A Decision-Making Framework to Support Urban Heat Mitigation by Local Governments

Abstract

Urban overheating can result in serious environmental, social and economic problems such as a poor outdoor thermal environment, heat stress, heat-related morbidity and mortality, increased peak electricity demand, and economic productivity loss. Extensive effort has been made by local governments to mitigate urban overheating, cool streetscapes and cities, and protect vulnerable people. However, there is uncertainty about which urban heat mitigation strategies (UHMSs) can provide better solutions for a specific urban context. Also, there is a compelling need for local governments to automate the decision-making process and optimise the combination of UHMSs to maximise the mitigation outcomes for their cities.This research aims to fill the gap by developing a novel decision-making framework that integrates artificial intelligence techniques with urban heat mitigation in the built environment. The novel decision-making framework comprises: the ontology-based knowledge representation of UHMSs and their relationships with urban contexts and performance assessment; sensitivity analysis of the environmental, social and economic performance of UHMSs; and genetic algorithm-based multi-objective optimisation of UHMSs. The novel decision-making framework enables an automated process of identifying appropriate UHMSs tailored to a specific urban context and generating the optimal combination of UHMSs to support decision making by local governments. The utilisation of the decision-making framework was investigated through Leppington and Green Square development cases in Sydney, Australia. The results showed that a set of optimal combinations of UHMSs with key planning and design variables was obtained automatically. For example, an optimum combination of 7 m Koelreuteria paniculata on street sides, greyish-white concrete on driveways, white concrete on roofs, and grey concrete on walls was automatically obtained for urban heat mitigation in Leppington. The most key planning and design variables are related to albedo since albedo had significant impacts on environmental, social, and economic performance in Leppington. The optimised combination of UHMSs led to the optimal environmental, social and economic performance, for example, an air temperature reduction of 0.9°C, land surface temperature reduction of 8.9°C, a heat-related mortality rate reduction of 6.4%, a UTCI improvement of 0.5°C, a reduction in the economic productivity loss of 2.2%, a reduction in electricity energy bills of 6.1%, and a reduced implementation cost of 22.8% for the Leppington case.This research incorporates artificial intelligence techniques into urban heat mitigation in the built environment to enable an automated process of decision making. The decision-making framework and its application in two urban development cases demonstrate a novel approach for local governments to automatically identify optimal combinations of UHMSs to tackle urban heat challenges in the local context. The research outcomes will advance interdisciplinary knowledge about the use of artificial intelligence techniques in the decision-making process for urban heat mitigation.