Regarding the determination of the area of ARD land, there are fundamentally two sources of information available: reports of activities (in countries where such reports exist) and monitoring of forest/non-forest change.
The activity reporting approach assumes that a well-defined set of human activities will determine ARD land and that information about these activities will be reliably reported and compiled. In some countries, for example, conversion of forest requires some form of legal permission. Because afforestation and reforestation activities require investments, it should be possible to create incentives for reporting of activities that would benefit the national carbon balance. An activity reporting approach is best if the ARD activities are well defined and are limited to changes to and from forest (e.g., the Land Use scenario) rather than changes within forest classes. An activity reporting approach cannot meet the data requirements for scenarios that include a transition from one forest class to another (e.g., the Degradation/Aggradation scenario).
Several techniques are available for monitoring forest/non-forest change. The simplest method is to identify the total area considered forest in 1990 and at the end of the commitment period. This information might be collected through independent samples at two points in time. A net increase in forest area represents an excess of afforestation and reforestation over deforestation; a net decline represents the reverse. This technique does not follow the fate of individual forest stands. Thus, it allows deforestation to the extent that it is compensated by afforestation and reforestation. A zero net change in forest area can be achieved by deforesting and afforesting the same amount of area. Because the two activities would be expected to contain very different carbon stocks per hectare, however, the zero net change in area would not result in a zero net change in carbon stocks. Therefore, an alternative technique that avoids this problem would be to identify specifically which areas were afforested, reforested, or deforested during the time period.
ARD activities can be monitored in a variety of ways. The two predominant methods are field sampling and remote sensing. These techniques are described in more detail in Section 2.4; both involve sampling the land (with either field plots or aerial/ orbital images) at multiple points in time and subsequently identifying locations (or sample plots) where land-use or land-cover changes indicate that ARD activities have occurred. ARD activities could be identified by an assessment of forest extent in 1990 and in 2012; those areas that change from forest to non-forest or from non-forest to forest during this interval could be deemed to have undergone an ARD event. An assessment conducted only at the endpoints of this interval, however, may miss many ARD events that could result in changes in carbon stocks. For example, a 1990 forest that is deforested in 1991, only to be reforested in 2010, is in forest condition at both interval endpoints, yet it may be counted as ARD land.
Other reports in this collection