Integrating mesoscale forcing and urban canopy modelling: Quantifying thermo-hygrometric variability in a Mediterranean coastal city (Bari, Italy)

Urban microclimates in coastal cities are shaped by the interaction between mesoscale atmospheric forcing and heterogeneous urban fabric. This study investigates the thermo-hygrometric variability of Bari (Italy), a Mediterranean coastal city, using an integrated multiscale framework combining canyon-level observations, ECOSTRESS land surface temperature (LST) data, and simulations with the Multi-Layer Urban Canopy Model (MLUCM BEP + BEM) forced by ERA5 reanalysis. Observations from eight canyon-level sensors reveal strong intra-urban variability, with daytime air temperature differences between coastal and inland sites reaching 5–6 °C under sea–land breeze conditions. ECOSTRESS LST confirms these spatial patterns, highlighting cooler coastal areas and localized hotspots associated with dense urban fabrics. Model performance was strongly controlled by the choice of meteorological forcing. Simulations forced with ERA5 sea grid points provided the lowest air temperature RMSE values in coastal areas (RMSE ≈ 1.8–2.0 °C), while inland sites show more heterogeneous behavior. Relative humidity showed a more site-dependent response than air temperature: sea forcing performed better at coastal locations, whereas coast and interpolated forcings provided lower inland errors. Compared with ERA5 air temperature, the modelling framework increases intra-urban variability by 14–18%, suggesting that urban canopy modelling can represent part of the microscale thermal variability associated with urban morphology. Relative humidity showed a less regular coastal–inland spatial distribution than air temperature and stronger site-to-site heterogeneity; ERA5RH-int tended to overestimate the observed spatial variability, whereas BEPBEM configurations generally underestimated it. The added value of the model was most evident in dense urban environments, where morphological effects are strongest. Conversely, model benefits were weaker in more vegetated or transitional sites and depended on forcing and statistical metric. The proposed framework provides a methodological basis for investigating thermo-hygrometric variability in Mediterranean coastal cities, with potential extensions toward outdoor thermal comfort assessment, urban heat mitigation, and building energy analysis.