M

Chlorine and bromine nitrate serve as temporary reservoirs for chlorine and bromine, taking them out of their ozone destruction cycles. (While theoretical studies suggest that other forms of bromine such as 02Br0N02 could in principle also act as reservoirs (Lee et al., 1999a, 1999b), there is no evidence at the present time that these are important under atmospheric conditions.)

NO, also interacts with the HO, cycle via the reaction of HOz with NO:

This reduces the H02 concentration, and hence its contribution to 03 destruction via reaction (8) of H02 with 03. In addition, the rates of reaction of H02 with CIO and BrO decrease with less HOz, which lowers the rate of 03 destruction through cycles outlined later.

N02 reacts with OH to form nitric acid:

(See Chapter 7.E for a discussion of the kinetics and Brown et al. (1999) for a recent study under low-temperature and low-pressure conditions representative of the stratosphere.) This reaction removes NO. from the ozone destruction cycle since HN03 does not readily regenerate active forms of NO, and is removed by transport to the troposphere followed by rainout and washout. Photolysis of HN03 to OH + NOz is

FIGURE 12.8 (a) Rates of removal of O, at 38°N in May 1993 due to NO,, (CIO, + BrO,), and HO, chemistry, respectively, as a function of altitude in the stratosphere (adapted from Wennberg et al., 1994); (b) 24-h average rates of removal of 03 as a function of altitude (adapted from Osterman et al., 1997).

Rate of removal (% per month)

03 Loss rate (molecule cm"3 s"1)

FIGURE 12.8 (a) Rates of removal of O, at 38°N in May 1993 due to NO,, (CIO, + BrO,), and HO, chemistry, respectively, as a function of altitude in the stratosphere (adapted from Wennberg et al., 1994); (b) 24-h average rates of removal of 03 as a function of altitude (adapted from Osterman et al., 1997).

Rate of removal (% per month)

relatively slow, although some model calculations find some sensitivity of the predicted ozone destruction to the HN03 photolysis rate (Jones et al., 1993).

As a result, increasing NOx emissions does not have a significant direct effect at lower altitudes as it does at higher ones but rather has indirect effects on the halogen and HOx cycles, which reduce the ozone destruction due to these species. The net result, then, is interference in these other ozone-destroying cycles, leading to an increase in ozone at these altitudes as seen in the model predictions in Fig. 12.7. (In the very low stratosphere, NOx can also produce 03 through the VOC-NOj chemistry discussed in Chapter 6.)

Figure 12.8b shows the contribution of the various cycles to 03 loss over a larger range of altitudes, deduced from a combination of measurements of OH, H02, CIO, and N02 and calculations using these measured concentrations (Osterman et al., 1997). ft can be seen that the NOx cycles dominate at altitudes from ~25 to 38 km, with HOx and, to a lesser extent, Ox (O + 03) and C10x being important above that. The NOx catalytic ozone destruction cycle has been proposed to be responsible for the low 03 concentrations observed at high latitudes in the Northern Hemisphere from 20 to 3f km during the summer (Brühl et al., 1998). The importance of various cycles at altitudes from 20 to 65 km and the importance of the chain length in determining these are discussed in detail by Lary (1997).

HSCTs also emit particles and S02, with the latter being oxidized to H2S04 and sulfate particles. Measurements of particle concentrations in the plume of the Concorde SST showed much larger particle concentrations than anticipated (Fahey et al., 1995a). Furthermore, a much larger portion of the S02 in the exhaust was oxidized to H2S04 particles than expected based on the OH levels measured in the exhaust plume (Hanisco et al., 1997), suggesting that there are some as yet unknown mechanisms of S02 oxidation in the plume.

Kärcher et al. (1996) suggest that this additional oxidation occurs on the soot particles that have been observed in the stratosphere and attributed to aircraft emissions (Pueschel et al., f992a; Blake and Kato, 1995). The oxidation of S02 on soot particles is known to occur in the troposphere as well (see Chapter 8.C.4). If the same is true of the exhaust from HSCTs, their emissions could lead to significant increases in both the number of particles in the lower stratosphere and as their associated surface area.

For example, modeling studies by Weisenstein et al. (1996) predict that the surface area of stratospheric particles could increase by as much as 75%, if 10% of the S02 is rapidly converted to H2S04. Yu and Turco

(1997) propose that the presence of ions in the exhaust may promote the nucleation of sulfuric acid aerosols if ~ 20-30% of the sulfur has been oxidized to H2S04.

