I missed this earlier this week from NASA, I got a bit distracted with other things.
Sixty two – that’s the new number from Hathaway on April 4th, have a look:
They write at NASA MSFC
Current prediction for the next sunspot cycle maximum gives a smoothed sunspot number maximum of about 62 in July of 2013. We are currently over two years into Cycle 24. The predicted size would make this the smallest sunspot cycle in nearly 200 years.
It’s quite a climbdown for Dr. Hathaway from his earlier predictions. Let’s give him credit for not trying to “hide the decline”.
Anthony,
Thanks for posting this. I’ve been checking it daily and the other ‘projections’ just
do not want to have anything to do with the data. Cheers.
NASA says “Predicting the behavior of a sunspot cycle is fairly reliable once the cycle is well underway (about 3 years after the minimum in sunspot number occurs ”
Predicting the sunspot cycle after it is completely over is even more reliable.
Perhaps in another 3 years NASA will be looking back one year and finally have a final “prediction”.
Interesting and useful and bears out what a lot of us have been saying.
However do bear in mind for ongoing cooling it is the odd cycles which count decisively. Even ones, eg SC24, which we are now in normally mean a cooler earth anyway. In odd ones Earth temp is best correlated with solar activity so a weak SC25 will be the ‘cooling clincher’. For such projections see slide 17 in the pdf in my submission to the UK Select committee enquiry into the extremely cold & snowy December 2010 crisis – short link- http://bit.ly/hEmBqG
Thanks, Piers Corbyn
I know Hathaway is trying to put on a good face and honor the observational data, but this succession of “adjusted forecasts” just makes me laugh.
“Predicting yesterday’s weather – – – tomorrow”
What is the opposite of “prescient”?
I think the previous number was too high, and I think this number is too low. But at least they have “bracketed” what I think it is likely to turn out to be.
I wonder on what basis Dr. Hathaway is making his predictions? Anyone know his methodology and why his predictions would be any better than a wild guess?
Note that Hathaway’s error band is so broad that any value between 30 and 90 would fit. He is almost certain to be correct on that.
pwl says:
April 9, 2011 at 6:41 pm
I wonder on what basis Dr. Hathaway is making his predictions? Anyone know his methodology and why his predictions would be any better than a wild guess?
Described here: http://solarphysics.livingreviews.org/open?pubNo=lrsp-2010-1&page=articlesu14.html in section 7.
“It’s quite a climbdown for Dr. Hathaway from his earlier predictions. Let’s give him credit for not trying to ‘hide the decline’.”
Not exactly. He was just wrong, that’s all. It’s okay for scientists to be wrong. Really.
Jim Cole says: “…What is the opposite of ‘prescient’?”
Not postscient, that’s for sure.
Leif Svalgaard says:
April 9, 2011 at 6:59 pm
Note that Hathaway’s error band is so broad that any value between 30 and 90 would fit. He is almost certain to be correct on that.
He’s definately got the low end sandbagged:
Take the previous low swing and the latest high swing, carry on until 01/2017 and you have a smoothed value of 30 for the next 5.7 years.
Jim Cole says: “…What is the opposite of ‘prescient’?”
A tad late-scient
Piers_Corbyn says:
April 9, 2011 at 5:30 pm
Has the UK Select committee responded to your submission yet?
I’ll take a Hathaway prediction over a Hansen prediction any day of the week.
Piers_Corbyn says:
April 9, 2011 at 5:30 pm
For such projections see slide 17 in the pdf in my submission to the UK Select committee enquiry into the extremely cold & snowy December 2010 crisis
simce you don’t label things very weel, ir is hard to figure out which one you are referring to. My guess would be this plot: http://www.leif.org/research/Piers2011.png
If so, you are somewhat economical with the truth. The blue curve is not what you pretend it is: the Wolf number from 1821 [even though you also label that point as 1892]. Poor style.
According to spaceweather.com, the current sunspot number (not smoothed) for April 8 is 84. But they’re mostly itsy-bitsy sunspots.
I still believe Hathaway is a smart and competent guy who labors under the direction, further up the line, of some world-class knuckleheads. I distinctly remember him being quoted as quite alarmed that “the great solar conveyor belt”, as he described it, had essentially ground to a halt at the beginning of the currently underway cycle. A month later he was carrying his bosses’ water with assurances that “everything on the sun is normal; nothing out of the ordinary going on here”.
littlepeaks says:
April 9, 2011 at 9:14 pm
According to spaceweather.com, the current sunspot number (not smoothed) for April 8 is 84. But they’re mostly itsy-bitsy sunspots.
Unfortunately there are several sunspot series out there. The once on spaceweather.com is the NOAA number. You have to multiply that by about 0.65 to get the International [official] Sunspot Number, so 84*0.65=55 would be the official SSN, which is what Hathaway is predicting.
Claude Harvey says:
April 9, 2011 at 9:20 pm
I still believe Hathaway is a smart and competent guy who labors under the direction, further up the line, of some world-class knuckleheads.
I know him well, he is smart and competent. But I don’t think his prediction is dictated from above. His prediction is based in a published and well-known formula that takes the actual data so far as input, and cannot be monkeyed with.
Re:Leif Svalgaard says:
April 9, 2011 at 9:42 pm
“His prediction is based in a published and well-known formula that takes the actual data so far as input, and cannot be monkeyed with.”
I have no reason to doubt your assessment concerning “the formula”. I continue to find the 180 degree turn of Hathaway’s public pronouncements early in the current solar cycle very odd. In a short period of time, he went from publicly declaring that something very much out of the ordinary was taking place on the sun to declaring the cycle was perfectly ordinary. At the time, I guessed that someone was leaning on the man. As it has turned out, the current cycle has been anything but business as usual and, it appears to me, has repeatedly defied predictions of the formula.
