Abstract. Using laser absorption spectrometry for the measurement of stable isotopes of atmospheric CO2 instead of the traditional isotope ratio mass spectrometry method decreases sample preparation time significantly, and uncertainties in the measurement accuracy due to CO2 extraction and isobaric interferences are avoided. In this study we present the measurement performance of a new dual-laser instrument developed for the simultaneous measurement of the δ13C, δ18O and δ17O of atmospheric CO2 in discrete air samples, referred to as the Stable Isotopes of CO2 Absorption Spectrometer (SICAS). We compare two different calibration methods: the ratio method, based on the measured isotope ratio and a CO2 mole fraction dependency correction, and the isotopologue method, based on measured isotopologue abundances. Calibration with the ratio method and isotopologue method is based on three different assigned whole-air references calibrated on the VPDB (Vienna Pee Dee Belemnite) and the WMO 2007 (World Meteorological Organization) scale for their stable isotope compositions and their CO2 mole fractions, respectively. An additional quality control tank is included in both methods to follow long-term instrument performance. Measurements of the quality control tank show that the measurement precision and accuracy of both calibration methods is of similar quality for δ13C and δ18O measurements. During one specific measurement period the precision and accuracy of the quality control tank reach WMO compatibility requirements, being 0.01 ‰ for δ13C and 0.05 ‰ for δ18O. Uncertainty contributions of the scale uncertainties of the reference gases add another 0.03 ‰ and 0.05 ‰ to the combined uncertainty of the sample measurements. Hence, reaching WMO compatibility for sample measurements on the SICAS requires reduction of the scale uncertainty of the reference gases used for calibration. An intercomparison of flask samples over a wide range of CO2 mole fractions has been conducted with the Max Planck Institute for Biogeochemistry, resulting in a mean residual of 0.01 ‰ and −0.01 ‰ and a standard deviation of 0.05 ‰ and 0.07 ‰ for the δ13C measurements calibrated using the ratio method and the isotopologue method, respectively. The δ18O could not be compared due to depletion of the δ18O signal in our sample flasks because of storage times being too long. Finally, we evaluate the potential of our Δ17O measurements as a tracer for gross primary production by vegetation through photosynthesis. Here, a measurement precision of <0.01 ‰ would be a prerequisite for capturing seasonal variations in the Δ17O signal. Lowest standard errors for the δ17O and Δ17O of the ratio method and the isotopologue method are 0.02 ‰ and 0.02 ‰ and 0.01 ‰ and 0.02 ‰, respectively. The accuracy results show consequently results that are too enriched for both the δ17O and Δ17O measurements for both methods. This is probably due to the fact that two of our reference gases were not measured directly but were determined indirectly. The ratio method shows residuals ranging from 0.06 ‰ to 0.08 ‰ and from 0.06 ‰ to 0.1 ‰ for the δ17O and Δ17O results, respectively. The isotopologue method shows residuals ranging from 0.04 ‰ to 0.1 ‰ and from 0.05 ‰ to 0.13 ‰ for the δ17O and Δ17O results, respectively. Direct determination of the δ17O of all reference gases would improve the accuracy of the δ17O and thereby of the Δ17O measurements.
Precision and accuracy of determination in the absorbance ratio method