Acetone–CO enhancement ratios in the upper troposphere based on 7 years of CARIBIC data: new insights and estimates of regional acetone fluxes

Acetone and carbon monoxide (CO) are two important trace gases controlling the oxidation capacity of the troposphere; enhancement ratios (EnRs) are useful in assessing their sources and fate between emission and sampling, especially in pollution plumes. In this study, we focus on in situ data from the upper troposphere recorded by the passenger-aircraft-based IAGOS–CARIBIC (In-service Aircraft for a Global Observing System–Civil Aircraft for the Regular Investigation of the atmosphere Based on an Instrument Container) observatory over the periods 2006–2008 and 2012–2015. This dataset is used to investigate the seasonal and spatial variation of acetone–CO EnRs. Furthermore, we utilize a box model accounting for dilution, chemical degradation and secondary production of acetone from precursors. In former studies, increasing acetone–CO EnRs in a plume were associated with secondary production of acetone. Results of our box model question this common presumption and show increases of acetone–CO EnR over time without taking secondary production of acetone into account. The temporal evolution of EnRs in the upper troposphere, especially in summer, is not negligible and impedes the interpretation of EnRs as a means for partitioning of acetone and CO sources in the boundary layer. In order to ensure that CARIBIC EnRs represent signatures of source regions with only small influences by dilution and chemistry, we limit our analysis to temporal and spatial coherent events of high-CO enhancement. We mainly focus on North America and Southeast Asia because of their different mix of pollutant sources and the good data coverage. For both regions, we find the expected seasonal variation in acetone–CO EnRs with maxima in summer, but with higher amplitude over North America. We derive mean (± standard deviation) annual acetone fluxes of (53 ± 27) 10−13 kg m−2 s−1 and (185 ± 80) 10−13 kg m−2 s−1 for North America and Southeast Asia, respectively. The derived flux for North America is consistent with the inventories, whereas Southeast Asia acetone emissions appear to be underestimated by the inventories