Ian H. Bryce writes at Jo Nova’s website:
The thing that intrigued me about the maximum temperatures is the high peaks, which occur at the peak of the odd solar cycles, and four other times, when we had strong El Nino events. (Most recently, three in four years) It is interesting to note that we did not have the Super EL Nino in 1998!
One wonders when our climate scientists graph global mean temperatures for tens of thousands of stations worldwide, that they “miss the wood for the trees.” I contacted the BOM some time ago about this phenomenon, but I have not had a reply yet. (Surprising?)
Read the entire fascinating article at Jo Nova’s website.
UPDATE: Willis finds some serious problems and posts in comments:
Not sure why I usually end up being the one to rain on the parade, but I’ve accepted my lot in life. Here is the Echuca data plotted against the peaks of the solar cycles, as measured by sunspot counts.
A couple things of note. First, he has misidentified the Cycle 11 peak, it happens earlier. Second, he is very vague about the timing of the cycles. Yes, the high years occurred during those cycles, but if we look at the actual peak year of each cycle, some happen two years before the peak temperatures, some three years before, some four years before, and some show no relation at all to the peak temperatures.
Sorry … but that’s the real data, and the sunspot/temperature correlation doesn’t hold up in the slightest.
SOURCES:
Willis,
You are using one weather station to make your point that Ian Bryce is relying on one weather station.to make his point. Wow!
Why not look at another record for Adelaide:
http://astroclimateconnection.blogspot.com.au/2012/06/singular-spectral-analysis-of-summer.html
ftp://ftp.bom.gov.au/anon/home/ncc/www/change/HQdailyT
Ninderthana says:
June 26, 2012 at 5:58 pm
Willis, You are using one weather station to make your point that Ian Bryce is relying on one weather station.to make his point. Wow!
It should hardly be necessary to even discuss this, as none of the solar activity indices have a clear 22-year cycle, so any rough 20-year correlation clear has other causes.
It seems the cherry’s are ripe this time of year yet again. What is interesting to note is that the solar cycles haven’t actually gotten any stronger yet when you look at the long term 30 year average maximum temperature data for Echuca you find that the average for the last 30 years is higher for every month, except November, than the total average for the length of tme the station has been there. This is consistent with warming across the rest of the country. http://reg.bom.gov.au/jsp/ncc/cdio/cvg/av
So if your assertion is that sunsppots are responsible for long term warming then it is incorrect because that would require solar cycles to be increasingly stronger, which they aren’t. If your assertion is that there has been no warming, well there’s no hope for you.
Leif Svalgaard says:
June 26, 2012 at 5:56 pm
Thanks Leif. I’m actually not looking for a “cycle” but rather to see if there is a consistent, short lived, observable, set of stimulus (the impulse) to which the system then responds. What intrigued me about Willis’ original Echuca plot and the latest for Deliniquin is the the sunspot maxima seem to partition the plots in such a way that from 1920 through 2000 each cycle represents a complete system response to some unique stimulus event at the beginning of the cycle. The question is what was the stimulus? Can it be identified? Can the system response be derived?
The result would be an empirically derived “pathway” for a highly complex system. This technique of analysis/modeling is similar to the way folks studying proteomics use experimentally measured metabolic pathways to partially understand complex cellular behavior.
Unlike the wet lab where biologists can carefully set up experiments and measure results, we don’t have that luxury with our climate system. This post stimulated the idea that maybe there were “natural” experiments occurring that could be identified and analyzed to provide empirically based climate system “pathways” that could be used to validate the veracity of current climate models.
techheadted says:
June 26, 2012 at 7:42 pm
I’m actually not looking for a “cycle” but rather to see if there is a consistent, short lived, observable, set of stimulus (the impulse) to which the system then responds.
The only short-lived ‘impulse’ that I know about is the polar field reversal near solar maximum, but it is not reasonable to think that that has any influence of any kind. This will, of course, not deter people from believing otherwise with their usual almost religious fervor.
Leif Svalgaard says:
June 26, 2012 at 7:58 pm
techheadted says:
June 26, 2012 at 7:42 pm
I’m actually not looking for a “cycle” but rather to see if there is a consistent, short lived, observable, set of stimulus (the impulse) to which the system then responds.
The only short-lived ‘impulse’ that I know about is the polar field reversal near solar maximum, and that happens every 11 years, not every 20.
