摘要：The IPCC Guidelines propose 3 Tier levels for greenhouse gas monitoring within the forest land category with a hierarchical order in terms of accuracy, data requirements and complexity. Due to missing data and/or capacities, many developing countries, potentially interested in the reducing emissions from deforestation and forest degradation scheme, have to rely on Tier 1 default values with highest uncertainties. A possible way to increase the credibility of uncertain estimates is to apply a conservative approach, for which standard statistical information is needed. However, such information is currently not available for the IPCC values. In our study we combine a recent global forest mask, an ecological zoning map and the pan-tropical AGB datasets of Saatchi and Baccini to derive mean forest AGB values per ecological zone and continent as well as their corresponding confidence intervals. Such analysis can be considered transparent as the datasets/methodologies are well documented. Our study leads to alternative Tier 1 values and allows the application of statistically-based conservative approaches. Our AGB estimates derived from Saatchi and Baccini datasets are 35% and 24% lower respectively than the IPCC values. When restricting the analysis to intact forest landscapes resulting ABG estimates derived from Saatchi and Baccini datasets get closer to the IPCC values with 13% and 1% differences respectively (underestimation). This suggests that the IPCC default values are mainly based on plots in mature forest stands. However, as tropical forests generally consist of a mixture of intact and degraded stands, the use of IPCC values may not properly reflect the reality. Finally, we propose to use the average composite of the Saatchi and Baccini datasets to produce improved alternative IPCC Tier 1 values. The values derived from such approach can easily be updated when newer and/or improved pan-tropical AGB maps will be available.
