Guest Posting by Ira Glickstein
Santa brought us a new Sunspot prediction to be added to NASA’s incredibly high series of at least five ill-fated predictions starting in 2006. NASA’s latest peak Sunspot Number for Solar Cycle #24 (SC24) is down 60% from their original, but it still seems a bit too high, judging by David Archibald’s recent WUWT posting that analogizes SC24 and SC25 to SC5 and SC6 which peaked around 50, during the cold period (Dalton minimum) of the early 1800’s.
According to Yogi Berra “It’s tough to make predictions, especially about the future.” Team leader Dr. Mausumi Dikpati of NASA’s National Center for Atmospheric Research and Solar physicist Dr. David Hathaway of the National Space Science & Technology Center have most likely learned that lesson well, having predicted, back in March 2006, that SC24 would start by the end of 2006 or early 2007 and would peak 30% to 50% higher than SC23, which would yield counts of 156 to 180. The latest prediction is 64 (I love their precision :^) but I predict it will have to be reduced further, kind of like an after-Christmas sale :^)
[NOTE added 28 Dec 9:45PM. See clarification comment by: John from CA, December 28, 2010 at 1:44 pm. I was mistaken in conflating NASA with NOAA in the graphic and discussion, wrongly assuming they coordinated their Sunspot predictions. The base chart, as labeled, is from NOAA but the predictions are from Dikpati and/or Hathaway at NASA, but later ones, on a NASA website, may be personal, not official. Thanks John from CA and sorry for my ignorance of government organization. Ira]
My graphic traces the downward progression of NASA Sunspot predictions, superimposed over NASA’s NOAA’s latest chart of actual Sunspot Numbers. SC23 is shown from its peak in 2000 to its demise in 2009, along with the rise of SC24 up to the latest November 2010 data. The red hoop, peaking at 90, is left over from their previous prediction and should be replaced by their new prediction in January. [Click graphic for larger version].
As indicated, SC23 peaked at a count of 120 around January 2000. It is instructive to read NASA’s March 2006 predictions (and somewhat humorous until you realize we paid for it). Some direct quotes [emphasis added]:
“The next sunspot cycle will be 30% to 50% stronger than the previous one,” [Dikpati] says… Dikpati’s prediction is unprecedented. In nearly-two centuries since the 11-year sunspot cycle was discovered, scientists have struggled to predict the size of future maxima—and failed. Solar maxima can be intense, as in 1958, or barely detectable, as in 1805, obeying no obvious pattern.
The key to the mystery, Dikpati realized years ago, is a conveyor belt on the sun…
Hathaway … explains: “First, remember what sunspots are–tangled knots of magnetism generated by the sun’s inner dynamo. A typical sunspot exists for just a few weeks. Then it decays, leaving behind a ‘corpse’ of weak magnetic fields.”…
“The top of the conveyor belt skims the surface of the sun, sweeping up the magnetic fields of old, dead sunspots. The ‘corpses’ are dragged down at the poles to a depth of 200,000 km where the sun’s magnetic dynamo can amplify them. Once the corpses (magnetic knots) are reincarnated (amplified), they become buoyant and float back to the surface.” Presto—new sunspots!
All this happens with massive slowness. “It takes about 40 years for the belt to complete one loop,” says Hathaway. The speed varies “anywhere from a 50-year pace (slow) to a 30-year pace (fast).”
When the belt is turning “fast,” it means that lots of magnetic fields are being swept up, and that a future sunspot cycle is going to be intense. This is a basis for forecasting: “The belt was turning fast in 1986-1996,” says Hathaway. “Old magnetic fields swept up then should re-appear as big sunspots in 2010-2011.”
Like most experts in the field, Hathaway has confidence in the conveyor belt model and agrees with Dikpati that the next solar maximum should be a doozy. But he disagrees with one point. Dikpati’s forecast puts Solar Max at 2012. Hathaway believes it will arrive sooner, in 2010 or 2011.
“History shows that big sunspot cycles ‘ramp up’ faster than small ones,” he says. “I expect to see the first sunspots of the next cycle appear in late 2006 or 2007—and Solar Max to be underway by 2010 or 2011.”
