Coupled Simulation of CO2, CH4, and N2O Fluxes from a Forested Wetland Using Data-Model Fusion Approach.

Fluxes of greenhouse gases (GHGs) from soils are likely to play a significant role as biotic feedbacks to climate change. Production and consumption of CH4 and N2O in soil are linked through heterotrophic dependence on fixed carbon sources for energy, but with contrasting effects of O2 as either essential substrate or potential inhibitor. However, most biogeochemical models use separate model versions for simulating SOM decomposition resulting in CO2 fluxes and the processes affecting CH4 and N2O emissions as it remains challenging to explain mechanistically these dynamics. Here we focus on a novel integration of measurement and modeling of CO2, CH4, and N2O, three key GHGs that are influenced by climate change, in a parsimonious modeling framework at the Howland Forest in central Maine. Howland serves as a close proxy for many boreal landscapes, where the mosaic of wetlands interspersed with upland positions provides a natural laboratory to study the influence of soil water content and oxygen supply on the interactions of CH4, N2O, and CO2 fluxes. For net CH4 and N2O fluxes in wetlands, we are attempting to simulate a saturated and unsaturated zone, using a frequency distribution of soil microsites with variable substrates and O2 concentrations. In most instances, the saturated zone is primarily anaerobic, but the unsaturated zone above it have a range of anaerobic and aerobic microsites, which play a key role for the net efflux of GHGs. Data streams of all three GHG fluxes are used to constrain the model in a Bayesian framework. Simulating all three GHGs simultaneously in a multiple constraints model afford greater confidence that the most important mechanisms are skillfully simulated resulting from improved parameterization and good process representation (i.e., getting the right answers for the right reasons).