce your formula (mine is J-S orbits)
– show us what your numbers mean in context of the Sun’s properties (mine are precisely defined)
– Produce a chart spanning period of at least 200 or more years.
If all of these are not shown in a way that can be verified, I shall assume and perhaps many others, that I have won this argument conclusively.
All the best in your endeavours.
NASA has just changed the name of the project from Solar Cycle 23 to Solar Cycle 24. I would love to have attended that meeting.
B.9 CAUSES AND CONSEQUENCES OF THE MINIMUM OF SOLAR CYCLE 24
Clarified March 10, 2009: All references to “Solar Cycle 23” have been updated to “Solar Cycle 24.” Reference in Section 1 to “Solar Cycle 22” has been updated to “Solar Cycle 23.”
See the changed text here:
Causes and Consequences of the Minimum of Solar Cycle 24
http://nspires.nasaprs.com/external/viewrepositorydocument/cmdocumentid=178281/B.9%20CCMSC_clarified.pdf
Talk about Freudian Slips, what Solar Cycle is it anyway? No wonder they can’t make predications!
Mike
Breaking News Update: Solar Cycle 24 has ended according to NASA.
NASA has just changed the name of the project from Solar Cycle 23 to Solar Cycle 24. I would love to have attended that meeting.
B.9 CAUSES AND CONSEQUENCES OF THE MINIMUM OF SOLAR CYCLE 24
Clarified March 10, 2009: All references to “Solar Cycle 23” have been updated to “Solar Cycle 24.” Reference in Section 1 to “Solar Cycle 22” has been updated to “Solar Cycle 23.”
See the changed text here:
Causes and Consequences of the Minimum of Solar Cycle 24
http://nspires.nasaprs.com/external/viewrepositorydocument/cmdocumentid=178281/B.9%20CCMSC_clarified.pdf
Talk about Freudian Slips, what Solar Cycle is it anyway? No wonder they can’t make predications!
But it gets even better. NASA has just declared that Solar Cycle 24 is over. Read the first paragraph in the above PDF:
1. Scope of Program
In 2009, we are in the midst of the minimum of solar activity that marks the end of Solar Cycle 24. As this cycle comes to an end we are recognizing, in retrospect, that the Sun has been extraordinarily quiet during this particular Solar Cycle minimum. This is evidenced in records of both solar activity and the response to it of the terrestrial space environment.
Obviously someone made an error when edition the text of the original document and did not catch their mistake. Quick, make your own backup copy of this “Great Moment in Science”.
Mike
vukcevic (03:03:52) :
If I reproduce your equation I get something totally different, what you show.
The formula is what is shown on the chart. In Excel you replace the ‘year’ with the name of the cell that contains the year, e.g. A5
The whole point of my little exercise was to show that there are only a few degrees of freedom and that just because to have a high correlation does not mean that you have captured the physics, as surely the parabola is not correct physics.
If one spends more than just a few minutes on this (like 15 or so) it is easy to produce a formula that has excellent correlation (94%) and even might express some of the physics behind the whole thing. On http://www.leif.org/research/Vuk-polar-fields-2.png
you can see one example. The red curve shows observed values. The formula is given on the chart. Replace the ‘y’ with the name of a cell in Excel that contains a year to reproduce the plot.
The two factors are
1) 215 + 68 * SIN(0.06052*(y-1966.67)+1.74
2) SIN(0.308*(y-1966.67)-0.96)
The second factor simply reflects the fact that the polar fields have a 20-year period. 2pi/0.308 = 20.4 years. The first factor simply reflects the fact that there is a long-term variation of solar activity. The values I have used correspond to a period of 104 years: 2pi/0.06052 = 103.82 and was chosen simple to conform to the observed trend (the red points).
The point, again, is simply to show that it is easy to get a high correlation and that a high correlation does not imply any physics.
Dr. Svalgaard
Let me first say, I admire your achievements in field of solar science, as well as your efforts to help out to everyone with their questions. I have also personally benefited from your advice, which I value greatly.
It gives me no particular pleasure to declare that you have not only failed to win (up to now one could have called it a draw), but finally lost this particular argument.
I do not wish to be too big-headed about it, I knew while sunspot formulae were to complex and to hard to prove, one way or the other, the polar field strength open the door, again thanks to you.
To anyone of scientific mind it is clear that an alternative formula argument has failed. Your initial attempt was abandoned (to me reasons are obvious), and a new formula has been assembled, which more and more looks like mine, but it does not make a case.
No serious attempt was made to answer crucial questions I put forward.
I attach a graphic representation of the attempt to discredit my effort.
http://www.geocities.com/vukcevicu/DrLSReply.gif
My formula had baptism of fire and survived it with honour.
As far as I am concerned this matter is now closed, but I would welcome help with hypothesis from anyone who may whish to contribute.
Thanks again.
Geoff Sharp (15:17:26) :
Thanks Carsten. I will look out for it.
Ok, I have updated my pages with the solar motion simulator programs, so you can download and run them both on either windows or linux:
SolarMotion (=”Sim1″)
http://arnholm.org/astro/sun/sc24/sim1/
SolarMotion2 (=”Sim2″)
http://arnholm.org/astro/sun/sc24/sim2/
For the windows version, you have to be using Win2000 or newer, no Win9x. Be sure to install the MS redistributable.
