Evolutionary design and morphological optimization of urban canyons as a mitigation strategy for urban heat island

The microclimate and intensity of the urban heat island are largely determined by urban morphology. Within this framework, this research implements a system that combines parametric modeling, evolutionary design, and the Non-Dominated Classification Genetic Algorithm-II (NSGA-II) to assess the impact of vertical densification in urban canyons to study urban form solutions that mitigate overheating for the vulnerable population. The NSGA-II was used to generate a set of solution scenarios whose alternatives differ in urban canyon geometric shape optimization based on three fitness objectives: (i) maximize the urban densification, (ii), maximize the aspect ratio (H/W) while taking the city's current urban policy restriction concerning thermal environment improvement, (iii) and minimize Universal Thermal Climate Index (UTCI). Finally, the optimal solutions for each fitness objective were selected independently using the "fitness rank" method, which ranks the solutions of the fittest to the least fit. The population size was limited to 500 scenarios, which were simulated for the hottest day of a typical year. The results of the scenarios show a reduction of up to 17% in the annual solar radiation that reaches the surface of the canyon and lower values of UTCI and mean radiant temperature of up to 1 degrees C and 5 degrees C respectively from temperatures above 30 degrees C, due to the increase in the shadow by a higher aspect ratio (H/W>2). The results obtained provide a perspective for the design and/or regulation of urban growth policies and instruments of ordering and territorial planning of urban centers.