A hybrid Python approach to assess microscale human thermal stress in urban environments

Abstract

Microclimate simulations are in high demand to assess the thermal impacts of urban design and vegetation changes. Modeling accurate microclimate dynamics in complex urban settings requires extensive computing power. A hybrid Python approach is introduced to simulate human thermal exposure (mean radiant temperature, MRT) and comfort (Universal Thermal Climate Index, UTCI) in cities. The proposed model combines various engines in Rhinoceros to account for interactions between urban surfaces, tree canopies, and the atmosphere. The model was validated in hot, dry Tempe, USA, using in-situ human-biometeorological observations and then applied to urban archetypes in Bologna and Imola, Italy. MRT and UTCI were simulated at five sites in Bologna, four in Imola, and four in Tempe, with varying building heights and canopy cover for the climatologically hottest week of the year (August 3–9). The model performed well with an RMSE of 5.4 °C, an index of agreement of 0.96, and outperformed existing models for tree-shaded sites. MRT and UTCI were driven mainly by shade from dense urban forms and trees. Highrise, medium-to-high tree canopy cover archetypes were the coolest concerning thermal exposure and comfort. Sites in Tempe exceeded the UTCI categories for Very Strong or Extreme Heat Stress independent of archetype. The model enables fast and accurate assessment of urban tree planting strategies.