Assessing the space-use efficiency of French cities by coupling city volumes with mobile data traffic

Abstract

In light of climate change, resource scarcity, and population growth, it becomes increasingly important to use the existing built-up space of our cities efficiently. However, the degree to which the available three-dimensional (3D) urban space is actually being utilized by human activities has never been studied systematically at a high spatial and temporal resolution. Here, we explore the space-use efficiency across 20 major French cities, measured by the dynamic occupancy rates of their volumes. The occupancies (number of people present) are predicted by a random forest model trained on fine-grained mobile data traffic from 2019, while the urban volumes are derived from 3D city models. The results show a surprisingly common ‘donut-like’ organization of cities, where the space occupancy is low in the city center, becomes high in the immediate surroundings, and then low again in the suburbs. This hitherto hidden regularity is associated with the distribution of urban amenities and reveals a potential to increase the utilization of under-used spaces especially in the city centers.