Characterizing Uncertainties in CH4 Leakage Flux Retrieval Using an Idealized Large Eddy Simulation (LES).

Understanding the global and regional CH₄ budget is vital for improving air quality and future climate projection, yet quantifying its sources remains highly uncertain. Recent advancements in remote sensing, such as the next-generation Airborne Visible/Infrared Imaging Spectrometer (AVIRIS-NG) and Hyperspectral Thermal Emission Spectrometer (HyTES), enable quantitative retrievals of column average methane mixing ratios at 2-5m spatial resolution, allowing for an integration of total methane amount within a CH₄ plume scene, denoted as IME, integrated methane enhancement. Previous studies have shown that this IME approach is an efficient mean to estimating the CH₄ emission rates from point sources (facility-sized) across a regional scale. The inversion of observed IME to the actual flux rates, however, is sensitive to atmospheric conditions during the overpass and still involve significant uncertainties. In this study, we use the model output from the idealized Large Eddy Simulation (LES) with prescribed surface winds, heat fluxes, and source leakage fluxes to simulate a realistic methane distribution in the atmospheric boundary layer over time, to explore the relationship between the IME and flux rates under various scenarios. We also applied the simulated 3-dimensional plume structures with instrument vertical sensitivity operators to produce pseudo-measurements for AVIRIS-NG and HyTES. Results show the non-linearity between the IME and the flux rates that are distinct under different values of surface wind speeds. In addition, by investigating the 2-dimensional plume shape in an airborne imagery perspective, we develop an algorithm using simple statistical metrics as well as modern machine learning techniques that might provide robust estimates of surface wind speed without the need of in-situ wind measurement at a CH₄ source. Our findings from the LES runs demonstrate that the IME together with its spatial pattern can disentangle the flux rate. This work provides a proof of concept that can contribute to a more accurate inversion of the flux rate under varying atmospheric conditions - a valuable information to guide field campaigns in the future.