hide
Free keywords:
-
Abstract:
We aimed to assess the feasibility of an affordable instrumentation, based on a non-dispersive infrared analyser, to obtain atmospheric CO2 mole fraction data for background CO2 measurements from a flux tower site in southern Finland. The measurement period was
November 2006–December 2011. We describe the instrumentation, calibration, measurements
and data processing and a comparison between two analysers, inter-comparisons
with a flask sampling system and with reference gas cylinders and a comparison with an
independent inversion model. The obtained accuracy was better than 0.5 ppm. The intercomparisons
showed discrepancies ranging from –0.3 ppm to 0.06 ppm between the measured
and reference data. The comparison between the analyzers showed a 0.1 ± 0.4 ppm
difference. The trend and phase of the measured and simulated data agreed generally well
and the bias of the simulation was 0.2 ± 3.3 ppm. The study highlighted the importance of quantifying all sources of measurement uncertainty.