Guest post by Guillermo Gonzalez
I recently happened upon the SORCE/TIM website and decided to look up the plot of the full total solar irradiance (TSI) dataset (http://lasp.colorado.edu/sorce/data/tsi_data.htm#plots)
The SORCE mission began collecting TSI data in February 2003.
I was curious to see if the variations in the TSI had begun to rise yet, perhaps indicating a start to cycle 24. Visual inspection of the SORCE TSI plot showed just the opposite – variations continue to decline in amplitude. If cycle 24 has started, there are no signs of it in these data.
We can be a bit more quantitative if we examine, instead, a plot of TSI variance with time. I produced such a plot using the daily average TSI data provided on the SORCE web site.
The red data are variance values calculated at two-week intervals. The blue curve is the smoothed data calculated in the same way as smoothed sunspot numbers (basically a 12-month running average). Note, the vertical axis is plotted on a logarithmic scale.
To compare the recent TSI variance trend with the previous sunspot minimum, I looked up the ACRIM2 daily average TSI data at: http://www.acrim.com/Data%20Products.htm
These data are plotted on the same scale as the SORCE data. The smoothed data show a minimum TSI variance near the beginning of 1996, some months before sunspot minimum (October 1996). Notice that the minimum value for the variance during the 1996 minimum was about an order of magnitude larger than the present TSI variance.
The SORCE web site quotes long-term 1-sigma precision (relative accuracy) of their TSI measurements to be 0.001%/yr. This corresponds to a variance of 2 ´ 10-4 W2 m-4. However, the precision should be considerably better than this on the 2-week timescale that I selected for calculating the variance. Unfortunately, I have not been able to locate a quote for the estimated precision of the ACRIM2 measurements. It would be worthwhile to know if the minimum TSI variance of the previous sunspot minimum measured by ACRIM2needs to be corrected for the instrumental precision.
Guillermo Gonzalez writes on his background:
I’m an astronomer, though my present title is associate professor of physics at Grove City College, PA. I wrote a paper (in Solar Physics) with Ken Schatten back in 1987 on predicting the next solar maximum with geomagnetic indices. That was my only contribution on anything having to do with the Sun-Earth connection, but I also got a letter published in Physics Today in 1997 wherein I urged readers to takethe Sun-Earth climate connection more seriously.
These days most of my research is on extrasolar planets.
UPDATE: I received a suggestion for an overlay via email from Terry Dunleavy and I’ve worked one up below. This was done graphically. I took great care to get the two lined up correctly. Note however that the datasets span different lengths of time, as you can note on the two timescales I’ve included on the combined graph. The vertical scale matches exactly between graphs though. – Anthony
UPDATE2: Here is another graphical comparison of the two TSI variance graphs, scaled to have a matching X-axis and appropriately aligned side by side. – Anthony
kim (08:17:20) :
It is not wrong to contemplate how climate hypersensitivity might be ameliorated.
It is certainly not wrong. But to evaluate this the contemplation has to move a bit forward to become a hypothesis or a theory that specifies and quantifies a mechanism. I’m all for this and am eagerly awaiting such quantification.
“Solar activity in the 20th century was no higher than in the 19th and 18th centuries”
Leif may be correct but his work supposedly proving this is not generally accepted.
kim (08:17:20) :
The path to understanding is diffuse. Some deduce, some dream up. It is not wrong to contemplate how climate hypersensitivity might be ameliorated. It’s the big objection to TSI being the climate driver. You yourself denigrate other solar manifestations as simply not having enough energy within them.
What are these mysterious manefstations of which you speak TSI, Magnetism, solar wind and ……….
J. Bob (07:51:35) :
“……..shows the presence of a strong 57 year cycle correlated to the sunspot activity and the east England temperature.”
There is a natural 53 year cycle as shown here:
http://www.geocities.com/vukcevicu/SSNanomaly1.gif
http://www.vukcevic.co.uk/ solar current link
.
>> But these are sychronouse with SSN so should show up
>>on an FFT at the 11, 22, 33,44, 55 year positions. There
>>is no peak (temperature) at these periods so magnetism
>>must be playing a minor role.
