The greenhouse effect is a warming of the earth caused by an increased accumulation in the atmosphere of trace gases which are capable of absorption in that part of the infra red spectrum which is normally transparent. Although the most important of these gases is water vapour, its effect is usually regarded as being primarily reliant on the accumulation of carbon dioxide. Water vapour is treated as a "feedback" to carbon dioxide concentration, and the other greenhouse gases, which include methane, nitrous oxide, sulphur oxides, CFCs and ozone have their effects converted to "carbon dioxide equivalent".
It is thus of major importance that we have reliable measurements of atmospheric carbon dioxide concentration, not only for recent years, but for the whole period that humans have been emitting carbon dioxide through the combustion of fossil fuels.
Since 1958 accurate measurements of atmospheric carbon dioxide concentrations have been carried out by Charles D Keeling, first at the Scripps Institute of Oceanography, La Jolla, California, then from 1974 at the Mauna Loa volcano, Hawaii, and subsequently by other scientists at many other places. Most of these sites are in remote places, so the measurements represent the well-mixed background. There is still no proper record of the carbon dioxide concentrations over industrial regions, or over forests and other cultivated regions.
As I have pointed out frequently (Gray 1998) background carbon dioxide as measured at remote sites has been increasing in the atmosphere at an almost linear rate of about 1.4ppmv per year ever since 1972. The rate seems to be unaffected by the large increase in emissions form combustion of fossil fuels over the period (4.4Gt in 1972 to 6.4Gt in 1995, an increase of 45%).
The significance of this behaviour is difficult to judge without a knowledge of atmospheric carbon dioxide concentrations from before the industrial era to the beginnings of the atmospheric measurements.
Measurements from this period have been made on air trapped in ice cores, drilled mainly from the Antarctic and Arctic, which have extended back as far as 160,000 years with the "Vostok" ice core. ( 78° 28’S, 106° 48’E; Barnola et al. 1991). Assumed atmospheric carbon dioxide concentrations between 1734 and 1958 have depended heavily on measurements made by Neftel et al. (1985) on the Siple ice core from the Antarctic (75° 55’S. 83° 55’W). These measurements indicated a constant concentration of carbon dioxide in the atmosphere of about 280ppmv before 1750, and a steady increase in an approximately exponential fashion since then, to join the modern measurements.
These ice core measurements have attracted much criticism, notably from Jaworoski (1996,1997).
It is difficult to believe that a chemically active gas such as carbon dioxide can remain unaffected by burial in ice for as long as 160,000 years. There are questions whether the carbon dioxide can diffuse through the snow, react with dust particles, or form clathrate compounds under pressure The technique of drilling, removal and preservation of the ice core is very difficult, and there are many things that can go wrong. The ice core must be protected from drilling mud and from entrance of air from the present atmosphere through cracks.. Fresh snow does not trap air until it has been buried for many years. There are accusations that this "adjustment" of age, estimated from diffusion measurements, has been deliberately chosen so that it will fit into the modern record.
Most of these objections have now been overcome, with a recent study on the Law Dome ice cores by Etheridge et al (1996, 1998), which involved collaboration between several of the previous teams. The three ice cores from East Antarctica (66°S, 112°E), were from a region with an unusually heavy snow accumulation, and many improved techniques from experience of previous studies were applied. The necessary adjustment for entrapped air was determined experimentally. The dating of the layers used three different seasonal parameters , the oxygen isotope ratio, electroconductivity, and hydrogen peroxide concentration. They were checked from the known dates of several volcanic eruptions.
The results from the years 1006 to 1978 are shown in the figure below.
<
http://www.john-daly.com/law-dome.gif>
They differ in several respects from the Siple ice core results:
There is no "pre-industrial" equilibrium carbon dioxide concentration. There was variability in the order of 10ppmv between 1006 and 1800. It is therefore not possible to calculate the forcing due to presumed combustion of fossil fuels by subtracting a presumed past equilibrium value from the current value.
There was a marked fall of about 9ppmv in carbon dioxide concentration during the period of the "Little Ice Age" (about 1550 to 1850) with a period of very slow growth from 1800 to 1850. The authors are of the opinion that this fall in carbon dioxide concentration was caused by the fall in temperature, rather than the other way about. The carbon dioxide concentration in the period after 1850, extending into this century would be influenced by recovery from the Little Ice Age..
Between 1935 and 1945 the atmospheric carbon dioxide concentration was constant, or even declined slightly. The reason for this is unknown.
The measurements fit well with the modern results from the South Pole, but the whole record, particularly the recent modern one, does not follow an exponential curve, fossil fuel emissions, or comply with any of the current models, (Enting and Lassey 1993) which were used as a basis for future projections by the IPCC (Houghton et al. 1996)
The authors admit that some effects may have been regional, or at least mainly in the Southern Hemisphere.
Discussion
It is back to the drawing board for carbon cycle models. Atmospheric carbon dioxide concentration varies in a manner which has not been predicted successfully by existing models. There was significant variability before there could have been a human contribution. This variability appears to have followed temperature changes, rather than being responsible for them .Although there has been an increase during the period of industrial development, the increase has not been uniform. Thus, the period between 1935-45 showed no change. The period since 1972, when the increase has been linear despite an increase of over 45% in emissions, suggests that there are new carbon sinks being established in the ocean and in the terrestrial biosphere to absorb the increases. This behaviour plays havoc with previous predictions of global warming, but it is difficult to know how long the present apparently stable rate of increase will continue.
