Algorithmic urban greening for thermal resilience: AI-optimised tree placement and species selection
Abdulrazzaq Shaamala, Tan Yiğitcanlar, Alireza Nili, Dan Nyandega
Abstract
Recent advances in artificial intelligence (AI) and metaheuristic optimisation have created new opportunities to address the growing challenges of urban heat and thermal discomfort. Among these, the strategic placement of urban trees has emerged as a promising intervention due to their capacity to moderate microclimatic extremes. However, existing approaches often rely on generic planting schemes that overlook the spatial complexity of urban morphology and the functional diversity of tree species. This study introduces a novel AI-based framework that combines Ant Colony Optimisation (ACO) with species-specific thermal traits and high-resolution simulations of the Universal Thermal Climate Index (UTCI) to optimise both tree placement and species selection at the neighbourhood scale. To evaluate the cumulative physiological benefits of these interventions, a new metric—the Bio-Thermal Gain Index (BTGI)—was developed to capture diurnal variations in thermal stress. Applied to a real-world suburban site and validated under extreme summer conditions, the framework achieved notable improvements: a 22 % reduction in areas exceeding 39 °C, an 18 % increase in thermally comfortable zones, and cooling benefits of up to 3.5 °C. This research advances a replicable, performance-oriented model for climate-responsive urban greening by uniting algorithmic intelligence with ecological precision. The proposed framework provides planners, designers, and policymakers with a scalable tool to enhance thermal resilience through informed, site-specific decisions on tree placement and species selection. • Optimised an AI-driven framework for urban tree placement and species selection • Introduced Bio-Thermal Gain Index to assess cumulative thermal comfort benefits • Achieved up to 3.5 °C cooling and 22 % reduction in extreme heat zones • Fused Ant Colony Optimisation with thermal index for spatio-temporal resilience • Demonstrated scalable performance-based model for climate-responsive greening