RESUMEN
Aircraft measurements were used to estimate the CO2 emission rates of the city of Rome, assessed against high-resolution inventorial data. Three experimental flights were made, composed of vertical soundings to measure Planetary Boundary Layer (PBL) properties, and circular horizontal transects at various altitudes around the city area. City level emissions and associated uncertainties were computed by means of mass budgeting techniques, obtaining a positive net CO2 flux of 14.7 ± 4.5, 2.5 ± 1.2, and 10.3 ± 1.2 µmol m(-2) s(-1) for the three flights. Inventorial CO2 fluxes at the time of flights were computed by means of spatial and temporal disaggregation of the gross emission inventory, at 10.9 ± 2.5, 9.6 ± 1.3, and 17.4 ± 9.6 µmol m(-2) s(-1). The largest differences between the two dataset are associated with a greater variability of wind speed and direction in the boundary layer during measurements. Uncertainty partitioned into components related to horizontal boundary flows and top surface flow, revealed that the latter dominates total uncertainty in the presence of a wide variability of CO2 concentration in the free troposphere (up to 7 ppm), while it is a minor term with uniform tropospheric concentrations in the study area (within 2 ppm). Overall, we demonstrate how small aircraft may provide city level emission measurements that may integrate and validate emission inventories. Optimal atmospheric conditions and measurement strategies for the deployment of aircraft experimental flights are finally discussed.