Note: See also the 1997 regime change in global climate data by the same method, here
I’m happy to report that something I recognized and reported back in 2008 related to solar data has been independently confirmed by another source, and was a surprise when it showed up in my inbox two days ago.
Readers may recall that for some time I’ve been pointing out a strange anomaly in the Solar Geomagnetic Index that occurred in October 2005. In a story I wrote on Feb 13th, 2008 titled Where have all the sunspots gone? I plotted the Ap data and pointed out the event.
I wrote then:
What is most interesting about the Geomagnetic Average Planetary Index graph above is what happened around October 2005. Notice the sharp drop in the magnetic index and the continuance at low levels.
This looks much like a “step function” that I see on GISS surface temperature graphs when a station has been relocated to a cooler measurement environment. In the case of the sun, it appears this indicates that something abruptly “switched off” in the inner workings of the solar dynamo. Note that in the prior months, the magnetic index was ramping up a bit with more activity, then it simply dropped and stayed mostly flat.
Since then, we’ve seen announcements like this:
BREAKING – major AAS solar announcement: Sun’s Fading Spots Signal Big Drop in Solar Activity
Posted on June 14, 2011 by Anthony Watts
“If we are right, this could be the last solar maximum we’ll see for a few decades,” Hill said. “That would affect everything from space exploration to Earth’s climate.”
See the official press release here – “All three of these lines of research to point to the familiar sunspot cycle shutting down for a while.”
Dr. Sam Outcalt : Emeritus Professor of Physical Geography, University of Michigan sent me this graph two days ago, showing his application of Hurst Rescaling to the Ap Solar Magnetic Index data. Using that method, he has independently identified the “step function switch off” I reported in Feb 2008:
The major regime transition is at the maximum of the integral at 2005.71, which corresponds to October 2005, the same date I identified.
Clearly the sun entered into a magnetic funk then, and has yet to come out of it.
We live in interesting times.
For more on Hurst ReScaling, see this paper: SIO_HurstReScale
UPDATE: As I expected he would, Dr. Leif Svalgaard takes exception to this characterization of the identification of October 2005 being a regime changepoint, saying:
While I agree that the sun is going quiet, the ‘step change’ is spurious. It is mainly due to a sporadic, single magnetic storm in September 2005: http://hirweb.nict.go.jp/sedoss/solact3/do?d=2005%2c09%2c04 and here is the next rotation: http://hirweb.nict.go.jp/sedoss/solact3/do?d=2005%2c10%2c01 You can find many such steps.
Such step changes happens all the time: http://www.leif.org/research/Ap-1844-now.png They are just weather, not climate.
I don’t think his analogy holds promise, because after the step change the “climate” of the solar dynamo stayed low, and then produced the lowest value in the record going back to 1844. See Leif’s graph (click to enlarge) which I’ve annotated:
While this is akin to the sustained drop starting in early 1871, clearly this was something new in the entire record.If we use smoothing to remove what Leif describes as weather noise, and magnify, we can see how this looks to be unique in the last century of data:
While it could be argued that this was a “weather” event, the facts remain that:
- In the 14 months prior to Oct 2005, the Ap index was on the rise
- It was a very sharp drop spanning a little over a month.
- It has not recovered to the average values in the preceding 14 months since then.
- Weather events are usually short term dynamics. The 2 rotation periods of the sun Leif refers to aren’t capable of maintaining the step change for the years following October 2005.
- Following the event, the Sun produced the lowest Ap value in the record in Dec 2008.
- Projections by others, including Livingston, suggest the sun has entered into a quiet magnetic state.
While I defer to Dr. Svalgaard’s overall superior knowledge on the dynamics of sun, and agree there are many sharp transitions in the Ap record, this looks to me to be a step change event of merit based on the factors listed above. I’ve yet to see a fully convincing explanation that this was a spurious event rather than a regime changepoint. But, I remain open to seeing such an explanation.
Leif Svalgaard says:
Spurious correlations [that furthermore violate the laws of physics] are not evidence.
…….
