Modelling how the relative spatial pattern of green within cities impact carbon flows

Cities - particularly the denser ones - are often depicted as sustainable systems that reduce the consumption of land and optimize energy use for buildings and transport. Cities, however - exactly because they concentrate human activities - constitute the main source of CO2 emissions into the atmosphere. Yet it is important to assess to what extent they can cope with their own emissions. We address this issue by examining how the internal spatial organization of cities can impact the flow of anthropogenic CO2 between their major sources - human activities, mainly on built-up lands - ; and their main storage infrastructures, with a focus here on urban green spaces and forests. Is it better to have a dense core with a peripheric green belt? Large green patches within the core centre? Or small and fragmented green spaces?

The objective of the present work is to tests whether the internal spatial organization of cities - in terms of green infrastructure characteristics and land use types - matters for evaluating carbon sequestration potentials of cities. Or whether these cities can simply be considered as single objects with a quantity of carbon emissions and a carbon sink capacity derived directly from aggregated land use data.

We model urban carbon flows for about 800 Functional Urban Areas (FUAs) of European cities using the Urban Atlas 2012 database. The model intends to characterize the emissions, their diffusion and sequestration within the urban regions as well as their exit from the analysed urban systems. The FUAs are classified along landscape metrics in order to derive a typology of urban forms focusing on the relative position of green and human activities and their characteristics. The analysis then investigates how the type of urban form can affect the level of efficiency of CO2 uptake. The efficiency is defined as the share of local emissions captured within the urban boundaries. The model will be validated using eddy covariance empirical data.