Endersbee compares 12 month averages of CO2 with a moving 21 year average of SST. Sorry, but you can enhance any spurious correlation with techniques like that…

Further the graph shows a good correlation, because it compares temperature and CO2 levels where both go up in the period 1985-2008. If you include the period 1900-1985 (ocean cooling 1945-1975, CO2 up) the correlation is far worse…
See: http://www.ferdinand-engelbeen.be/klimaat/klim_img/temp_emiss_increase.jpg

One need to compare yearly averages with yearly averages. The correlation then is far less, the causation in part right, but limited: some 3 ppmv/°C. That is based on the influence of temperature on the variation around the trend during the 1992 cooling (Pinatubo) and the 1998 warming (El Niño).

You should definitively read my page about several reasons why the increase of CO2 in the atmosphere is largely from the emissions (there is a small addition due to the temperature increase 1900-2000, but that is reversing now):
http://www.ferdinand-engelbeen.be/klimaat/co2_measurements.html

The mass balance is the ultimate proof of the contribution of humans to the increase of CO2 in the atmosphere. As long as more is emitted than there is increase, there is no net addition from nature. That not all emissions are absorbed by the oceans/vegetation is a matter of process equilibrium: one need a driving force to push CO2 from the atmosphere into the oceans(/vegetation), which is only possible if the pressure of CO2 in the atmosphere (pCO2atm) is higher than the average pCO2 of the oceans. Thus levels in the atmosphere must go up first. The higher the pressure difference, the more CO2 is absorbed.

With increasing emissions, the pressure difference will go up, but never will reach a level where all emissions are absorbed. If we should stop the increase and the emissions should stay at a certain level, then after some time, the pressure difference may increase to a value where as much CO2 is absorbed by the oceans as is emitted by humans. Thus a new equilibrium is reached. This is typical for a simple first order (physical absorption) process.

Another quite solid indication is the decline of 13C in the atmosphere and upper oceans. Vegetation can’t be the source of extra CO2 (and the cause of the decline), as the oxygen deficit points to more vegetation uptake than decay. The oceans can’t be the source of extra CO2, as that has a zero to positive impact on 13C levels, while we see a decline. Moreover, CO2 levels in the upper oceans increase too and the pH is declining.

Thus there is only one known source of the increase: human emissions…

My impression is that several skeptics don’t like the idea that we are responsible for the increase in the atmosphere, just because if that weren’t true, that would falsify the whole AGW theory…

The different stations all have QC for their own data, which rejects data for daily, monthly and yearly averages, based on objective criteria. Some are general: large spikes to either side (values outside 3 sigma of the previous series of values) and instrument malfunction (calibration with standard mixture outside the tolerance). Others are site specific: land side wind (Barrow, La Jolla), upslope wind (Mauna Loa), etc… Daily and monthly averages in the NH may differ 10-15 ppmv from SH averages, due to seasonal influences, yearly averages are all within 5 ppmv of each other.

Differences between subsequent measurements usually are within tenths of a ppmv, there is no QC adjustment or comparison with measurements of other stations other than that all use calibration mixtures which are calibrated to the same central standard. Thus while all stations show the same trends (and nearly the same increase each year compared to each other), there is a bias in the trends, where the SH stations lag the NH stations. This points to a CO2 source in the NH. The ITCZ forms a barrier which hinders the tranport of CO2 between the hemispheres. That is the reason that there is a current (and slightly growing) difference of about 5 ppmv in the SH trends vs. the NH trends.

Even if one includes all outliers, that doesn’t change the trend with more than a tenth of a ppmv. In fact, it is a luxury problem: there are so many good data that we may choose the best available. One can afford to reject any suspect data (for whatever reason), as that doesn’t affect the only thing we are interested in: the trend of CO2 in the atmosphere on longer term.

