Locating trees to mitigate outdoor radiant load of humans in urban areas using a metaheuristic hill-climbing algorithm – introducing TreePlanter v1.0

Abstract

Mean radiant temperature (Tmrt) is a frequently used measure of outdoor radiant heat conditions. Excessive Tmrt, linked especially to clear and warm days, has a negative effect on human wellbeing. The highest Tmrt on such days is found in sunlit areas, whereas shaded areas have significantly lower values. One way of alleviating high Tmrt is by planting trees to provide shade in exposed areas. Achieving the most efficient mitigation of excessive Tmrt by tree shade with multiple trees requires optimized positioning of the trees, which is a computationally extensive procedure. By utilizing metaheuristics, the number of calculations can be reduced. Here, we present TreePlanter v1.0, which applies a metaheuristic hill-climbing algorithm on input raster data of Tmrt and shadow patterns to position trees in complex urban areas. The hill-climbing algorithm enables dynamic exploration of the input data to position trees, compared with very computationally demanding brute-force calculations. The hill-climbing algorithm has been evaluated with a static greedy algorithm that positions trees one at a time based on ranking and is expected to always find relevant locations for trees. The results show that the hill-climbing algorithm, in relatively low model runtime, can find positions for several trees simultaneously, which lowers Tmrt substantially. TreePlanter, with its two algorithms, can assist in optimization of tree planting in urban areas to decrease thermal discomfort.