Hey Pikachu;
As usual, I disagree with Gothmog. A significant correlation between CO2 and temperature would imply (not "prove") that one is controlling the other. You would need that correlation if global warming is true. I know we disagree about how much other factors affect that relationship: paradoxically, you believe that aerosols will destroy the correlation in the short term, while at the same time, you insist that natural variants like aerosols will not protect us from anthropogenic global warming! I am intrigued by how global warming can be "massive" and yet "weak" or "statistically undetectable", depending on which would benefit GW proponents at the time!
Gothmog is partially correct. A strong regression coefficient would not
prove that CO2 drives temperature up. However, no correlation
would "disprove" (actually strongly undermine) global warming. That's what I am interested in, not what such an analysis would
prove, but what it could potentially
disprove. (This is why performing my test would not be in Gothmog's best interests. He has nothing to gain, and potentially everything to lose. Perhaps by offering the possibility of concession, I can increase interest in the test?)
Or in other words, if global warming is real, then more CO2 should cause temperatures to go up. And we are pumping more CO2 into the environment right now...therefore there should be an observed increase in surface temperature. This would not "prove" global warming because there is a possibility of coincidence, though unlikely. However, for global warming to be true, as long as CO2 concentrations increase temperature
must increase. If there is a
known and
quantifiable reason for temps not to increase, then I will soften the demand and allow its inclusion into the plot. But there will be consequences. If we have to include other variables to make it work, there should be a concession that there are other climatic factors that could potentially overpower GW. Also, I require that models not be used in this analysis. All data must be derived from observations of the real world.
It is my opinion that the
consistent refusal to comply with this test constitutes ideological protectionism, and is evidence that GW is non-falsifiable. I've learned more about the debate since I got here, the remote sensing data tempts me over to the GW side. (When you move Gothmog away from models, the science he presents is quite good.)
But until this observational issue is dealt with, I cannot concede the debate with a clear conscience.
Now, the plot above is from a politically biased site and probably should not be trusted at face value. But the data that is plotted is linked to above, and is from US government sources that tend to be sympathetic to global warming. If anyone happens to have a stats packages like SAS they could generate an expression of how confident we are that 2004 is warmer than 1978. Scuffer pointed out that I won't get a 99% or 95% CI. At this point, though, I've given up on that request and would be happy to just to know what CI we can get for the hypothesis, and whether the hypothesis, when tested statistically, is accepted or rejected. This would be a good start, I think, to fulfilling my concerns about GW in the real world, and would actually bring more data into this debate, the quality of which is in steady decline (and I take a lot of the responsibility for that).
Anyway, here is a global temperature anomaly dataset:
http://vortex.nsstc.uah.edu/data/msu/t2lt/tltglhmam_5.1
Along with some other links:
Here's a 2003 article about solar forcing:
http://www.csmonitor.com/2003/0717/p12s02-stss.html
A paper in Proceedings of the National Academy of Sciences
http://www.pnas.org/cgi/content/full/97/23/12433
Geophysical, archaeological, and historical evidence support a solar-output model for climate change
Charles A. Perry* and Kenneth J. Hsu
* U.S. Geological Survey, Lawrence, KS 66049; and Tarim Associates, Frohburgstrasse 96, Zurich, Switzerland 8006
Contributed by Kenneth J. Hsu, September 5, 2000
From the abstract:
Extrapolation of the model into the future suggests a gradual cooling during the next few centuries with intermittent minor warmups and a return to near little-ice-age conditions within the next 500 years. This cool period then may be followed approximately 1,500 years from now by a return to altithermal conditions similar to the previous Holocene Maximum.
From the paper:
Current global warming commonly is attributed to increased CO2 concentrations in the atmosphere (3). However, geophysical, archaeological, and historical evidence is consistent with warming and cooling periods during the Holocene as indicated by the solar-output model. The current warm period is thought to have not reached the level of warmth of the previous warm period (A.D. 800-1200), when the Vikings raised wheat and livestock in Greenland. Therefore, the magnitude of the modern temperature increase being caused solely by an increase in CO2 concentrations appears questionable. The contribution of solar-output variations to climate change may be significant.
Emphasis mine.
Hmm, looks like their model is different from GothMog's.