For the South Pole, we don’t have that luxury, as besides a brief continuous period, only biweekly flask samples are taken, which is enough to measure the trend. Even there one sees sometimes one flask from a triple with spikes of +50 ppmv, unexplainable…

Over this period we had positive PDO, ENSO, and NAO. As water warms, it outgasses CO2. As surface water warms up, it will dump excess CO2 into the atmosphere meaning that the resulting increase in CO2 is caused by the ocean warming and not the other way around. The rate of increase in atmospheric CO2 looks *nothing* like the rate of increase in human CO2 emissions but does look identical to the rate of increase of global SST.

To test that idea, lets have a look over the next few years as temperatures cool. We currently have a cold PDO, neutral ENSO, and cold NAO. Global SST should drop or possibly go negative. And as we have seen, the rate of CO2 increase has, in fact, declined even with the adjustment. As we watch it over the next 12 to 24 months, I fully expect it to continue to track with global SST anomaly.

“So every place has its own specific problems. Barrow has low readings in summer when the wind is from land side (tundra), La Jolla also if the wind is from land side. Mauna Loa sometimes from fumaroles but more frequently in the afternoon from upwind conditions. Despite that, from the 8600 measurements per year (Mauna Loa), some 60% are good enough to give a profile of what happens with CO2 in the atmosphere.”
Its that part “good enough to produce a profile of what happens with CO2…” with which I have a problem. I still do not understand why 60 percent is good enough, but more importantly why are the other 40 percent not good enough. Why can we not have nearer 100 or 90 percent. The Alps and Rockies aren’t inaccessible. Since when was 40% failure rate good enough? How about the middle of the Sahara? More importantly, how does one know a good sample when one sees one? What are the criteria for rejection?

“The 10 base stations in use all are situated in or at the edge of oceans and/or at high altitude, where there is little variation due to vegetation and/or other local sources/sinks. After quality control, the remaining trends all are identical and all within 5 ppmv of each other.”
So the samples are chosen (by QC) that are within 5ppmv of each other? And this difference is 3 times the quantity that the CO2 has been said to rise on an annual basis? It sounds to me like the difference is far greater than 5ppmv if one allows all the data to be included. Hence my question as to why 60% is good enough.

The difference is mainly from near surface to altitude and between the NH and SH. Maybe the highest parts of the Alps (Jungfraujoch) or one of the Rockies can be better, but the question is if it is necessary, because the existing stations do work good enough.”

Again, “do work good enough” on whose authority? Who decided that this was good enough?

“Since the Mauna Loa measurements started, CO2 levels increased with 20%, which is a little more than “P” in a lake.”
But who determines that the 20% is from humans?

“If all emissions should have remained in the atmosphere, the increase would have been 35%. Maybe there are other causes of the increase, but where have the emissions gone if not partly into the atmosphere, the rest into oceans and vegetation?”
Why could not all of the emissions have gone into the oceans and the vegetation?

“Again, admitting that we are the cause of the…” it should be a matter of proof, not admission. I have never confused the issue of accuracy of atmospheric CO2 data with human accountability for the twentieth century’s modest warming trend. It is obvious that the latter statement can be falsified without the data accuracy being questioned. This is not an excuse to ignore the data accuracy.
I am grateful that you took the time to reply to my questions.

There are no ideal places to measure “global” CO2, but there are several places which may be deemed good. The best place in fact is the South Pole, far from vegetation and far above and far away from the only volcano in the (wide) neighbourhood. If it only was not so cold there… You may imagine the technical/personal difficulities one encounters in the Austral winter to take samples there.

So every place has its own specific problems. Barrow has low readings in summer when the wind is from land side (tundra), La Jolla also if the wind is from land side. Mauna Loa sometimes from fumaroles but more frequently in the afternoon from upwind conditions. Despite that, from the 8600 measurements per year (Mauna Loa), some 60% are good enough to give a profile of what happens with CO2 in the atmosphere.

The 10 base stations in use all are situated in or at the edge of oceans and/or at high altitude, where there is little variation due to vegetation and/or other local sources/sinks. After quality control, the remaining trends all are identical and all within 5 ppmv of each other. The difference is mainly from near surface to altitude and between the NH and SH. Maybe the highest parts of the Alps (Jungfraujoch) or one of the Rockies can be better, but the question is if it is necessary, because the existing stations do work good enough.

