Generative design optimization of tree distribution for enhanced thermal comfort in communal spaces with special reference to hot arid climates.

The quality of the communal outdoor environment is crucial for enhancing the urban quality and the well-being of its residents. These spaces are essential for providing more opportunities for social interaction and leisure. However, in hot arid climates like Egypt, achieving optimal outdoor thermal comfort remains a challenge. Accordingly, more comprehensive methodologies are highly needed to improve the research-based design of landscape parameters and components for developing outdoor thermal comfort performance using an iterative design exploration process that employs AI-driven software. These applications, help designers in solving multi-objective design quandaries through the generation and evaluation of numerous design options. Therefore, this study explores the efficiency of generative design tools in optimizing tree distribution based on mutation evolution to enhance outdoor thermal comfort, providing a dynamic, iterative approach that adapts to diverse urban morphologies. The methodology adopts a simulation-based analysis for framing this study, which is classified into three main phases. Firstly, analyze the current environment for specific outdoor spaces with different settings in Madinaty, New Cairo (fully clustered with buildings neighborhood, semi-clustered neighborhood, fully open neighborhood). Secondly, a generative design tool with a Dynamo evolutionary algorithm is utilized to optimize the tree distribution across the communal areas of these three spaces considering the current built environment. Lastly, testing thermal comfort using Grasshopper and Ladybug simulation to assess the Universal Thermal Climate Index (UTCI) between the base case scenarios and the optimized scenarios to validate the generative design tool. Results indicate tangible improvements across the three different neighborhoods. In the Clustered Neighborhood area, the optimized design with 33 trees resulted in a lower UTCI (with an arithmetic mean of 37.55 °C) compared to the base case with 43 trees (38 °C). In the Semi-Clustered Neighborhood area, the optimized design with 45 trees highly improves the UTCI (38.01 °C), compared with the base case with 27 trees (39.40 °C). Lastly, for the Fully Open Neighborhood area, the optimized design with 25 trees achieved a slightly improved UTCI (39.55 °C) over the base case of 31 trees (39.60 °C).