Mobile App Traffic Reveals 3 Park Types in 45 Parisian Green Spaces

Urban parks are vital for public health, but quantifying their actual functional use has remained difficult—limited to small surveys or coarse movement data. Researchers from Nokia Bell Labs, INRIA, University of Cambridge, and Orange introduced a novel method: they refined mobile base station coverage using antenna azimuths to isolate mobile traffic strictly within park boundaries vs. surrounding areas. Using a large-scale dataset of per-app mobile network traffic across 45 parks in Paris, they tested two competing hypotheses—Central-City (dense areas breed multifunctional parks) vs. Socio-Spatial (parks reflect neighborhood routines). The result: parks have distinctive traffic signatures, differing from their urban surroundings and from each other.

The analysis revealed three functional park types: Lunchbreak parks (heavy midday traffic, productivity apps), Cultural parks (arts, entertainment apps, evening peaks), and Recreational parks (sports/leisure apps, weekend spikes). Centrally located parks exhibited more diverse app usage and strong temporal variation, supporting the Central-City hypothesis. Suburban parks, conversely, showed app preferences tied to neighborhood income, with high-income areas leaning toward productivity apps and lower-income areas toward entertainment—aligning with the Socio-Spatial view. This passive, large-scale approach opens new possibilities for urban planners to understand how green spaces serve their communities and to design parks that meet diverse needs.