As I said before, the oceans act as a temperature capacitor and smooth out any small fluctuations in solar energy forcing (through the agencies of magnetic flux, solar wind or cosmic ray cloud formation). The minimum movement time for temperature changes appears to be about 8 years.
Thus small fluctuations, like an individual 11-year sunspot cycle, will not really show up in the temperature record, and only the larger Gleissberg cycles (88 and 166-year cycles) really have an effect.
.
vukcevic (00:50:58)
An abrupt falloff of white-light faculae occured starting July, 2008 and continues to Dec 31, 2008. What it is doing right now I can’t tell you until the next data quarter comes in.
The Maunder also demonstrated (by meticulous observers of the time) a distinct and abrupt falloff, though it was in Sunspots.
There being more white-light faculae than sunspots during deep minimums (the record so far shows me this), this would be the last observable sign of activity on the Sun. If it goes, there’s nothing left but flux, and you can’t see that.
.
>>The aa-index is wrongly calibrated. There are two errors:
>>1) a jump up in 1957
>>2) a change of observers in ~1937
>> http://www.leif.org/research/Analysis%20of%20K=0%20and%201%20for
This ‘paper’ is the biggest load of cods I have ever seen. It’s central thesis is that, quote:
“Since aa is too low, it seems possible that many K1 values were misclassified as K0”.
In other words, the data looks wrong (too low) therefore the measurements are suspect. Thus the author will attempt to adjust the presumed errors (that are too low), to make early aa geomagnetic values the same as modern ones, and thus eliminate any perception of a solar influence on global warming!!!
One might also speculate that we are currently witnessing a peak in the sunspot cycle, with some 150 pots a day on the surface of the Sun, it is just our measurements that are at fault, which is why we think we are at a sunspot minimum!!
Clearly, the true answer to this ‘problem’ is that the early 20th century aa geomagnetic data appears to be low SIMPLY BECAUSE IT WAS LOW. The data is correct!
The aa data was low because the sunspot activity was low. And because it was low, global temperatures were lower too. Thus the Sun is the major forcer of climate change.
Q.E.D.
I have never seen an era in the West for fiddling the data like this one. We are almost getting to the stage where will will emulate the Soviet grain harvest figures (the most fiddled data in the history of mankind).
.
Where is it saved the energy which comes from the sun, where is that little pink pig we used to save our coins? (hope it is not with flu :))
Is it the “seven seas”?, if so, how is it going the earth’s current economy?
Gwrs. say no, they have a rather gaseous pig: CO2, but when you put coins in it they just fell through…
I think we need an economist to fix our economy or, at least, to tell us what is it ahead.
In any case what we do not need is a black magician to scare us.
MartinGAtkins (06:19:15) :
The annual 7% oscillation is not detectable because over the thousands of years it has been going on, all feed backs are fully utilized.
The same can be said for the solar cycle(s). But more to the point. A feedback is a process or a mechanism. Please list all the feedbacks that are fully utilized for the annual oscillation.
You would be able to detect the variation of the annual TSI if you removed all our atmosphere along with the water. A back of an envelope calculation tells me it would be about 7%.
Please share with us the back of the envelope calculation details [can’t be that hard as it fits on the back of an envelope], thanks.
This is a factor I have yet to see anybody address with regards to the effect of the solar minimum. Could this be a contributing factor in why the Earth may cool during low solar activity?
First, we know that we are seeing that the Atmosphere has shrunk due to the solar minimum:
“During the first months of CINDI operations the transition between the ionosphere and space was found to be at about 260 miles (420 km) altitude during the nighttime, barely rising above 500 miles (800 km) during the day. These altitudes were extraordinarily low compared with the more typical values of 400 miles (640 km) during the nighttime and 600 miles (960 km) during the day.”
http://www.sciencedaily.com/releases/2008/12/081215121601.htm
The diameter of the earth at the equator is 7,926.41 miles (12,756.32 kilometers).
But, if you measure the earth through the poles the diameter is a bit shorter – 7,901 miles (12,715.43 km). Thus the earth is a tad wider (25 miles / 41 km) than it is tall, giving it a slight bulge at the equator.
