Sensitivity analysis as a tool for systematic reductions in greenhouse gas inventory uncertainties

A greenhouse gas emission inventory consists of a large number of input parameters, many of which have high uncertainties. The Kyoto protocol will require accurate emission data. It is, consequently, important to reduce the inventory uncertainty by improving the input parameters and methodologies in a cost-efficient manner. A key parameter is defined as one that has significant effect on the total emissions or trend and their uncertainty. Key parameters may be ranked according to their contribution to total emissions and trend uncertainty. Quantitative uncertainty estimates of emissions are not available in many countries. In order to evaluate key parameters in such inventories a simplified approach based on thresholds is proposed. Various methods of sensitivity analysis have been applied to Norwegian inventory data. The simplified threshold approach gives insight into the inventory and identifies the key parameters. More sophisticated methods of sensitivity analysis assessments are, however, useful in order to seek specific improvements. The key parameters for determination of total emissions are the large and uncertain sources. Smaller emission sources may be key parameters for the trend determination if their source level is strongly increasing or decreasing.