Recommendation system for climate informed urban design under model uncertainty

Abstract

We develop a novel recommendation system for optimal urban design from a socio-economic and Outdoor Thermal Comfort (OTC) perspectives. Currently, urban planners and designers do not have quantitative tools or methods at their disposal to incorporate various important effects into consideration in a systematic way. Our framework is based on risk measures which quantify and take into account various important design criteria such as OTC, spatial use in the form of exposure maps and the investment cost of the design. Our framework combines those criteria into a spatio-temporal risk measure to assess the performance of candidate urban design options. This is used as the basis for our optimal design problem formulation. We then formulate the problem as an optimisation problem which is easy to solve and has a clear interpretation. The objective of this paper is therefore twofold: on the one hand, we develop technical knowledge and methodologies to assist urban planners develop and prioritize competing urban design strategies. On the other hand, we demonstrate the importance of incorporating the uncertainty in climate models into the utility function when making policy decisions. To illustrate how our framework can be used in practice we present a real-world study, which is based on a set of urban design strategies that aim to improve the OTC of a specific site in Singapore. The ENVI-met micro-climate model has been used in order to calculate the spatio-temporal OTC process. We show how our framework can assist decision-makers make more informed and interpretable choices on how to select the optimal design option and where to allocate best their investment/resources.