Taking the average of the diameter at the Equator and the Poles, I get 12736 km (rounded) km.
If we look at Earth’s atmosphere at a plane perpendicular to the solar rays from the sun, we see there is an amount of light that strikes the surface, but there is a portion of that solar radiation that passes through the atmosphere around the edges of the perimeter of the earth. Lets define the amount of sunlight as that just passes through the atmosphere, without directly striking the earth, as the Earth’s “atmospheric halo”.
Earth Atmos Diam Radius area sunlight
sq km
Surface Diameter 12736 0 12736 6368 127400409
Atm Diam current 12736 1600 14336 7168 161421250
Atm Diam “typical” 12736 1920 14656 7328 168707983
Halo Current 34020841
Halo Typical 41307574.3
Difference 7286733.31
Plugging the numbers above (by averaging day and night altitudes), we see that in a “typical” atmospheric height, we have 41,307,574 sq km of sunlight that passes through that halo. During the current solar minimum, we have 34,020,841 sq km of sunlight passing through that halo, or a reduction of 7,286,733 sq km of sunlight that no longer passes through Earth’s atmospheric halo. While these values represent current compared to “typical” heights of atmosphere, we do know that during periods of extremely high solar activity, the upper layers of the atmosphere may extend even further above the surface, and increase the amount of “atmospheric halo” that could potentially trap even more solar radiation.
While much of that solar radiation may escape back out of the atmosphere without striking Earth, how much of it would be absorbed, and how much of an effect would that have on global average temps?
What is also interesting to note, is that as the Earth rotates, the portion of the atmosphere in the halo at the equator would move quickly, but the halo at the poles, particularly during the equinoxes, would remain fairly constant.
My question is for the scientists out there, has there ever been any significant research done in calculating how much more solar radiation is trapped in the fringes of this “atmospheric halo” during periods of high solar activity versus low solar activity?
ralph ellis (09:52:11) :
Clearly, the true answer to this ‘problem’ is that the early 20th century aa geomagnetic data appears to be low SIMPLY BECAUSE IT WAS LOW. The data is correct!
It would be nice if it were so, but, unfortunately it isn’t. That aa is wrong in the past is now generally accepted. The detailed argument and some references can be found in section 5.3 of this paper:
http://www.leif.org/research/2007JA012437.pdf
The first note I referred you to is a first attempt to uncover why aa was too low before 1937.
.
>>It is not only insignificant, it is also wrong. Solar activity in the 20th
>>century was no higher than in the 19th and 18th centuries, while
>>temperatures were:
>> http://www.leif.org/research/Napa%20Solar%20Cycle%2024.pdf
Another load of cods from Leif, I feel.
Here he notes a difference between the Greenwich sunspot numbers (prior to 20th century) and those from Zurich – and notes reasons why the Zurich team might have made such errors. But to make amends, he add a fiddle factor not simply to the Zurich data, but to both sets of data !! Thus, at a stroke, he has managed to equalise the early data and the later 20th century data – thus eliminating any increase in solar activity.
Like his geomagnetic data, there seems to a desire here to deny that solar activity has increased over the last 3 centuries.
In actual fact, the variability of sunspot observation techniques is well known and already allowed for in the graphs that NASA (and others) provide. This is the ‘k’ factor, which is applied to each data set.
In addition, groups of sunspots are the biggest factor in calculating sunspot numbers, not individual spots. These are multiplied by an arbitrary figure of 10, to produce ‘individual’ sunspot numbers. Thus the fear of many people, that modern techniques are counting more (smaller) sunspots, is largely negated. A group is classified as 10 spots, no matter how many spotlets and specks are in that group.
Thus NASA are happy to acknowledge that there HAS been an increase in sunspot numbers in many of the cycles over the last 3 centuries.
http://www.ngdc.noaa.gov/stp/SOLAR/SSN/image/annual.gif
How sunspot numbers are calculated:
http://www.ngdc.noaa.gov/stp/SOLAR/SSN/ssn.html
http://science.nasa.gov/headlines/y2000/ast05apr_1m.htm
P.S. A word of advise to Leif. Try to explain your self better and use less jargon in your papers. You might think this makes you look intelligent, but it actually makes your papers look like the ramblings of a deranged pseudo-scientist.
