2.3.6.2. Duration of Sequestration

Carbon sequestration in forest and other types of land cover is potentially reversible because carbon contained in terrestrial ecosystems is vulnerable to disturbances such as wildfires or pest outbreaks, as well as subsequent changes in management that would return some or all of the sequestered carbon to the atmosphere in addition to what would have been released if the sequestration activity had never taken place.³ This situation contrasts with the case of avoided fossil fuel emissions because fossil fuels left in the ground in a given year will not be accidentally released in a subsequent year, even if the emission reduction activity itself is of a limited duration. For example, suppose that a homeowner replaces an incandescent bulb with a compact fluorescent bulb, avoiding one ton of emissions over the life of the compact fluorescent. This benefit is not reversed even if an incandescent bulb is installed at the end of the fluorescent's useful life.

It is important to understand that the logic behind considering either fossil fuel or forest as "permanent" is not based on the assumption that specific atoms of carbon will remain in the ground or in the forest forever. Instead, the effect of delaying for 1 year a given amount of fossil fuel burning or a given amount of deforestation will be to delay the release of carbon from the barrels of oil that would be burned or hectares of forest that would be deforested in subsequent years. To the extent that the emission displacement propagates forward until the end of the time horizon, the result is a "permanent" saving. Suppose that each ton of carbon is labeled ton1 through tonn, and a mitigation project avoids 1 t of emissions in year 1. As Table 2-4 illustrates, ton1 is emitted in year 1 in the baseline scenario but is emitted in year 2 in the mitigation scenario. By the end of the time horizon (year n), n tons have been burned in the baseline scenario, but only n-1 tons have been burned in the mitigation scenario. The savings would not be permanent, on the other hand, if avoiding 1 t of emissions in year 1 results in additional emissions in any subsequent year before the end of the time horizon because of resource exhaustion or price feedbacks.

The potential reversibility of biotic carbon sequestration implies that substituting credit for LULUCF activities for reductions in fossil fuel emissions carries a risk of increasing long-term atmospheric CO2 concentrations (Lashof and Hare, 1999). Fearnside (1999b) has argued, however, that this reasoning applies only to silvicultural plantations, and within the category of plantations it applies only to their role in carbon sequestration (as distinct from fossil fuel substitution). Lashof and Hare's argument is that, by allowing countries to emit more carbon from fossil stocks into the active carbon pool (biosphere + atmosphere), the increases in biotic carbon stocks that have been encouraged under the Kyoto Protocol as carbon offsets have a risk of subsequent release into the atmosphere through natural changes (which fossil carbon stocks do not have), and reduce the options available for future responses in the forest sector because the capacity of these options to absorb carbon will have been saturated.

Fearnside (1999b) has argued that, in the case of avoiding tropical deforestation, the result is more like reducing fossil fuel carbon emissions than it is like carbon sequestration in plantations. Carbon stocks in areas of high-biomass old-growth forest, such as those in the moist tropics, are very unlikely to be allowed to regenerate to their present levels if these forests are cut down. Therefore, some of the carbon released by deforesting these areas is just as permanent an addition to what might be called the "most active carbon pool" (i.e., atmospheric carbon plus carbon in rapidly cycling stocks such as plantation biomass) as is release of fossil carbon. Activities by Annex I countries under the Kyoto Protocol to offset fossil fuel carbon emissions by helping tropical forest countries avoid deforestation keeps carbon out of this "most active pool" in the same way that avoiding fossil carbon emissions would and thus avoids carbon releases that would be just as irreversible as fossil fuel combustion (Fearnside, 1999b). These assumptions can break down if the area of remaining forest is small enough that it could be exhausted within the time horizon under consideration. If a country runs out of forest (or accessible or unprotected forest) within the time horizon, no net reduction in the atmospheric CO2 concentration would accrue, although there may still be a benefit from the delayed emissions (see Section 2.3.6.3).

Concern regarding the duration of sequestration arises from the fact that forests are subject to degradation or destruction by forces such as climate change, extreme weather events under current climate regimes, insect outbreaks, diseases, and the entry of loggers or deforesters. These events can release all or part of the carbon contained in the affected forests. Table 2-5 shows the effect on carbon benefits in a hypothetical deforestation avoidance case. Here the hectares of forest cleared each year lose biomass from generalized degradation of the forest. Although a full hectare is gained by the cascading effect (as in the fossil fuel case in Table 2-4), the displaced hectare also degrades, proportionally reducing the benefit. In the example in Table 2-5, the total emission in the baseline scenario is 2.8 t C; in the mitigation scenario the total emission is 2.4 t C. If no degradation had taken place, the totals would have been 4 and 3 t C, respectively.

One approach to accounting for net impacts on the atmosphere is to treat removals and emissions as separate events. In this case, a full 1 t of credit would be awarded for each ton of carbon emissions avoided or removed in the year that it occurs; there would be an ongoing liability to account for any subsequent release of that carbon, however. The separate-events approach would require special attention in the context of the Kyoto Protocol: Any credit for carbon removed from the atmosphere by, for example, reforestation under Article 3.3 would have to be balanced by accounting for subsequent emission of that carbon regardless of the cause of the emission. In other words, once land enters the accounting system, full carbon accounting would have to be applied, even if it were to mean accounting for activities (e.g., forest degradation) that might not otherwise be included in the accounting system. A similar continuing liability would be required for project-based activities if unaccounted releases to the atmosphere are to be avoided. If full credit is awarded when removals occur, credits would have to be retired for any subsequent release of this carbon, even if it occurs after the end of the formal project period in a country that does not otherwise have quantified emission limits.

IPCC Homepage