Accurately predicting terrestrial net methane (CH₄) fluxes from wetlands depends on multiple physical, biological, and chemical mechanisms that are poorly understood, oversimplified, or missing in regional and global biogeochemical models. The large uncertainty of CH₄ fluxes and the challenging aspects of modeling them are driven, to a large degree, by the small-scale spatial and temporal heterogeneity of CH₄ fluxes, the complex coupling between aboveground and belowground processes, and the complexity of meteorological, hydrological, ecological, and microbial processes that affect these fluxes. We aim to improve the quantitative understanding of the key processes that affect methane emissions at a high (patch level, vertically detailed) spatial resolution, and translate this understanding to improve the modeling of coastal wetland CH₄ fluxes for E3SM Land Model (ELM v1).

Eddy Covariance, chambers, and peepers’ measurements are taking place in Old Women Creek (OWC) wetland site in Ohio and in three Coastal wetland in Louisiana. ELM simulations of daily CH₄ fluxes matched observation in Louisiana site while they were significantly lower than observations in OWC. We hypothesize that the high fluxes in OWC are due to the high dissolved carbon and nutrient influx the upstream agricultural watershed. Downstream carbon export is not accounted for in ELM, or most other global land models. Ongoing work focuses on incorporating wetland vegetation types into the model plant functional types. This will allow simulation of wetland plant biogeochemical and physical dynamics (such as seasonal leaf area, photosynthetic capacity, CH₄ transport through aerenchyma, etc.) at the level of the eco-hydrological patch type, potentially resulting in a more accurate prediction for methane flux in wetlands.