Climate-informed urban design via probabilistic acceptability criterion and Sharpe ratio selection

We develop a new framework for selecting an urban design which performs best from an Outdoor Thermal Comfort (OTC) perspective, while taking into account the uncertainty in the OTC preference of individuals. To this end, we first present and develop the notion of Probabilistic Acceptability Criterion (PAC) which is a new method to quantify people’s satisfaction of OTC values, based on data collected from a survey. We develop the PAC for both regression and classification models which are most common statistical analysis methods in the literature. Next, based on the PAC, we develop a new approach to scoring each of the urban designs, based on Binary Probabilistic Impact Function, which extends the widely used deterministic impact function. We show that the score is a random variable which follows a Poisson-Binomial distribution and characterise its parameters. We then use those results and present a new approach for scoring of the urban designs that is based on the Sharpe ratio, which is a widely used metric in financial applications. Our framework is the first model which provides urban designers the ability to evaluate the quality of their urban designs from an OTC point of view, while taking the uncertainty into account in a holistic and rigorous way. We illustrate our framework by applying it to a real case study in Singapore.