Leif,
How about my shot at this. You mention there being no difference in temperature due to solar variation (well that’s not exactly what you said, but is maybe how I am reading it), when in fact there is. Throughout the year, there is a difference and the order of ~4K between the high global mean and the low global mean. So there is an anual oscillation that coincides with distance from the sun. The oceans (or as I like to call it, the world’s thermal inertia) even out the temperature as best as it can, as a result temperature lags TSI by about six months eyeballing it.
Of course, I would be remiss if I did not mention that the T variation only accounts for ~ 60% of the TSI variation. How do I explain the other 40%. Pixie dust (It’s as good as CO2 isn’t it.) Well, maybe not really. If I had to guess (and in the face of my overwhelming lack of proof, I do), I would say that the upper few meters or less of the ocean store 60% of the heat from TSI to be released over the course of the year. The other 40% is lost to deep ocean mixing and the increased heat loss that would occur from the larger delta T between the atmoshphere and space and quite possibly some other mechanism that I have not yet thought about.
I have no way to prove any of the following, but this would be my “pet hypothesis”. During the LIA, there was some sort of random/external negative forcing event (cosmic dust cloud, period of increase volcanic activity, extra measure of climate pixie dust, etc…) that cooled the earth by 1-3K. Over that last couple centuries, depending on when you think it ended, the earth has been playing catchup. Because of the thermal dampening effects of the ocean, that I so ineptly attempted to conceptualize above, I think that the world is only able to warm up a little bit each year in terms of mean surface temperature. There is simply too much inertia in the deep oceans for it to occur any faster.
That being said, I’ve come to agree with you that the changes in TSI are too small to produce a meaningful change in temperature over the course of a year, or even two or three. I’m open to feedbacks, even ones based on the sun, but temperatures are not rising or falling at an extreme rate year over year, and definitely not at a rate that is going to affect me in my lifetime (and quite possibly my children’s lifetime either.)
LB
ralph ellis (09:52:11) :
The data is correct!
Perhaps a somewhat more accessible exposition would be this Seminar talk I recently gave at the University of California, Berkeley:
http://www.leif.org/research/Seminar-SPRG-2008.pdf
.
>>It would be nice if it were so, but, unfortunately it isn’t. That aa
>>is wrong in the past is now generally accepted.
Generally accepted by Svalgaard, I am sure.
But this is simply an extension of the fiddling that has been going on lately, just like the recent elimination of the Medieval Warming period.
http://scienceandpublicpolicy.org/
See the video, of how the Medieval Warm Period was eliminated by the UN and other Green sympathisers.
Do you believe the deletion MWP should be ‘generally accepted’, Leif? ~snip~
L Bowser (11:25:09) :
Throughout the year, there is a difference and the order of ~4K between the high global mean and the low global mean. So there is an anual oscillation that coincides with distance from the sun.
This matches my estimate of 5K, so we don’t disagree on that. My point was that people claim that there are ‘feedbacks’ that amplify the 0.04K caused by the solar cycle but leaves the 5K alone. All I want to know is what those ‘feedbacks’ are, how they work, and how their effect varies with the time scale involved [e.g. a graph or equation that elucidates that].
>>That aa is wrong in the past is now generally accepted.
>>The detailed argument can be found in section 5.3 of this paper:
>> http://www.leif.org/research/2007JA012437.pdf
Again the data does fit your theory, so the data needs amending. It was Einstein who said “If the facts don’t fit the theory, change the facts” – but I think you take this too literally.
I expect you can also amend the Maunder Minimum to have the same geomagnetic activity and number of sunspots as the 20th century mean, if you try hard enough, but I would no believe that either.
Ralph
ralph ellis (11:19:56) :
Another load of cods from Leif, I feel.
Here he notes a difference between the Greenwich sunspot numbers (prior to 20th century) and those from Zurich – and notes reasons why the Zurich team might have made such errors.
