By Paul L. Vaughan, M.Sc.
The amplitude of Earth’s zonal winds is modulated by the solar cycle. Here’s a concise visual update based on the latest data:
LOD’ = rate of change of length of day
Data
ftp://ftp.iers.org/products/eop/long-term/c04_08/iau2000/eopc04_08_IAU2000.62-now ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/COSMIC_RAYS/STATION_DATA/Monthly_data/moscow.tab
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Jeff C (December 26, 2011 at 8:39 am) wrote:
“I have to agree with the bulk of the comments, this has to be one of the worst posts in WUWT history. Frankly, it’s embarrassing. Assuming it actually does mean something, how about explaining it for those of us that can’t read the author’s mind?
Anthony must not be back from his Christmas break. Better to post nothing than something like this.”
With serious financial backing I could do a lot more. I don’t have a lot of free time. Consider dropping the snark and being thankful for observation-based insights into nature that are volunteered. Who else here would (or even could) do the calculations and share them? Let’s see if Leif can reproduce the calculations.
Leif Svalgaard (December 26, 2011 at 9:45 am) posted
” http://www.leif.org/research/LOD-Excess-and-Change-daily.png “
You have the data. Now are you going to do the analysis?
Paul Vaughan says:
December 26, 2011 at 9:53 am
You have the data. Now are you going to do the analysis?
Just did, and it shows no solar cycle effect. It is impossible to reproduce what you did as you do not describe it clearly. This is all one has to go by: “The amplitude of Earth’s zonal winds is modulated by the solar cycle. Here’s a concise visual update based on the latest data:
LOD’ = rate of change of length of day. 0.5a LOD’ 11a power.”
Mark says….”some of us don’t have PhDs in anything.”
Don’t worry Mark it only means they have a certificate in intellectual masturbation.
Leif Svalgaard (December 26, 2011 at 9:42 am)
“On the plot at the right is the time series and 91-day and 365-year means. At the left is the FFT power spectrum with daily resolution. There are clear 1 year and 0.5 year lines, as well as the various tidal lines around 27-30 days [and their harmonics at half, one-third, etc of that], but no evidence of power near 10 years.”
Now that you’ve done your quick introductory look, it’s time to get more serious about exploring the details of the geometry. Plain FFT can’t handle this job. I’ve warned of it’s limitations repeatedly in past WUWT discussions (and again above). The analysis NEEDS to be windowed at a TUNED window width.
The power’s not at 11 years. It’s at 6 months (but you can find 11 year power in the temporal evolution of the 6 month power).
Isolate the semi-annual variation. Then let a grain = 0.5 year Morlet extent = 7pi (~11 year) complex wavelet resonate with it. Should take 10min to reproduce my results EXACTLY. If you don’t isolate the semi-annual variation first, you’ll get different results. I actually suggest doing the analysis both ways and thinking carefully about why the results come out different.
I need to go now. I work today and the next 4 days.
AFAIK, LOD is well correlated with Pacific trade winds, which dictate the ENSO conditions, which.. the rest is history.
Paul Vaughan says:
December 26, 2011 at 10:19 am
Leif Svalgaard (December 26, 2011 at 9:42 am)
Now that you’ve done your quick introductory look, it’s time to get more serious about exploring the details of the geometry.
There is no geometry involved that can be explored with time series analysis.
Plain FFT can’t handle this job. I’ve warned of it’s limitations repeatedly in past WUWT discussions (and again above).
If there is a signal strong enough to warrant your unqualified claim “The amplitude of Earth’s zonal winds is modulated by the solar cycle” plain FFT is perfectly capable of finding it.
The analysis NEEDS to be windowed at a TUNED window width.
Gives you mostly a circular argument. And the time series is short enough that variations of the solar cycle length doesn’t matter.
The power’s not at 11 years. It’s at 6 months (but you can find 11 year power in the temporal evolution of the 6 month power).
Yet you claim “The amplitude of Earth’s zonal winds is modulated by the solar cycle”
Isolate the semi-annual variation.
