Sensitivity analysis of a wetland methane emission model based on temperate and arctic wetland sites

摘要: Abstract. Modelling of wetland CH4 fluxes using soil emission models is used to determine the size this natural source on local global scale. Most process formation and soil-atmosphere transport processes operate a plot For large scale modelling (regional scale) upscaling type model requires thorough analysis sensitivity these parameter uncertainty. We applied GLUE (Generalized Likelihood Uncertainty Analysis) methodology well-known model, Walter-Heimann as implemented in PEATLAND-VU model. The tested data from two temperate sites one arctic site. tests include experiments with different objective functions, which quantify fit results data. indicate that 1) most cases capable estimating better than an estimate based avarage, but does not clearly outcompete regression data; 2) reproducing larger (seasonal) temporal variability data, small-scale (daily) variability; 3) strongly sensitive parameters, 4) parameters determining oxidation vegetation, temperature microbial population. method also allowed testing several smaller modifications original conclude plot-based feasible, considerable improvements will result improvement vegetation