The global forest C pool of tropical forests is about 990 Pg (Pg = 1015g), with about a third in above and below ground live and dead vegetation and two-thirds in soil (O horizon and mineral soil to 1 m depth). Tropical forests contain over 60% of the world's forest C pool in vegetation, but only 33% of the soil C pool. In contrast, boreal forests contain about 18% of the world's vegetation C pool but 53% of the soil C pool. However, large uncertainties exist in these estimates caused by uncertainties in the extent of forests and the proportions that are mature, degraded, or disturbed, and the C densities of the corresponding forest vegetation and soil.
The current estimate of the net C flux resulting from tropical deforestation is a source of 1.6 + 0.4 Pg C/yr, and a sink of 0.7 + 0.2 Pg C/yr from regrowth and expansion of boreal and temperate forests. The uncertainty around these C fluxes is also high, caused not only by the uncertainty associated with biomass and soil C densities, but also by the uncertainty in the rates of deforestation, degradation, disturbance, expansion, and regrowth. This approach matches, in a GIS, maps of the spatial distribution of changes in land cover and land use at the subnational scale (from FAO) with similar ones of biomass and soil C densities.
The global estimate of the cumulative amount of C that could be conserved and sequestered by forestation, slowing deforestation, and natural and assisted regeneration is about 6087 Pg over the period 19952050. This is equivalent to about 1215% of the projected total fossil fuel emissions over the same period. The greatest potential exists in the tropics (80%), followed by the temperate zone (17%), and the boreal zone (35). Management practices presented for C conservation and sequestration are compatible with other forestry objectives. The cumulative cost for such a program could range from US$250300 billion, with an average unit cost of about US$45/Mg C.
A system to monitor or track the C flux resulting from changes in the cover, use, and management of the world's forests must be capable of (1) distinguishing between state and change in land cover (LC - the physical appearance of the land surface), land use (LU - the purpose for which land is used by humans), and land management (LM - how a given land use is managed); (2) accounting for net C accumulations (sinks) or losses (sources) in vegetation and soils brought about change in LC, LU, and LM; (3) estimating errors and range of variation in the C flux; (4) predicting the C flux into the near term future, and (5) monitoring for compliance with regulatory targets and policy actions under the UN Framework Convention on Climate Change. The scientific basis for developing such a system is well understood. Its implementation is hampered by a lack of data on the key components for many tropical countries, technical and political problems in data acquisition, and a lack of understanding of the socioeconomic forces that drive change in LC/LU/LM.
Key words: carbon cycle, tropical forests, mitigation.
Correspondence: Sandra Brown, US Environmental Protection Agency & Dept. of Forestry, Univ. of Illinois, 200 SW 35th St., Corvallis, OR 97333, USA
Telefax: +1-503-7544799
E-mail: british@mail.cor.epa.gov