Abstract
The numerical and experimental techniques are widely used for urban air flow studies. As spatiotemporal scales increase, these models face an extended range of limitations, from the experimental setup sizing/equipment constraints to expensive physics-based computational modeling. Reduced order models (ROMs) are introduced as an alternative or a complementary tool to current practices for identifying urban airflow characteristics. This paper investigates the implementation of a group of ROMs on an opensource experimental dataset. These models follow two major steps dimensionality reduction and airflow feature computation. The results show a significant reduction in the computational cost with similar accuracy achieved when studying urban airflow characteristics to those from experimental testing. These models can be further used in urban climate analysis.
