Absolute and Relative Uncertainty of Model Generated Soil Greenhouse Gas Emissions.

Rigorous methods have recently been developed to quantify uncertainty in model generated greenhouse gas emissions reported in national greenhouse gas emission inventories. Monte Carlo analysis involving probability distribution functions are used to quantify uncertainty due to not precisely knowing model inputs. Model outputs are compared with observations to quantify uncertainty in model structure. Uncertainty in absolute levels of GHG emissions ranges from 30% to greater than 100%, depending on spatial and temporal scales. However, there is some evidence that relative uncertainty based on comparing simulated and observed treatment differences may be narrower. We compare absolute and relative uncertainty in soil GHG emissions from various cropped systems in the US.