Eddy covariance measurements of CO2 isotopologues with a quantum cascade laser absorption spectrometer

Stable isotope isotope measurements of the biosphere-atmosphere CO2 exchange can be used to constrain carbon budget estimates on different spatial and temporal scales. Eddy covariance (EC) measurements can provide a direct measure of the isotope flux ratio of CO2 and thus indicate the ecosystem fingerprint on the isotope budget of the atmosphere. For the first time we report on EC flux measurements of stable CO2 isotopologues over a forest canopy. We used a quantum cascade laser absorption spectrometer (QCLAS, Aerodyne Research Inc.) for the simultaneous measurement of (CO2)-C-12, (CO2)-C-13 and (COO)-O-18 isotopologue fluxes with a sampling rate of 10Hz over the Lageren mixed forest site in Switzerland during an intensive field campaign in 2008. The total CO2 flux measured with the closed-path QCLAS matches well the half-hourly fluxes measured with a conventional open-path infrared gas analyzer (R-2=0.988). The O-18 isotope discrimination is strongly reduced after precipitation events, reflecting O-18 isotope exchange of CO2 with water in various water pools of the ecosystem. The (CO2)-C-13/(CO2)-C-12 flux ratio is surprisingly constant on a diurnal timescale, but the substantial random noise in the half-hourly EC flux ratio puts a limit on resolving diurnal variations of only a few permil. Hence, we explore which component in our EC set-up is limiting the precision of the measured flux ratios. On the one hand we use data simulations to estimate the error contribution from the QCLAS instrument. On the other hand different EC data processing procedures are examined to quantify the uncertainty inherently associated with turbulent flux measurements and calculations. We show that the random error associated with EC data processing is of the same order of magnitude as the flux ratio uncertainty from the instrument precision. Therefore, a direct determination of ecosystem discrimination on hourly or diurnal time scales with the EC technique remains a challenge. (C) 2011 Elsevier B.V. All rights reserved.