Partitioning of net CO exchanges at the city-atmosphere interface into biotic and abiotic components.

Eddy covariance (EC) method has been used to measure CO fluxes over various ecosystems. Recently, the EC method has been also deployed in urban areas to measure CO fluxes. Urban carbon cycle is complex because of the additional anthropogenic processes unlike natural ecosystems but the EC method only measures the net sum of all CO sources and sink. This limitation of the EC method hinders us from the underlying processes of the carbon cycle, and it is necessary to partition the net CO fluxes into individual contributions for a better understanding of the urban carbon cycle. Here we propose a statistical method to partition CO fluxes into individual components by extending the method of Menzer and McFadden (2017).•Statistical method is proposed to partition CO fluxes into gross primary production, ecosystem respiration, anthropogenic emissions from a vehicle and building.•This method uses eddy fluxes and footprint-weighted high-resolution land cover data with temporal subsets that a few components can be negligible.•New partitioning method produces reliable individual components of the urban carbon cycle when compared to inventory data and typical biotic responses to environmental conditions.