Individual scenarios are considered independent entities in the database. Clearly, in practice, individual scenarios are often related to each other and are not always developed independently. Some are simply variants of others generated for a particular purpose. Many "new" scenarios are designed to track existing benchmark scenarios. A good example is the set of IS92 scenarios, especially the "central" IS92a scenario, which was often used as a reference from which to develop other scenarios. A further consideration is that not all scenarios are created in an equal fashion. Some are the result of elaborate effort, which includes extensive reviews and revisions; others are simply the outcome of input assumptions without much significant reflection. Some are based on extensive formal models, while others are generated using simple spreadsheets or even without any formal tools at all.
Numerous factors influence future emissions paths in the scenarios. Clearly, demographic and economic developments play a crucial role in determining emissions. However, many other factors are involved also, from human resources, such as education, institutional frameworks, and lifestyles, to natural resource endowments, technologic change, and international trade. Many of these important factors are not documented in the database, and sometimes not even in the respective scenario reports and publications. Some are neither quantified in the scenarios nor explicitly assumed in a narrative form.
For this analysis, a simple scheme is used to decompose the main driving forces of GHG emissions. This scheme is based on the Kaya identity (Kaya, 1990; Yamaji et al., 1991), which gives the main emissions driving forces as multiplicative factors on one side of the identity and total CO2 (or GHG) emissions on the other. It multiplies population growth, per capita value added (i.e., per capita gross world product), energy consumption per unit value added, and emissions per unit energy on one side of the identity, and total CO2 emissions on the other side (Yamaji et al., 1991)3; it is a specific application of a frequently used approach to organize discussion of the drivers of emissions through the so-called IPAT identity that relates impacts (I) to population (P) multiplied by affluence (A) and technology (T), (see Chapter 3 for a more detailed discussion). The same approach can be used for other emissions such as SO2. However, the driving forces might be different for some species of anthropogenic emissions.
Apart from its simplicity, an advantage of analysis that uses the Kaya identity to decompose emissions into four main driving forces is that it facilitates at least some standardization in the comparison and analysis of many diverse emissions scenarios. This decomposition is very useful because it indicates where to seek differences in scenario assumptions that may account for differences in the resultant GHG emissions (Alcamo et al., 1995). However, the identity is not used here to suggest causality. An important caveat is that these driving forces are not independent of each other; in many scenarios they explicitly depend on each other. For example, scenario builders often assume that high rates of economic growth lead to high capital turnover. This favors more advanced and more efficient technologies, which result in lower energy intensities. Sometimes a weak inverse relationship is assumed between population and economic growth. Thus, the scenario ranges for these main driving forces are not (necessarily) independent of each other. (See also the discussion of the relationships between the main scenario driving forces in Chapter 3.)
In the following sections, scenario ranges are presented for each of the four factors in the Kaya identity that represent the main (energy-related) emissions driving forces: population, gross world product, energy consumption, energy intensity (energy per unit of gross world product) and carbon intensity (CO2 emissions per unit of energy). The ranges for CO2 and SO2 emissions are presented first because they represent the "dependent variable" in the Kaya identity. These are followed by scenario ranges for the other factors in the decomposition that represent the "independent variables" (main emissions scenario driving forces) in the identity. This sequence was chosen to present the main scenario driving forces because it corresponds to their representation in the Kaya identity; it does not imply a priori any causal relationships among the driving forces themselves or between the driving forces and CO2 emissions.
Four complementary methods of analysis are used:
The main findings of this scenario analysis are reported in Nakicenovic et al. (1998a).
Other reports in this collection