In any event, I’m pleased to hear knowledgeable confirmation of my “smart and competent” guesstimate.
To me 55 is quite a lot different from 62…
Piers_Corbyn says:
April 9, 2011 at 5:30 pm
Interesting and useful and bears out what a lot of us have been saying…..
However do bear in mind for ongoing cooling it is the odd cycles which count decisively. Even ones, eg SC24, which we are now in normally mean a cooler earth anyway. In odd ones Earth temp is best correlated with solar activity so a weak SC25 will be the ‘cooling clincher’.
Just remember no matter how vocal the static, observational evidence always trumps theory!
Query re slide 17 vertical axis, I realise the Wolf curves are qualitative in nature, but do you have any feel for the magnitude of tropospheric temperature reduction we may be in for?
At least with predictions for solar cycles we don’t have to wait 20, 50, or 100 years to find out the scientists and their computer models are wrong. The irony of course isn’t that David Hathaway was wrong, but rather that he was completely and utterly wrong – in 2006 the prediction was that SS24 would be the most intense of the last 400 years and now it’s one of the least intense in the last 200 years.
Time to develop some better computer models!
The grand old duke of Huntsville, Alabama
He marched them up to the top of the hill (the highest ever),
And he marched them down again (the lowest in 200 years).
And when they were up, they were up,
And when they were down, they were down,
And when they were only halfway up
They were neither up nor down.
but up in the mountains of Montenegro, the rebel isn’t budging
http://www.vukcevic.talktalk.net/NFC7.htm
Dr. Hathaway’s earlier prediction was contradicted by observations, so he changed his prediction. This is “science”. If he had applied “Mike’s ‘Nature’ [clever technique]” to hide the divergence between his prediction and observed data, this would not have been “science”. Thanks for the illustration of the difference, Dr. Hathaway!
Best,
Frank
Hathaway’s uncertainty margin for the predicted maximum is 26.
This far in the solar cycle, there exist other statistical methods with somewhat smaller uncertainty margins. See http://users.telenet.be/j.janssens/Engzonnecyclus.html#Bendel
The v20-method predicts a maximum of 78+/-20.
One should also take into consideration the rather large uncertainty in the timing of the predicted maximum (using Waldmeier’s rule, at least 16 monts). See my 05 March 11-comments on my main page.
Jan Janssens says:
April 10, 2011 at 1:14 am
http://users.telenet.be/j.janssens/Engzonnecyclus.html#Bendel
The v20-method predicts a maximum of 78+/-20.
Jan, there is good evidence that the Wolf sunspots number is too small before 1945. See http://www.leif.org/research/SIDC-Seminar-12Jan.pdf and http://www.leif.org/research/SIDC-Seminar-14Sept.pdf
Try to multiply all sunspot numbers before 1945 by 1.2 and repeat the analysis. Do the R-squares improve?
Janssens site linked above provides a very good insight into the projection methods that currently exist. The reason for keeping up-to-date forecasts is that for some fields (not climate, but space missions, communications etc) use the projections to calculate insurance risk, shielding requirements, fuel requirements, redundancy, etc. The people using the projections presumably realise that is is subject to revisions, and today would be keen to use an up-to-date projection.
Does anyone know if the same standard for SSN’s is used in comparison to records going centuries back? I’m wondering if their is inflation, somewhat like increased numbers of named storms in the Atlantic due to the advantage of satellites.
Point being, are there sun spots counted now that would have never been seen centuries past?
Puckster says:
April 10, 2011 at 2:27 am
Does anyone know if the same standard for SSN’s is used in comparison to records going centuries back?
http://www.leif.org/research/SOHO23.pdf
Leif Svalgaard says:
April 10, 2011 at 1:55 am
Try to multiply all sunspot numbers before 1945 by 1.2 and repeat the analysis. Do the R-squares improve?
In East Europe we use to multiply all achievements since 1945 and the West’s failures by 2. It produced ‘very agreeable’ R^2 correlation between the East’s achievements and the West’s failures.
What about L&P? Haven’t seen any updates lately?
littlepeaks says:
April 9, 2011 at 9:14 pm
According to spaceweather.com, the current sunspot number (not smoothed) for April 8 is 84. But they’re mostly itsy-bitsy sunspots.
If you want to look at the international sunspot numbers for
March 2011, see:
http://sidc.oma.be/products/ri_hemispheric/
When Dave Hathaway and others are talking about the charting of
sunspot numbers they’re referring to the monthly averages.
See Leif Svalgaard says @April 9, 2011 at 9:42 pm, above.
For a graphical comparison of the current cycle with the three previous
ones see:
http://www.solen.info/solar/cyclcomp.html
Vuk etc. says:
April 10, 2011 at 3:01 am
Leif Svalgaard says:
April 10, 2011 at 1:55 am
Try to multiply all sunspot numbers before 1945 by 1.2 and repeat the analysis. Do the R-squares improve?
In East Europe we use to multiply all achievements since 1945 and the West’s failures by 2. It produced ‘very agreeable’ R^2 correlation between the East’s achievements and the West’s failures.
At least Vuk has a sense of humor. This was funny, let alone that it matched my sentiment as I was reading. (“Sure, fudge enough, and the R^2 will improve.”)
