Scaling traffic variables from sensors sample to the entire city at high spatiotemporal resolution with machine learning: applications to the Paris megacity

This is a Preprint and has not been peer reviewed. This is version 1 of this Preprint.

Add a Comment

You must log in to post a comment.


There are no comments or no comments have been made public for this article.


Download Preprint


Xavier Bonnemaizon , Philippe Ciais, Chuanlong Zhou, Simon Ben-Arous, Steven J Davis, Nicolas Megel


Road transportation accounts for up to 35% of carbon dioxide and 49% of nitrogen oxides emissions in the Paris region. In an effort to mitigate those emissions, local authorities have implemented a climate plan that includes measures such as converting car lanes into cycle paths and enforcing strict low emission zones. Moreover, the COVID-19 crisis has had a notable impact on citizens' behavior and traffic patterns. There was a sharp decline and subsequent recovery in traffic during spring 2020, followed by an extended period of restrictions in late 2020 and 2021, leading to decreased traffic levels. However, estimates of city traffic patterns are often incomplete and of coarse spatio-temporal resolution, even where extensive networks of sensors exist. Here, we use a machine learning approach to assess data from 2086 magnetic road sensors in the city of Paris, resulting in a comprehensive dataset of hourly traffic flow and road occupancy covering 6846 road segments from 2018 to 2022. Our model captures flow and occupancy on an hourly and road segment basis with a Symmetric Mean Absolute Percentage Error of 37% and 54% respectively, allowing us to create a new map of hourly transportation patterns in Paris.



Transportation Engineering


Road Transportation, Traffic monitoring, carbon dioxide emissions, Pollutants, COVID-19, bottom-up approach, machine learning, open data, Paris, Urban areas


Published: 2024-04-10 17:59

Last Updated: 2024-04-11 00:59


CC-By Attribution-NonCommercial-NoDerivatives 4.0 International