A mesoscale CFD framework for extreme temperature mapping in urban environments

This study presents a combined statistical and numerical framework for mapping extreme temperatures in a complex urban environment. The municipality of Genoa, a coastal city in Italy, was used as a case-study example. Historical temperature data from 14 meteorological stations were analyzed using the Generalized Extreme Value (GEV) distribution to establish a robust statistical characterization of extreme temperatures across the region. A detailed three-dimensional model of the city, incorporating topographic and land-cover information, was developed for conducting Computational Fluid Dynamics (CFD) simulations using OpenFOAM. Steady-state Reynolds-Averaged Navier-Stokes (RANS) simulations were performed to reproduce the thermal conditions associated with extreme summer and winter events. The simulated temperature fields were validated against observations from the weather stations, showing good overall agreement. The resulting datasets were used to generate high-resolution maps of extreme maximum and minimum temperatures, highlighting the spatial variability temperatures within the city, that is seen to be locally affected by terrain features like topography and texture. The proposed methodology demonstrates the potential of integrating statistical extreme value analysis with CFD modeling to support evidence-based urban planning and the development of targeted extreme temperature mitigation and adaptation strategies, also in the context of climate change, in complex urban environments.