Could it be this contributed to the southern Indian ocean cooling a bit?
http://www.letterdash.com/HenryP/marion-island-assessment-of-climate-change-in-the-southern-indian-ocean-due-to-the-increase-in-greenhouse-gases
Bruckner8 says:
April 10, 2011 at 3:36 am
“In East Europe we use to multiply all achievements since 1945 and the West’s failures by 2. It produced ‘very agreeable’ R^2 correlation between the East’s achievements and the West’s failures.”
At least Vuk has a sense of humor. This was funny, let alone that it matched my sentiment as I was reading. (“Sure, fudge enough, and the R^2 will improve.”)
But no sense of science (as we all know). Multiplying everything by 2 does not change the correlation nor R^2.
You gotta love a scientist who changes an opinion on the basis of evidence and reality. Admirable. I wish that more climate scientists did this – but it seems to not matter to them so much…..
LearDog says:
April 10, 2011 at 6:10 am
You gotta love a scientist who changes an opinion on the basis of evidence and reality.
It’s pretty hard to hide sunspots, compared to hiding a few disagreeable tree rings or thermometers.
The lowest 10.7 cm flux value is approximately 65 units. The highest value is about 280 units.
The present Sun’s output is 110 units therefore:
(110-65)/(280-65) = 20.9%
of the normal peak during a regular 11 year solar cycle (minimum at 2005 + 5.5 years = 2010.5). Now, of course, the Sun being a variable star has the length of the solar cycle vary. This level of energy input corresponds to a decrease in Earth temperature of about .08 degree C / 2.5 years (20% x 1.0 degree C/2.5 years).
The oceans are giving up their heat with a result of +0.6C down to -0.1C on the Earth’s Global temp. This will continue until the Sun becomes active. We are at the Global temperature equivalent to the 1970s. In 2.5 years, the Global temperature will be equivalent to the 1900s.
Leif Svalgaard says:
April 10, 2011 at 5:52 am
But no sense of science (as we all know). Multiplying everything by 2 does not change the correlation nor R^2.
Vuk said since 1945, implying to bring the correlation in line with one before 1945, in which case your remark is out of place.
Sorry, math error.
(20% x 0.1 degree C/2.5 years)= 0.08.
Hahahathaway is just a survivor!
Re:Leif Svalgaard says:
April 9, 2011 at 9:42 pm
..“His prediction is based in a published and well-known formula that takes the actual data so far as input, and cannot be monkeyed with.”
Leif
Any significant changes to the “formula” as a result to the 23-24 cycle change and the slow and low progress of 24 to date? Based on the early estimates of Cycle 24 (your’s and some others at the low end of the scale and Hathaway’s et al toward the high end) seems some formulas are more accurate than others.
PS: Wish the world had a lot more Hathaways. We need “scientists” not “psyentists”.
harrywr2 says:
April 10, 2011 at 7:13 am
It’s pretty hard to hide sunspots, compared to hiding a few disagreeable tree rings or thermometers.
The Sun may be hiding the spots at times, e.g. during the Maunder Minimum and recently causing the Livingston&Penn effect…
See slide 14-17 of http://www.leif.org/research/Eddy-Symp-Poster-2.pdf
Bob, observational evidence never trumps mechanism. Else we would still be fearing and providing sacrifice to female cycles as harbingers and prevention of crop destruction. I still remember, in my youth, reading articles that postulated, based on “confirmed” observational evidence, menstrual cycles were tied to lunar stages. I wonder what the score is historically: observation versus theory. My hunch is that in all of scientific history, plausible mechanism destroyed observation many more times than the other way around.
Just one example: Read up on how the periodic table was completed before there was observational evidence of some of the elements now listed.
I have been interested in the sunspot cycle for a number of years now and I suspect some kind of Earth climate effect but still am not convinced that any of the proposed mechanisms explain how it would work.
Solar irradiance measures do not seem to vary enough from low points and high points in cycle or from low cycles to high cycles to have any major effect on climate. Galactic cosmic ray theories haven’t been convincing either. On the other hand, we clearly seem to have periods of high solar activity associated with warming (Medieval Warm Period) and periods of low activity associated with cooling (Little Ice Age). Our current period of high activity and warming, of course, has the complicating issue of being also a time of increasing green house gases and so can’t be used to argue one way or the other as to relative effects of GHGs vs solar influence.
It would seem that this solar cycle and perhaps the next ought to be fairly definitive in providing some idea of relative effects of the two.
Is there some other theory out there that I am missing? What are others thoughts on this?
Piers_Corbyn says:
April 9, 2011 at 5:30 pm
However do bear in mind for ongoing cooling it is the odd cycles which count decisively.
Piers opened my eyes to something that is often overlooked. The true solar cycle is 22 years; for 11 years the earth and sun’s magenetic field are aligned, the other 11 years they are opposite.
The observation that we get different weather patterns when the magnetic fields are aligned as compared to when they are opposite is largely ignored in climate scince.