Gross did that and his plot [Figure 4] shows no solar cycle signal.
http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/28080/1/95-0060.pdf
He describes the method thus: “annual and semi-annual period that are of interest in this study. They have been isolated from the total LOD signal by singular spectrum analysis (SSA), a data adaptive technique wherein a time series is decomposed into a set of orthonormal basis functions (called temporal empirical orthogonal functions) that are determined by the time series itself rather than being constrained to bc, for example, functions of sines and cosines”
Juraj V. says:
December 26, 2011 at 10:22 am
AFAIK, LOD is well correlated with Pacific trade winds, which dictate the ENSO conditions, which.. the rest is history.
Indeed it is and Gross finds precisely that. Furthermore he shows that the dedacal signal in LOD is caused primarily by core-mantle boundary processes.
(posted also on the other thread by mistake)
Important question here is:
What drives ‘global’ temperatures (Land and Land & Oceans) ?
There are two major drivers:
1. The AMO with principal the ‘9 year’ oscillation, which is not related to the sunspot cycle (but not totally immune to it) , for more details see:
http://www.vukcevic.talktalk.net/theAMO-NAO.htm
see also:
http://berkeleyearth.org/pdf/berkeley-earth-decadal-variations.pdf
for the ‘9 year’ and importance of the AMO to global temps.
2. the Solar magnetic (Hale) cycle, not the sunspot one.
On this one I am writing more (on lines of the above linked as ‘theAMO-NAO’)
Hope everyone had a good Xmas
Sorry left out the all important link:
http://www.vukcevic.talktalk.net/TS.htm
Paul Vaughan says:
December 26, 2011 at 10:19 am
The power’s not at 11 years. It’s at 6 months (but you can find 11 year power in the temporal evolution of the 6 month power).
Yet you claim “The amplitude of Earth’s zonal winds is modulated by the solar cycle”. Now, if you restrict yourself to claim that the [tiny] semiannual variation’s amplitude is controlled by the solar cycle, then we can make a simple test that everybody can understand [and that FFT can pick up]: we make a synthetic LOD’ series with an annual cycle of 1 plus a semiannual cycle of 0.5 on average, but the latter modulated by an 11-yr cycle of amplitude 1. For this series we can calculate the power spectrum and compare it with the actual spectrum: http://www.leif.org/research/LOD-Change.png
The result of the modulation is a splitting of the 1/2 yr peak [right side]. No such splitting above the noise is seen in the actual spectrum. We would actually expect a tiny solar cycle modulation [splitting] of both the annual and the semiannual peaks because of the 0.1K temperature variation caused by the solar cycle change in TSI. You can, if you like, regard some of the minute peaks down in the noise as such side-lobes, but that’s a hard call.
Once again, if ANY intrinsic driver is capable of over-riding a solar modulation, the solar theory is DOA. As it does with CO2 theories. It completely boggles the mind that the truly powerful drivers are summarily ignored. The search for the single speck of liquid water on Mars is but child’s play compared to your search for a solar connection.
“The amplitude of Earth’s zonal winds is modulated by the solar cycle”
And ???
What are the implications in this.
Leif
Why didn’t you include a 183 days smoothing window?
Leif Svalgaard (December 26, 2011 at 12:42 pm)
“The result of the modulation is a splitting of the 1/2 yr peak [right side]. No such splitting above the noise is seen in the actual spectrum. We would actually expect a tiny solar cycle modulation [splitting] of both the annual and the semiannual peaks because of the 0.1K temperature variation caused by the solar cycle change in TSI. You can, if you like, regard some of the minute peaks down in the noise as such side-lobes, but that’s a hard call.”
Now you’re thinking’s getting a little more clear but solar excitation doesn’t behave like a clean orbital cycle. The power’s smeared around and you can’t track it down with plain FFT. The pattern I’m showing is a property OF THE MEAN over 1 solar cycle. THAT is what you’re failing to comprehend (so far).
lgl says:
December 26, 2011 at 1:44 pm
Why didn’t you include a 183 days smoothing window?
The 31-day window was to suppress the solar and lunar tides but leave longer periods basically intact, the 365-day window was just because I was curious. Since the issue was the semiannual wave, both 183 and 365 days would effectively suppress the semiannual wave, so little to be learned if including the 183 day window. You can ignore the 365-day window if it bothers you.
