2012-03-15 20:38:21Friss-Christensen and the solar cycle length and Global warming
Klaus Flemløse






Under develeopemt -  text will follow soon



Stauning.P(2011): Solaractivity-climate relations: A different approach. J.Atm.Solar-Terr.Phys. 73 p. 1999-1012.

Laut. Peter(2003):Solar activity and terrestrial climate: an analysis of some purported correlations. J.Atmos.Solar-Terr.Phys 65,  801– 812

Friss-Christensen, E. Lassen (1991): Length of the Solar Cycle: An indicator of Solar  Activity Closely Associated with Climate, Science, New Series, Vol.254,No5032.(Nov.1,11991),pp.698-700

2012-03-15 22:06:14
Ari Jokimäki


In case you need them, here are some references on the issue.

2012-03-16 08:05:40
John Cook


Klaus, great spot re the Stauning paper, valuable addition to SkS. This post needs the actual figures from the paper displayed in the post - that picture is worth 1000 words.

However, the bit about Svensmark filing for scientific misconduct seems out-of-place and unnecessary, I'd strike that paragraph. Ditto for the speculationg about Friis-Christensen - it's not up to SkS to speculate whether F-C has followed up his own work. I'm also uncomfortable about Jekyll/Hyde metaphor & comments on his staff - it makes the science personal and drowns out the important science in Stauning's paper.

2012-03-16 09:00:22E. Friis-Christensen et al about solar cycle length and global warming - declawed version
Klaus Flemløse


E. Friis-Christensen et al about solar cycle length and global warming

In the scientific debate on global warming and the sun/cosmic rays relations, Friis-Christensen et al have produced several fascinating and inspiring ideas. New and alternative ideas are always welcome and they deserve recognition for their contribution. Not all ideas have passed the test by the scientific community. However, it is clear that they have started a scientific debate and as well as a blog debate on solar influence and cosmic rays.

Friis-Christensen et al investigated the relation between solar cycle length and global temperature in a paper from 1991. They found a close relation between solar cycle length and global temperature for the years 1850-1985. As I understand it, they have been doing a data mining exercise and found this relation. However, their result could not be supported by scientific evidence. Data mining is OK in some cases, but there is a risk to detect flawed relations. The result was consequently challenged by Peter Laut in 2003.  Peter Laut has publicly criticized Friis-Christensen for adjusting data in an improper way. As a consequence there was at heated debate. 

Recently P. Stauning has produced a paper where he has investigated and updated the same graphics, now from 1850 to 2005, and corrected the calculation errors of Friis-Christensen et al. The result of this update is that the close relation from 1850 through 1985 has stopped and from 1985 the solar cycle length curve and the temperature curve are diverging.

This can be seen in the paper by P. Stauning on page   figure 5, page 2003.

This shows that the hypothesis based on data mining - put forward by Friis-Christensen et al – had been falsified. There could be other explanations for a relation between global temperature and the sun, so the debate on sun/climate relations will continue, but one of the branches of sun/climate debate has now been cut off.

The Friis-Christensen (1991) paper has been used again and again by the climate deniers without any protest from Friis-Christensen et al. The paper by P. Stauning will hopefully prevent the deniers from using the Friis-Christensen graphs in the future.

The paper by P. Stauning is an investigation in general of the sun/climate relations and the Friis-Christensen case is only a minor part of his interesting paper. P. Stauning’s paper deserves attention from readers of Skeptical Science.

However, I would like to express my thank you to Friis-Christensen and his team for their contribution to climate science in the past. Hopefully he and his team will publish both verifications and falsifications in the future.  


Stauning,P (2011): Solar activity-climate relations: A different approach. J.Atm.Solar-Terr.Phys. 73, 1999-2012.

Laut,Peter (2003):Solar activity and terrestrial climate: an analysis of some purported correlations. J.Atmos.Solar-Terr.Phys 65, 801– 812


Friis-Christensen, E. Lassen (1991): Length of the Solar Cycle: An indicator of solar activity closely associated with climate, Science, New Series, Vol.254, 698-700



2012-03-16 11:13:06
John Cook


I've added the graphic to your blog post above. That's one helluva strong graph!

2012-03-16 14:41:58
Sarah Green

That's a fantastic figure.

The legend terms "min-to-min" and "max-to-max" are not obvious to me. Can they be explained in the caption?

Likewise, what exactly is "solar cycle length"? and does it have units? 


"between solar cycle length and global temperature in a paper from 1991"


"between solar cycle length and global temperature in their 1991 paper."

"Recently P. Stauning has produced a paper where he has investigated and updated the same graphics,.."