More likely something neither solar or the climate officialdom whish to know
http://www.vukcevic.talktalk.net/Aa-CAM.htm
Leif Svalgaard (July 4, 2012 at 9:23 pm) wrote:
“The signature in Ap is spurious, just a coincidence, as there are many other Ap-steps not coinciding with any solar changepoints..”
I suggest you take a lesson from Dr. Judith Curry and restrict yourself to expert commentary in areas where you actually have expertise.
For example, when Tomas Milanovic wrote a guest post at Climate Etc. about ergodicity, Dr. Curry was honest & forthright in admitting the shortcomings of her knowledge in that area.
With all due respect:
Your misunderstandings &/or deliberate distortions of elementary stats fundamentals are misguiding naive members of this community.
You can’t get away with arguing
1 + 1 = 1984 (Law of Authoritative Ignorance &/or Deception).
In the case at hand, you are AGAIN wrong AT THE LEVEL OF ABSOLUTE LOGIC. This time it has happened in a sufficiently simple context that more audience members than usual are capable of independently catching you red-handed and calling you on it.
If you want to ague that 1+1 does not equal 2 then there is no basis for sensible communication and you are destroying trust at such a fundamental level that it is straight-up creepy.
Reread what I wrote above more carefully ( http://wattsupwiththat.com/2012/07/04/another-regime-change-indication-this-time-in-solar-data/#comment-1024740 ).
vukcevic says:
July 5, 2012 at 2:26 am
More likely something neither solar or the climate officialdom wish to know
More likely something nobody needs to know.
Paul Vaughan says:
July 5, 2012 at 3:58 am
“The signature in Ap is spurious, just a coincidence, as there are many other Ap-steps not coinciding with any solar changepoints..”
I suggest you take a lesson from Dr. Judith Curry and restrict yourself to expert commentary in areas where you actually have expertise.
It just so happens that I do have expertise in the meaning and physics of Ap. Do you?
So as I said: The signature in Ap is spurious, just a coincidence, as there are many other Ap-steps not coinciding with any solar changepoints..
Is it just me, or does anyone else picture Sheldon when reading Leif’s comments?
so what is the correlation with Earth’s albedo?
that is the 900 pound gorilla in the room that provides the equivalent to solar tsi variations, can vary dramatically, and is potentially subject to all sorts of internal and external influences via its primary contributor – clouds.
The Ghost Of Big Jim Cooley says:
July 5, 2012 at 4:30 am
Is it just me, or does anyone else picture Sheldon when reading Leif’s comments?
where is my Amy?
cba says:
July 5, 2012 at 4:36 am
so what is the correlation with Earth’s albedo?
Hardly any that I can see. Here http://www.leif.org/research/Albedo-Since-1984.png is a plot of the Albedo since 1984 [from http://www.leif.org/research/Palle_Earthshine_2008.pdf ] . The black curve is from ISCCP FD data, while the blue curve is the albedo measured from earthshine on the Moon. The ordinate is the albedo anomaly [or simply difference] relative to the average 29%, so a value of +1 means 30%, and so on.
Leif Svalgaard (July 5, 2012 at 4:21 am)
“It just so happens that I do have expertise in the meaning and physics of Ap. Do you?”
I do not falsely claim expertise in physics.
You do falsely project expertise in stats.
At the level of absolute logic you have erred above. That is a very specific & absolutely accurate grievance which you cannot escape.
Leif Svalgaard says:
July 5, 2012 at 4:16 am
More likely something nobody needs to know
so naïve can be fooled forever by the ‘status quo elite’
climate change is:
nothing to do with the sun
nothing to do with the Earth
everything to do with CO2
anything else is “spurious” !
(first 5 commandments)
For some even many it is fine, but for me not, I don’t by the ‘status quo’ science.
Clear cut spectral response of sun-earth-climate indices shows differently:
http://www.vukcevic.talktalk.net/Aa-CAM.htm
-Nature abhors coincidences; it is ruled by cause and consequence-
It’s time to realize, as old Winston said
“You can fool all of the people some of the time, and some of the people all of the
time, but you can’t fool all of the people all of the time.”