There has been indications for some time now that the early Sunspot numbers are off by 25-40%. There are the “ramblings of a another deranged pseudo-scientist”, Peter Foukal:
http://www.leif.org/research/Foukal-F107-Rz.pdf
It is from:
GEOPHYSICAL RESEARCH LETTERS, VOL. 25, NO.15, PAGES 2909-2912, AUGUST 1, 1998
and reads:
Extension of the F10.7 index to 1905 using Mt. Wilson Ca K spectroheliograms
by Peter Foukal
Abstract. The F10.7 index provides a daily record of solar microwave emissions, which vary in rough proportion to the projected area of bright magnetic structures called plages and network, and also sunspots on the sun’s disk. The daily observations used to form the index only began in 1947. Recently, we digitized the archive of daily Ca K spectroheliograms obtained at Mt. Wilson Observatory between 1905-1984, and measured the area variations of plages and enhanced network, on these photographic plates. We calibrated these variations against the F10.7 index between 1947-1984, so we are able to construct a full-disk proxy of F10.7 extending back to 1905. The behavior of this extended index indicates that UV irradiance levels achieved near the peaks of sunspot cycles 15, 16, and 17 between 1915-1945, were 25-40% higher than would be estimated from behavior of the Zurich sunspot number, Rz.
—————
Instead of the Sun pumping out more UV in 1915-1945 we tend to believe that the Sunspot Number calibration is what needs to be changed [to be higher]. I’m, in fact, working with the Mount Wilson observers right now to submit a paper suggesting just that, based on their CaK data.
– – –
Re: bill (04:07:47)
I’m still not convinced that you are aware of the limitations of FFT. It’s a useful tool, but it is just one tool in the box – the one that comes right after time-plot & scatterplot during a quick overview of some new time series …..then the analysis starts…..
Also, I’m not sure if you interpreted Micky C (MC) (06:36:54) judiciously.
Micky certainly shared a great post, which should help us realize the failure of our education system to ensure that each & every human being is aware of the nature of a good variety of complex phenomena. At this stage, it is actually difficult to imagine what could be more important (long-term) than getting solid courses on chaos & fractal geometry into school systems worldwide – to help break civilization’s increasingly-fatal slavery to misguided belief in linearity where none exists.
– – –
Peter Taylor (05:13:36)
“[…] so these ocean cycles, land cycles, solar cycles – they do not line up easily and require a more sophisticated analysis using time lags and harmonics […]”
Good to see a comment acknowledging that the real game is not one of simplistic wiggle-matching. [If it was, the analysis would be done/over/finished/complete….]
– – –
Re: Basil (07:01:02)
The problem with global averages is that the spatial averaging destroys important signals (that are out-of-phase across sharp-gradients, for one example).
A thorough analysis needs to consider a range of scales in both time & space (…and that can include global/annual resolution).
– – –
Leif Svalgaard (07:59:19)
“The important result is that geomagnetic activity in 1845-1875 is comparable to the past 30 years, while temperatures are not.”
I have to challenge you on this Leif. We are not dealing with a simple bivariate system. If the role of the hydro cycle in climate variations was fully understood, you might have a better case with an argument of this nature.
– – –
Re: Dell Hunt, Michigan (10:25:24)
Thanks for pointing this out:
“Boundary Between Earth’s Upper Atmosphere And Space Has Moved To Extraordinarily Low Altitudes, NASA Instruments Document” (Dec. 16, 2008)
http://www.sciencedaily.com/releases/2008/12/081215121601.htm
In light of analyses I performed a year ago (while investigating severe statistical errors that made their way into public policy), this story does not surprise me.
Perhaps the people responsible for investigating mechanisms need to spend less time saying “no mechanism” and more time investigating mechanisms.
ralph ellis (12:03:12) :
Again the data does not fit your theory, so the data needs amending.
Well, other groups [as referenced] agree that the aa-index needs amending, so we just have to take it from there.
Paul Vaughan (12:30:35) :
“Leif Svalgaard (07:59:19)
“The important result is that geomagnetic activity in 1845-1875 is comparable to the past 30 years, while temperatures are not.”
I have to challenge you on this Leif. We are not dealing with a simple bivariate system.
