Contaminant Dispersion in Oklahoma City - High Fidelity Simulations and Comparison with Magnetic Resonance Measurements

The intentional or accidental release of a contaminant in an urban environment has the potential to be harmful for a large population. However, predicting the dynamics of the contaminant plume is complicated due to the complex geometry of the urban buildings. Many studies have been performed that provide velocity and concentration data at a few discrete points within the complex flow. This work leverages a rich four-dimensional (three-spatial and one-temporal) measurement taken with magnetic resonance imaging (MRI). The experiment was conducted on a 1:2206 scaled model of downtown Oklahoma City within a water tunnel and provides contaminant concentration and three components of velocity at 1.6 million locations and 12 times. This dataset is compared against LES simulations performed with the same geometry and boundary conditions. The one-to-one comparison between experimental measurements and simulations allows us to 1) explore the physics of the dispersion in the urban environment and 2) assess the LES model's predictive capabilities. In particular, velocity statistics, contaminant isosurfaces, and flow intricacies near the buildings are investigated.