High resolution Turbulence modelling to improve complicated methane emissions observed from Satellite Imagery

With the growing interest to identify and quantify methane (CH₄) emissions from various sources around the globe, the development and launches of satellite imagers tracking CH₄ plumes have accelerated during the past decade. Thanks to high-resolution images collected by PRISMA (30-m resolution), Sentinel-2 (20-m resolution), or Tanager-1 of Carbon Mapper (30-m resolution), it is now possible to sample CH₄ plumes at small scales to enable source attribution and quantification at lower detection levels. However, high-resolution images also come with limitations, due to the dominance of small-scale turbulence physics near the source. Therefore, Large Eddy Simulation (LES) modelling becomes necessary to resolve the turbulence, leading to more robust emission quantification methods. The Fire Dynamics Simulation (FDS) model offers unique capabilities by allowing to introduce infrastructures, terrain topologies, and roughness, individual trees (incl. leaves, branches, tree shapes), surface temperature gradients, velocity of the gas release, velocity of the leaked air, and to simulate the dynamics of the plume at a high resolution near the source and coarse resolution away from the source (adaptive mesh refinement). Using such high-resolution LES modelling allows us to simulate the spatial structure of the CH₄ plumes, at various distances from the source, under different meteorological conditions (forced by wind measurements or re-analysis products). We aim here to (i) define more rigorously the effective wind speed used by the Integral Mass Enhancement (IME) method, (ii) to determine the sensitivity to external parameters that affect the plume dynamics, and (iii) to analyze complicated plumes (excluded in current IME analyses) thanks to our LES simulations offering additional detections to current monitoring systems (e.g., MARS).

 

More specifically, we compare and evaluate the importance of all the environmental conditions (topography, obstacles) and the characteristics of the source (temperature, velocity, height) affecting the dynamics of the turbulent CH₄ plumes to determine the most favorable conditions and the uncertainties in IME estimates diagnosed from satellites and aircraft measurement campaigns. We also determine the sensitivity of the emissions to the effective wind speed simulations as a function of distance to the source and its comparison with existing wind speed calculation methods (e.g., extrapolation from 10-m wind speed measurement) used in current mass balance approaches. This study leads us to discuss the use of high-resolution LES simulations in mass balance calculations to produce more reliable estimates of facility-level sources around the globe.