Validation of MLUCM BEP+BEM: Assessing urban microclimate across mesoscale and microscale dynamics

To accurately simulate urban climate under extreme conditions and assess its evolution in the context of climate change, a one-dimensional model, MLUCM BEP+BEM, has been developed. This model integrates the vertical turbulent diffusion approach proposed by Santiago and Martilli (2010) with the Building Effect Parameterization (BEP) from Martilli et al. (2002) and the simplified Building Energy Model (BEM) introduced by Salamanca et al. (2009). Enhancements in turbulent length scales for dissipation and eddy coefficients have been incorporated, accounting for atmospheric stability following the methodology of Bougeault and Lacarrere (1989).
Designed as a standalone model, MLUCM BEP+BEM operates with atmospheric forcing applied at the top boundary, its design enables offline simulations and simplifies its integration in mesoscale models. Additionally, it includes representations of urban greenery and street trees, following the approaches of Zonato et al. (2021) and Stone et al. (2021), respectively, while also introducing new terms for temperature and humidity fluxes.
This study validates MLUCM BEP+BEM using observational data from the Urban-PLUMBER project, assessing its capability to simulate surface-atmosphere fluxes in a suburban area of Preston, Melbourne, Australia. Comparisons with similar models highlight its very good performance in representing urban climate dynamics. Moreover, the intra-urban sensitivity of the parameterization is examined through a case study of Bari, a mid-sized Mediterranean city, evaluating the interplay between mesoscale forcings and urban microscale processes during an intense heatwave. The model, forced by ERA5 reanalysis data, was compared with temperature measurements from sensors placed in different city neighbourhoods. The findings indicate that mesoscale influences play a dominant role in shaping urban microclimate simulations, with urban geometry exerting a secondary effect. 

This work is supported by ICSC – Centro Nazionale di Ricerca in High Performance Computing, Big Data and Quantum Computing, funded by European Union – NextGenerationEU (CUP F83C22000740001).