Reduced Order Modeling of Fluid Dynamics in Large Scale Urban Domains

The main objective of this project is to investigate the feasibility of implementing a reduced-order model for weather simulations in urban areas. The effectiveness of reduced order models is strictly dependent on the application. Therefore, a proper campaign of simulations is required to evaluate its feasibility in the context of large-scale weather simulations in the urban scale. This project bridges the fields of computational mathematics and aerospace engineering to evaluate the potential of reduced-order modeling in large-scale urban simulations.

To achieve this objective, several steps must be taken: - Literature review on state-of-the-art results and data of atmospheric boundary layer for wind and temperature (in urban areas). - Generate an appropriate 3D mesh of the computational domains representing urban areas. - Formulate realistic boundary conditions based on available data. - Set up the computational fluid dynamics (CFD) pipeline for large scale simulations of incompressible flows (coupled with temperature) in a supercomputer. - Postprocessing of the CFD solutions - Explore the solution vector space to construct a reduced-order model. - Estimate the associated errors.