In addition, the peak daily temperature always occurs about the same time of day, that is to say, the lag of temperature with respect to solar is stable.
There are 2 main factors affecting the time of the daily maximum temperature.
One is season, which determines the average angle of incidence of the suns rays, and hence the amount of insolation. The lower the angle of incidence the earlier in the day the maximum temperature occurs.
The other is cloud cover. Generally, clouds decrease solar insolation more than they decrease OLWR, and hence result in an earlier maximum temperature.
The peak in the Delininquin graph just before 1900, shows the effect of the Federation Drought in Victoria. Long periods of hot dry weather. Melbourne set a record daily average temperature in 1898 that still stands.
Leif Svalgaard says:
June 26, 2012 at 8:18 pm
The only short-lived ‘impulse’ that I know about is the polar field reversal…
I see. I think the key is to look for a “set” of stimulus that trigger a new “steady state” in the system (assuming that the Climate system has many equivalent energy states that can be reached by a small, relatively speaking, perturbation).
Time to crunch some numbers!
(from JoNova webpage)
Joanne Nova June 27, 2012 at 2:31 am •
Vukcevic, can I post that graph as well? It’s worth sharing.
Yes, please do, I have tried to post reply at your blog, but for some reason isn’t getting through.
This corner of Australia from Adelaide to Melbourne has some interesting features. The 100 year record 1900-2000 is long enough to observe natural changes, while this temperature record is important, if correct (it is not the absolute values but decadal rate of change) which has highlighted and also confirms that natural events (of temperature is only a hazy proxy) change direction only for exceptionally good reasons.
correction:
also confirms that natural events ( temperature is only a hazy proxy) change direction only for exceptionally good reasons.
At cross-ENSO Schwabe-timescale, the solar cycle spatiotemporally modulates zonal westerly winds — i.e. it cyclically ( http://i46.tinypic.com/2yw7711.png ) drives variations in this pattern:
200hPa Wind:
http://i52.tinypic.com/zoamog.png
200hPa Wind — Polar View:
http://i52.tinypic.com/cuqyt.png
Zonal Wind Vertical Profile:
http://i51.tinypic.com/34xouhx.png
It’s dead simple.
Solar-Terrestrial-Climate Weave
http://i49.tinypic.com/2jg5tvr.png
CAUTION: Obscurantism on this subject by climate discussion “leaders” has reached absolutely intolerable levels.
Processing the data Mr. Eschenbach has linked to gives me a completely different result than ‘dry as a bone’. For those who decide to look at it, the blue line is smoothed maximum temperature record and red line is smoothed sun sunspot number with each even cycle flipped to negative. The record starts at 1882 and ends 2010. And I hope you’ll excuse me for my lack of Excel graphing skills.
http://imageshack.us/photo/my-images/401/tempsun.png/
Of course, using Mr. Eschenbach’s approach all of that is ‘lost in spaghetti’.
@Kasuha (June 28, 2012 at 6:19 am)
http://i41.tinypic.com/29zxus7.jpg — from:
Tisdale, Bob. (2009). A closer look at the ERSST.v3b Southern Ocean data. http://bobtisdale.wordpress.com/2009/04/17/a-closer-look-at-the-ersst-v3b-southern-ocean-data/
@Scottish Sceptic (June 25, 2012 at 4:32 pm)
Schwabe-nonstationarity implies cross-correlation can’t yield an optimal estimate in this context, so I recommend careful diagnostics and cross-comparisons with other approaches that deal more gracefully with nonstationarity.
Kasuha says:
June 28, 2012 at 6:19 am
Thanks, Kasuha. As you can see in both your graph and mine, sometimes the solar leads the temperature, with a lag of one to four years. That alone would make a theory that the sunspot cycle is a cause of the temperature changes in Echuca quite suspect—why would the lag vary so much? And why would it be so long?
But the real problem is that sometimes the temperature leads the sunspot cycle, again by up to four years. How do you explain that? Does a rising temperature in Echuca cause the sun to have more spots?
Next, you have ignored a huge issue. You need to be very careful when you smooth data, because it leads to spurious correlations. See my post here for more insight into the problem, which it appears you don’t even know exists.
Finally, when data is smoothed, the autocorrelation of the resulting curves greatly reduces the significance of the results. You can use the method of Quenouille to calculate an “effective N” to remove the effect of autocorrelation …
Let us know how significant the correlation of the two lines is after applying the method. Because when I do what you did, and I calculate the significance, I find a p-value of around 0.10 to 0.14 depending on the amount of smoothing, meaning that the relationship between the two (temperature and sunspots) is NOT STATISTICALLY SIGNIFICANT.
