Within the scope of the Aveiro STEAM City project, an air quality monitoring network was installed in the city of Aveiro (Portugal), to evaluate the potential of sensors to characterize spatial and temporal patterns of air quality in the city. The network consists of nine sensors stations with air quality sensors (PM10, PM2.5, NO₂, O₃ and CO) and two meteorological stations, distributed within selected locations in the city of Aveiro. The analysis of the data was done for a one-year measurement period, from June 2020 to May 2021, using temporal profiles, statistical comparisons with reference stations and Air Quality Indexes (AQI). The analysis of sensors data indicated that air quality variability exists for all pollutants and stations. The majority of the study area is characterized by good air quality, but specific areas—associated with hotspot traffic zones—exhibit medium, poor and bad air quality more frequently. The daily patterns registered are significantly different between the affected and non-affected road traffic sites, mainly for PM and NO₂ pollutants. The weekly profile, significative deltas are found between week and weekend: NO₂ is reduced on the weekends at traffic sites, but PM10 is higher in specific areas during winter weekends, which is explained by residential combustion sources.

Accidents involving hazardous materials during transport are rare but can have severe consequences for human health. This study evaluates accidental railway chemical release scenarios along the Leixões railway line in Matosinhos, Portugal, to assess the influence of diurnal and seasonal meteorological variations on toxic vapour cloud extent and population exposure. Daytime and nighttime, as well as seasonal, scenarios were developed using historical weather data, and a modelling framework integrating the ALOHA^® dispersion software with geographic information system (GIS) tools quantified potential human impacts. Nighttime plumes with low wind speeds of 2.87 m/s and high atmospheric stability were wider and more elongated, particularly in low-risk zones and in confidence areas, resulting in higher population exposure. Daytime plumes with stronger winds of 4.24 m/s and enhanced vertical mixing were more compact and affected fewer individuals. Seasonal patterns further influenced dispersion, with winter and autumn producing moderately elongated plumes with substantial exposure, spring yielding the smallest footprints, and summer generating the largest horizontal plumes, especially in low-risk zones, leading to the highest number of exposed people. These findings highlight the importance of meteorological variability and site-specific conditions in risk assessment, emergency planning, and scheduling of hazardous material transport to minimise population exposure.

Ambient air pollution remains a major public health concern, contributing to millions of premature deaths worldwide according to the World Health Organization. Regional air quality assessments are commonly performed using chemical-transport models that require substantial computational resources due to their detailed representation of atmospheric processes. This study explores the feasibility of applying the second-generation dispersion model URBAIR^® as a computationally efficient alternative for long-term regional air quality simulations. URBAIR^® was implemented for three European case studies within the DISTENDER project to simulate particulate matter (PM₁₀) and nitrogen dioxide (NO₂) concentrations for 2018 under different spatial and temporal resolutions. Model performance was assessed against background monitoring stations and compared across grid configurations. The results show that the model successfully reproduces annual mean concentration patterns, particularly in urban areas, with R² values ranging mostly between 0.2–0.6, RMSE between 16–36 µg.m⁻³, and mean bias from −8 to 5 µg.m⁻³, indicating overall acceptable statistical performance. Within the specific configurations evaluated in this study, increasing spatial resolution was not consistently associated with improved model performance. However, because spatial resolution covaried with other factors including meteorological temporal resolution, domain characteristics, and monitoring station density, the present analysis does not allow the independent effect of spatial resolution to be isolated. Moreover, a key limitation of the modeling approach is the absence of chemical transformation processes, which may affect the representation of secondary pollutants. Overall, the dispersion-based modeling framework substantially reduces computational demand and input complexity, proving suitable for long-term exposure and climate-related applications when annual average concentrations are the primary objective. In future studies, the modeling approach should be applied to other case studies to consolidate the findings of this exploratory work so that it may contribute to sustainability-oriented decision making by facilitating regional assessments of air quality and potential health impacts related to climate change.