Strange? The prediction is the same as the current sunspot number (about). How many of the past 23 cycles are like this in year 3? I would expect that the middle would be higher. The sudden spike looks like an anomalous thing. So why would you define the end by the sudden spike? Without the spike we would have gone for a smaller final count. And the spike: is it real or an artefact of how the spots are being counted …
While it has been conclusively proven that solar variance does not drive climate change, even Wikipedia seems to recognize things are not quite as they have been:
“The level of solar activity during the past 70 years is exceptional — the last period of similar magnitude occurred over 8,000 years ago. The Sun was at a similarly high level of magnetic activity for only ~10% of the past 11,400 years, and almost all of the earlier high-activity periods were shorter than the present episode.[27]”
http://en.wikipedia.org/wiki/Solar_variation
an interesting graph from wikipedia, demonstrating that there is no connection between solar activity and the earths average temperature:
http://en.wikipedia.org/wiki/File:Carbon14_with_activity_labels.svg
I thought everyone predicted a small 25, so Corbyn is right probably as usual. Also David Archibald predicted a very low 24 I think it was 40-60 so spot on. Its interesting to see that the AGW skeptics have been spot on since predicting things 2 years ago whereas nearly ALL the AGW’ers have been way off except for last years temps, which I concede.. I also have grave reservations they way CT calculates NH ice. The feeling is that the baseline is adjusted to keep below anomalies. Unfortunately the ice is constant and not getting any lower, although NCDC makes a marvelous job of making the ice appear to get lower every year non-stop .
Tried to share this on Facebook .. it was blocked! … sent them a nasty message .. I am not amused ..
I believe it is time to rid myself of Facebook all together.
I am not at all surprised that SC24 appears to be “small”. In the 1950’s I did some work on the relative spottedness of the north and south hemispheres using the Greenwich records with backward extension using drawings by Schwabe and Sporer. The results were published in a paper in Monthly Notices ((MNRAS, 115,4,1955). In the paper it was noted that there was a marked correlation between the activity of a particular cycle and the change of relative spottedness during that cycle between the northern and southern hemispheres as represented by the ratio (N-S)/(N+S) (either in terms of spot counts or spot areas). For example, a very active cycle tended to be associated with a positive slope (south dominant at the beginning and north at the end of the cycle) and vice versa. During the decline of SC23 this slope was markedly negative and at minimum the activity flipped to north dominant in the new cycle. North has remained dominant on average up to the present (i.e. for 2-3 years). This would suggest that SC24 will not be a very active cycle and could indeed be one with a very low maximum.
Leif Svalgaard says:
April 9, 2011 at 6:59 pm
“Note that Hathaway’s error band is so broad that any value between 30 and 90 would fit. He is almost certain to be correct on that.”
It would appear from this, almost unbelievably in light of your education, that you don’t know the meaning of either “error band” or “smoothed sunspot number”.
From the referenced webpage,
“These predictions are for “smoothed” International Sunspot Numbers. The smoothing is usually over time periods of about a year or more so both the daily and the monthly values for the International Sunspot Number should fluctuate about our predicted numbers. The dotted lines on the prediction plots indicate the expected range of the monthly sunspot numbers.”
These values are not “error bands”.
Astronomy Picture of the Day
April 10, 2011
http://apod.nasa.gov/apod/image/1104/sunspotloops_trace_898.jpg
It was a quiet day on the Sun. The above image shows, however, that even during off days the Sun’s surface is a busy place. Shown in ultraviolet light, the relatively cool dark regions have temperatures of thousands of degrees Celsius. Large sunspot group AR 9169 from the last solar cycle is visible as the bright area near the horizon. The bright glowing gas flowing around the sunspots has a temperature of over one million degrees Celsius. The reason for the high temperatures is unknown but thought to be related to the rapidly changing magnetic field loops that channel solar plasma. Large sunspot group AR 9169 moved across the Sun during 2000 September and decayed in a few weeks.
Leif Svalgaard says:
April 10, 2011 at 5:52 am
“But no sense of science (as we all know). Multiplying everything by 2 does not change the correlation nor R^2.”
As anyone who had a sense of science would know, that depends on just what “everything” means. Do you?
Leif Svalgaard says:
April 10, 2011 at 7:33 am
“The Sun may be hiding the spots at times”
That is the most profound statement you have ever made. I agree, the Sun hides at least half of its spots at any given time.
Do any of the sunspot measurements take size and intensity into account?
I’m curious on that point because the old manual measurements would have missed a lot of what gets counted today.
My apologies if this has been covered (just post the link and I’ll happy reading away).
Try this in Vegas. I said “hold.” No, “hit, me.” No, “black.” I said, “red.”
iEdward says:
April 10, 2011 at 8:45 am
…. Its interesting to see that the AGW skeptics have been spot on since predicting things 2 years ago whereas nearly ALL the AGW’ers have been way off except for last years temps, which I concede…..
Edward
Could you possibly tell me why AGWers would particularly want to over-estimate the SSN. It occurs to me that if the sunspot number is low while global temperatures remain high it helps their argument considerably. The fact that David Archibald and lots of others (including Leif Svalgaard) predicted a weaker sunspot cycle is irrelevant to the AGW debate – if global temperatures don’t respond as expected (by some). I’m far from convinced they will.
Dr. Lurtz says:
April 10, 2011 at 7:23 am
We are at the Global temperature equivalent to the 1970s. In 2.5 years, the Global temperature will be equivalent to the 1900s.
Or it may hold the ’70’s values until SC24 is spent and rolls down into the next minimum. Then the bottom will fall out.
…something completely different:
Anthony said “I missed this earlier this week from NASA, I got a bit distracted with other things.” We all hope everything is going well with your wife’s surgery and recovery, Anthony!
fabron says:
April 10, 2011 at 7:24 am
Vuk said since 1945, implying to bring the correlation in line with one before 1945, in which case your remark is out of place.
Eastern Europe was part of the ‘West’ until taken over by Communism. So his remark only makes sense for years after 1945.
Pascvaks says:
April 10, 2011 at 7:31 am
Any significant changes to the “formula” as a result to the 23-24 cycle change and the slow and low progress of 24 to date? Based on the early estimates of Cycle 24 (your’s and some others at the low end of the scale and Hathaway’s et al toward the high end) seems some formulas are more accurate than others.