I say again, you simply can’t just throw up a couple of lines, eyeball them from across the room, and declare that their relationship is significant. That’s why statistics was invented, and you neglect it at your peril.
For all of these reasons, along with the fact that (as I showed above) the nearest long-term temperature record to Echuca shows no signs of the Hale sunspot cycle, I say that you are looking at spurious correlation.
w.
Willis,
You have completely ignored my Singular Spectral Analysis result for the median maximum temperature for Adelaide at:
http://astroclimateconnection.blogspot.com.au/2012/06/singular-spectral-analysis-of-summer.html
You can recreate these results for yourself by downloading the data for Adelaide at:
ftp://ftp.bom.gov.au/anon/home/ncc/www/change/HQdailyT
and do the SSA.required.
You have also not addressed the paper by Thresher (2002) which completely disproves your contention that the Echuca result is somehow spurious.or unique.
It’s amazing what you can not see when your wearing blinders! Just because you don’t want a result, it doesn’t mean that it will go away.
Ninderthana says:
June 28, 2012 at 5:43 pm
You are not that important in my universe, Ninderthana. I didn’t “completely ignore” your graph. In fact I’ve never seen it, and have no clue about it. Perhaps you mentioned it on one of my threads. If so, I never got to it. In your imagination I’m shying away from or ignoring the truth you are bravely and loudly declaiming.
But in reality, you and your little graphs and citations haven’t been important enough to even appear on my radar.
And now, having seen your graph, I find it singularly uninteresting and unimpressive. Does the moon have an effect on the weather? Sure, every sailor knows that. Does that have anything to do with the Hale cycle? Well, by gosh, the new moon occurring at the moon’s perigee repeats every 20.3 years, and that’s kinda sorta near the length of the Hale cycle, that’s amazing, so I guess that must prove that the Hale cycle rules the weather …
Are you serious?
Again, this is the first I’ve heard of the Thresher paper, so of course I have not “addressed” it, how could I? And I’m sure that there are volumes of other stuff by both scientists and cyclomaniacs out there that I haven’t addressed. If that upsets you, well, sorry, that’s your business.
Oh, piss off with your nasty insinuations that because I haven’t read what your Majesty thinks is important that I’m wearing blinders. You are a most unpleasant person, Neanderthal, you should truly step back, take a deep breath, and start over. You’re not getting any points with me by your ugly lack of common politeness.
And by the way, it’s “when you’re wearing blinders”, not “when your wearing blinders”.
w.
PS—OK, now I’ve looked at Thresher, and I’m still unimpressed. He even says:
Yeah, both the periods and the amplitudes are different, that sure establishes his claims beyond doubt, that’s the kind of solid results that count in science.
Finally, does the Hale sunspot cycle affect the weather? I’d say there are datasets for which that is true, although the signal is weak, it seems like a second order effect … but I’d also say that Echuca is not one of those datasets, unless you think that the effect can somehow precede the cause, and that the lag from cause to effect can vary by up to ± four years between cause and effect.
And if you think that, then I can’t help you. If you think that the cause can either precede or follow the effect by a lag that can vary anywhere between one and four years, then you need to think hard about what you are calling causation, because you are not using the common scientific meaning.
Willis, your “understanding” of statistics in the context of climate is hopeless.
Thank you Willis for looking at my graph and for looking at Thresher (2002). I apologize for the last statement that I made in my previous post. It was rude and it implied that you were not open to new ideas on this topic. Clearly this is not true. I hope you will accept my apology.
You are correct in pointing out that Thresher did his best to point out the uncertainties of his results.
You are also correct in pointing out that their is ambiguity between the Lunar 20.3 year Perigiean/Perihelion tidal cycle and the 22 +/- 1 year Hale cycle.
The problem lies in the fact that the climate system is complex and surface temperatures are a poor diagnostic of the system. Many factors contribute to long term temperature records including the Sun and the Moon [amongst others].
The truth is that I was searching for a Lunar tidal signal in the long term median maximum summer (DJF) temperature record for Adelaide. I was surprised that I found a strong ~ 22 year signal.
