The primary atmospheric source of soot or black carbon particles is combustion of fossil fuels and biomass burning at the Earth's surface, with total emission values near 12 Tg C yr-1 (Liousse et al., 1996). This value exceeds reasonable estimates of the aircraft source of black carbon by several orders of magnitude (Bekki, 1997). For example, aircraft are estimated to have emitted 0.0015 to 0.015 Tg C as soot into the atmosphere in 1992 [with EI(soot) of 0.01 to 0.1 g C/kg fuel] (Friedl, 1997; Rahmes et al., 1998). As in the case of sulfate aerosol, deposition and scavenging of black carbon near surface sources creates large vertical gradients in the lower atmosphere, with soot concentrations falling by 1 to 2 orders of magnitude between the surface and the lower stratosphere (Penner et al., 1992; Cooke and Wilson, 1996; Liousse et al., 1996). A possible meteoritic source of soot in the lower stratosphere has been considered but is not well quantified at present (Chuan and Woods, 1984).
Figure 3-11: Latitude and altitude distribution of annually averaged increase in soot mass density calculated with the AER 2-D model, assuming a 1992 aircraft fuel-use scenario and a soot emission index of 0.04 g C/kg fuel. Contour values are in ng m-3 (adapted from Weisenstein et al., 1997).
Few direct measurements of soot abundance are available in the upper troposphere and lower stratosphere. The most extensive measurements in these regions are from aircraft impactor measurements (Pueschel et al., 1992, 1997; Blake and Kato, 1995). The accuracy of such measurements depends on knowledge of the impactor for small soot particles. The results (Figure 3-10) are considered to represent a lower limit for soot number and mass because of size-selective sampling and because of scavenging of soot by background aerosol particles. Features of the measurements include a large gradient between the Northern and Southern Hemispheres and large variability with altitude at northern mid-latitudes. The large vertical variability of soot at northern mid-latitudes cannot be explained by the accumulation of aircraft soot emissions. With typical soot mass densities observed to be approximately 1 ng m-3, soot is estimated to represent approximately 0.01% of the stratospheric aerosol mass (Pueschel et al., 1992). In other sampling flights over southern Germany, measurements of absorbing mass (probably soot) at 8-12 km altitude and partly within cirrus clouds showed concentrations above 10 ng m-3, with higher values correlated with local aviation fuel consumption (Ström and Ohlsson, 1998).
Modeling studies of soot distribution in the upper troposphere and lower stratosphere differ on the importance of aircraft sources of soot. Some model simulations suggest that ground-level sources of soot (12 Tg C yr-1) are as important as aircraft sources in the upper troposphere and predict maximum soot values there [2 to 5 ng m-3 in Liousse et al. (1996) and 10 to 50 ng m-3 in Cooke and Wilson (1996)] that exceed observed values near 200 hPa at northern mid-latitudes (Pueschel et al., 1997). Other model results show that current aircraft could be a noticeable source of the soot near the tropopause at northern mid-latitudes (Bekki, 1997; Danilin et al., 1998; Rahmes et al., 1998). Tracer simulation results from the AER 2-D model (Section 3.5.1) were multiplied by EI(soot) of 0.04 g/kg fuel (Döpelheuer, 1997) to estimate the global distribution of soot. The results (Figure 3-11) show maximum values of 0.6 ng m-3 near 12 km at northern mid-latitudes. These values are in the middle of the range of other model results and are in the range of observed values (Figure 3-10). However, because the fleet-mean EI(soot) is uncertain and may range from 0.01 to 0.1 g/kg fuel, the effective range of the maximum in Figure 3-11 is 0.15 to 1.5 ng m-3. At 20 km, fuel tracer simulations show aircraft-induced zonal mean soot perturbations to be approximately 100 times smaller than maximum observed values (Figure 3-10).
Observations of soot in the upper troposphere and lower stratosphere are too limited to provide an estimate of any long-term changes in soot concentrations in those regions. The possible consequences of heterogeneous reactions on soot (Bekki, 1997; Lary et al., 1997) are discussed in Chapter 2.
During winter in the polar regions, low temperatures lead to the formation of polar stratospheric cloud (PSC) particles, which contain H2SO4, HNO3, and H2O (e.g., WMO, 1995; Carslaw et al., 1997; Peter, 1997). PSCs activate chlorine, leading to significant seasonal ozone losses in the lower stratosphere, particularly in the Southern Hemisphere (WMO, 1995). PSC formation may be enhanced by the atmospheric accumulation of aircraft emissions of NOx, H2O, and sulfate, as well as through direct formation in aircraft plumes in polar regions (Section 3.2 and Chapter 4). If aircraft emissions change the frequency, abundance, or composition of PSCs, the associated ozone loss may also be modified (Peter et al., 1991; Arnold et al., 1992; Considine et al., 1994; Tie et al., 1996; Del Negro et al., 1997). The effects of subsonic aircraft emissions on PSCs and stratospheric ozone are expected to be smaller than those of similar emissions from supersonic aircraft because subsonic emissions occur in the 10- to 12-km region, whereas supersonic emissions will most likely occur in the 15- to 20-km region. Ambient temperatures in the 10- to 12-km region are usually too high (> 200 K) for PSCs to form with available H2O and HNO3, and ozone and total inorganic chlorine concentrations are much lower than near 20 km.
The impact of the subsonic fleet on PSC formation has not been well studied. The results of the fuel tracer simulation discussed in Section 3.3.4 can be used to estimate the increase of PSC SAD as a result of aircraft emissions of H2O and NOx. Assuming an EI(H2O) of 1,230 g/kg, EI(NOx) of 15 g/kg, complete conversion of NOx to HNO3, and formation of NAT particles at threshold temperatures, AER model results for a 1992 subsonic fleet show an additional condensation of HNO3 on NAT particles ranging from 0.02 ppbv at 60°N to 0.12 ppbv at 85°N at 20 km in January. These values provide an increase of 0.08 mm2 cm-3 in PSC SAD at 20 km and 85°N, assuming a unimodal distribution of PSC particles with diameter of 1 mm. The increase in spatial extent of PSCs both vertically and latitudinally is small in the model. PSC increases are very sensitive to background temperature, H2O, and HNO3 values and will differ considerably among models. The increases are not likely to significantly alter ozone changes in polar winter because the SAD increases are much less than typical values of 1-10 mm2 cm-3 calculated for PSC events, and satellite data observations show that the probability of PSC formation below 14 km in the Arctic is generally very low (< 1%) (Poole and Pitts, 1994).
An important caveat related to the assessment of additional PSC formation as a result of aircraft emissions is that plume processes are not included. Global models generally assume that aircraft emissions are homogeneously distributed in a model grid box that is much larger than an aircraft plume. The consequences of this assumption have not yet been fully evaluated. In one model study, reactions on PSCs did not affect ozone chemistry in a subsonic plume at northern mid-latitudes in April (Danilin et al., 1994). A further caveat is that estimated PSC changes from aircraft emissions have not accounted for projected cooling of the stratosphere, which may enhance PSC formation.
The chemical implications of increased PSC formation for ozone chemistry and atmospheric composition are further discussed in Chapter 2. The effects of future aircraft fleets on additional PSC formation and subsequent ozone response are presented in Chapter 4.
Other reports in this collection