Semi-continuous automated Δ47 measurements of atmospheric CO2

Stable isotope measurements of atmospheric CO₂ are a powerful tool for partitioning contributions of different CO₂ sources and sinks. In addition to the conventional tracers δ¹³C and δ¹⁸O, the “clumped isotope” tracer Δ₄₇ can improve the distinction between high- and low-temperature sources of atmospheric CO₂ in urban studies. Despite its potential, Δ₄₇ measurements of atmospheric CO₂ remain sparse, particularly from long-term observations. One reason for this may be the high effort of manually processing samples for the measurement of Δ₄₇ in atmospheric CO₂, with high precision analysis typically require several hours per sample. Here, we present an automated preparation line coupled with a dual inlet isotope ratio mass spectrometer (MAT253+). This setup enables automated extraction and purification of atmospheric CO₂ and measurement of approximately five atmospheric CO₂ samples per day with sample preparation time of about 90 minutes. Over a 10-month period, the system achieved a reproducibility of ∼ 0.005 ‰ for δ¹³C, ∼ 0.01 ‰ for δ¹⁸O, and ∼ 0.011 ‰ for Δ₄₇.

Regular measurements using this setup provided insight into the temporal change in atmospheric Δ₄₇ in the semi-urban area of Heidelberg (Germany). In addition to the technical challenges, also the scientific interpretation of atmospheric Δ₄₇ data is not straightforward, because this “clumped isotope” tracer exhibits nonlinear behavior during air-mass mixing. Consequently linear extrapolation approaches such as the traditional Keeling plots can yield biased source signature estimates. We therefore present a thorough correction procedure applicable to cases where CO₂ enhancements are too small to allow a direct nonlinear fit.