Tall tower eddy covariance measurements of CO2 fluxes in Vienna, Austria

Abstract

This study reports on three years (2018–2020) of turbulent CO2 fluxes measured at 144 m above the city of Vienna, Austria using an eddy covariance system installed on the A1 Arsenal radio tower. Overall CO2 flux data availability after quality control filtering (49%) was comparable to that observed by other studies of urban and suburban eddy covariance measurements. Average monthly CO2 fluxes followed seasonal trends in temperature-based proxies of heating demand, with the highest fluxes observed in the winter months. When analysing separate trends for the two dominating wind directions, this seasonality was most pronounced for the flows from the north-west when flux footprints overlapped with Vienna's most populated and urbanised districts and fluxes. For flows from the south-east, when flux footprints covered areas of built-up and industrial land yet were dominated by green urban spaces and arable land, the seasonality was much less marked and average monthly CO2 fluxes were generally lower, particularly in winter. Weekday-weekend patterns as well as weekday diurnal variations in CO2 generally followed respective trends in local traffic counts; however, flux correlations with hourly traffic counts on weekends and public holidays were much weaker. Weekend/public holiday diurnal trends in CO2 fluxes indicated an influence of vegetation fluxes on net CO2 flux variations between midday and late afternoon in spring and summer. Furthermore, variations in weekend/public holiday CO2 fluxes from night-time to morning periods suggested a partial night-time decoupling between turbulent exchange at 144 m and the net fluxes at the surface. While the influence of CO2 advection could not be quantified, the pronounced morning flux peaks indicated that CO2 released to the atmosphere during these partially decoupled, night-time periods accumulated to some extent and was subsequently mixed vertically with the morning increase in boundary layer depth. Indeed, the calculated annual fluxes suggested that positive and negative 30-min biases were largely offset when averaging over longer time intervals. The 2018 flux of 10.89 kt CO2 km−2 was in close agreement with the annual CO2 emissions estimate of 10.19 kt CO2 km−2 derived from an official emissions inventory (excluding large point source emissions). A similar annual flux was observed in 2019; however, the substantially lower flux in 2020 likely reflected emissions reductions associated with the COVID-19 pandemic. Comparison of normalised daytime CO2 fluxes between years indicated that the initial lockdown led to a ca. 50–60% reduction in net CO2 emissions from Vienna's most populated and urbanised districts. This study thus provides further evidence that defensible estimates of urban CO2 fluxes can be achieved with tall tower eddy covariance.