"Recently P. Stauning updated and expanded the original data,..."

et al --> et al.,

2012-03-16 17:51:46
Glenn Tamblyn


Wow. Killer graph. So just to bullet-proof things, is SH and Global also available. Also second what Sarah says - need to explain what min-min and max-max mean in the context of the temperature record. Perhaps a caption under the graph would do it.


Bye Bye Solar Cycle Length.

2012-03-16 18:32:19
Rob Painting

Klaus, I agree with the other comments - that is one powerful graph for scuttling the solar/cosmic ray memes. 

2012-03-16 19:56:48Reply to Sarah and Glenn
Klaus Flemløse


Hi Sarah 

I have tried to explain legend the min-to-mim/max-to-max and cycle length.

Any suggestions to a shorter description is well come.

Hi Glenn

P.Stauning has honly showd the figures for NH. 

2012-03-20 08:50:33
Klaus Flemløse


Link to blogpost


2012-03-20 10:34:18Made some tweaks
John Cook


There were some broken images that I deleted from your post. If everyone is happy with this, I'll add it to the blog schedule.

2012-03-20 11:04:51Versus the Temperature v Solar Activity graph
Same Ordinary Fool


There is a similar divergence (although earlier, in 1960, and with less of a dive) in the TSI graph [Fixed Skeptic Argument #1, "it's the sun"].  I had remembered the details, but many readers won't.  They'll just remember the shape, the similar shape.  And be confused.  If not now, then they certainly will later, when they eventually learn that there are two of them. 

Is it worth briefly mentioning this similar graph, and providing a link, to avoid this confusion? 

2012-03-21 03:41:20
Klaus Flemløse


OK from me.

2012-03-21 12:21:23Issues with the original Friis Christensen graph
Tom Curtis


As an introductory remark, I like the post and think it should be published.  However, I will take this opportunity to raise an issue of my own to see if it is valid, and of interest to you for posting purposes.

I have never been able to take Friis-Christensen's theory seriously because of a simple fact check I performed on his graph.  That fact check was to compare the distance between data points, ie, the time between each solar cycle, with length of the solar cycles as indicated on the Y-axis of the graph.  Given a normal understanding of what the solar cycle is, ie, the length of time between sunspot maxima (or minima) the two values should be identical.  Instead what I found was substantial discrepancies.  For example I arrived at these values for the first ten points on this version of his graph:

X-axis value X-axis Cycle Y-axis Cycle


1865.6756756757 4.1277641278 11.6543778802 7.5266137524 7.3468902501
1872.8992628993 7.2235872236 11.4700460829 4.2464588594 4.4307906566
1877.371007371 4.4717444717 11.6543778802 7.1826334084 6.9983016112
1889.2383292383 11.8673218673 11.8525345622 -0.0147873051 -0.2129439871
1896.2899262899 7.0515970516 11.465437788 4.4138407364 4.8009375106
1901.1056511057 4.8157248157 11.4516129032 6.6358880875 6.6497129723
1907.2972972973 6.1916461916 11.2626728111 5.0710266194 5.2599667116
1912.285012285 4.9877149877 11.0599078341 6.0721928464 6.2749578233
1918.8206388206 6.5356265356 10.8755760369 4.3399495012 4.5242812985

The x-axis value is the year of the solar maximum or minum (it is not specified).  The X-axis cycle is the duration since the last value.  The Y-axis Cycle is the duration of the cycle as indicated on the Y-axis.  The second last column is the difference between the two values assuming the Y-axis indicates the length of the cycle that has just occurred, while the last value shows the difference assuming the Y-axis indicates the length of the cycle which is about to occur.  

The determination of values was made by a pixel count at high (3x) magnification, and would have an error less than 10%. I always measured to the trailing edge of data points so there date of occurence may have a larger error, but the length of cycles will be consistent and have a smaller error as a result.  Therefore error in determining the values cannot account for the obvious discrepancies.  The values are also discrepant with the assumption that data points represent half cycles.

I am not sure if this is relevant to your post.

Certainly a quick check of your graph showed only slight discrepancies which could well be due to measurement error by me.  There also may be some perfectly reasonable explanation for the discrepancies I have noticed.  However, if there is such a good explanation, no-one has yet been able to provide it to me.

My questions are,

Does the discrepancy exist in the values used by Friis-Christensen, or just an artifact of poor graphs?

Do you know a good explanation?

If they are not an artifact, and there is no good explanation, do you want to tackle this issue in your post?

2012-03-21 19:44:39
Klaus Flemløse


Hi Tom

I have not investigated the details in the paper, but I konw that Peter Laut has done this:


The blog post is a only a summray of the findings of P.Stauning. 

Hope you might find the answers here.