and there likely will never be. 3 Preindustrial concentrations of carbon dioxide, for example, were in the neighborhood of 275 parts per million volume (ppmv). They had risen to 368 ppmv by 1999. Under a variety of scenarios this concentration rises to anywhere from 500 to more than 700 ppmv over the course of the twenty-first century.4
The natural carbon cycle governs the relationship between emissions and concentrations. Anthropogenic emissions originating from net changes in land-use and fossil fuel oxidation initially enter the atmosphere but are eventually partitioned between the atmosphere and the ocean. While the oceans ultimately take up much of the net release, a fraction of any net emission remains in the atmosphere for more than a millennium. As a consequence, the preindustrial level of 275 ppmv concentration of CO2 is no longer accessible in the present millennium without reversing the net flow from fossil fuel oxidation and land-use change.
Stabilizing the concentration of CO2 in the atmosphere therefore implies that net emissions ultimately fall to zero from present levels, which are in excess of 6 petagrams of carbon per year (PgC/yr). 5 It is cumulative emissions that matter. Therefore, the particular time path of emissions will depend on the concentration to which atmospheric CO2 is limited. It also means that to satisfy the cost-effectiveness objective of the FCCC, emissions may rise before finally declining. Emissions trajectories consistent with five alternative concentration limits are shown in Figure 1.1. Key characteristics associated with emissions paths consistent with CO2 concentrations limits are shown in Table 1.1.
Understanding the key drivers of historic and future carbon emissions is critical to developing policies to control emissions. Yoichi Kaya developed a simple analytical framework to understand the relationship between population, economic activity, energy, and emissions. He observed a simple identity that has significant analytical power:
C = (C/E)*(E/Y)*(Y/N)*N.
In this equation, C = carbon emissions per year, E = energy consumption per year, Y = GNP per year, 6 and N = population. It implies that the rate of change in carbon emissions is simply the sum of the rates of change of carbon intensity (C/E), energy intensity (E/Y), per capita income (Y/N), and population growth.
By using the above equation, historical trends can be examined and future analysis dissected. Figure 1.2, Figure 1.3, Figure 1.4 through Figure 1.5 show population, GNP per capita, energy intensity, and carbon intensity for historical and forecast years. Data for forecast years are taken from Pepper et al. (1992), which docu-
3There are multiple greenhouse-related gases. These include water vapor, carbon dioxide, carbon monoxide, methane, nitrous oxide, odd nitrogen compounds, the chlorofluorocarbons and their replacements, and aerosol compounds. Carbon dioxide is the most important human-released greenhouse gas from the perspective of potential change in future climate. Its principal source of emissions is fossil fuel use, however, land-use change in general and deforestation in particular also play important roles.
4See for example IPCC (1996a) and Nakicenovik et al. (2000), although the latter gives only emissions and cumulative emissions calculations.
51 Pg = 1 billion tonnes.
6In this equation GNP refers to “gross world product.” Therefore we use GNP to mean the value of all new final goods and services produced in the world in a given year. We continue to use the acronym GNP to avoid confusion with the concept of global warming potential (GWP), which is also found in the literature.