GUEST POST BY FRANCIS MASSEN
Biography of Francis Massen here.
Atmospheric CO2 mixing ratios vary with latitude, regional specificities (as sea-side, continental or urban location) and time. Daily values may differ by more than 40%, as can be seen in the first figure showing CO2 and wind speed time series at the meteorological station “meteoLCD” in Diekirch, Luxembourg (semi-rural environment).
Figure 1: CO2 and wind speed values from 14 to 21 March 2010 in Diekirch, Luxembourg.
The causes of these hefty variations are many: magnitude of wind speed, periods of boundary layer inversions, changing plant behavior (photosynthetic CO2 absorption or CO2 emissions by plant respiration) and human activity. When the boundary layer is well mixed up, CO2 levels tend to a reproducible minimum, which represents the regional background level and may even be close to the published global mean CO2 mixing ratio. The mixing up of the near ground layer is essentially caused by the wind: a plot of CO2 versus wind speed often has a typical boomerang shape, as shown in the second figure.
Figure 2: Plot of CO2 versus wind-speed using the values of Fig.1.
The background level can be thought as being the CO2 mixing ratio that would exist if wind speed was infinite. Simple visual inspection, or better, fitting the data to an exponential function of the type CO2 = a + b*exp(-c*windspeed) delivers this asymptotic CO2 level. In the example above, the model is statistically significant (R2 = 0.64) and suggests a regional background of 390 ppm (to be compared for instance to the 2009 seasonal corrected mean Mauna Loa level of approx. 388 ppm).
The CO2 versus wind speed plot can also be used as a first step to validate historic CO2 measurements, made by chemical methods. One example is the very careful measurements done from 1939 to 1940 by W. Kreutz in the town of Giessen, Germany. Kreutz used a chemical gas analyzer having an accuracy better than 1.5% and also recorded the wind speed. The exponential fit points to a very high asymptotic level of 398 ppm, well in excess to the consensus value of 310 ppm derived from the ice-cores.
This CO2 background problem is studied in a peer reviewed paper I presented with E. Beck as coauthor (Beck is a specialist of historical CO2 measurements) at the online conference Klima2009 organized in Nov. 2009 by the University of Applied Sciences of Hamburg, Germany. Our paper “Accurate estimation of CO2background level from near-ground measurements at non-mixed environments” was rewarded “Best Paper” among the 103 contributions. A slightly edited version will be published in an upcoming book “Social, Economic and Political Aspects of Climate Change” (editor: Prof. Walter Leal, Springer Verlag).
The text of the original Klima2009 paper can be found here