Simulating urban air quality in Augsburg, Germany, using the PALM model to understand health impacts of changing traffic emissions in future urban transportation landscapes.

Understanding the interplay between urban air quality, traffic emissions, and the impact of future urban transport landscapes on cardiometabolic health outcomes is critical for sustainable urban development. This study employs the high-resolution Parallelized Large-Eddy Simulation Model (PALM) to model air quality in the Augsburg region, focusing on spatial and temporal dynamics of traffic-related pollution. By integrating Agent-Based Modeling (ABM) using MATSim, detailed emission data from vehicle traffic based on mobility is incorporated to quantify its contribution to overall urban air pollution. The PALM configuration includes a fine-grained grid resolution, dynamic representation of urban microclimates and urban topography to realistically simulate pollutant dispersion in complex terrain. Key objectives of this study include estimating individual exposure to health-relevant pollutants like NO2, PM2.5, ultrafine particles (UFP), as well as metal compounds in particulate matter across Augsburg and its neighborhoods, with particular attention to residential areas near traffic hubs. Spatially resolved exposure assessments are refined using population density, mobility data, and activity patterns to evaluate demographic-specific risks, particularly for vulnerable groups such as children and the elderly. Temporal analyses explore diurnal variations in traffic emissions, identifying peak pollution periods during rush hours and seasonal fluctuations. Preliminary results highlight pollution hotspots near major roads and intersections, with expected higher concentrations during peak traffic times. Initial maps reveal significant spatial heterogeneity, indicating heightened exposure levels for populations residing in traffic-dense zones. Anticipated contributions include a robust quantification of traffic's role in air quality deterioration and spatial visualization of pollutant distributions. Future work will focus on developing machine learning-based parameterizations to simplify interactions between traffic emissions and urban background pollution, facilitating scalable predictions for diverse urban settings by further investigating the non-linear health effects of air pollution. This research provides insights for urban planners and policymakers to improve air quality and health in traffic-affected areas.