If you took the trouble to look at the Hathaway paper I have already linked to a couple of times, you’ll see that the formula is a description of the current cycle based solely on observed solar activity so far since the minimum.
ferd berple says:
April 10, 2011 at 8:19 am
The observation that we get different weather patterns when the magnetic fields are aligned as compared to when they are opposite is largely ignored in climate science.
And for good reason. As there is no convincing evident or mechanism for that. BTW, the alignments are from solar max to solar max.
Doug Proctor says:
April 10, 2011 at 8:27 am
And the spike: is it real or an artefact of how the spots are being counted …
It is real. In a weak cycle such spikes are commonplace, e.g for cycle 14: http://www.leif.org/research/SC14.png
ferd berple says:
April 10, 2011 at 8:30 am
“The level of solar activity during the past 70 years is exceptional — the last period of similar magnitude occurred over 8,000 years ago.
No, it is not: see e.g. Figure 10 of this peer-reviewed paper: http://www.leif.org/research/2009JA015069.pdf or http://www.leif.org/EOS/muscheler05nat_nature04045.pdf or http://www.leif.org/EOS/muscheler07qsr.pdf or
http://www.leif.org/EOS/2009GL038004.pdf
“It is not what you know that gets you in trouble, but what you know that ain’t so”.
Glenn says:
April 10, 2011 at 9:08 am
“The dotted lines on the prediction plots indicate the expected range of the monthly sunspot numbers.”
If you would care to actually learn about how their method works, you would find that the dotted lines represent the 5% and 95% percentiles of the prediction [plus/minus two standard error bars]. We would, indeed, expect individual values to fall in that range.
Glenn says:
April 10, 2011 at 9:15 am
“But no sense of science (as we all know). Multiplying everything by 2 does not change the correlation nor R^2.”
As anyone who had a sense of science would know, that depends on just what “everything” means. Vuk said: ” use to multiply all achievements”. The all indicates ‘everything’ in my book.
Glenn says:
April 10, 2011 at 9:23 am
That is the most profound statement you have ever made. I agree, the Sun hides at least half of its spots at any given time.
Actually, not. We can see the whole surface of the Sun these days, by several methods. But it would be reasonable for you to comment on science not on misunderstood banalities [this and previous posts as well]. You bring nothing to the table.
TRM says:
April 10, 2011 at 9:43 am
Do any of the sunspot measurements take size and intensity into account?
I’m curious on that point because the old manual measurements would have missed a lot of what gets counted today.
Unfortunately, yes. Bigger spots are counted by the sunspot reference station [Locarno] with a weight of up to 5 depending on size. You can see that clearly here: http://www.specola.ch/drawings/2010/loc-d20100927.JPG look at some of the groups [they have numbers]. e.g. number 94 is shown to have 3 spots in the table in upper right while the smaller spot in 98 is counted with weight 2. Group 97 has three small spots [counted as 1 each], while the two bigger spots are counted as three each, for a total of 9 for the group.
All other solar observers do not do this, but since the SIDC [and indirectly NOAA as they try not to stray to far away from SIDC/0.6] uses Locarno as their reference station [which all observers are calibrated to], the weighting creeps in in the overall count.
There are ways to correct for this [as one should], see e.g. http://www.leif.org/research/SIDC-Seminar-12Jan.pdf
The r2-values only improved for the Waldmeier-relation, from 0.59 to 0.63. This reduces the uncertainty in the maximum timing by 1 month.
However, for the v20 and R25 predicton methods, the r2-values decreased by resp. 0.06 and 0.05, increasing the uncertainty of the prediction by 2 (resp. 22 and 20). With the changed SSN, the v20-method now predicts a SC24-maximum of 90+/-22.
This result pretty much summarizes what one does by increasing all sunspotnumbers prior to 1945: Because the maxima prior to 1945 increase by 20%, the trendline also shifts up by about 20%, whereas the slope decreases (only) by 4-7%.
That September-workshop promises to tackle very interesting topics. Apart from corrections to the past sunspotnumbers, one could also reflect on the influence of the instantly available high-resolution images on the SSN as determined by the solar observers, and one could at the same time ponder on what constitutes a sunspot.
Leif Svalgaard says:
“The Sun may be hiding the spots at times”
Glenn says:
April 10, 2011 at 9:23 am
That is the most profound statement you have ever made. I agree, the Sun hides at least half of its spots at any given time.
“Actually, not. We can see the whole surface of the Sun these days, by several methods.”
Uh, that isn’t attributable to the Sun, Leif. I’m disappointed in you. You would apparently think that all spots should be counted and relevant to the sunspot count. I doubt that would being much to the table.
Leif Svalgaard says:
“Vuk said: ” use to multiply all achievements”. The all indicates ‘everything’ in my book.”
It is sad that you would mischaracterize what another said, Leif. Just downright sad.
Vuk said
“In East Europe we use to multiply all achievements since 1945 and the West’s failures by 2.”
You left out an important part of “everything”, actually *half* of everything. Why, oh why?
Leif Svalgaard says:
April 10, 2011 at 10:58 am
Glenn says:
April 10, 2011 at 9:08 am
The dotted lines on the prediction plots indicate the expected range of the monthly sunspot numbers.
“If you would care to actually learn about how their method works”
I did, posted it, and you just quoted part of it. The values of the dotted lines are not the range of the *smoothed* number prediction. That means that were the smoothed sunspot prediction to be 30 or 90, their prediction of the smoothed sunspot number and maximum would be, well, wrong. I’ll repeat part of the quote:
“The dotted lines on the prediction plots indicate the expected range of the monthly sunspot numbers.”