Leif Svalgaard (July 5, 2012 at 4:21 am)
“So as I said: The signature in Ap is spurious, just a coincidence, as there are many other Ap-steps not coinciding with any solar changepoints..”
You’re wrong on absolute logic. There’s no basis for communicating with you if you will not admit 1+1=2.
Reread what I wrote above more carefully ( http://wattsupwiththat.com/2012/07/04/another-regime-change-indication-this-time-in-solar-data/#comment-1024740 ).
If you want to make a sensible argument, the only option is to contest whether the changepoint is significant in the long-term vs. the short-term. Interpreting it as significant at the scale of 1 solar cycle is eminently sensible, accurate, & RIGOROUSLY DEFENSIBLE. HOWEVER (and I think this is what has you all bothered), it’s way, way, way too soon to interpret it as being significant at long timescales. Rather than simply say this, you have opted to engage in obfuscation, furthering undermining the basis for trust.
Paul Vaughan says:
July 5, 2012 at 5:31 am
I do not falsely claim expertise in physics.
You do falsely project expertise in stats.
Interesting difference there between ‘claim’ and ‘project’…
The behaviour of Ap is not a question of statistics. On a finer time resolution there is nothing special about October 2005. The specific ‘step’ is caused by a single, sporadic magnetic storm in September 2005 overlaying the general decline in geomagnetic activity that happens in every solar cycle.
At the level of absolute logic you have erred above.
Tell us what is the difference between ‘absolute logic’ and ordinary logic?
vukcevic says:
July 5, 2012 at 5:33 am
you can’t fool all of the people all of the time.
You seem to labor very hard to try to do just that…
Paul Vaughan says:
July 5, 2012 at 5:47 am
If you want to make a sensible argument, the only option is to contest whether the changepoint is significant in the long-term vs. the short-term.
Which changepoint? The Ap-one is spurious as there are many abrupt changes in Ap.
it’s way, way, way too soon to interpret it as being significant at long timescales.
It is just a coincidence and is not significant on any time scale.
Rather than simply say this, you have opted to engage in obfuscation, furthering undermining the basis for trust.
I’m not fishing for trust, and I have simply said repeatedly that the Ap-step has no significance that I can see [having expertise in Ap and its long-term behavior].
Paul Vaughan says:
July 5, 2012 at 5:31 am
You do falsely project expertise in stats.
The behaviour of Ap is not a question of statistics. On a finer time resolution there is nothing special about October 2005. The specific ‘step’ is caused by a single, sporadic magnetic storm in September 2005 overlaying the general decline in geomagnetic activity that happens in every solar cycle.
Perhaps a [clearly labeled] graph will help you over the hump: http://www.leif.org/research/October-2005-Non-Event-Fine.png
Since Ap is caused by the solar wind, the variation of its magnetic field [B] and speed [V] could also show you that there is no ‘changepoint’ in the solar wind either: http://www.leif.org/research/October-2005-Non-Event.png The arrows point to October 2005.
Correlation is not causation, but it is often a good clue as to where to look.
gopal panicker says:
July 4, 2012 at 3:47 am “”thats all good….but has there been any change in solar output””
Depends, . . . . .The data only goes back about “”The earliest surviving record of sunspot observation dates from 364 BC, based on comments by Chinese astronomer Gan De in a star catalogue.[7] By 28 BC, Chinese astronomers were regularly recording sunspot observations in official imperial records.[8]””
according to the “wickedpedia” http://en.wikipedia.org/wiki/Sunspot
So that means we have been “measuring”, compared to the estimated age of the sun, for just an instant.
I feel “confident” most would agree that there have been changes in solar output . . . what is “debate-able” is: how much change, was that change significant, and that ever populular question . . . were those changes due to human made causes???? . . .
An afterthought: Unusual activity of the Sun during recent decades compared to the previous 11,000 years
http://www.ncdc.noaa.gov/paleo/pubs/solanki2004/solanki2004.html
Laurie Bowen says:
July 5, 2012 at 10:52 am
An afterthought: Unusual activity of the Sun during recent decades compared to the previous 11,000 years
Very likely didn’t happen: http://www.leif.org/research/The%20long-term%20variation%20of%20solar%20activity.pdf
@Leif Svalgaard
1+1=1984. Loud & clear. (/sarc)
Paul Vaughan says:
July 5, 2012 at 7:10 pm
1+1=1984. Loud & clear. (/sarc)
Shirking away meaningful contributions, again.
