Hi Jennifer,
Following on from the discussion at 'Carbon Dioxide versus Temperature' I have done two plots first is the normalises annual mean CO2 growth rate with annual fossil fuel usage the second is normalised CO2 mean growth rate compared with sea surface mean temperature anomalies.
First FYI the provenance of the data in some the actual data is ftp and current that causes safari to crash the links to actual data are in the page.
For fossil fuel usage: http://cdiac.ornl.gov/trends/emis/tre_glob.htm
For mean annual growth rate of CO2: http://www.esrl.noaa.gov/gmd/ccgg/trends/
For mean annual sea surface temperature: http://www.cru.uea.ac.uk/cru/data/temperature/#datdow
The covariance for this is .63. The motivation for comparing annual growth rate with usage is that if there is a relationship the difference if any will be due to what is actually put into the atmosphere this assumes (quite wrongly of course) that all the other sources and sinks are inactive. So the caveat here is there will be much more actually influencing CO2 concentrations and the correlation could well be meaningless or due to common factors if there is indeed a link. Bottom line is that apart from the general trend (which leads to the relatively high covariance) the growth rate varies much more than growth in usage and the CO2 peaks and troughs don't match and I expect that if the data is detrended the covariance will be much smaller. I don't have time to check this as I have to finish packing and putting stuff out for council clean up.
This has a better covariance of .73 (correlation is the same) but as I suspected we don't see any lag. This is because there is a single data point at each year for each series and any lag less than year is likely to be completely obliterated. Since the CO2 levels have an annual cycle superimposed on the long term trend any such lag will be buried in the "noise". However we do have a physical (chemical) link with partial pressure of atmospheric CO2 and concentration in solution that is also temperature sensitive this needs more work than I have time for at the moment. There is however this http://www.rocketscientistsjournal.com/2006/10/co2_acquittal.html . Where he plots concentration versus temperature difference from the vostok ice core, below I plot temperature versus concentration difference which is better I don't know yet and I did it this way because that's the way I had the data loaded I'll look more closely when I get back [from Bellengen].
Looks similar to Jeffry Glasman's maybe it makes no difference but I'll have to convince my self of that on.
Cheers,
Jan Pompe
There about 100 places where air is collected in flasks about once a week and shipped back to Boulder, Colorado to be analysed for CO2 and other things. This avoids a bunch of
problems such as having instrument calibration drift at different
places. The US program is the most comprehensice, but Japan, Germany, France, Australia and others regularly measure CO2.
However, most of these are measuring CO2 at the surface and since almost all the sources and sinks are at the surface, the surface concentrations will vary more than the average over the atmosphere as a whole. If you are trying to analyse changes that
happen over only a few years, your statistics needs to take into account the difference between surface average and average over the whole atmosphere.
"The only station for measuring CO2 is that of Mauna Loa, Hawaain Islands.
Are there other CO2 measuring stations on the earth?"
Luke has told you numerous times, we have our own monitoring station at Cape Grim. It agrees with the Mauna Loa record.