Evidence of variability of atmospheric CO2 concentration during the 20th century
Geo-Ecological Seminar University of Bayreuth, 17th July 2008 (see here)
Ernst-Georg Beck, Dipl. Biol
Summary of the presentation (printable PDF available here)
In 1958 the modern NDIR spectroscopic method was introduced to measure CO2 concentrations in the atmosphere [Beck 2007]. In the preceding period, these measurements were taken with the old wet chemical method. From this period, starting from 1857, more than 90,000 reliable CO2 measurements are available, with an accuracy within ± 3 %. They had been taken near ground level, sea surface and as high as the stratosphere, mostly in the northern hemisphere. Comparison of these measurements on the basis of old wet chemical methods with the new physical method (NDIR) on sea and land reveals a systematic analysis difference of about minus 10 ppm.
Wet chemical analyses indicate three atmospheric CO2 maxima in the northern hemisphere up to approx. 400 ppm over land and sea since about 1812. The measured atmospheric CO2 concentrations since 1920 –1950 prove to be strongly correlated (more than 80 %) with the arctic sea surface temperature (SST).
A detailed analysis of the Atlantic Ocean water during the arctic warming since 1918 – 1939 by Wattenberg (southern Atlantic ocean) and Buch (northern Atlantic ocean) indicates a very similar state of the Atlantic Ocean (pH, salinity, CO2 in water and air over sea etc.) These data show the characteristics of the warm ocean currents (part of global conveyor belt) at that time, indicating a strong CO2 degassing from the Atlantic Sea, especially in the area of Greenland/Iceland and Spitsbergen. More than 360 ppm had been measured over the sea surface.
In 2004 Polyakov published evidence for a multi-decadal oscillation of the ocean currents in the arctic circle, showing a warm phase (strong arctic warming during 1918 –1940 with high temperatures in the Iceland/Spitsbergen area) similar to the current situation, and a cold phase (around 1900 and 1960). Today the Iceland/Spitsbergen area is known for a strong absorption of CO2.
This multi-decadal heating of the oceanic CO2 absorption area and larger parts of the Northern Atlantic Ocean was followed by an increase of the atmospheric carbon dioxide concentration to approx. 400 ppm during the 30s and approx. 390 ppm today. The abundance of plankton (13C) and other biota supports this view.
Conclusion:
Atmospheric CO2 concentration varies with climate, the sea is the dominant CO2 store, releasing the gas depending on multi-decadal changes of temperature.
See several supporting graphs here: 180 Years of atmospheric CO2 Gas Analysis by Chemical Methods
NOTE: While this paper presents some interesting findings, I and others do have some concerns with the older chemically derived sources of CO2 concentration data, which are obtained from a chemical analysis process that has some significant variability. Beck seems to think he has accurate enough data from these methods, and he has another essay on the process of chemical analysis of CO2 gas concentration in the atmosphere here. I’d recommend reading it also. My personal opinion on this paper is that I’m “on the fence with it”, and I encourage a rigorous debate, pro and con. – Anthony
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Corrected link:Beck’s web cite links to numerous historical papers and analyses which should keep most critics happily(?!) reading for many hours.
Reply: I think I fixed it now~Charles the moderator
The above ( Joel Shore (10:41:27)) makes for interesting browsing. http://www.esrl.noaa.gov/gmd/ccgg/carbontracker/tseries.php?type=mr
Notice: “If the modeled CO2 exceeds the tolerance by more than a factor of 3, the observation is not used in our system and colored blue in this figure.”
More CO2 data sets:
Four Pacific-area sites: CMDL Hourly Average Surface Carbon Dioxide (DSI-3273) but have to ask for data? — and corresponding weather data: CMDL Hourly Surface Observations (DSI-3274) .
Various observing sites: NOAA ESRL Observation Sites >> Listing by Project. Click on the Interactive Visualization to browse graphs from many locations; data is in FTP archive. Did not check documentation to see if any data is discarded.
Monthly Atmospheric CO2 Mixing Ratios from the NOAA CMDL Carbon Cycle Cooperative Global Air Sampling Network, 1968-2002.