Who’s right? Time will tell. Either way, a storm is coming.
Did Dikpati and Hathaway honestly believed they had cracked the Sunspot code that had eluded science for two centuries? In hindsight, we all know they were wrong in their heady predictions of a “doozy”. (A doozy, according to Webster is “an extraordinary one of its kind”. NASA expected SC24 to be extraordinarily intense. But it is shaping up to be extraordinarily weak, so they at least get credit for using the correct word :^)
But, were they being honest? Well, Hathaway had long been aware of the relationship between Sunspot counts and climate, writing:
Early records of sunspots indicate that the Sun went through a period of inactivity in the late 17th century. Very few sunspots were seen on the Sun from about 1645 to 1715. … This period of solar inactivity also corresponds to a climatic period called the ‘Little Ice Age’ when rivers that are normally ice-free froze and snow fields remained year-round at lower altitudes. There is evidence that the Sun has had similar periods of inactivity in the more distant past. The connection between solar activity and terrestrial climate is an area of on-going research.
Is it possible that their prediction was skewed to the high side by the prevalent opinion, in the Inconvenient Truth year of 2006, that Global Warming was “settled science”. Could it be that they felt pressured to please their colleagues and superiors by predicting a Sunspot doozy that would presage a doozy of a warm spell?
It seems to me that NASA has a long history of delayed Sunspot predictions, particularly when the trend was downward. They seem to have waited until the actual counts forced them to do so.
Have a look at the graphic. SC23 SC24 [thanks Steeptown December 27, 2010 at 11:37 pm] was supposed to start by early 2007, but it did not. Yet, it took them until October 2008 to revise their prediction of a later start and lower peak (137) and then they dropped it further in January 2009 (predicting a peak of 104 to occur in early 2012).
I am not any kind of expert on Sunspots, yet it was clear to me, nearly two years ago, that 104 was way too high so I predicted a peak of 80 and moved the date of that peak to mid-2013. NASA eventually reduced their peak to 90, and just this month down to 64, and they moved the peak date to mid-2013. My latest prediction is 60, to occur in early 2014, but I believe I may still be a bit too high.
With apologies to Pete Seeger:
Where have all the sunspots gone? NA-SA search-ing,
Where have all the sunspots go-ne? NASA don’t know.
Where have all the sunspots gone? Global Cooling, anyone?
Will NASA ever learn? Will NA-SA ev-er learn?
Where has all the carbon gone? Green-house gas-es,
Where has all the carbon go-ne? Come down as snow!
Where has all the carbon gone? Heating houses, everyone,
Will NASA ever learn? Will NA-SA ev-er learn?
Where has Global Warming gone? Point not tip-ping,
Where has Global Warming go-ne? Its gonna slow.
Where has Global Warming gone? Normal seasons of the Sun,
Will NASA ever learn? Will NA-SA ev-er learn?
tallbloke says:
December 29, 2010 at 9:16 am
“He calibrated other peoples data against the geomagnetic and auroral data, nothing to do with threshold or specks. Accept it.”
What geomagnetic data? You told me yesterday on the Trenberth thread that there are no geomagnetic data for the 1804-1817 period.
The Dalton minimum was more than just 1806-1817. For that interval he used auroral counts which were calibrated against before and after geomagnetic records. An additional point is that the data is so sparse that our knowledge of what activity actually was is very poor [I estimate can be off by 50% either way].
Leif Svalgaard says:
December 29, 2010 at 9:23 am
The Dalton minimum was more than just 1806-1817. For that interval he used auroral counts which were calibrated against before and after geomagnetic records. An additional point is that the data is so sparse that our knowledge of what activity actually was is very poor [I estimate can be off by 50% either way].
Thanks for the clarification. If activity was low, then presumably the auroras that were seen must have been well to the north. Who recorded them from where? Scandinavia?
Great cross-discussion between Leif and Geoff. As a layman I have no basis to judge who is correct, except that Leif seems more courteous, which I appreciate.
Which counting method is “right” is less important than the accuracy of the relative magnitudes of historical and current Solar Cycle counts. In particular, I would like your expert opinions of David Archibald’s Figure 9, analogizing SC3, SC4, SC5 and SC6 (1777 thru 1821, prior to and during the Dalton minimum) to SC22 and SC23 and, by implication predicting that SC24 and SC25 may peak at around 50, which is what triggered me to start this topic thread.
