2.2.4. A Schemata for Forest Definitions

In the context of the foregoing discussion, any set of definitions of the critical terms forest and ARD clearly introduces its own set of issues and questions. It is instructive to look systematically at the options on the basis of two primary defining criteria for forest: land use and a measure of land cover or biomass (Figure 2-2). A definition of forest could require that a patch of land:

1. Meet both requirements (i.e., current land use is forestry and land cover or biomass exceeds a threshold, here labeled the "LU+LC" option)
2. Require only that either one or the other of the criteria is met ("LU or LC")
3. Require only one or the other of the requirements ("LU" or "LC").

Given one of these definitions for forest, deforestation can then be broadly defined as the conversion of a forest to a non-forested state, and reforestation (and afforestation) can be defined as the conversion of a non-forested area to a forest. The scenarios in Chapter 3 analyze the implications of these definitions. The scenarios also examine the implications of treating the harvest/ regeneration cycle of operational forestry as deforestation followed by reforestation or as reforestation alone. Under such an interpretation (shown as D2 and R2 in Figure 2-2), deforestation and reforestation are regarded as activities that result in land change (possibly of short duration) but without reference to subsequent land-use change.

Figure 2-2: Alternative definitions of afforestation (A), reforestation (R), and deforestation (D) in terms of land cover and/or land use. Land-cover indicator (e.g., canopy cover) is plotted against time. In each panel, the dots represent the A, R, or D activity.

The possible states of a patch of land can be formalized as shown in Table 2-2. The rows indicate the land-use criteria, and the columns are land-cover criteria. Four cases are distinguished. The two diagonal elements (1 = forest, 4 = non-forest) are unambiguous classifications under all criteria.

The status of lands falling in the off-diagonal categories (2 and 3) depends on definition chosen. Neither category 2 nor 3 would be included under the "LU+LC" definition. Both categories would be included with the "LU or LC" definition. Category 3 would be captured by an "LC" definition, and 2 by an "LU" definition.

Areas that meet the land-cover but not the land-use criterion (category 3) include significant carbon pools. This circumstance is especially true for sparsely treed grazing lands (savanna and woodlands) that may have canopy covers of less than 20 percent. If the threshold for canopy cover is set too high to include these lands, many sparsely treed areas will fall outside the forest definition (appearing instead under category 4).

Areas that fulfill the land-use but not the land-cover criterion (category 2) include service and recreation areas within a broader forested estate. Most national definitions of forest include these areas, with adjustments being made for their extent in calculating forest production statistics. In some nations, the legally defined forest estate includes extensive areas that are not currently forested. The carbon content of such areas will usually be relatively low, and they are likely to have little significance to the Protocol. If these areas were to be accepted as forests under the Protocol, however, devising effective definitions for deforestation and reforestation would be difficult.

A significant issue in finding a simple, unambiguous definition of a forest is the treatment of young stands that are regenerating after disturbance or harvesting. Such stands may not have reached the canopy cover or carbon threshold but may be expected to do so if allowed to continue their growth.
In practice, distinguishing such stands from similarly affected lands that cannot-or will not be permitted-to regenerate will be difficult. Delaying recognition of such land as forest until crown cover develops could create accounting complications. These problems would be greatest in stands that are disturbed or harvested near the end of a commitment period: Such cases may require a retrospective analysis to be performed, with no guarantee that the necessary data were collected.

The complications are exacerbated by the wide variation in regeneration rates among different forest vegetation types. A regenerating boreal forest stand may require decades for its canopy cover to reach a definitional threshold. Again, specific amendments to definitions can be developed (see Chapter 3).

IPCC Homepage