Optimization of tree positioning to maximize walking in urban outdoor spaces: A modeling and simulation framework

Abstract

• Developed an agent-based model to simulate tree planting effect on walking frequency. • Created a model based on bivariate gaussian distribution to map MRT change from trees. • Proposed a variable BSE to quantify how likely a point will be shadowed by a building. • Analyzed how spacing between trees, direction of buildings, and BSE affect walking. • Framework can be used as a tool to find optimal tree distribution to maximize walking. Trees can improve the walkability of urban outdoor spaces, hence studies have been conducted to elucidate the relationship between tree distribution and walking frequency. However, there is still a lack of tool that could exploit these relationships and optimize tree distribution to maximize walking. This study develops a modeling and simulation framework to optimize the distribution of urban trees to maximize walking. The framework features an agent-based model that simulates Mean Radiant Temperature as a function of tree position using a mathematical model based on a bivariate gaussian distribution. Environmental conditions are sensed by Human agents which generates comfort levels and affects walking decisions. To optimize tree distribution, a Genetic Algorithm is used to find the tree distribution that generates maximum walking frequency. The simulation results indicate that spacing between trees, direction of neighboring buildings, and Building Shadow Exposure (BSE) affect walking frequency. Applying the framework in an urban area in the Philippines showed that the optimal planting strategy is to plant trees at a 14 m spacing on all sidewalks except which have buildings to the south and only on locations with BSE less than 30.17. The proposed framework can be a powerful tool for urban tree planting projects.