National forest inventories: Prospects for harmonised international reporting

Despite the important differences in inventory estimates resulting from the use of different national definitions, variables, and variable thresholds, prospects for developing procedures leading to compatible estimates amongst countries are generally positive. Analyses of national definitions and responses to questionnaires distributed by COST Action E43 indicate that definitions tend to be based on the same rather small set of inventory variables. For example, national definitions of forest all focus on minimum area, minimum cover, minimum width, and minimum height, although the thresholds for these variables differ considerably among countries (Vidal et al. 2008). Important differences among these national definitions relate mostly to kinds of lands with tree cover that are considered forest for reporting purposes: for example, managed versus unmanaged forest land, inclusion or exclusion of forested park and leisure lands, inclusion or exclusion of forest lands whose tree cover consists primarily of non-native species, and inclusion or exclusion of permanently non-stocked areas within forest land (Cienciala et al. 2008). For reporting to the UNFCCC, estimates for carbon pools are to be reported separately for different carbon pools and within pools for lands whose land use category has remained unchanged and for lands whose land use category has changed (IPCC 2003). European countries participating in COST Action E43 report estimates by the same carbon pools as the USA, although estimates for individual carbon pools are obtained by aggregating estimates for different constituent components (McRoberts et al. 2009b). Estimation by unchanged and changed land use categories requires techniques for distinguishing between the two classes. Among COST Action E43 countries, 70% use sampling techniques alone or in combination with other techniques for this purpose, 32% use map analyses alone or in combination with other techniques, and 18% use both sampling techniques and map analyses (Cienciala et al. 2008). For estimating carbon stock changes, two approaches are used: estimation based on differences between successive inventories, characterized as the stock change method, and estimation of biomass increment and removals separately from the inventory, characterized as the default method. Among COST Action E43 countries, the two approaches are used in approximately equal proportions. As an example, estimates for the deadwood component are obtained by 74% of countries using ground measurements whereas 37% of countries use predictive models.