IPCC Working Group I (WGI) SAR (Houghton et al., 1996) confirmed the importance of tropospheric ozone as a greenhouse gas. Ozone is produced in the troposphere in a complex chain of reactions that involve the ozone precursors nitrogen oxides (NOx), non-CH4 hydrocarbons or volatile organic compounds (NMVOCs), and CO. Therefore, it is important to explore possible future developments of emissions of these substances to analyze the evolution of tropospheric ozone levels.
NOx are released through fossil fuel combustion (24 MtN per year around 1990), natural and anthropogenic soil release (12 MtN per year), biomass burning (8 MtN per year), lightning (5 MtN per year), NH3 oxidation (3 MtN per year), aircraft (0.4 MtN/year), and transport from the stratosphere (0.1 MtN per year). These figures are mean estimates within a range; for fossil fuel combustion, aircraft emissions, and stratospheric input the ranges may be as narrow as 30%, but for natural sources the ranges may be up to a factor of 2 (Prather et al., 1995). The uncertainties in the estimates are illustrated by comparison of the detailed emissions inventory of Olivier et al. (1996) with the 1994 IPCC estimates - while global total emissions estimates by source are very similar, at the regional level emissions estimates show pronounced differences, particularly in Asia.
Fossil fuel combustion in the electric power and transport sectors is the largest source. Emissions from fossil fuel use in North America and Europe have barely increased since 1979 because fossil fuel consumption leveled off and air quality abatement was enacted, but in Asia emissions are believed to increase by 4% annually (Prather et al., 1995). As a result of the first NOx Protocol in Europe, NOx emissions in Europe had decreased from 1987 levels by 13% in 1994, but the European Union is unlikely to meet its target of the 5th European Action Plan of a 30% reduction (EEA, 1999). An important reason is that it is difficult to abate NOx emissions in the growing transport sector. Perhaps critically, there are significant differences between the characteristics of abatement of SO2 and NOx emissions. While both substances have regional acidification effects, a priority for SO2 abatement is induced by its important local health effects. Also, whereas SO2 emissions relate closely to the type of fuel, NOx emissions are more dependent on the combustion technology and conditions.
Few scenarios for NOx emissions exist beyond the studies for Europe, North America, and Japan (IS92 scenarios are a notable exception). New scenarios, such as those by Bouwman and van Vuuren (1999) and Collins et al. (1999) often still use IS92a as a "loose" baseline, with new abatement policies added as they were introduced in the OECD countries after 1992, according to current reduction plans (CRP). Collins et al. (1999) also explore a maximum feasible reduction scenario, in which European NOx emissions decrease by 60% by 2015 and North American emissions by 5%. In the related CRP scenario of Bouwman and van Vuuren (1999), NOx emissions in the developing countries are assumed to decrease also (by more than 10%) by 2015. These studies, however, should be used with care as the authors developed their somewhat arbitrary scenarios primarily for atmospheric chemistry analysis; they are not based on an in-depth analysis of the characteristics of the emissions sources and potential policies in the various regions outside the OECD.
Prather et al. (1995) estimates the total global emissions of CO at 1800 to 2700 MtC per year in the decade before 1994. The most important of the approximately 1000 TgC anthropogenic sources are technological (300 to 550 MtC per year) and biomass burning (300 to 700 MtC per year). Technological sources dominate in the northern hemisphere, and include transport, combustion, industrial processes, and refuse incineration. Biomass burning dominates in the southern hemisphere, and includes burning of agricultural waste, savanna burning, and deforestation. The detailed, geographically explicit EDGAR database (Olivier et al., 1996) has similar emissions estimates for CO. Other sources are biogenics (60 to 160 MtC per year), oceans (20 to 200 MtC per year), and oxidation of CH4 (400 to 1000 MtC per year) and of NMVOCs (200 to 600 MtC per year). To a large extent, this oxidation may be considered anthropogenic in origin because many emissions sources of CH4 and NMVOCs are of an anthropogenic nature.
Global emissions estimates of NMVOCs are also very uncertain. Prather et al. (1995) indicate a global total for anthropogenic NMVOCs of about 140 MtC per year, from road transport (25%), solvent use (14%), fuel production and distribution (13%), fuel consumption (34%), and the rest from uncontrolled burning and other sources. The EDGAR inventory by Olivier et al. (1996) suggests that global emissions may be higher (178 MtC per year) because of higher estimates of emissions from energy production and use. As with CO, emissions in the northern hemisphere are dominated by transport and industry, while in the southern hemisphere biomass and biofuel burning is often the dominant source. In Europe, emissions of NMVOCs are controlled under the 5th Environmental Action Programme of the European Union and the VOC Protocol of the UN Convention on Long-Range Transboundary Air Pollution. However, for reasons similar to those for NOx , the current reduction of 11% with respect to 1990 levels and 15% with respect to 1987 levels suggests that the planned reduction to 30% in 1999 may not be reached (EEA, 1999). As a consequence, threshold values for ozone continue to be exceeded in Europe.
No long-term global scenarios for emissions of NMVOCs and CO were identified beyond IS92, which assumes increasing emissions. The important role of biomass combustion in these emissions means that a scenario with low carbon emissions because of an increased used of biomass energy does not automatically lead to low emissions of NMVOCs and CO. Also, emissions trends are influenced significantly by assumptions as to the type of combustion or other conversion technology (e.g. gasification) deployed in the future. If biomass fuel is used in modern large power plants or boilers, or is to be converted into modern energy carriers, CO emissions will be almost negligible compared to those of traditional uses. As with sulfur, however, it seems plausible that with rising incomes, abatement of the ozone precursors may be initiated in non-OECD regions to address local and particularly regional air pollution (photochemical smog). Since control of these substances is more difficult than that of sulfur, it may not be implemented until later.
Other reports in this collection