If the analogy holds, SC24 may peak as late as 2015, which is over a year later than any NOAA/NASA/Hathaway prediction I have yet seen, and four years later than some of the original predictions.
tallbloke says:
December 29, 2010 at 9:34 am
Who recorded them from where? Scandinavia?
Sweden
Ira Glickstein says:
December 29, 2010 at 9:42 am
Great cross-discussion between Leif and Geoff. As a layman I have no basis to judge who is correct, except that Leif seems more courteous, which I appreciate.
Which counting method is “right” is less important than the accuracy of the relative magnitudes of historical and current Solar Cycle counts. In particular, I would like your expert opinions of David Archibald’s Figure 9, analogizing SC3, SC4, SC5 and SC6 (1777 thru 1821, prior to and during the Dalton minimum) to SC22 and SC23 and, by implication predicting that SC 24 and SC25 may peak at around 50, which is what triggered me to start this topic thread.
Our knowledge of SC5 is very poor and uncertain [perhaps up to 50% off], so no good conclusions can be drawn. And, you are correct in stating that the relative magnitudes are what is important. The sunspot counts before 1877 [by Wolf and others] are not suited on their own to provide good relative magnitudes. Wolf discovered a neat [and we know today, correct] way of calibrating the sunspot counts, using the influence of solar far Ultra Violet on the ionosphere and the resulting geomagnetic variations. Wolf, of course, did not know the physics behind this, but got the method right anyway. This is an example of a correlation without a mechanism [at the time].
Leif Svalgaard says:
December 29, 2010 at 9:53 am
Sweden
Are the records publicly accessible?
Geoff Sharp seems to think that a modern 150 mm telescope will “see more”. Frankly, this is irrelevant. I used to make visual observations in the 1960s and report them to the BAA. I found that in good seeing conditions the granulation of the solar surface was just visible in a 2.5 cm telescope and really stood out in a 7.5 cm telescope, the problem lay in deciding visually what was a small pore and what was merely an artifact of the granulation. My problem with the observations during the Maunder Minimum lies with the type of telescope in use then. Non-achromatic refractors with resultant abysmal resolution and reflectors with mirrors made of speculum metal which would rapidly deform with solar heating (particularly a secondary mirror which would be subject to enhanced heating?) In my active time as an amateur observer I experimented with a simple non-achromatic refractor and found that only the largest sunspot groups could be seen. Has anybody attempted the reconstruction of a 1700s telescope and used it for solar observations?
tallbloke says:
December 29, 2010 at 10:11 am
Are the records publicly accessible?
Google ‘swedish auroral catalog rubenson’
Thanks Leif for your prompt reply. I have had another look at the high and low 2007 predictions of the SC24 panel. As I understand it, the vote was nearly evenly split: 6 votes for the low (90) prediction, which you supported, and 5 votes for the high (140). The bottom figure shows the high peaking towards the end of 2011 and the low peaking in 2012.
If, as you say, SC5 is up to 50% off, it could really be as high as around 75 (which would be a feather in your cap for your prescient 2005 prediction of 70) or as low as 25. I have not been able to find the predicted peak year for your 2005 prediction. What is it and what is your latest and greatest prediction? In your May 2010 paper you say “Who knows?” for SSN and I could not find a date for the peak. Am I correct to assume the predicted peak year will be 2014 or later?
Ira Glickstein says:
December 29, 2010 at 9:42 am
I would like your expert opinions of David Archibald’s Figure 9, analogizing SC3, SC4, SC5 and SC6 (1777 thru 1821, prior to and during the Dalton minimum)
Wolf himself when he compiled his sunspot series has this to say about SC5:
a. Arago, Herschel, Fritsch, Flaugergues saw 1801-1802 ‘rich groups.
b. in 1803-1804 this richness was extraordinary.
c. Fritsch saw in 180201803 often more than 50 large spots.
d. Eimbeke states that he has never seen as persistent and often occurring spots as in 1803.
e. Huth says that he had never seen as many and as large spots as in 1804.