Fabron & Glen thanks. It really isn’t that important, so let’s go back to science.
Why not just add 1300 to all SSN?
That way, no matter what happens, the SSN will be just as insignificant as TSI variations, and nobody will ever bother to observe or analyse them again due to triviality.
Why waste time on a constant?
Set it & forget about it.
Here on WUWT, the NASA Sunspot Prediction Roller Coaster tracks Hathaway’s numbers from March 2006 (156 to 180) through December 2010 (64). His high precision is all out of proportion to his evident lack of accuracy.
Back in January 2009, when he was predicting 104, I predicted a nice round 80, and in December 2010, when he was saying 64 I predicted a nice round 60. Now he is down to 62. Seems like we are playing leapfrog, but I’ll stick with my nice round 60.
By the way, unlike him, I do not claim any expertise in this area, nor am I funded by your tax dollars :^)
John Finn says:
April 10, 2011 at 9:55 am (Edit)
“Could you possibly tell me why AGWers would particularly want to over-estimate the SSN. It occurs to me that if the sunspot number is low while global temperatures remain high it helps their argument considerably. The fact that David Archibald and lots of others (including Leif Svalgaard) predicted a weaker sunspot cycle is irrelevant to the AGW debate – if global temperatures don’t respond as expected (by some). I’m far from convinced they will.”
It’s typical of their magical thinking, John. They subconciously know that we are right but think that if they lie about it enough that things will come out true: if they keep counting more and more sunspots that aren’t there, that Earth will magically keep warming to fulfill their disasturbationist fantasies.
Glenn says:
April 10, 2011 at 11:42 am
You would apparently think that all spots should be counted and relevant to the sunspot count. I doubt that would being much to the table.
Indeed, yes, for studying the Sun that is what one should do. Once we get solar magnetographs in orbit on the backside, we will observe the full soar surface and our calculations of the heliospheric field to expect at Earth, or at other places where such information is needed will improve.
“In East Europe we use to multiply all achievements since 1945 and the West’s failures by 2.”
You left out an important part of “everything”, actually *half* of everything. Why, oh why?
Because that is what was said, otherwise it would read: we used to multiply all achievements since 1945 by 2 and correlate with the West’s failures. Nobody needs your sad ‘why, oh why’ lamenting.
Glenn says:
April 10, 2011 at 11:54 am
I did, posted it, and you just quoted part of it. The values of the dotted lines are not the range of the *smoothed* number prediction. That means that were the smoothed sunspot prediction to be 30 or 90, their prediction of the smoothed sunspot number and maximum would be, well, wrong.
Not wrong, just as in all science uncertain. But you did not read about their method. Read their paper http://www.leif.org/EOS/1999JA900313.pdf carefully, especially the discussion of Figures 16 and 17. Figure 16 estimates the standard deviation, and 17 explains that the dashed lines denote the 5% and 95% percentiles: “The dotted lines are placed at plus and minus two standard deviations”. This is the usual definition of two-standard deviation error band, within which one would expect 95% of individual monthly values to fall. Perhaps, you shouldn’t comment on things you do not understand. What was that again: ‘sad, so sad’.
rbateman says:
April 10, 2011 at 1:07 pm
Why not just add 1300 to all SSN?
Because we are not silly about this.
mikelorrey says:
April 10, 2011 at 1:54 pm
if they keep counting more and more sunspots that aren’t there, that Earth will magically keep warming to fulfill their disasturbationist fantasies.
“they” are not counting sunspots. Solar activity is kept track of by careful and dedicated observers for a lot of practical purposes [apart from studying the Sun].
Seems to me that there are 2 civilizational dangers these days;
Iran lighting off a nuke on the East coast or an X-C flare cooking our geese.
Ira Glickstein, PhD says:
April 10, 2011 at 1:35 pm
Back in January 2009, when he was predicting 104, I predicted a nice round 80, and in December 2010, when he was saying 64 I predicted a nice round 60. Now he is down to 62. Seems like we are playing leapfrog, but I’ll stick with my nice round 60.
In Hathaway’s description of what he is doing http://www.leif.org/EOS/1999JA900313.pdf he says: “Finally, on the basis of our analysis presented here, we predict that cycle 23 will have a maximum amplitude near 150 (slightly higher if expressed as a 13-month running mean) with the maximum occurring midway through the year 2000. Activity for the next 3-4 years (1999-2002) should be typical of the maximum phase condition [cf. Wilson et al., 1998b]. Although this prediction should not change much as the cycle progresses[because we are well into the cycle – the paper was written in March 1999], we update the prediction every month as new data becomes available and post it on a world wide web site at http://science.nasa.gov/…”
This is what should be done: updating all the time when new data becomes available. If you are just at the beginning of a cycle, so with little data to work with] the uncertainty will be large. As the cycle progresses, the prediction becomes better and better. This is no ‘roller-coaster’, but just correct application of data as good science should be done.
Vuk etc. says: “Fabron & Glen(n) thanks. It really isn’t that important, so let’s go back to science.”
Amen. Thanks, Vuk. The snark to content ratio was getting a bit thick.
Leif Svalgaard says:
April 10, 2011 at 3:04 pm
Why not just add 1300 to all SSN?
Because we are not silly about this.
Indeed, I do get why you are doing this. It’s just that I see the end result being full of unintended consequences, the chief result being increasing uncertaintly as the older records are adjusted. It won’t be that way at first, but I guarantee that, over time, the tendency to assume will grow.
You want a standardized record, but none really exists.