I’m not sure what you are challenging. I was only making a statement about the data, not its interpretation. Are you challenging my statement that temperatures the past 30 years are not comparable to what they were in 1845-1875? That would mean that you would say that the temperatures 1988-2008 were not different from 1845-1875. Perhaps that is so, but I think you won’t find many takers of that.
Re: Leif Svalgaard (12:48:56)
To reiterate:
“If the role of the hydro cycle in climate variations was fully understood, you might have a better case with an argument of this nature.”
Paul Vaughan (13:08:33) :
“If the role of the hydro cycle in climate variations was fully understood, you might have a better case with an argument of this nature.”
I [obviously] have no idea of what you are hinting at [repeating it three or more times won’t help]. I was only making a statement about the data, not its interpretation. Are you challenging my statement that temperatures the past 30 years are not comparable to what they were in 1845-1875? That would mean that you would say that the temperatures 1988-2008 were not different from 1845-1875. Perhaps that is so, but I think you won’t find many takers of that.
.
>>The behavior of this extended index indicates that UV irradiance
>>levels achieved near the peaks of sunspot cycles 15, 16, and 17
>>between 1915-1945, were 25-40% higher than would be estimated
>>from behavior of the Zurich sunspot number, Rz.
You’re making this up as you go along, aren’t you?
Your very own 2008 paper said that the Zurich data was correct from 2000 back to 1875, but now you cite evidence that it was incorrect between 1915 and 1945. But if it was incorrect in the 1900s, then this evidence negates the whole of your arguments in your Dec 2008 paper.
So who is right, you or Foukal – you cannot both be right?
The answer is clear, you are both wrong and there is no need to readjust the historical sunspot and geomagnetic data. Counting sunspots is hardly a dark art, after all.
Q.E.D.
Ralph
Hello to all and apologies Anthony if this post is a little laboured. The reason for my last post and I believe it touches on a lot of what Leif talks about is that the oceanic and atmospheric dynamics appear to have, or possibly must have, a complex coupling matrix between natural bounded oscillation patterns, attractors you could call it, with input frequencies.
Now if you go and look up ‘drift waves plasmas’ or ‘anomalous resistance’ in plasmas, which is very interesting, you will see that time and again people were able to stimuate an acoustic oscillation using an electromagnetic oscillation (either RF or a simple duty cycle in what is a d.c. power supply).
What happens in the plasmas I was talking about is that the electrons carry the information about the anode and magnet d.c. supplies (i.e. the duty cycle). If you hook the return lines up you see the oscillation. Its normally at 100s KHz. I’ll also clarify that the anode voltage looks like white noise in the low kHz region, where over a certain power level up pops the acoustic oscillation, which then also changes in frequency slightly (like 17 to 20 kHz for example) as power is increased until you get some amazing Cantor dust interference and plasma turbulance. What happens is the resistance goes negative in the plasma for a period of time. Yes I will repeat that: negative or anomalous resistance. The VI curve inverts. Power supplies go crazy.
Now the reason behind this is for low power the electrons dictate the current flow, but as they interact with ions, they give them a little jiggle. For a certain amount of power the ions can’t just vibrate at random (thermal background) but they have to vibrate somehow and this is dictated by the acoustic effect modified slightly with electron temperature. When there is too much power the ions are moving so much that they cause the current carried by the electrons to be frustrated and so they become the dominant carrier. But only for a while. What ensues is a kHz Predator Prey type behaviour which I can guarantee does not have a fixed FFT. It is a real visible chaotic effect replete with scaled self similarity. But then more power and suddenly its all still and back to the start
The coupling matrix appears quite broad band so it is not so much about exact synchronous resonance as well. There’s a nice 1973 paper in Phys Rev Letts about this (I don’t have the exact ref on my home PC). Another factor is that once the behaviour starts it actively changes the input i.e. it feeds back on itself changing the input frequency characteristics and the coupling matrix. Fascinating stuff.
So consider that it may be that the Earth’s climate patterns have coupling frequencies within certain bands but then they actively change depending on their current state. If the coupling is linked to geomagnetic or TSi it may not show up as a direct correlation; the dynamic evolution may need to be mapped.
Of course it may be simpler: it may be based around CO2 as the driver. Just maybe not as linear as the current models assume.