AI potential in urban environmental modelling

Abstrakt

ZÁKLADNÍ ÚDAJE: Resilient Society | Application of AI | Climate Change/Energy: Programme and Abstracts. Saxon Academy of Sciences and Humanities in Leipzig and Czech Academy of Sciences, 2025. [Resilient Society | Application of AI | Climate Change/Energy: Second Czech-German Scientific Symposium. 22.09.2025-23.09.2025, Leipzig]. ABSTRAKT: Urban areas are increasingly vulnerable to climate change impacts, particularly heatwaves. Related changes will directly affect the thermal comfort of dwellers. Previous studies revealed interesting patterns, daytime heat stress increases more uniformly across the city, whereas nighttime heat stress exhibits greater spatial heterogeneity. They are driven by factors such as shading, land cover or material properties. However, the influence of the factors above on outdoor thermal comfort, a precise understanding of all effects and interactions or their future prediction, remains insufficiently quantified. The current most accurate models, due to the complexity of urban surroundings, are based on the large-eddy simulation principle. Although this approach enables us to simulate an urban environment in the units of meters, it requires extreme computational cost, this represents a potential for deep learning models designed to predict human thermal comfort in complex urban environments.