Accuracy in city CO 2 inventories can be evaluated using independent approaches based on atmospheric measurements. Here, CO 2 fluxes observed with eddy covariance (EC) at 144 m above the city of Vienna are compared against two official inventories ( Bundesländerluftschadstoff-Inventur , BLI; Emissionskataster der Stadt Wien , Emikat). The two datasets were integrated (spatial downscaling of BLI according to Emikat; temporal extrapolation of Emikat according to BLI) to produce respective maps of annual emissions, which were further augmented with bottom-up estimates of human respiration, and additionally resolved according to 30 min temporal proxies. For comparison against EC, the spatially-explicit inventories at annual and half-hour scales were intersected and weighted by flux footprints estimated with two different models (Kljun et al. (2015), K; Kormann & Meixner (2001), KM). An encouraging degree of similarity in relative inter- and sub-annual trends between the inventories and the EC fluxes was observed. In terms of emission levels, agreement between inventories and EC, expressed by the inventory-flux closure ratio (IFCR), ranged from 0.58 to 0.83. Generally, IFCRs were insensitive to temporal resolution of the comparison and the stringency of flux filtering. Instead, IFCRs varied according to the combination of footprint and inventory, with the lowest IFCRs observed for the BLI sampled with KM, and the highest IFCRs observed for the Emikat intersected with K. The variation in IFCRs underscores the need for further research and standardisation on flux footprints yet also provide meaningful indications on the accuracy of the two inventories, which may have implications for local climate policy. • Two official CO 2 inventories resolved spatially (ha) and temporally (annual and half-hour) • Spatio-temporal emissions augmented with BU estimates of human respiration • Comparison with tall tower CO 2 flux measurements via two footprint models • Range of inventory-flux agreement comparable with previous studies • Results nonetheless indicate likely bias in traffic emissions of one of the inventories
Fasano et al. (Sun,) studied this question.