Based on that Ap index graph, you could say that there is some longer-term cycle in play of 120 – 140 months. If that were so, the plot would only cover one and half of such cycles.
Plenty of other places in the graph where a sudden, sudden decrease in index takes place. In particular, if you took the plot from month 40 to month 80, it’d be pretty similar to months 150 – 200.
I’m not saying you’re wrong, or that I know what’s going on, just looking at the graph and seeing what it tells me……..
Rhys Jaggar (12:11 July 6) refers to longer term cycles. I don’t know if he refers to the Hurst or the raw Ap diagrams but if the former this was a recognized problem with this type of rescaled cusum analysis when it was in vogue in the hydrology literature in the 1970s. Because the graph uses the sample mean it is forced to zero at both ends and this imposes spurious long waves into the trace. Maybe things have changed but statistical significance tests of the time did not allow for this departure from true random walk and so led to false claims for the existence of long waves.
@Leif Svalgaard (July 5, 2012 at 7:36 pm)
Work obligations limit the time I have to volunteer comments here.
I don’t take your commentary on stats seriously.
FACT:
If you take the integral of second order central differences, you’ll shift the wave a 1/4 cycle.
This SHOULD be obvious to EVERY reader here.
Since it is NOT (very unfortunately), you’ve managed to get the wool over innocent eyes yet again. Anthony’s strong common sense appears to have led him to some vague intuition about this. Regrettably, I don’t have to the time to develop a little course module to straighten every innocent (potentially gullible) reader out.
I suggest everyone read the intermediate-level-stats comments P. Solar has volunteered to vukcevic over here:
a. http://wattsupwiththat.com/2012/07/03/detecting-regime-shifts-in-climate-data-the-modern-warming-regime-ended-in-1997/#comment-1024136
b. http://wattsupwiththat.com/2012/07/03/detecting-regime-shifts-in-climate-data-the-modern-warming-regime-ended-in-1997/#comment-1024282
For more advanced readers, see here:
1. http://wattsupwiththat.com/2012/07/03/detecting-regime-shifts-in-climate-data-the-modern-warming-regime-ended-in-1997/#comment-1024301 (LCAM, CLT, TTG = tropical temperature gradient)
2. http://wattsupwiththat.com/2012/07/03/detecting-regime-shifts-in-climate-data-the-modern-warming-regime-ended-in-1997/#comment-1024320 (ACC, aggregation criteria)
3. http://wattsupwiththat.com/2012/07/03/detecting-regime-shifts-in-climate-data-the-modern-warming-regime-ended-in-1997/#comment-1023997 (HCS)
Now: Back to work developing a customized nonstationary anharmonic wavelet analyzer to explore hierarchically coupled frequency shifts in QBO & solar-terrestrial data…
Best Regards to All.
Paul Vaughan says:
July 6, 2012 at 9:31 am
FACT: If you take the integral of second order central differences, you’ll shift the wave a 1/4 cycle.
And what has that to do with whether the October 2005 ‘step’ in Ap is spurious?
@Leif Svalgaard (July 6, 2012 at 9:55 am)
In the integral of 2nd order central differences there are going to be 2 changepoints EVERY CYCLE.
You’re attempting to authoritatively direct a narrative based on a step in the data to obfuscate the meaning of a changepoint in the integral.
This will fool naive followers (e.g. Pamela Gray), but the costly tradeoff of your ongoing obfuscatory commentary is the undermining of trust from more advanced readers.
Don’t waste anymore of my time.
It is often said (by ‘cranks’ and others) that there may be a link between volcanic and solar activity, I couldn’t find any using a volcanic index I constructed some time ago, since this is a thread about Ap, I decided to have another go:
http://www.vukcevic.talktalk.net/Ap-VI.htm
Not much in it, but at least it is an improvement on what I got last time.