Anthony,
When I studied Beck’s previous work, a small point that particularly impressed me was the subtle correlation of CO2 concentrations with moon phases. It shows up in Beck’s reconstructions of antique data as well as the modern data. The ubiquity of the lunar signature helped persuade me to believe in the reality of his findings.
I’m looking forward to reading this new paper.
– Erik
I’ve also heard that the diurnal range of CO2 over heavily-vegetated areas can be well in excess of 2:1
James (09:03:41) : gave this link.
http://www.biokurs.de/treibhaus/180CO2/bayreuth/bayreuth1e.htm
I f you go to this and click on the graph, you go to more graphs. At the bottom of the last one is a clickable link, “Keeling on accuracy”
“Keeling´s statements on pre-1958 chemical CO2 measurements
* The mean value that Theodore de Saussure found was roughly 0.04% by volume, which I will put in modern units as 400 parts per million by volume (ppmv). This value was much less than von Humboldt had found, but still in considerable error.
* De Saussure’s Memories, published in 1830, nevertheless ushered in a period of increasingly precise measurements of atmospheric carbon dioxide culminating in some nearly correct measurements in the 1880s by a Belgian named Jules Reiset.
* Unfortunately, none of the other investigators of carbon dioxide reproduced the seasonal cycle even approximately correctly, casting doubt on all of their data.
* After the 1880s, interest in carbon dioxide diminished for reasons which I have not been able to establish. Indeed, a quick and easy but not very precise 19th century technique of measurement, called Pettenkofer’s method, became the most common method for measuring carbon dioxide, so that measurements actually became less precise than the best of those of the 19th century. The 20th century data were generally higher than the correct concentrations, although of course this wasn’t known at the time.
* My first measurements were made with only a brief prior investigation of the scientific literature.
* In all cases I found the concentration of CO2 in the afternoon to be close to 315 ppmv. I came to the conclusion that the published 20th century data generally were in error and that the CO2 concentration, when not locally influenced, was nearly the same at least from the equator to the northern limits of the United States, perhaps everywhere.
Speech: Winner of the Second Blue Planet Prize (1993);
http://www.af-info.or.jp/eng/honor/bppcl_e/e1993keeling.txt”
Sure of himself, isn’t he?
Is the raw data available for modern day measurements?? If 80% of those are rejected, it would be interesting to see the min/max values
“dC/dt = -a*(C-Ceq) + Em”
These very many CO2 fluences, e.g., the CO2 burned (from the biomass sink) to create CaO, the CO2 release from the carbonate (the geologic sink), the CO2 returned to produce cement (from the ground water sink and atmospheric sink), all this is captured in the above equation?
You need to work with vector math not scalar math, otherwise nothing certain you might know about this moment’s state can be used to gain insight into the next moment. The ‘sinks’ are in dynamic exchange with multiple others continuously. Silly scalar math requires tagging your C atoms.
Now, Spencer, some months back here, measured the 13C/12C variance in the Mauna Loa seasonal and long-term trend signals by means of an F-Test. The variance was the same! Therefore, whatever its limitations, the biogenic and anthropogenic fluences were not significant. What, pray tell, does that tell you?
Keeling maintains the seasonal signal is biogenic, he is wrong about this as he is wrong about the 19th century averages.
One major author on CO2 variations is: P. Bawkin
Some samples of CO2 variation with time and altitude are shown by:
A multiple-scale simulation of variations in atmospheric carbon dioxide using a coupled biosphere-atmospheric model Melville E. Nicholls, A. Scott Denning, Lara Prihodko, Pier-Luigi Vidale, Ian Baker, Kenneth Davis, and Peter Bakwin JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 109, D18117, doi:10.1029/2003JD004482, 2004
e.g., Fig 14, Fig 16, Fig. 17.
Seems Altitude is a major item to add to Beck’s data set where it can be identified in the reports.
I would like to draw your attention that the means exist to have global maps of CO2 concentrations:
http://www-airs.jpl.nasa.gov/index.cfm?fuseaction=ShowNews_DynamicContent&NewsID=10
Unfortunately this information starts and stops in July 2003, despite ” . I do not know whether they are waiting for a commercial use of the information and keeping it private, or the information contradicts orthodoxy.
