Uncertainty Quantification of Global Net Methane Emissions from Land Ecosystems Using Mechanistically-Based Biogeochemistry Models

Methane is widely considered as the second most powerful greenhouse gas. A large amount of it is emitted from natural wetlands and is consumed in upland ecosystems. To date, the quantification of its natural sources and sinks as well as their net budget using process-based biogeochemistry models still remains large uncertainties. Here we use an extant methane biogeochemistry model and multiple in situ and satellite data of methane fluxes and wetlands and upland distributions as well as different climate data, to quantify these uncertainties. We found that, driven by CRU data, the global wetland methane emissions are 152 and 189 Tg CH₄ yr^-1 by using a static wetland distribution data and a dynamical inundation data, respectively, during the period of 2000-2012. This study presents detailed uncertainty sources and magnitude of global methane flux estimates in different ecosystems, and indicates that serious considerations should be taken when choosing wetland distribution data and forcing climate data. In addition, this study also highlights the need of: 1) more in situ flux data to better constrain the model, 2) more accurate dynamical wetlands and upland and inundation distribution data; and 3) more accurate climate data, to improve the global land methane biogeochemistry model, a critical component of Earth System Models.