Urban Climate Risk Mitigation via Optimal Spatial Resource Allocation

Decision makers (DMs) who are involved in urban planning often required to allocate finite resources (say, money) improve outdoor thermal comfort (OTC) levels a region (e.g., city, canton, country). In this paper, for the first time, we address following two questions, which directly related requirement: (1) How can statistical properties of spatial risk profile an area from OTC perspective be quantified, no matter index DM chooses use? (2) Given profile, how much and where should levels? We answer these fundamental questions by developing new rigorous mathematical framework as well class models models. Our approach is based on methods machine learning: first, surrogate model that provides both accuracy tractability developed via regression analysis. Next, incorporate imperfect climate derive index. present concept spatio-temporal aggregate (STAR) measures their properties. Finally, building our derivations, develop algorithm resource allocation, useful DMs modern portfolio theory. implemented tool used it illustrate its operation practical case large-scale Singapore using WRF model.