Another contributing factor may be that more highly oxidized sulfur-containing compounds than S02 are formed in the aircraft engines and emitted in significant amounts. For example, modeling studies (R. C. Brown et al., f996a, 1996b; Karcher and Fahey, 1997) suggest that if a significant portion (~35%) of the total sulfur emissions are in the form of S03, it would form H2S04 through the rapid reaction with gaseous water (see Chapter 8). Condensation of H2S04 to form new particles or on existing soot particles could act to further catalyze S02 oxidation.

Such catalysis on particles is potentially important since heterogeneous reactions on such particles are now recognized as playing a key role in the chemistry of the stratosphere. A number of nitrogen-containing species such as N205, C10N02, Br0N02, and H02N02 are now known to react on particles to form nitric acid, which ties up the NOx and ultimately removes it from the stratosphere. Hence particles act to remove oxides of nitrogen from the stratosphere, lowering the predicted effects of HSCT NOx emissions compared to the gas-phase only case (e.g., see Pitari et al., 1993; Weisenstein et al., 1993; Bekki and Pyle, 1993; and Considine et al., 1995). However, the same chemistry leads to increased 03 destruction by halogens and HOx. Randeniya et al. (1996a, 1996b), for example, suggest based on modeling calculations that Br0N02 hydrolysis may be particularly important since it occurs during the day. In contrast, N2Os formation occurs at night (because of the rapid photolysis of its precursor N03 during the day) and its removal by hydrolysis can be limited ultimately by its rate of formation, which involves the relatively slow 03-N02 reaction. As discussed in detail in Section D, Br0N02 hydrolysis generates HOBr, which rapidly photolyzes to OH + Br, thus increasing HOx while tying up NOx as HN03 in the particles.

The direct destruction of 03 by its reaction on soot particles generated by aircraft at midlatitudes has also been proposed, but given the large uncertainties in the mechanism and kinetics of this reaction, it is not clear that this will prove to be significant (Bekki, 1997; Lary et al., 1997).

HSCT emissions may also interact with polar stratospheric clouds, PSCs, in much the same way as with particles (Pitari et al., 1993). That is, reaction of a number of nitrogenous species on PSCs leads to the formation of HN03, which can remain adsorbed on or in the PSC. The larger cloud particles sediment to lower altitudes in the stratosphere, redistributing NO , or into the troposphere, permanently removing NOx from the stratosphere (so-called denitrification; Fahey et al., 1990). An additional issue with respect to PSCs is the addition of water vapor from the HSCT exhaust, as well as NOx. Peter et al. (1991) estimate that the emissions of NOx and H20 from HSCTs could lead to a doubling in the occurrence of Type I PSCs and as much as an order of magnitude increase in the occurrence of Type II PSCs (see later for a description of Type I versus Type II PSCs).

There are several important points with respect to the effects of any future HSCT emissions. First, ozone concentrations at a particular location and time depend not only on the local chemistry but on transport processes as well. In the lower stratosphere, transport processes occur on time scales comparable to the rates of ozone formation and loss so that taking into account such transport is particularly important. However, in the middle and upper stratosphere, production and removal of 03 are much faster than transport so that a steady state exists between these two processes.

Second, as already discussed, recent advances in understanding stratosphere-troposphere exchange suggest that even though HSCT emissions would occur primarily in the 20-km region, there are mechanisms for transporting these into the middle and upper stratosphere in the tropics, leading to more ozone destruction than otherwise might have been anticipated. For example, one 3-D model simulation predicts that about 15-25% of the exhaust released in the region from 30°N toward the pole is transported into the tropics (Weaver et al., 1996). However, field measurements of NO>( and 03 and their ratio in the lower stratosphere show a steep gradient near the tropics. This is expected if the "tropical pipe" model discussed earlier applies, where there is a barrier that reduces the exchange between the tropics and midlatitudes (Fahey et al., 1996).

Third, there are uncertainties in the actual emissions estimates. While the few in-flight measurements of NOx emissions that have been carried out on subsonic and supersonic aircraft to date show NO, emissions in good agreement with those predicted from ground-based tests (e.g., see Zheng et al., 1994; Schulte and Schlager, 1996; and Fahey et al., 1995a, 1995b), there are significant uncertainties in other emissions, particularly the particles as discussed earlier.

Figure 12.9 shows some model-calculated percent changes in total column ozone due to a HSCT fleet that was projected in 2015 assuming the emission goal of EINOi = 5 g of N02/kg of fuel was met (Stolarski et al., 1995). These calculations compare the change in 03 due to this fleet compared to a completely subsonic fleet in that year using the three different models for which predicted altitude changes were shown in Fig.

0 0

Post a comment