f. Huth, Bode, Flaugerguess mention large spots in 1805
g. first in 1807 did the spots begin to abate
h. Fritch, Bode, Gruithusen and Ende agree that around 1810 the sun only had few spots and those were very small.
i. Fritch counted in 1817 often more than 100 spots per day, several naked-eye spots
Based on these observations of sunspots Wolf gave SC5 a rather high count of 75 in his list of 1874. When Wolf got Rubenson’s auroral catalog, he decided in his 1882 list to reduce [based on aurorae, not sunspots] the size of SC5 to 47.5 and hence was born the Dalton minimum.
Leif Svalgaard says:
December 29, 2010 at 6:22 am
The original telescope has been in continuous use since 1856 to this day.
The daulton ended around 1830, why use a scope that was constructed in 1856 and is superior to other earlier scopes, the event was over 25 years before this scope was constructed, how can the Maunda Min be compared using this scope.
Messier’s favorite instrument was a 32-feet FL, 7.5-inch aperture Gregorian reflector with mag. 104x. Bailly has computed that the effective aperture of this instrument was equivalent to a 3.5-inch refractor. Even worse was the situation for the old Newtonian reflector he came over with from Delisle, which was an 8-inch but as effective as a 2.5-inch refractor only, so it was little used, although it seems this was the “original” instrument at Hotel de Cluny, Messier’s observatory. Later he preferred to use several 3.5-inch (90 mm) achromatic refractors, which were usually about 3.5 feet long and magnifying 120 times. He selected to use these scopes because they were the best accessible instruments for him.
It remains to state that all of Messier’s instruments could probably not compete with a modern 4-inch refractor
http://i446.photobucket.com/albums/qq187/bobclive/Wolf-Telescope.png
Lief and Ira,
Sorry if I was a bit hard nosed about this post — I’d hate to see NASA discounted for something they had nothing to do with and, though I’m not a big fan of NASA’s news department, I am a big fan of their and our other amazing science agency accomplishments.
In the face of it, I thought I’d share an antique family recipe that is best served with a hot cup of coffee or a good glass of red wine.
Anise Toast
Ingredients:
6 eggs
1 cup sugar
1 cup cake or bread flour
2 tsp Anise seeds
pinch of salt
Instructions:
6 eggs separated: beat whites stiff — set aside; beat yokes ’til light yellow and then gradually add 1 cup sugar and 2 tsp Anise seeds until throughly mixed.
Carefully fold the the egg whites into mixture.
Finally, add 1 Cup of either sifted cake or bread flour and a pinch of salt.
Lightly fold into greased pan(s) and bake at 325 degrees F for 25 minutes or until done (use a tooth pick to test)
Let stand until cool and cut in the pan into strips (bread knife works best). Remove from the pan, cover with a dish towel, and let dry overnight on a rack. Toast in the oven the next day (sides slightly brown) and serve.
Happy New Year
Leif Svalgaard says:
December 29, 2010 at 10:40 am
Based on these observations of sunspots Wolf gave SC5 a rather high count of 75 in his list of 1874. When Wolf got Rubenson’s auroral catalog, he decided in his 1882 list to reduce [based on aurorae, not sunspots] the size of SC5 to 47.5 and hence was born the Dalton minimum.
Clearly you believe the count should be higher. Did Wolf’s original count of 75 influence your prediction for sc24?
RE:
John from CA says:
December 29, 2010 at 11:04 am
Note: The pan should be 8″ x 11″ x 2″ or a bit larger but not deeper. Strips are pan width and cut to preference (1/2″, 3/4″, or 1″)
Leif Svalgaard says:
December 29, 2010 at 9:53 am
Sweden
Are the records publicly accessible? 8 brb
————–
Leif,
I warmly agree with the above comment of John F. Hultquist.
Today I learned from you the existence of solar cycles existing in stars similar to our sun’s ~ 11 yr cycle (actually it is a ~22 yr cycle, n’est-ce pas?). I thank Berényi Péter (@ur momisugly December 28, 2010 at 12:02 am) for asking you the great question about it.
