Effect of Water Vapor on the Destruction of Ozone in the Stratosphere Perturbed by C1X or Pollutants

We describe results of a self-consistent one-dimensional coupled flow calculation for Ox, NOx, HO(cid:127), CIX, H:O, H:, CH4, H:O:, and N:O densities between 10 and 120 km. Our results agree well with observations for the normal mid-latitude atmosphere over this altitude range. We have varied CIX, NO(cid:127), and H:O independently in our model. We show that the effect of depletion of ozone by CIX is to remove ozone preferentially above 30 km and to lower the altitude of maximum ozone density. This leads to enhanced solar heating of the lower stratosphere and tropopause and suggests the possibility of an increased flux of water into the stratosphere. We show that increasing water vapor in the stratosphere greatly enhances the rate of destruction of O8 by CIX and also causes an increase in the rate of destruction of O8 in the NO(cid:127)-perturbed atmosphere.

Below 10 km we have assumed the amount of ozone present to be 0.54 X 10 •a cm-:, a quantity taken from the tables of    I  i  i  I  I I I I  I  I  I  I  I I I I  I  i  i  I i I !

RESULTS: STANDARD AND CIX-PERTURBED ATMOSPHERE
In Figure 1 we show some of the profiles of species obtained from our calculations for standard conditions and, where appropriate, for high and low values of k:8. Figure 2 shows standard ozone profiles for both values of k:8. Also shown in Figure 2 is the effect of adding 8 ppb CIX on the ozone densities. The preferential removal of ozone at high altitude is obvious. As a result of this perturbation, the altitude of the ozone maximum is decreased, leading to an increase in the solar and infrared heating rates below the ozone maximum [Ramanathan et al., 1976]. For example, when k:s is high and 15 ppb of CIX is present, the ozone maximum is lowered from 24 to 22 km, and the solar heating rate at 10, 15, and 20 km is 13%, 16%, and 19% larger, respectively. Ramanathan et al. [1976] have indicated that this kind of change in the ozone distribution might increase the temperature at the tropopause by several degrees. The saturation vapor pressure doubles for a 4øK rise in tropopause temperature and increases by an order of magnitude for a 15øK rise. Thus in the presence of CIX the water vapor concentration at the tropopause would increase [Stanford, 1973]. In the NOx-perturbed stratosphere, however, the ozone density profile is affected more uniformly [McElroy et al., 1974] and thus could lead to a decrease of water vapor concentration at the tropopause [Ramanathan et al., 1976].
To assess the effect of changing the amount of water vapor in the stratosphere on the ozone distribution, we have repeated the steady state calculations for water vapor mixing ratios at l0 km ranging from 10-: to 30 ppm. When the water vapor mixing ratio at the tropopause is lower than about 1 ppm, the water vapor present in the stratosphere will be produced mostly from methane because oxidation of methane is an important source of water vapor above the tropopause [Hunten and Strobel, 1974;Liu and Donahue, 1974a]. We show in Figure    Our results for the new rates k• and k•, compared to those for the old k• and k•, are that I ppb of CIX reduces the total amount of O8 by 0.6% compared to 2.2%, 3 ppb of CIX reduces O8 by 1.9% compared to 6.4%, and 5 ppb of CIX reduces'Os by 3.6% compared to 10.4%. These comparisons are made for the larger value of k•. In case the smaller value is more nearly correct the reduction in the integrated O= abundance for a given combination of CIX and water vapor concentrations could be much greater, as can be seen in Figures 2, 4,  Figures 4 and 5 also demonstrate the principal points that we wish to make in this paper. The effects produced by changing H,O are significant. In particular, when the CIX pollution level is I ppb, the percentage ozone reduction goes from 0.6% to 2.9% if the amount of H,O is tripled (9 ppm at 10 km) and to 9.7% if it is increased by a factor of 10 (for the high value of kl•). Again for the low value of kl• a much more dramatic effect occurs, i.e., a 6.8% reduction for I ppb of CIX and 3 times normal H,O and a 16.6% reduction for 10 times normal H,O. Of course if the stratosphere were to dry out for some reason, the changes would occur in the opposite sense.
The reason that water vapor has such a great influence is that it is converted by reaction with O(1D) into odd hydrogen in the stratosphere. All forms of odd hydrogen destroy ozone catalytically in their own right, but in addition to this direct effect there is a special indirect effect on ozone destruction.

EFFECT OF H:O ON NOx-PERTURBED STRATOSPHERE
We have also considered the effects of NO•, and H:O produced by the 'standard' SST models, at 20 km [Grobecker, 1974]. In Figure 12 we show our results for the globally averaged reduction in Oa resulting from 100 SST's emitting 9 X l0 • NO cm -: s -• between 19 and 21 km. If the stratosphere contains 3 ppm of water vapor, the reduction is 1.22% for the high value of k•s and 0.66% for the low value. This means 1.83% and 1% reduction of Oa in the hemisphere in which the NO•, is dominantly emitted, a result that agrees well with previous calculations (summarized in the publication by the National Academy of Sciences [1975]). Note, however, that adding wa-  . This exercise cannot duplicate the large effect of HO•, on Oa that we find below 24 km. Our disagreement with Crutzen [1974b] may be due to different reaction rate constants. We note that Rao-Vupputuri [1974] also finds that increased H:O leads to decreased 08.
In a sense, our result is disturbing because it suggests that the ozone layer is necessarily unstable against runaway destruction from perturbations (even natural ones that may have occurred in the past) allowing water into the stratosphere. Such a conclusion is not warranted, however, for the nature of the perturbation, the exact nature and location of the change in ozone density, the changes in temperature accompanying the perturbation, and the temperature dependences of all important rate constants must all be considered. Thus a perturbation like the one we describe here resulting from chlorocarbons, which removes the top of the layer of O8 and causes the temperature to rise below 25 km and fall above 25 km, will cause HO•, to have different effects at high and low altitudes. decreases with increasing temperature. Thus as the temperature increases at low altitude, the destruction of ozone from HO•, and from NO•, tends to be augmented, but the opposite is true at high altitude. As a consequence, the ozone will tend to recover and the opacity to grow at high altitude, leading to a decrease in temperature below. Effects such as these certainly must be taken into account in a complete time-dependent theory of the phenomena we are describing in this note.

CONCLUSION
This paper draws attention to the very large influence of water vapor in the stratosphere on the catalytic destruction of ozone, particularly by odd chlorine originating from photolysis of chlorocarbon. A change in water vapor concentration, either an increase or a decrease, could conceivably result from changes in the environment affecting the so-called cold trap for water vapor, the production of methane, or both. The destruction of ozone by pollutants is particularly sensitive to the amount of H20 present in the stratosphere in the case of odd chlorine produced by photolysis of chlorocarbon because of the reConversion of HCI to Cl by OH. The effects considered depend strongly on the value of the rate constant for the reaction of OH with HO2 producing water vapor, because it determines the amount of odd hydrogen in equilibrium with a given amount of wafj•r vapor.