In this July 2003 map one can see differences in concentration of 15ppm over the globe, a few percent .
It is a pity we do not have all the months since then.
Anthony,
When I first stumbled over Beck’s datas, I questioned – at least some of – them. But, BUT, generally, one of the the ‘old’ methods, escecially by titration according to Pettenkofer was known to me.
It was one of the methods I had to use when I had to do the practice for an Chemical Engineer in the ’70ties. And the method, a known volume of air, at a known/measured temperature, at a known/measured air presusre has not that much errors. The air was ‘washed’ in an alcalic solution of Kaliumhydroxide and/or
Natriumhydroxide (The alcalic solution did have a known amount of one or both
of these substances). It was added a coloric indicator, then a acid solution was added, drop by drop. The amount of acid solution used until the color of the indicator changed, could give good to excellent results to measure the CO², measured as H²CO³.
To get rid of errors (SO³ and NO² within air, use different alcalic solutions, different
concentrations and different alcalic substances, because the behavior of CO², NO² and SO³ differs). The results were usually in a plus/minus 1.2 error frame, when
did properly and when done by at least halve a dozen anaylysisses.
TIDBIT: Last time when I did it was about ’74, changed my profession then to EDP
, now IT. So some of my explanaitions may be not that correct in my rememberance.
So, whoever at this forum, could enhance it and/or correct it, I would sincerely appreciate it.
Personnaly, I do live approx 27 kilometers driveway to the ‘Justus von Liebig’ -University, Giessen, Germany. I took two days off, to check some of the datas with the librarythere. At least I can say, Becks datas from Pettenkofers’ measures are not ‘Jim-Hansen-ed’. Pettenkofer did extremely good bookkeeping, when he did his measures, location, wind direction, season, distance from Giessen and Wetzlar, the towns in this area at that time … and so on.
Best Regards
KlausB
Re: “After the 1880s, interest in carbon dioxide diminished for reasons which I have not been able to establish.”
Could the answer his perplexity be the 30 year cooling trend that kicked off around 1880? How early was CO2 driving temperature originally speculated?
Gary Gulrud,
Please don’t make me pass for an idiot! I explicitly said that you can use a set of coupled differential equations. But the CO2 in air IS a scalar, at least in a very good approximation because the time it takes to mix the atmospheric concentration is quite small (less than a year), so the differences are small. To be precise, let’s just say that C is the average CO2 concentration in air.
There are many so-called “box models” for the carbon cycle. I worked on this extremely simple model (where all the sinks are condensed into a single one) after considering those complex models, and attempting to build one myself. The problem with those complex models, just like with GCMs by the way, is that you soon get bogged down in the complexity, and you don’t have the real world data to support your model. So you start guessing and “parameterizing”, and soon you’re completely lost in your little virtual world.
So after going through that phase, I decided to come back to simplicity. Because all the data I have is CO2 in air, CO2 emissions, and temperatures. Nevertheless, what I found is that I can, with this very simple model, get a very good match with those three data sets. So in a way, that means that despite the complexity of the carbon cycle, it’s average behavior is quite simple.
For what it’s worth, while doing this work, I read about 100 scientific papers on the carbon cycle. So please don’t assume that I’m just making this up.
I think there is some merit in what Beck has done. But faced with different, conflicting data sets, you always have to make a choice. Those old measurements were either right or wrong. I think if you make a model that is consistent with the data, especially with independent data, it improves your confidence in both data and model. If he comes up with a physical and plausible model for the carbon cycle that reconciles his data with other variables, like temperature, for example, good for him!
This whole thread seems to cry out for tele-connecting these CO2 monitoring sites so that NASA could determine the “correct” value for all reports.
klausB
What would be interesting is to reproduce those measurements today: same method, same location, etc., and see the result.
addendum to my July 25th, 1312:40
When I looked for ‘Pettenkofer’ & ‘titration’ & ‘CO²’, on WIKIPEDIA,
about two years ago, there was still a link to his method. If I remember correctly, of
course. Now it’s gone.
