For CO2 emissions from fossil fuel combustion and most other categories of emissions from Annex I countries, the Kyoto Protocol would require that Parties reduce emissions by some specified amount below emissions levels in 1990. (We note for completeness that some countries are in fact permitted an increase in emissions and that, for some countries and some GHGs, an alternate reference year is permitted.) In application to Article 3.4, this requirement would suggest that the carbon stock change during the commitment period be compared with the carbon stock change during the base period (1990).
In mathematical terms, the change in stocks over time is equal to net emissions, so net emissions are the first derivative of the stocks (the first derivative of stocks gives the rate at which stocks are changing). The change in net emissions over time is the second derivative of stocks, which indicates the rate at which net emissions are changing. The opening line of Article 3.3 suggests that, in fact, the second derivative of stocks is of interest; this phrase conflicts with later portions of Article 3.3, however, and has now been officially rendered meaningless by a decision at COP4 interpreting Article 3.3 (UNFCCC, 1998b). Nonetheless, some observers believe that our interest should be in the second derivative of stocks. This conflict has been referred to as the "gross-net disparity" (IGBP, 1998).
For Parties to whom Article 3.7 applies (i.e., Parties for whom land-use change and forestry constituted a net source of GHG emissions in 1990), aggregate emissions and sinks from land-use change in 1990 are to be included in the 1990 reference. Subtraction of base-year emissions from commitment-year emissions then yields a true measure of the change in net GHG emissions from the base period to the commitment period if the same list of activities is included in both calculations-limited by the "since 1990" stipulation. If the change in carbon stocks in 1990 (or other base year) is not taken as the reference for Article 3.4 activities, and if some of the activities brought in under Article 3.4 were significant sources of GHG emissions in the base year, a qualification such as that in Article 3.7 might be considered. If some of these Article 3.4 activities are defined as land-use changes, Article 3.7 may be interpreted to apply.
One of the procedural problems with this approach has to do with year-to-year variability in the natural part of the global carbon cycle, particularly in the terrestrial biosphere (see Chapter 1). Another difficulty is that retrospectively determining the rate of change in some stocks may not be technically feasible-restricting this approach to application only in the future. Although this approach was rejected in Kyoto, it would most nearly treat emissions from the biosphere in the same way that emissions from fossil fuel burning are treated, while retaining the linkage to carbon stocks.
Another issue that has been raised with regard to this approach is the possibility that best efforts could still result in a net rate of removals that shrinks with time. As an activity continues over time, it could use all of the land available (conservation tillage, for example, might become adopted on virtually all cropland) and/or begin to fill up the ecologically available carbon stocks (a soil achieves a general equilibrium at a certain SOC level, or a large region of forests matures and growth rates decline); thus, the rate of change would begin to decline compared to an earlier period. This issue may be more immediately germane to project-level accounting-which is more likely to contain a few soil situations or ecosystem successional phases-but it could become important at the Party level if aggressive implementation of effective activities were continued for several decades. The result is that a Party with a declining rate of carbon sequestration would appear to be a source from an accounting standpoint even though it was a net carbon sink.
Other reports in this collection