METHANE CYCLE AND ATMOSPHERIC OXIDATION CAPACITY SINCE THE LAST GLACIAL MAXIMUM, A REVIEW OF MODELLING STUDIES

Ice core data indicate a large increase in greenhouse gas concentrations (CO2, CH4, N2O) between last glacial maximum and the Holocene. The change in methane concentration can be due to changes in its sources (mostly wetlands) and/or its major sink (oxidation by OH). Through its sink, the methane cycle is related to other atmospheric species controlling the oxidation capacity of the atmosphere: (CO, NOx, VOCs). Therefore atmospheric chemistry modelling is required to simulate the methane budget. As for methane, the major sources of OH-controlling species are located on continental surfaces: vegetation emissions (VOCs), soil emissions (NOx), biomass burning emissions (CO, NOx, VOCs). NOx also have a major atmospheric source from lightning.

A number of 1D, 1.5D or 2D atmospheric chemistry studies followed the first publication of glacial/interglacial signal for methane. They all converge to suggest that the change in methane concentration is mostly due to its sources rather than its sink. However, these studies are based on somewhat crude estimates of emissions for all species involved. We should note that modelling methane sources and sinks that allow to simulate observed methane concentrations represents a strong constraint on the plausible model results. Other ice-core data such as CO concentrations or CH4 and CO isotopes can further help constraining the model scenarios.

To the author's knowledge, there is no published study of the LGM atmospheric chemistry with a 3D model. A complete study with currently available modelling tools would involve an important effort for establishing global maps of chemical species emissions, and require a vegetation model, a wetlands model, and a fire model with physically-based parameterizations of chemical species emissions, together with a chemistry-transport or chemistry-climate model. The study of past methane cycle can provide a mean for validating the earth-system models currently under development. Key validation data include indicators of past vegetation, wetland extent, and fire activity. Although there are palaeo indicators for all these data, a global database is available only for past vegetation. The presentation will include information about on-going work known from the author.