Emissions of carbon dioxide and other greenhouse gases from human activity are increasing the concentrations of these gases in the atmosphere. The Earth is expected to warm as a result, with consequences that are potentially highly disruptive to human societies. Reductions in the use of fossil fuels and in rates of deforestation worldwide will reduce emissions of CO2, but atmospheric concentrations will continue to increase unless emissions are reduced by more than 60% (about 4.5 billion tons of carbon annually). Reforestation seems to offer one of the few means for reducing the atmospheric concentration of CO2 over periods as short as human generations. We report here an approach for evaluating the potential for reforestation to help stabilize or even reduce the concentration of CO2 in the atmosphere. Reforestation is defined broadly to include tree plantations, natural regrowth of secondary forests, and the practice of agroforestry. Our premise is that human use of the land has generally reduced woody biomass and that such lands have a potential for reaccumulating carbon if appropriately managed. We used published ground studies together with global vegetation index data from the NOAA 7 satellite to estimate current land cover in tropical regions.

Then, superimposing this map of current land cover over maps depicting the distribution of vegetation cover prior to human disturbance, we obtained an estimate of about 3200¿106 ha in the tropics (almost 60% of the total land area considered) where woody biomass had been decreased, and where carbon might again be sequestered. We calculated the amount of carbon that could be withdrawn from the atmosphere and stored in woody biomass if several management options were implemented. Biomass accumulations were determined from forestry statistics. Application of the data on biomass to the areas suitable for accumulation of carbon yielded an estimate of potential accumulation of 160--170 Pg carbon, an amount equivalent to the accumulation of carbon in the atmosphere since the start of the industrial revolution, or to about 25 years of fossil fuel emissions at current rates. Estimates of both area and potential accumulation of carbon were crude, probably not better than ¿50%. They are useful for suggesting the role that tropical lands might play in stabilizing atmospheric concentrations of CO2, but they should not be used to suggest specific management options in individual countries. As maps with higher spatial resolution become available, however, the method should provide more precise estimates overall and in specific locations.