5.5.2. Sulfur

Aerosols result from complex atmospheric processes in which sulfur emissions play an important role. Besides sulfur, other substances, like NO_x, ammonia, and small particles from the burning of fossil fuels and biomass, are involved in these processes.

Global anthropogenic sulfur emissions are estimated to range between 65 and 90 MtS in 1990 (Houghton et al., 1995; Benkovitz et al., 1996; Olivier et al., 1996; WMO, 1997). Reviews of most recent inventories, given in Smith et al. (2000) and Grübler (1998), indicate a most likely value of 75 ± 10 MtS. These reviews draw on a large body of literature sources and sulfur inventories (in particular the EMEP and CORINAIR inventories for Europe, NAPAP for North America, and the most recent inventories available for Asia, including Akimoto and Narita (1994), Foell et al. (1995), and Kato (1996)). Anthropogenic emissions add to natural sulfur flows, which are estimated to range between 4 and 45 MtS (Houghton et al., 1995). Pepper et al. (1992) adopted an intermediary, constant natural sulfur flux of 22 MtS for the IS92 scenario series.

Even with a comparatively good agreement on global sulfur emission levels, important uncertainties remain at the sector and regional levels (discussed in more detail in Alcamo et al. (1997) and Grübler (1998)). The main sources of uncertainties are a lack of detailed inventory data (especially for developing countries outside Asia, but also for the non-European part of Russia), uncertainties in sulfur contents of fuels (especially coal) in many regions, and the use of different base years for development of sulfur inventories. For instance, inventories and scenario studies for China and Centrally Planned Asia give a range of sulfur emissions that differ by more than a factor of two (8.4 to 18 MtS) for the year 1990 (Grübler, 1998).

Base-year differences in the available data sources are especially important because regional sulfur emissions trends have changed drastically in the past decade. Although they decreased strongly in Europe and North America as a result of sulfur control policies, they increased rapidly in Asia with growing energy demand and coal use. For instance, between 1980 and 1995 sulfur emissions declined by 59% in Western Europe and Russia (albeit for entirely different reasons - environmental policy limiting sulfur emissions in Western Europe versus a massive economic depression in Russia), by 37% in Eastern Europe, and by 36% in North America (ECE, 1997). Conversely, emissions in China rose rapidly, from an estimated 6.6 MtS in 1985 to 9.1 MtS in 1994, or by 38% (Sinton, 1996; Dadi et al., 1998). These diverging emission trends and their rapid changes also require a continuous updating of available gridded sulfur emission inventories (e.g., Dignon and Hameed, 1989; Spiro et al., 1992; Benkovitz et al., 1996; Olivier et al., 1996) that in, some instances, still rely on outdated 1980 emissions data.

Global base-year (1990) sulfur emission values from the SRES models range from 63 to 77 MtS, with the addition of 3 MtS from international shipping⁹. This difference reflects the existing uncertainty in sulfur emission estimates, particularly at the regional level. The range is within the range of values given by global inventories.

Model differences at the regional level are even larger, which reflects the greater uncertainty of emission inventories at this level, particularly outside the OECD countries. To standardize sulfur emissions, the number of SRES reporting regions was increased to six regions, by splitting Latin America from the ALM region and Centrally Planned Asia and China from ASIA. Important differences in economic development status and resource endowments lead to different patterns of sulfur emissions across all SRES scenarios. Regional emissions were standardized (see Box 5-1) and then aggregated to the global level. Global standardized base-year emissions for 1990 for the SRES scenarios are equal to 70.9 MtS, in line with the literature range of global emission inventories given above. The regional sulfur emission profiles were also used to generate spatially gridded emission patterns (see Section 5.6.2 below).

Concerning future emissions of sulfur, the SRES scenarios reflect recent literature and trends of sulfur control scenarios, as well as the conclusions from the 1994 evaluation of the IS92 scenarios (Alcamo et al., 1995). Despite considerable scenario variability, all scenarios portray similar emission dynamics - at various future dates (between 2020-2030 and 2070, depending on the scenario and its underlying storyline), global SO₂ emissions reach a maximum level and decline thereafter (Figure 5-12). By 2030 sulfur emissions range between 40 and 160 MtS¹⁰, by 2070 between 20 and 165 MtS, and by 2100 between 10 and 95 MtS. Emission trajectories of the SRES scenarios reflect a combined impact of different scenario driving forces (local air quality concerns, structural change in energy supply and end-use, etc.), which lead to a gradual decline in sulfur emissions in the second half of the 21^st century.

Importantly, all SRES scenarios are sulfur-control scenarios only and do not assume any additional climate policy measures. There is, however, an indirect effect of GHG emission reduction from sulfur-control policies that result in energy conservation and inter-fuel substitution from high sulfur to low sulfur fuels (e.g., from coal to gas).

5.5.2.1. A1 Scenario Family

The A1 family of scenarios covers most of the range of the 40 SRES scenarios (Figure 5-13a). The A1B-AIM marker is in the middle of the range and its trajectory is similar to those of many other scenarios - a rapid increase in the near term followed by a decline. Increasing fossil fuel use in developing countries combined with low levels of SO₂ controls produces the near-term increase in emissions. After 2025, per capita incomes reach levels at which countries place more emphasis on the environment, resulting in emission controls on SO₂ (see Section 4.4.10). These controls, combined with a transition from fossil fuels to non-fossil energy, result in declining emissions. More detailed information as to how SO₂ emissions are treated in the A1B-AIM scenario is provided in Box 5-3.

Another A1 family scenario, A1G-MiniCAM, shows similar behavior for the same reasons, although the increase and subsequent decline in emissions is less pronounced. The MiniCAM model has a higher level of SO₂ controls, but also a greater fossil fuel use, resulting in slightly higher SO₂ emissions in 2100.

The A1T, A1G, and A1C scenario groups assume various roles for the development of energy resources and technologies (the coal-intensive A1C scenario group, the oil- and gas-intensive A1G group, and the A1T scenario group with accelerated non-fossil technology diffusion). Consequently, they span a very wide range of future sulfur emissions, ranging from 20 MtS by 2100 in the A1T-MESSAGE scenario to 83 MtS in the A1C-AIM scenario.

IPCC Homepage