The observers of today are alive and thier means & intentions discovered.
Not entirely true those who are gone, and much of what they wanted those who followed to know is now lost.
I dare say that the dumbing down that takes place today has good reasons behind it.
It’s the lesser of two evils, there being 2 sides to the coin.
If there is no reasoning around to prevent you from convincing others to adjust the past which cannot be 100% verified, there will also be no reason many years from now to prevent others from doing the same. Even worse, data handling that has been subject to adjusting has a nasty habit of losing the original.
Do you understand where I am coming from?
rbateman says:
April 10, 2011 at 3:45 pm
You want a standardized record, but none really exists.
The observers of today are alive and their means & intentions discovered.
Not entirely true those who are gone, and much of what they wanted those who followed to know is now lost.
Already Wolf discovered [and used] an objective and ‘intention-free’ method of getting a standard record, namely the amplitude of the diurnal variation of the magnetic needle. This method has no long-term instrumental calibration problems because an angle does not require calibration, 10 arc minutes are 10 arc minutes and could easily be measured with precision 200 years ago. So, it is possible to standardize the record. There are, of course, always detail that must be cleared up. The magnetic needle is also sensitive to electric currents in the ground so small local differences exist, but we can measure today at the same locations as in the past and correct unambiguously for that. The overhead current depends on the conductivity of the ionosphere which in turn depends on the overall intensity of the Earth’s magnetic field. That field is slowly changing, but again we know by how much, so can also correct for that [small] effect. So, a revised sunspot number will not drift away into the sky as time goes on, but will become better and better as we improve our understanding of how to correct for those [again: small] environmental effects. The magnetic data is not lost, much has been digitized and the rest will follow [and there are active projects aimed at that going on].
I agree, Leif, that new data must be considered and applied to established theory. If the theory is right, the predictions will be rapidly refined.
On the other hand, Hathaway’s initial predictions said SC#24 would be 35% higher than SC#23 and subsequent predictions, based on new data, continually downgraded the peak Sunspot number to the point they now say SC#24 will be 50% lower than SC#23. Doesn’t that show there was something wrong with the theory on which Hathaway is basing his predictions? Do you agree? Hathaway has also changed the predicted peak date from mid-2010 to late-2013, an adjustment of three years for an nominal 11-year cycle. I initially said it would peak in mid-2013 and I now think it will peak in 2014.
Good theories produce relatively stable predictions. While I congratulate Hathaway on his forthright admission that new data required him to change his predictions, tacitly admitting they were way too high and peaked way too soon, his “conveyor belt” theory has not panned out very well.
Your theory and predictions have fared much better than Hathaway’s. I do not believe you have had to change your predictions much, if at all, based on new data. The whole idea of predictions is to predict the future, far enough in advance of the event to make a difference. Does that not prove your theories are closer to the truth than Hathaways?
Leif Svalgaard says:
April 10, 2011 at 3:04 pm
Glenn says:
April 10, 2011 at 11:42 am
You would apparently think that all spots should be counted and relevant to the sunspot count. I doubt that would being much to the table.
“Indeed, yes, for studying the Sun that is what one should do.”
Hey Leif, news alert: The sunspot count is and has been a *visible* count from the Earth.. You apparently think everyone should accept any harebrained thing you say, but that just isn’t reality.
Leif Svalgaard says:
April 10, 2011 at 3:04 pm
Glenn says:
“In East Europe we use to multiply all achievements since 1945 and the West’s failures by 2.”
You left out an important part of “everything”, actually *half* of everything. Why, oh why?
“Because that is what was said”
No, Leif, that was only half of what was said. [snip. gratuitous insult. ~dbs]]
Leif Svalgaard says:
“Not wrong, just as in all science uncertain. But you did not read about their method.”
Yes, Leif, predictions can be wrong. And I did read about their method. Would you like me to post it again?
Ira Glickstein, PhD says:
April 10, 2011 at 4:18 pm
On the other hand, Hathaway’s initial predictions said SC#24 would be 35% higher than SC#23 and subsequent predictions, based on new data, continually downgraded the peak Sunspot number to the point they now say SC#24 will be 50% lower than SC#23. Doesn’t that show there was something wrong with the theory on which Hathaway is basing his predictions?
No, it just shows the danger [as is already known] of early prediction.
Good theories produce relatively stable predictions.
Hatahway’s prediction is not based on any theory, but on the assumption [based on previous cycles] that the shape of the cycle determines its size: if it starts out slow, it will not have time to reach a big cycle once it is time for maximum [Waldmeier’s rule]. Basically what he does is to find a few parameters [by fitting to observed data] that describes the shape of the cycle [so far], then applies that to find the size.
his “conveyor belt” theory has not panned out very well.
first, it was not his theory, second it is not used for his prediction [actually better to call it a forecast, as he is just doing what weather forecasters do].
Your theory and predictions have fared much better than Hathaway’s. I do not believe you have had to change your predictions much, if at all, based on new data.
My prediction depends on the polar fields in a few years before minimum and will thus change slightly if the polar fields change over that time [it is one of the strengths of the prediction that it can adjust]. Fortunately, observations show that the polar fields once well-established don’t change much, and after the minimum, the polar fields begin to decrease because of new-cycle flux getting to the poles, so they can no longer be used for forecasting of the current cycle, hence the prediction is ‘frozen’ at that point.
The whole idea of predictions is to predict the future, far enough in advance of the event to make a difference. Does that not prove your theories are closer to the truth than Hathaways?
My theory is useful some three years before the next cycle. Hathaway’s is not applicable at all until you are some time into the cycle. But once there, his, of course, with time becomes better than mine, because it takes into account the actual evolution of the real data.
