Computational fluid dynamic simulation data for methane plume dispersion and Bayesian uncertainty quantification in vehicle-based emission estimates

This dataset contains computational fluid dynamics (CFD) simulation outputs used to quantify uncertainty in methane emission estimates derived from vehicle-based atmospheric measurements. The data support a Bayesian inference framework developed to characterize uncertainties in emission rate estimates obtained using Gaussian plume dispersion models combined with concentration measurements and anemometry. The dataset provides representative instantaneous flow fields from Detached Eddy Simulations (DES), which were used to evaluate model error and inform the likelihood functions within the Bayesian framework. Simulations were conducted using OpenFOAM on high-resolution computational grids (~22 million cells), employing a hybrid turbulence modeling approach with large eddy simulation (LES) in the freestream and Reynolds-averaged Navier–Stokes (RANS) near surfaces . Synthetic turbulence was applied at the inlet, and simulations were initialized on a coarser mesh before being remapped to the final grid and advanced over multiple flow-through times. The dataset includes spatially resolved three-dimensional flow field snapshots stored in VTK format, and time-resolved probe data collected at various locations throughout the simulation domain. Probe datasets include velocity vectors, pressure, and methane concentration (treated as a passive scalar), enabling detailed temporal analysis of plume dynamics. The simulation outputs were used to evaluate the variability and uncertainty in inferred emission rates and to assess the performance of the Bayesian framework