Since the Mauna Loa measurements started, CO2 levels increased with 20%, which is a little more than “P” in a lake. If all emissions should have remained in the atmosphere, the increase would have been 35%. Maybe there are other causes of the increase, but where have the emissions gone if not partly into the atmosphere, the rest into oceans and vegetation?

Again, admitting that we are the cause of the increase has nothing to do with admitting that CO2 has a huge influence on temperature, that are totally separated items.

Thanks, Ferdinand.

Flagging outliers is a part of quality control, as we are interested in more or less global averages and trends, not in local volcanic outgassing, local sugarcane field CO2 use or data from failing equipment. Local CO2 in vegetation (or volcanoes) may be of interest for biologists and scientists who want to study the carbon cycle in more detail, but there are better places to do that: midst the fields and/or forests, where some 400+ stations over the world are at work, or at the mouth of fumaroles…

Are there better places to measure atmospheric CO2 than the top of a large volcano? maybe the top of an “Alp”? Or a” rocky”? That’s like choosing to measure pollution near a motorway and claiming this measure is the norm for the rest of the countryside simply because the wind was blowing in the right direction for a safe measurement. Whilst we all understand the need for meaningful statistics and pristine curves question the validity of the data before it even reaches the QC stage. My guess is that measuring from ML is an exercise in proving that CO2is well mixed when comparisons are made with other stations in other regions.

To paraphrase Crosspatch, and quote Piers Corbin,
“if you “P” in a lake, does the level go up? The accurate answer is “yes” but can we measure it?“
Let alone the question of whether it is responsible for modern day warming.

I had a lot of discussion with Beck about his historical data. Besides the accuracy, the main problem is that most measurements were taken at places with huge local sources/sinks. And there is not one series which continues (or repeated) measurements at the same place over a longer period. Thus his “trend” is the result of a mix of places with completely different local circumstances at different times.

Some better indication of pre-Mauna Loa CO2 data is from ice cores and firn. Law Dome has a quite high resolution (5 year averages) and even more interesting, there is an overlap of about 20 years between the ice core gas age and atmospheric CO2 data from the South Pole. This shows that the South Pole CO2 data and the ice core are in line with each other within the accuracy of the ice core measurements (+/- 1.2 ppmv). See:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/law_dome_sp_co2.jpg
http://www.ferdinand-engelbeen.be/klimaat/klim_img/law_dome_co2.jpg

Two out of three Law Dome ice cores show a 1 ppmv peak around 1938 (sorry, remembered that it was around 1942). This may give a false impression of a dip around 1945, but it is within the error margin of the measurements…

Even the graph of Mauna Loa trends shows that the rate of increase in the 1960′s was smaller than in current times. It is not a straight line! If you go to the data themselves (see http://www.esrl.noaa.gov/gmd/ccgg/trends/index.html#mlo ), you will see that the yearly increase in the 2000′s is about a threefold of the increase in the 1960′s. The same for the emissions: there is about a threefold increase of emissions between the 1960′s and the 2000′s.

In fact the increase of CO2 in the atmosphere follows the accumulated emissions in such a lockstep (56% of the emissions remain – as mass – in the atmosphere), that it is near impossible that this is due to any natural process. A natural process that follows the emissions with such an accuracy simply doesn’t exist. And temperature has a much worse correlation with increasing CO2 levels.

For the graphs see:
http://www.ferdinand-engelbeen.be/klimaat/klim_img/temp_co2_acc_1900_2004.jpg
http://www.ferdinand-engelbeen.be/klimaat/klim_img/acc_co2_1900_2004.jpg

Maybe use CO2, that would be fighting CO2 with CO2. It is heavier than air. It might effectively sequester itself too and also would prevent future fires unless it leaked away. A lot of weasel words, to be sure.

a very powerful point you make. Thanks for patiently explaining it. Not everyone has steel trap minds.

Yet the “global temperature” has fluctuated over periods longer than a year, the 60s -70s stands out. Would this not affect the level of “well mixed” CO2 enough to register on this graph? Or is it just a coincidence that fluctuating CO2 levels affected by varying temperature happen to match changes in emissions.