“Ein Schelm, der Arges dabei denkt” – “Honi soit qui mal y pense”
According to the most picturesque legend, it was established to commemorate an incident in which Edward was dancing with Joan of Kent, Countess of Salisbury, when one of her blue garters dropped to the floor. As bystanders snickered, Edward gallantly picked up the garter and put it on his own leg, admonishing the courtiers in French with the phrase that remains as the order’s motto “Honi soit qui mal y pense” (“Shame to him who thinks evil of it,” or, more popularly, “Evil to him who evil thinks”). Britannica 2001
Regards
Klausb
Interesting information on pre-industrial CO2 levels from Prof. Zbigniew Jaworowski
Chairman, Scientific Council of Central Laboratory for Radiological Protection
Warsaw, Poland http://www.john-daly.com/zjiceco2.htm
Excerpts from Statement written for the US Senate Committee on Commerce, Science, and Transportation
March 2004
1)“…glaciological studies are not able to provide a reliable reconstruction of CO2 concentrations in the ancient atmosphere”
2) “The data from shallow ice cores, such as those from Siple, Antarctica[5, 6], are widely used as a proof of man-made increase of CO2 content in the global atmosphere, notably by IPCC[7]. These data show a clear inverse correlation between the decreasing CO2 concentrations, and the load-pressure increasing with depth (Figure 1 A) . The problem with Siple data (and with other shallow cores) is that the CO2 concentration found in pre-industrial ice from a depth of 68 meters (i.e. above the depth of clathrate formation) was “too high”. This ice was deposited in 1890 AD, and the CO2 concentration was 328 ppmv, not about 290 ppmv, as needed by man-made warming hypothesis. The CO2 atmospheric concentration of about 328 ppmv was measured at Mauna Loa, Hawaii as later as in 1973[8], i.e. 83 years after the ice was deposited at Siple.
An ad hoc assumption, not supported by any factual evidence[3, 9], solved the problem: the average age of air was arbitrary decreed to be exactly 83 years younger than the ice in which it was trapped. The “corrected” ice data were then smoothly aligned with the Mauna Loa record (Figure 1 B) , and reproduced in countless publications as a famous “Siple curve”. Only thirteen years later, in 1993, glaciologists attempted to prove experimentally the “age assumption”[10], but they failed”
3) “ The modelers ignored the evidence from direct measurements of CO2 in atmospheric air indicating that in 19th century its average concentration was 335 ppmv[11] (Figure 2) . In Figure 2 encircled values show a biased selection of data used to demonstrate that in 19th century atmosphere the CO2 level was 292 ppmv[12]. A study of stomatal frequency in fossil leaves from Holocene lake deposits in Denmark, showing that 9400 years ago CO2 atmospheric level was 333 ppmv, and 9600 years ago 348 ppmv, falsify the concept of stabilized and low CO2 air concentration until the advent of industrial revolution [13]”.
Don’t satellites pointed at the planet have spectrometers?
Could they not be used to measure CO2 concentrations?
“But the CO2 in air IS a scalar, at least in a very good approximation because the time it takes to mix the atmospheric concentration is quite small (less than a year), ”
The peak in 14C production, at high latitude in high stratosphere leads surface concentration peak by 60 years, whereas tropospheric residence estimates for CO2 reported by Seagalstad range from 5 to 12 years. How is the excellence of mixing determined?
I did not set out to make of you and idiot (although I do do that sort of thing) I see no conceivable hope for simplicity sans accuracy and your reasoning is not giving me a warm fuzzy.
So what do you think of Spencer’s F-Test result?
Francois Ouellette
“So after going through that phase, I decided to come back to simplicity. Because all the data I have is CO2 in air, CO2 emissions, and temperatures.”
That gives you correlation but doesn’t get you climate sensitivity. I think this is what Spencer is getting at. Localized variations in CO2, air temperature, ocean temperature, cloud cover, convection, evaporation, slope of trends point to differing sensitivities.
The simplification and averaging and smoothing of CO2, emissions, and global temperature mask the true sensitivity. For example, positive PDO periods increase the rate of temperature trend, negative phases decrease it, and the models don’t take that into consideration–all they see is the over-simplified correlation of more CO2 and increasing temps.