Now, the current Hathaway predictions are different than his geomagnetic precursor method used before the minimum. This is not always appreciated, see http://www.leif.org/research/Predicting%20the%20Solar%20Cycle%20(SORCE%202010).pdf The difficulty with that method is the ‘pick the right peak’ [see the paper]. Once you do, it works reasonably well. Unfortunately for Hathaway he picked the wrong one. Possibly encourgaed to do so by the Dikpati et al. high prediction.
Glenn says:
April 10, 2011 at 4:27 pm
Hey Leif, news alert: The sunspot count is and has been a *visible* count from the Earth.. You apparently think everyone should accept any harebrained thing you say, but that just isn’t reality.
Yes, but we would really wish to know all of solar activity all over the surface, and one day we will.
Glenn says:
April 10, 2011 at 4:30 pm
No, Leif, that was only half of what was said. [snip. gratuitous insult. ~dbs]
You have a reading disability or what? Haven’t you wasted enough of everybody’s time on this?
Glenn says:
April 10, 2011 at 4:33 pm
Yes, Leif, predictions can be wrong. And I did read about their method. Would you like me to post it again?
You only read then dumbed down version for public consumption [but is seemed appropriate for you]. Their paper in the abstract clearly says: “This Combined Solar Cycle Activity Forecast gives, as of January 1999, a smoothed sunspot maximum of 146 plus/minus 20 at the 95% level of confidence for the next cycle maximum”. do us all a favor and educate yourself a bit: here is the paper again: http://www.leif.org/EOS/1999JA900313.pdf To prove that you have read it, tell us what the first word of the text on each page is.
Is it purely a coincidence that the recent upward spike in sunspot number comes at a time when the lower troposphere spiked up on Dr. Spencer’s site?
>>Leif Svalgaard says: April 10, 2011 at 3:23 pm
>>In Hathaway’s description of what he is doing ,
>>we update the prediction every month as new data becomes available
If the model only uses data from the current cycle, then how did they make a prediction before the cycle started?
Also, it is quite obvious that the Sun has cycles that are longer than just one 11-year cycle, and to not utilise any of these other factors in the model for sunspot prediction demonstrates a complete ignorance about the Sun and its inner workings. In fact, it is hardly a model at all. You would do better with a pencil and a squinted eyeball. Perhaps an ouija board might help.
.
Ref – Leif Svalgaard says:
April 10, 2011 at 4:49 pm
Dr. Ira Glickstein, (April 10, 2011 at 4:18 pm) thank you for your comments/questions. Dr. Leif Svalgaard, thank you for your comments and answers. My poorly worded comment/question earlier regarded what both of you just covered so clearly. Again, thank you both. Much clearer now 😉
Ralph says:
April 11, 2011 at 2:09 am
If the model only uses data from the current cycle, then how did they make a prediction before the cycle started?
1) it is not a model, just a description of the shape of the current cycle
2) the prediction before the cycle started was based on a completely different method, namely the observation that variations of the geomagnetic field depend on the state of the solar wind which in turn is influenced by the solar polar magnetic field. Empirically it is found that geomagnetic activity often has a peak just before solar minimum and that the size of that peak is a predictor of the next cycle. Problem is when you have more than one peak: which one to choose? Hathaway chose the wrong one.
“”””” Piers_Corbyn says:
April 9, 2011 at 5:30 pm
Interesting and useful and bears out what a lot of us have been saying.
However do bear in mind for ongoing cooling it is the odd cycles which count decisively. Even ones, eg SC24, which we are now in normally mean a cooler earth anyway. In odd ones Earth temp is best correlated with solar activity so a weak SC25 will be the ‘cooling clincher’. For such projections see slide 17 in the pdf in my submission to the UK Select committee enquiry into the extremely cold & snowy December 2010 crisis – short link- http://bit.ly/hEmBqG
Thanks, Piers Corbyn “””””
Piers, only reason I can conjure up as to why odd/even cycles would have different solar activity consequences, would be the magnetic factor. Sun’s magnetic field switches every cycle, while earth’s does not, so net local magnetic field should have an odd/even assymmetry. This presumably has a cosmic ray/charged particle effect on cloud formation.
George E. Smith says:
April 11, 2011 at 11:49 am
This presumably has a cosmic ray/charged particle effect
The cosmic ray intensity does have a weak asymmetry between od/even cycles [or rather from solar max to solar max]. See Slide #7 of http://www.leif.org/research/Historical%20Solar%20Cycle%20Context.pdf
where the peaks marked with circles are higher and more ‘peaked’ than the intervening GCR maxima, but the effect is second order and its climate effects [if any] would not be noticeable.
Since I cannot understand climate change without understanding solar dynamics, I feel that it is my duty to sit at the feet of the great oracles like Hathaway and learn why SSN predictions have more oscillation that the SNN themselves. I think I’m ready: I just bought a dozen books on the subject…and a Ouija board. I’m also thinking about buying a really warm coat.
Leif Svalgaard says:
April 11, 2011 at 12:54 pm
Then there is a great disparity opening up between the cooling that Solar/Galactic science allows for and the actual cooling observed.
If the end result of very low solar activity is history repeating itself, did we learn anything?
rbateman says:
April 12, 2011 at 3:46 am
Then there is a great disparity opening up between the cooling that Solar/Galactic science allows for and the actual cooling observed.
Just showing they have nothing to do with each other.
If the end result of very low solar activity is history repeating itself, did we learn anything?
Some people will never learn anything.