The linux versions have been built with Kubuntu 8.04. You need to extract the .tar.gz file and run the .install script in sudo mode. Then start it using
$ solarmotion (for sim1)
$ solarmotion2 (for sim2)
vukcevic (09:39:12) :
To anyone of scientific mind it is clear that an alternative formula argument has failed.
I have failed to get across that both formula fail, as they are just numerology. The sun doesn’t work like that, following strict cycles. You can compare the two formulae here http://www.leif.org/research/Vuk-polar-fields.png as well as the sunspot number curve shifted half a cycle. Neither of the formulae are any good [e.g. they both fail in 1965]. The point that I failed to get across is that a high correlation over a limited interval does not in itself ensure that there is any physical reality behind the formula. But, I’ll say good riddance to your departure on this matter.
Leif Svalgaard (13:11:47) :
vukcevic (09:39:12) :
You can compare the two formulae here http://www.leif.org/research/Vuk-polar-fields-3.png
Leif Svalgaard (13:11:47) :
The sun doesn’t work like that, following strict cycles.
And for the same reason that is why doing frequency tests on recurring periods of grand minima is rather pointless. As I have shown there is a recurring 172 yr AVG grand minima period, but the modulation of that period can vary significantly based on the strength of the Jupiter/Saturn position of the era.
There is a cycle, but like most things in nature is subject to controlling factors.
http://landscheidt.auditblogs.com/
Geoff Sharp (14:53:46) :
based on the strength of the Jupiter/Saturn position of the era.
You have still not defined and quantified the ‘strength’ in terms of angular momentum, which is all the Sun can see, and which cannot have any effect anyway.
Carsten Arnholm, Norway (12:15:39) :
Thanks Carsten, but sorry to be a pain but the sim1 is returning an error and wont start(this application has failed to start because the application configuration is incorrect). I also have trouble with sim2 (it draws diff curves in relation to yours shown on your site). I have the ms redistributable installed but not sure what what to do as far as win2000 is involved, I am running XP pro.
The information is invaluable but finding it hard to install. If you have time one day could I make a request. It would be great to fill in the missing time frames on your sim1 page, so we have a complete history back to the start of the Oort. It would be a great reference.
If I manage to get it installed correctly I will do it and post it as a reference on Carl’s Blog.
Leif Svalgaard (15:09:59) :
Geoff Sharp (14:53:46) :
based on the strength of the Jupiter/Saturn position of the era.
You have still not defined and quantified the ’strength’ in terms of angular momentum, which is all the Sun can see,
I have, just not in the terms you expect. AM causes the disturbance which is what needs to be quantified, the AM value during the disturbance tells us nothing. Currently this is observed on Carl’s graph and I have summarized the disturbances into 4 categories. Type “A” weak and strong & Type “B” weak and strong. This is also observed in Carsten’s Sim1 and Sim2 programs in the form of altered retrograde motions etc.
http://users.beagle.com.au/geoffsharp/typeab.jpg
These categories are more than sufficient when working with grand minima strength, but I will progress to a more analytical quantification in the future to keep the boffins at bay.
http://users.beagle.com.au/geoffsharp/carsten.jpg
Geoff Sharp (14:53:46) :
……………..
There is a cycle, but like most things in nature is subject to controlling factors.
You are absolutely right. This is clearly shown in:
http://www.geocities.com/vukcevicu/CycleAnomalies.gif
controlling formula, identifying all anomalies since 1600’s, is based on orbital harmonics (4S & J+U, numbers are rounded off). Pronounced minima around 1800, 1910 and a possible 2020, can also be seen as identical paths on the Carsten’s charts. The original formula was published some 5 years ago:
http://xxx.lanl.gov/ftp/astro-ph/papers/0401/0401107.pdf
Geoff Sharp (17:24:58) :
I have, just not in the terms you expect. AM causes the disturbance which is what needs to be quantified
disturbance in what?
vukcevic (23:58:32) :
Geoff Sharp (14:53:46) :
There is a cycle, but like most things in nature is subject to controlling factors.
You are absolutely right. This is clearly shown in:
Can we have some comments on why the controlling cycles are so different for the nobrainer and vuk charts? Astronomical cycles tied to the planets should stay unchanged for thousands [if not millions] of years to the accuracy of solar cycles.
Leif Svalgaard (12:05:11) :
vukcevic (23:58:32) :
Geoff Sharp (14:53:46) :
……………
Can we have some comments………
My knowledge on these matters is severely limited. Formulae I developed take the most basic and simplistic view of the Sun plus two major planets, as if they were in a single plane, planets with perfect circular orbits. My understanding is, that is not the case, and their heliocentric latitude/longitude relationship is infinitely variable.
99.9% of the current scientific establishment rejects possibility of any suggested connection, and in the 0.1% left may be very few with any interest in the matter, and of those who do have, may not whish to jeopardise their good name in associating themselves with such ideas. Possibly someone with sufficient knowledge and interest, may take a plunge and work out theory if physics does allow for it. For now it is not science, just a speculation and no more.