This paper explores the urban carbon cycle from the natural sciences perspective, identifying key knowledge gaps and priority areas for future research. The combination of large, concentrated carbon fluxes and rapid change makes cities key elements of the carbon cycle and offers the potential for them to serve as “first responders” for climate action. Estimates of urban‐scale carbon fluxes are significantly more uncertain than at larger spatial scales, in part because past studies have mostly avoided local/urban scales where the mix of anthropogenic and natural fluxes is complex and difficult to observationally isolate. To develop effective emission reduction policies, we need to understand emission sources and how they may be changing. Such improved quantification and understanding of underlying processes at the urban scale will not only provide policy‐relevant information and improve the understanding of urban dynamics and future scenarios, but will also translate into better global‐scale anthropogenic flux estimates, and advance our understanding of carbon cycle and climate feedbacks across multiple scales. Understanding the relationship between urbanization and urban carbon flows requires intellectual integration with research communities beyond the natural sciences. Cities can serve as interdisciplinary process laboratories that are sufficiently constrained in both spatial and governance scale to support truly integrated research by the natural sciences, social sciences, and engineering. A thoughtfully crafted science research agenda that is grounded in sustained, dense observations relevant to estimating urban carbon fluxes and their controlling processes and is focused on a statistically significant sample of cities will advance our understanding of the carbon cycle.

Urban areas are expanding, changing the structure and productivity of landscapes. While some urban areas have been shown to hold substantial biomass, the productivity of these systems is largely unknown. We assessed how conversion from forest to urban land uses affected both biomass structure and productivity across eastern Massachusetts. We found that urban land uses held less than half the biomass of adjacent forest expanses with a plot level mean biomass density of 33.5 ± 8.0 Mg C ha^-1. As the intensity of urban development increased, the canopy cover, stem density, and biomass decreased. Analysis of Quercus rubra tree cores showed that tree-level basal area increment nearly doubled following development, increasing from 17.1 ± 3.0 to 35.8 ± 4.7 cm² yr^-1. Scaling the observed stem densities and growth rates within developed areas suggests an aboveground biomass growth rate of 1.8 ± 0.4 Mg C ha^-1 yr^-1, a growth rate comparable to nearby, intact forests. The contrasting high growth rates and lower biomass pools within urban areas suggest a highly dynamic ecosystem with rapid turnover. As global urban extent continues to grow, cities consider climate mitigation options, and as the verification of net greenhouse gas emissions emerges as critical for policy, quantifying the role of urban vegetation in regional-to-global carbon budgets will become ever more important.

Urban areas are directly or indirectly responsible for the majority of anthropogenic CO₂ emissions. In this study, we characterize observed atmospheric CO₂ mixing ratios and estimated CO₂ fluxes at three sites across an urban-to-rural gradient in Boston, MA, USA. CO₂ is a well-mixed greenhouse gas, but we found significant differences across this gradient in how, where, and when it was exchanged. Total anthropogenic emissions were estimated from an emissions inventory and ranged from 1.5 to 37.3 mg·C·ha⁻¹·yr⁻¹ between rural Harvard Forest and urban Boston. Despite this large increase in anthropogenic emissions, the mean annual difference in atmospheric CO₂ between sites was approximately 5% (20.6 ± 0.4 ppm). The influence of vegetation was also visible across the gradient. Green-up occurred near day of year 126, 136, and 141 in Boston, Worcester and Harvard Forest, respectively, highlighting differences in growing season length. In Boston, gross primary production—estimated by scaling productivity by canopy cover—was ~75% lower than at Harvard Forest, yet still constituted a significant local flux of 3.8 mg·C·ha⁻¹·yr⁻¹. In order to reduce greenhouse gas emissions, we must improve our understanding of the space-time variations and underlying drivers of urban carbon fluxes.