Also, I learned about the interesting dynamic interaction of a group of scientists on NOAA’s Solar Cycle 24 Prediction Panel. Thanks for the insight into the solar science community; it adds to the credibility of science.
Leif – Please note that I am in SF and still have that bottle of Scandinavian white lightning in my daughter’s freezer here. Should I drop it by your place? Hope there is not a charge for late tuition payment. : )
Happy Holidays to all . . . . but especially to Anthony & his family and of course the WUWT band of merry moderators!!
John
Brian Carter says:
December 29, 2010 at 10:17 am
Geoff Sharp seems to think that a modern 150 mm telescope will “see more”. Frankly, this is irrelevant.
A 80mm refractor with good lenses and good viewing has a resolution of 1051 kilometers on the solar surface, a 150mm refractor can see down to 558 kilometers. Locarno and Catania count specks that are 700 kilometers across, the older telescopes cannot do this. But yes I agree the old lenses would have also struggled.
There is no such thing as “good viewing” wrt. refractors, other types of telescopes, or astronomical observation in general. A “good viewing” sounds like a TV-program experience.
The proper term is “seeing”, it describes the amount of turbulence in the Earth’s atmosphere. Poor seeing corresponds to much turbulence. The seeing is a function of weather, location and other factors, but independent of the telescope.
It is also incorrect to say that a telescope with a given aperture has a given resolution. The resolution depends on the magnification, i.e. the focal length and optical quality of the eyepiece.
And finally, with visual observation, the final resolution also depends on the observer. Registering details requires training.
Robuk says:
December 29, 2010 at 10:49 am
It remains to state that all of Messier’s instruments could probably not compete with a modern 4-inch refractor
Interesting, but irrelevant, because the standard is still the old Fraunhofer 80 mm refractor. Observers strive to match their observations to what the 80 mm could see.
John from CA says:
December 29, 2010 at 11:04 am
tallbloke says:
December 29, 2010 at 11:05 am
Clearly you believe the count should be higher.
For starters it should be 20% higher because of Waldmeier. I’m not so sure the auroral data are as good as Wolf thought. This is something I’m looking into with the hindsight of our modern understanding of the aurora. What strikes me is that all those seasoned observers who had just witnessed a strong cycle 4 tell us us cycle 5 had extraordinarily ‘richness’ of spots. That doesn’t sound like a Grand Minimum to me.
Did Wolf’s original count of 75 influence your prediction for sc24?
Of course not, the Sun doesn’t remember that far back. Our paper http://www.leif.org/research/Cycle%2024%20Smallest%20100%20years.pdf explains exactly on what our prediction is based.
John from CA says:
December 29, 2010 at 11:39 am
John Whitman says:
December 29, 2010 at 3:00 pm
Leif – Please note that I am in SF and still have that bottle of Scandinavian white lightning in my daughter’s freezer here. Should I drop it by your place? Hope there is not a charge for late tuition payment. : )
We were just in town celebrating our wedding anniversary. You are welcome to drop by any time.
Carsten Arnholm, Norway says:
December 29, 2010 at 4:18 pm
Geoff Sharp seems to think that a modern 150 mm telescope will “see more”. Frankly, this is irrelevant.
Also irrelevant because the sunspot counts from 150 mm telescopes match those obtained from the original 80 mm, so theory aside, empirical evidence says the telescope doesn’t matter, once it is good enough i.e. magnification x64.
The proper term is “seeing”, it describes the amount of turbulence in the Earth’s atmosphere. Poor seeing corresponds to much turbulence. The seeing is a function of weather, location and other factors, but independent of the telescope.
And finally, with visual observation, the final resolution also depends on the observer. Registering details requires training.
it takes several years to learn how to observe sunspots. Here are typical ‘learning curves’ for Wolfer and another assistant: http://www.leif.org/research/K-Factor-Learning-Curve.png
10 years seems to be a good number.
After much searching, Landscheidt predicted SC 24 maxima at 2011.8 ±0.16 years (around 2 months) with R<80.
http://bourabai.narod.ru/landscheidt/extrema.htm So Leif have you read Landscheidt's paper?
In Lanscheidt's words: A forecast experiment could help to decide whether this is correct. Those are the kind of words I expect from someone looking at all the angles.
