Inverse modeling of CH4 emissions over Europe, Part I: modeling tools, input data sets and forward modeling evaluation

According to an European Commission report, the global anthropogenic CH4 emissions could increase up to 100 % by 2050, leading to a situation in which O3-related premature mortalities linked to CH4 emissions as well as crop damage events from CH4-induced O3 concentrations would be more often observed. On the other hand, given that the atmospheric CH4 budget over a region is strongly dependent on its terrestrial and aquatic CH4 sources, inverse models appear as a powerful tools in identifying critical areas that can be submitted to emission mitigation strategies. In this regard, an inverse modeling system of CH4 emissions based on state-of-the-art atmospheric modeling tools is being developed at the Department of Geoscience at Aarhus University, in Denmark. The forward modeling component is based on the Weather Research and Forecasting (WRF) model version 4.3 coupled to a module for greenhouse gases. The CH4 emission fluxes from anthropogenic sources are based on the Emission Database for Global Atmospheric Research (EDGAR) model version 6, while the CH4 emission fluxes from biogenic sources are calculated online based on the Vegetation Photosynthesis and Respiration Model. Sets of 96-hour simulations for different study periods in 2018 and 2019 are conducted over a 30-km modeling domain covering most part of Europe. The first 24 hours of each simulation are discarded as spin-up time, while the remaining 72 hours will be used as inputs for the backward modeling component. Finally, the last 24 hours of each simulation are used to evaluate the model performance against CH4 total-column estimates from the TROPOspheric Monitoring Instrument (TROPOMI) onboard the Copernicus Sentinel-5 Precursor satellite. Model results also include the relative contribution from anthropogenic and natural sources, as well as the combined effect of the background concentrations and plume intrusions coming from regions outside the modeling domain.