2011-11-02 23:33:19Dodgy solar vs temperature graph
John Cook


Someone emails me about a dodgy solar graph:

I have noticed a few cases of this graph floating around http://climatechange.procon.org/view.resource.php?resourceID=003801

It was also in a movie that faked graphs. http://www.youtube.com/watch?v=0BmWy6Nsm6I such as some graphs covered in youtube videoshttp://www.youtube.com/watch?v=N2B34sO7HPM and http://www.youtube.com/watch?v=MYgX7tSGbmQ

[Note, I have only looked at climate science informally and only recently (not a scientist, etc).]

The arctic temp part of the graph appears to match this http://www.arctic.noaa.gov/report08/images/essays/atmosphere/a1.jpg (from a trusted source), but I could not find solar irradiance graph to match the one in that picture.

skepticalscience does have this graph http://www.skepticalscience.com/images/Temp_vs_TSI_2009.gif and this onehttp://www.skepticalscience.com/pics/Solar_vs_Temp_basic.gif and perhaps others showing solar irradiance. It is quite different.

If the graph mentioned at the top is perverted, is that picture worth an "argument" myth bust or can that example be added in the discussion of some existing one?

Is there a blog on this already? An argument? I searched a bit but didn't find it.

Anyone know where that solar data comes from?

2011-11-03 00:48:07
Tom Curtis


The only source that I know of is the fake paper distributed with the OISM petition:



The authors are Arthur Robinson, Noah Robinson and Willie Soon.

The data is referenced to:

Soon, W. (2005) Geo phys i cal Re search Let ters 32, 2005GL023429


Hoyt, D. V. and Schatten, K. H. (1993) J. Geo phys i cal Res. 98, 18895-18906.

2011-11-03 00:57:55
Dikran Marsupial
Gavin Cawley

One trick sometimes used is to present the integral of TSI solar minus some baseline.  If you choose the threshold nicely you can get a plot that looks a bit like the temperature time series.  I have seen this done before, but I cant remember where.  I suspect though in that case the labelling of the y-axis is probably bogus.


Having said which, the Hoyt TSI reconstruction does actually look quite like that already (although I could only find the data up to 1980, I suspect that the problem lies with the post-1980 data, where the fit doesn't appear anything like as good.


It ocurrs to me that the graph implies that Arctic temperatures are extremely sensitive to very small changes in TSI, which raises the question "why are Arctic temperatures so sensitive to solar forcing, but not CO2 radiative forcing?".

2011-11-03 03:28:39comment
Robert Way


I'm gonna be honest John. This could be close to the real graph... I'll show you something This is between us of course between us because its from a publication I have under review but this is a subarctic region and the relationship between temperature (black) and solar activity (grey) averaged over a 22-year solar cycle...

I think it is important to note that regions like the arctic can be more impacted and more sensitive to certain factors than other regions... I think climate sensitivity is somewhat latitude dependent and it is temperature dependent... higher sensitivity at higher temp... etc...

Notice the divergence at the end of this graph. (Remember Pinatubo cause a big dip in the 1990s)

2011-11-03 14:02:08
Tom Curtis


Robert, the main difference between your graph and the OISM graph is that it shows the 1980 peak in solar activity as the weakest rather than the strongest of three most recent half sunspot cycles (11 years).  Had it shown the correct values as given by PMOD, it would, like you, show a stong divergence after 1980.  What is more, there is good reason to think the significant correlation before the 1980's is at least partly coincidental.  Specfically, to attribute the pattern to solar forcing only, we need to neglect the hiatus in large volcanoes in the early 20th century, the significant reduction in aerosol emissions due to the great depression, the significant increase in Black Carbon emissions during the early 1940's, the significant increase in aerosol emissions from 1950 to 1970 and the hiatus thereafter, not to mention the rise in CO2 concentrations.

2011-11-03 20:41:23
Dikran Marsupial
Gavin Cawley

also worth noting that there is more than one TSI reconstruction, and it is possible that this is a cherry pick, perhaps plotting all reconstructions against the data would reveal that.