Modeling Patterns of Population Exposure With a Multiscale Framework Utilizing LOTOS-EUROS Chemical Transport Model,...

The present study proposes a high-resolution, multi-component framework to estimate human exposure to air pollution in tropical urban environments, addressing complex topography and limited historical mobility and exposure data. This framework is applied to the city of Medellín in the Aburrá Valley in Colombia. To assess population exposure to particulate matter, numerical simulations, mobility and toxicity data were integrated. The LOTOS-EUROS Chemical Transport Model, driven by the Weather Research and Forecasting (WRF) model, was used to simulate air quality at a spatial resolution of 1 km  $\times$  1 km for the year 2019. Mobility dynamics were derived from the Origin-Destination Survey (ODS), comprising over 180,000 trips across 240 traffic analysis zones. The utilisation of this data set facilitated the estimation of time-weighted exposure levels for diverse demographic groups during weekdays. The morpho-chemical properties of the particulate matter samples were analyzed to determine the presence of metals and carbonaceous fractions due to significant health implications. The exposure model is a weighted model that integrates exposure time, ${\text{PM}}_{2.5}$ and ${\text{PM}}_{10}$ concentration, and cytogenotoxic indicators. The application of cluster analysis to the available data, resulted in the identification of areas of elevated health risk, indicating that the central and southern zones, approximately 60% of the metropolitan population and main highway corridors, exhibited mean particulate matter exposure levels that exceeded 40 μg/m³ during peak hours, thus surpassing the WHO air quality guidelines. These zones demonstrated the highest levels of exposure, as indicated by cluster significance levels, suggesting more epidemiological studies and public health interventions.