I, for one, wish that Mr. Sharp would keep a more civil tongue. His interjections don’t help his argumentation.
Geoff Sharp says:
December 29, 2010 at 3:38 pm
A 80mm refractor with good lenses and good viewing has a resolution of 1051 kilometers on the solar surface, a 150mm refractor can see down to 558 kilometers. Locarno and Catania count specks that are 700 kilometers across, the older telescopes cannot do this.
Daytime seeing limits the resolution to 1-3″, typically 1.5″ on a very good day. Only on rare occasions [1% of the time] does the seeing allow a resolution of 1″. This depends on the atmosphere, not on the telescope. 1″ is 700 km, 1.5″ is 1500 km. Schaefer made the definitive study of this http://www.leif.org/EOS/Schaefer-1993ApJ-411-909.pdf
He points out that: “Even in modern times, all sunspot counting is done with the human eye and a small telescope [he recommends 50 mm at high magnification]. So the primary record of solar activity since 1610 is based on the human eye seeking sunspots at the limits of vision” [ultimately determined by the seeing]
Carsten Arnholm, Norway says:
December 29, 2010 at 4:18 pm
It is also incorrect to say that a telescope with a given aperture has a given resolution. The resolution depends on the magnification, i.e. the focal length and optical quality of the eyepiece.
I am assuming the end magnification is 64x. My point is that two telescopes set up with 64x magnification and differing aperture sizes will have different resolution. You know this to be true, but it is not getting through to Leif, even though it has been explained many times. I have also had this confirmed from the head sunspot recorder at Locarno who is now retiring after 5o years of service.
I have two telescopes of my own that are setup similar (70mm & 110mm). The 110mm definitely sees more.
KBK says:
December 29, 2010 at 6:20 pm
I, for one, wish that Mr. Sharp would keep a more civil tongue. His interjections don’t help his argumentation.
Dont worry Leif knows how to dish it out, perhaps you have not been around long enough to witness. But I will take your comments onboard.
Ira Glickstein says:
December 29, 2010 at 10:36 am
Am I correct to assume the predicted peak year will be 2014 or later?
We did not attempt to predict the peak date. Our paper mentions 2010 but that was a ‘nominal’ date [obtained by adding 10 to the previous peak date 2000], so not really a prediction. 2014 sounds like a good peak date, but the whole notion of a peak date is somewhat mushy. I think there is a good chance that SC24 will look something like SC14. Here is SC14: http://www.leif.org/research/SC14.png
Try to find a meaningful peak date…
Geoff Sharp says:
December 29, 2010 at 6:44 pm
I am assuming the end magnification is 64x. My point is that two telescopes set up with 64x magnification and differing aperture sizes will have different resolution.
This is irrelevant as the seeing is limiting what you can see. Bill Livingston with the biggest solar telescope in the world [a solar image 4 feet across] does not see more spots than the standard 80 mm refractor.
I have two telescopes of my own that are setup similar (70mm & 110mm). The 110mm definitely sees more.
As Schaefer explains it takes years of training to seek out sunspots [and it is a bit garder on the smaller scope], and you are not nearly there. What is the magnification of the two scopes? And the filters?
You ignore the fact that Locarno [times 0.6] reports the same as SIDC which reports essentially the same as Keller/Friedli [apart from SIDC being a tad too low].
I have also had this confirmed from the head sunspot recorder at Locarno who is now retiring after 5o years of service.
Yet his count is substantially identical to Keller/Friedli with 80 mm. Did you ask the question correctly? If you asked: does a 150 mm telescope have greater resolution than a 80 mm telescope? the answer might be yes. If you had asked: given that seeing limits what you can see to 2″, do you see more with a bigger telescope? the answer would be no.
I realize that you may not understand the subtlety of seeing and telescopes, but you should be able to compare simple numbers. So take Locarno and compute the count for several months. Then take SIDC and do the same. You can even take Keller/Fridli [from here ]. If you do not wish to do this then just look at the graphs I have already shown you [where I did it for you]. One more time: 0.6 * raw Locarno = 0.6 * raw Keller = SIDC, so count(150 mm) = count(80 mm) because for both the seeing is the limiting factor.