Agent-Based Modeling of Paris Low-Emission Zone Impacts: Scale Effects and Behavioral Adaptations

Low-emission zones (LEZs) have become a key policy instrument in many European cities to address urban air quality challenges. Their effectiveness depends largely on positive behavioral adjustments among users of high-emitting vehicles, such as vehicle replacement or shifts toward public transport. However, some users may respond with environmentally unfavorable behaviors, including rerouting (and thus lengthening their trips) to avoid entering the LEZ. This study presents a comprehensive agent-based modeling assessment of the metropolitan Paris LEZ (ZFE-m) policy using MATSim. It incorporates three major forms of behavioral adaptation: vehicle replacement, modal shift toward car alternatives, and route modification. The simulation indicates that 27% of affected agents—those using vehicles banned within the LEZ—shift to alternative modes: walking (5%), cycling (7%), or public transport (15%). Meanwhile, 10% adjust their routes to bypass the restricted zone. Air quality improvements are driven primarily by the ban on Euro 4 diesel vehicles. The simulation reveals substantial reductions in pollutant emissions, particularly particulate matter (PM2.5): -74.2% within the LEZ, -62.2% across the Métropole du Grand Paris (MGP), -45% in the Versailles Grand Parc (VGP), and -41.3% across Île-de-France. Reductions in NOx emissions are mixed with -17.8% in the LEZ, -15.2% in the MGP, +6% in the VGP, and -6.8% across Île-de-France.