Statistical partitioning of a three-year time series of direct urban net CO2 flux measurements into biogenic and anthropogenic components

Eddy covariance flux measurements are increasingly used to quantify the net carbon dioxide exchange (F-C) in urban areas. F-C represents the sum of anthropogenic emissions, biogenic carbon release from plant and soil respiration, and carbon uptake by plant photosynthesis. When F-C is measured in natural ecosystems, partitioning into respiration and photosynthesis is a well-established procedure. In contrast, few studies have partitioned F-C at urban flux tower sites due to the difficulty of accounting for the temporal and spatial variability of the multiple sources and sinks. Here, we partitioned a three-year time series of flux measurements from a suburban neighborhood of Minneapolis-Saint Paul, Minnesota, USA. We segregated F-C into one subset that captured fluxes from a residential neighborhood and into another subset that covered a golf course. For both land use types we modeled anthropogenic flux components based on winter data and extrapolated them to the growing season, to estimate gross primary production (GPP) and ecosystem respiration (R-eco) at half-hourly, daily, monthly and annual scales. During the growing season, GPP had the largest magnitude (up to - 9.83 g C m(-2) d(-1)) of any component CO2 flux, biogenic or anthropogenic, and both GPP and Rem were more dynamic seasonally than anthropogenic fluxes. Owing to the balancing of Reco against GPP, and the limitations of the growing season in a cold temperate climate zone, the net biogenic flux was only 1.5%-4.5% of the anthropogenic flux in the dominant residential land use type, and between 25%-31% of the anthropogenic flux in highly managed greenspace. Still, the vegetation sink at our site was stronger than net anthropogenic emissions on 16-20 days over the residential area and on 66-91 days over the recreational area. The reported carbon flux sums and dynamics are a critical step toward developing models of urban CO2 fluxes within and across cities that differ in vegetation cover. (C) 2017 Elsevier Ltd. All rights reserved.