High-resolution modelling of methane plumes: validation and sensitivity experiments to explore UAV and satellite observations

With the rapid expansion of high-resolution satellite imagers collecting methane plume images across the globe, emissions assessment has been performed using Gaussian plume approaches, mass-balance estimations, or flux divergence methods. But many plumes of methane sampled by space borne imagers present complex features due to topography, to the presence of infrastructures, or to discontinuous turbulent structures inherent to the near-surface atmospheric dynamics. We present here the results from a Computational Fluid Dynamics (CFD) model to study and to analyse methane emissions from point sources using the Fire Dynamics Simulation (FDS) model. High-resolution models like FDS allow to include the terrain characteristics, buildings, canopy cover, and use the real-time weather (re-analyses or field measurements of 3-D wind conditions) to simulate the observed plumes. In addition, FDS enables the study of the non-linear relationship between emissions and the height of the methane release, the temperature of the decompressed gas, and its mass flow rate. Controlled released experiments of methane measured by drone are used as a starting point to evaluate our FDS simulations and to perform various sensitivity experiments applied to real-cases (with terrain information from SRTM30, temperature and wind from meteorological ERA-5 re-analyses). We defined the optimal model physics configuration and domain characteristics based on two UAV campaigns (TADI campaigns in 2019 and 2021). The second part of our work is based on controlled releases measured by the high-resolution PRISMA satellite, covering a wide range of methane concentrations under various atmospheric conditions to evaluate the performances of FDS when simulating pressure-weighted columns. We conclude here that high-resolution atmospheric simulations outperform current approaches when analysing irregular plumes due to uneven terrains, buildings, and canopy. This method helps to improve the quantification of even small methane leaks from different point sources such as oil and gas storage facilities.