What really convinced me that a 22 year signal was present was the fact that I found sub-harmonics [with confidence > 99.0 %] at 22.3 years/3, 22.3 years/6, 22.3 years/9, and 22.3 years/10 [and other sub-harmonics with confidence > 95.0 %].
This result lead me to Thresher (2002) paper which I believe does have something to say about the presence of a bi-decadal cycle in the Southern Hemisphere mid-latitude wind speeds.
Willis Eschenbach you are a very nasty man. For years John Gribbin has been my favourite pop science writer and his “The Strangest Star” a book I loved to read again and again. Now you’re telling me it’s all BS and Gribbin was R.O.N.G. spells Wrong. It’s a harsh transition from delusion to disillusion. Sob I’m heartbroken.
@Ninderthana
The next thing to do is window your analyses.
Embrace the power of Central Limit Theorem.
Where a hair-splitter sees a raised floor of noise, CLT sees pure signal.
Regards.
Ninderthana (June 29, 2012 at 3:08 am) suggested:
“The problem lies in the fact that the climate system is complex and surface temperatures are a poor diagnostic of the system.”
No “problem” in well-constrained aggregate.
Sun talks to Earth through fan:
http://i48.tinypic.com/2yydr92.png
Integral’s coherent with climate:
http://i45.tinypic.com/2nbc3dw.png
Consistent with Tsonis framework.
vukcevic (June 26, 2012 at 12:06 am) wrote:
“Hi Paul
Dickey and his colleagues did lot more research since 1997, by 2006 he may have found out that his calculations are good but conclusions could be wrong, and by 2009 that a revision may be necessary.”
She (not he).
I have already alerted the WUWT community on at least 2 occasions that Dickey’s recent suggestions (2011) are inconsistent with observation.
I have written to Dr. Dickey requesting information that is not publicly available. She has not replied.
I have also written to Dr. Jackson at ETH Zurich requesting information that is not publicly available. He has not replied.
If Dr. Jackson can find time to supply the info I’ve requested, I’ll be empowered to show the community how you’re getting some of your graphs. I’ve already figured out what you’re doing.
I sincerely hope Dr. Dickey will find time &/or permission to respond to my request. I deeply appreciate her research and I respect that she is not free to reveal all that she knows. Vukcevic, you may find the following of interest:
Dickey, J.O.; Marcus, S.L.; & de Viron, O. (2003). Coherent interannual & decadal variations in the atmosphere-ocean system. Geophysical Research Letters 30(11), 1573.
http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/11255/1/02-3203.pdf
As you read her articles, be aware that the pictures say things that have NOT been summarized in words. I wish you enjoyable, careful “reading”.
Best Regards.
Paul Vaughan says:
June 28, 2012 at 11:21 pm
Here’s a surprise, its another content-free, idea-free, and citation-free attack from Paul … truly, Paul, such meaningless outbursts do your reputation no good. You may be 100% right, but saying what you said just makes folks point and laugh.
If you have a scientific objection to what I said about statistics, you really should spell it out and back it up, because as it is, you just look like a sore loser …
w.
Ninderthana says:
June 29, 2012 at 3:08 am
You are a gentleman, sir, and your apology is most gracious.
Indeed. Look, I’m not opposed to the idea that the cycles of the sun, moon, and planets affect the climate. I’m just allergic to overblown claims of causation based on cycle-mining. Yes, the lunar Perigiean/Perihelion cycle is close to the Hale cycle, but in my world that means nothing.
So I’d like nothing more than to find incontrovertible evidence that say the orbits of Jupiter and Saturn affect the rain in Spain … it’s just that such evidence seems to be sketchy, incomplete, and subject to fading into and out of existence as the time passes. For a while a clear signal is there, as in the Echuca temperatures … and then it goes off the rails and the solar changes start trailing the temperature changes instead of leading them.
That’s the problem I have, not with the underlying idea that the heavenly bodies might affect the climate, but that the data for it doing so is soooo poor and occasional … I mean, if it were a strong effect, we’d see it everywhere, but we don’t. That means either it’s a very weak effect or that there is no effect at all. I suspect the former, but we just don’t have the data to decide that.
So my argument against Echuca above is not an argument that the Hale cycle is not visible in some weather somewhere sometime. Instead, it is an argument that the Echuca temperature data (smoothed or not) do not show a statistically significant correlation with the Hale cycle, and in fact sometimes precede and sometimes trail it, which makes it a very poor example for making the argument that the Hale cycle affects the weather.
My best to you,
w.