The first thing that got me was a philosophical thing: a plot of the rate of change of the length of the day. Give me a few hours for a few beers and I’ll work that out and then get on to the rest of the post.
@Leif Svalgaard (December 26, 2011 at 10:49 am)
Gross is studying a DIFFERENT statistic. After more than a full year you still do not realize the importance of tuning the window width. A complex wavelet can pick signals out of very messy data IF THE PROPERTIES ARE THERE ON AVERAGE. I’m trying to teach you something here.
Leif Svalgaard (December 26, 2011 at 10:49 am)
“Furthermore he shows that the dedacal signal in LOD is caused primarily by core-mantle boundary processes.”
NASA now acknowledges climate’s role.
Paul Vaughan says:
December 26, 2011 at 2:00 pm
Leif Svalgaard (December 26, 2011 at 12:42 pm)
Now you’re thinking’s getting a little more clear but solar excitation doesn’t behave like a clean orbital cycle.
I’m clarifying your obscure verbiage. And if you use a fixed window size [extent] also assume a clean cycle.
The power’s smeared around and you can’t track it down with plain FFT. The pattern I’m showing is a property OF THE MEAN over 1 solar cycle.
Yet you plot what looks like yearly data [at least very smooth] in your Figure 1.
THAT is what you’re failing to comprehend (so far).
What you are failing to show.
Pamela Gray (December 26, 2011 at 12:54 pm)
“Once again, if ANY intrinsic driver is capable of over-riding a solar modulation, the solar theory is DOA.”
Like the day & year? Might as well make your narrative that simple.
Leif Svalgaard (December 26, 2011 at 12:42 pm)
“the 0.1K temperature variation caused by the solar cycle”
And how exactly is that distributed spatially over the globe in your abstraction?
Paul, you are most unkind to the many with gratuitous insults and a superior attitude – how only you can perform the complex calculations, etc. etc. It is amazing that seemingly intelligent people (as we have seen with a number of arrogant CAGW stars and yourself) seem unaware that with the internet they are putting their wares before not just the whole world but a concentration of the better minds that are attracted to a site like this. Arrogance makes the CAGW priests think they can pull the wool over everyone’s eyes with suprisingly unsophisticated parlor tricks. Many commenters have given you good advice, the pithiest was that of:
John Slayton says:
December 26, 2011 at 8:31 am
““If you can’t make sense out of what you’re reading and it seems like so much nonsense, I beg of you, consider the possibility that it may be so much nonsense.”
Paul, you are not a magician, you are supposed to be informing not performing. If the majority of the commenters on this site are not ‘getting it’ the failure is yours. I’m afraid I agree with those who feel that the objective is to be obtuse and unintelligible. You may not know that a superiority complex is the very gravest form of inferiority complex.
Leif Svalgaard (December 26, 2011 at 10:02 am)
“It is impossible to reproduce what you did as you do not describe it clearly.”
NOT TRUE.
ALL of the info needed is given.
Morlet complex wavelet
grain = 6 months
extent = 7pi ~= 11 years
LOD’ = rate of change of length of day
ftp://ftp.iers.org/products/eop/long-term/c04_08/iau2000/eopc04_08_IAU2000.62-now
If you were a student writing one of my exams your score would be 0. As I suspected you don’t understand the stats and you can’t do the calculations.
You have all the info you need. Roll up your sleeves and learn how to escape the cross-scale exploratory limitations that have held you back. I’m trying to help you.
@Gary Young Pearse (December 26, 2011 at 2:17 pm)
Gary, the calculations are sound. There’s no sensible reason to tolerate Svalgaard’s suggestions otherwise. Just do the calculations.
Leif Svalgaard (December 26, 2011 at 2:08 pm)
“I’m clarifying your obscure verbiage. And if you use a fixed window size [extent] also assume a clean cycle.”
Do the calculations for a variety of extents and you’ll learn something about FOCUS. The way my software works, I have dial, so I can tune manually. Doing this for some familiar datasets that one knows well is a helpful way to gain intuition – and soon enough with a bit of experience one can pre-tune. Just vary the extent. I know you know to vary the grain. There are academics sending out the same message: vary the extent. They realize features are being overlooked. Do the calculations and you’ll realize they are sound. I’m not here to BS.