Abstract: Mathematical modeling is one of the primary tools for assessing the quality of local and regional air. However, this approach has some limitations in representing smaller scales, not adequately capturing specific local features such as buildings, roads, green areas, and especially in deficient planned urban environments. This limitation becomes more evident when analyzing the dispersion of pollutants and their direct influence on pedestrians, where the correct interpretation of urban canyons poses a challenge. In this context, the street-network model, MUNICH, emerges as a tool that incorporates local urban characteristics, using a box network to simulate the spread and distribution of pollutants, with special attention to better representing urban canyons. The MUNICH model stands out by considering urban flow and capturing concentrations more efficiently in roadways, although it has limitations in representing pollution at a regional scale, such as in a metropolitan area. Additionally, the model needs to incorporate the effects of background concentration as an essential approach to enhance the accuracy of simulations. In this study, the MUNICH model was applied to evaluate the influence of varying dimensions of urban canyons on the dispersion and transport of nitrogen dioxide (NO2) in a neighborhood in the city of Vitória, Espírito Santo, Brazil. This area has significant emission sources, especially steel industries and heavy traffic. The enhancement of urban canyons within the urban canopy for this study was conducted in the Jardim Camburi neighborhood, located in the municipality of Vitória. Data provided by the Global Local Climate Zones (LCZ) and the cartographic database made available by the municipality of Vitória were used. These datasets provide a more homogeneous representation of the urban canopy and a more realistic depiction, respectively. The use of these two databases revealed differences in geometric complexities, as the local dataset provided by the municipality represented reality more accurately, while the second database (LCZ) considered a more homogeneous height. Thus, it was possible to understand the influence of these elements on NO2 dispersion in real urban environments. This study also sought to comprehend the impact of temporal variations in NO2 emissions, considering local differences, providing information for improving air quality management in a complex urban environment like Jardim Camburi and, consequently, other areas with similar urban structures. The results demonstrate how an accurate representation of urban canyons in the model affects wind speed, modifying vortex formation and flow patterns, thereby directly influencing pollutant dispersion. The findings also underscore the importance of accurately representing urban canyons in pollutant dispersion (in this case, NO2), especially in poorly planned urban areas where significant variations in building height make modeling accuracy a parameter that requires special attention. Therefore, careful consideration of the geometric complexity of the location and its more precise and realistic incorporation into the model is essential for accurate simulation and reliable data generation, supporting decision- making in the management of air quality in complex urban environments.

Keywords: Street-network, air pollution, pollutant transport, air pollution modeling, MUNICH model.