At the risk of being pounded over the head with a twobyfour, I’m seeing the sun as the basic driver–no matter the fact TSI is less variable than thought. CO2 is an amplifier not the driver and the oceans control the distribution and amount of heat. How much CO2 amplifies is the big question. It may not be very much considering quiet sun and cool oceans with rising CO2 currently results in cooler global temps.
So, basically, simplifying temp, CO2, and emissions is fine for some purposes, but nothing is going to beat getting down and dirty with the details to understand the processes occurring on a more regional than global basis.
We try to find an average temperature of the earth by making as many measurements in as many places, with as standardized procedure and equipment, as possible. I think this is reasonable, ignoring the excessive attempts to ADJUST the data.
Why then is CO2 measured at only a few SPECIAL locations with procedures to EXCLUDE data?? Isn’t this the ultimate cherry picking??
Can’t you just see Hansen trying to tell us that only a few locations that he picked would be used to measure the BACKGROUND temperature and he would only use readings within certain parameters using his personally preferred method??
I think Hansen is a LOON, BUT, his GISS is perfect compared to the CO2 hoax!!
Francois Ouellette, very interesting and I find myself in agreement with most of what you say (I don’t disagree with the rest. I just don’t know).
Perhaps you could help answer a question. It seems plants with adequate temperature, water and nutrients are limited in their growth by CO2 concentrations.
Trees represent the bulk of the Earth’s terrrestial biomass. Trees grow relatively slowly and generally large trees create more additional biomass than small trees (such as more leaves).
We know that there has been a significant increase (6%) in plant biomass over recent years. It appears plants take up more CO2 at higher concentrations and there will be a lag due to the tree growth effect. My guess would be between 20 and 50 years before maximum bio uptake is reached.
My question is, Has anyone quantified this and what do the models assume?
The reason I ask is that it appears possible that bio activity alone will limit CO2 concentrations, irrespective of human emissions.
Syl,
take a look at this data file from SOURCE.
http://lasp.colorado.edu/sorce/tsi_data/daily/SORCE_L3_TSI_DM_V0008_20030225_20080717.txt
It has 2 columns of intensity. One marked 1 AU and one marked Earth Distance. Notice how much variance there is in the Earth Distance data compared to the standardized 1 AU data. This is what is actually hitting the upper atmosphere and varies at much higher levels due to the distance change. Right now the southern hemisphere summer gets the highest radiation and northern hemisphere summer gets the lowest. The southern hemisphere has more water so absorbs more of the energy it receives. Due to differences in surface and water absorption, clouds… this energy variance does NOT average out over the year.
Francois & Gary
Yes the CO2 concentrations are scalar AND the CO2 fluxes are vectors (or tensor fields). Both trends and flows need to be examined in evaluating the reasonableness of the data Beck has compiled.
klausB
Thanks for noting your experience in that CO2 titration and observation “Pettenkofer did extremely good bookkeeping, when he did his measures, location, wind direction, season, distance from Giessen and Wetzlar, the towns in this area at that time … and so on.”
That is a good reminder of the Germanic trait of systematic thoroughness and penchant for accuracy. There appears no basis for facially dismissing that quantitative accumulated data. That data needs to be accounted for and the differences with the ice core data and modern spectroscopic data need to be clearly understood.
All:
Look at the embedded information on the first graph for the Kreutz samples from the Giessen weather station 1939-40.
“The Giessen data
He conducted 64 000 measurements in 4 differnet altitudes using a high precision gas analyser invented by P Schuftan, Accuracy: 1,5%. The yearly average for 1939/40 was 385 ppm. Country air with lowest CO2 was in average 372 ppm. The seasonal amplitude was 54 ppm which was roughly 20 ppm more than the data measured at the coast of the Baltic sea at that times. Most interesting is that the about monthly variations correlate with the lunar phases (peak on full moon)”
This series represents 75% of the data samples surveyed.
Klaus B: Is this data available locally? Is there a continuing sampling program locally at present.
Keith
Speaking of CO2, we know that NASA has satellites that can check CO2 over the entire earth and update the data daily, much the same was that UAH checks temperatures. Why can we NOT have access to that data? Maybe it contains inconvenient truths.