The study of the terrestrial carbon cycle is currently data limited. One approach to increasing the density of data over the continents is to instrument eddy-covariance flux towers with well-calibrated CO2 mixing ratio measurements. More than two hundred such towers are currently being operated at continental sites around the globe. Most of these towers, however, while measuring CO2 mixing ratios at high frequency, continuously, and with good relative precision, do not have carefully calibrated long-term mixing ratio measurements. Similarly it has been thought that mixing ratio measurements in the atmospheric surface layer, the lowest portion of the atmospheric boundary layer, would be too close to strong sources and sinks to be useful for studying the carbon cycle via atmospheric budget or inverse studies. Methods exist, however, for both precise calibration of flux tower mixing ratio measurements and careful interpretation of surface layer data.
A relatively low-cost, high-precision CO2 mixing ratio measurement system has been developed in collaboration with NCAR-ATD, to support inverse analyses of the terrestrial carbon balance at regional to continental scales. The systems use LICOR-820 non-dispersive infrared analyzers rather than the more expensive, but lower noise LICOR-6262 or LICOR-7000, since fast time response is not required for this application and the noise of the LICOR-820 can be reduced sufficiently by averaging over two minutes. Two nafion driers are used, ensuring that the difference in water vapor concentration between the dried sample and the moistened calibration gases is less than 300 ppm (corresponding to an error in the [CO2] measurement of 0.1 ppm). Flow control, such that the flow rate changes by less than 4 cc/min between the sample air and calibration gases, is achieved using a mini-regulator. Leak tests are automated and the systems are temperature controlled.