Tuesday 03.02.26

at 15:00 in Peka Gallery, 2nd floor in Amado Building

Urban built environments that are responsive to microclimatic conditions are essential for mitigating the impacts of climate change and urbanisation on building energy use and outdoor thermal comfort. Although urban microclimate modelling tools enable designers to assess how built environments influence local climate, they remain computationally intensive, particularly for annual simulations. This study presents a Machine Learning–based framework for developing metamodels that can predict urban microclimate conditions around buildings up to 1,350 times faster than the actual tool. This framework primarily introduces new methodologies for generating metadata. It also offers predictors that describe position and distance of multiple urban objects from each point near building surfaces. The seminar will detail the framework and demonstrates the metamodels developed for Tel Aviv. 

Contact information: Naga | 055-9728686 | mail: nagam@campus.Technion.ac.il