The Space Shuttle's Impact on the Stratosphere

Launch of spacecraft using solid rocket motors leads to release of gaseous and particulate matter in the stratosphere. Concem over these emissions, particularly chlorine, goes back to the Climatic Impact Assessment Program (Hoshizaki, 1975). The buildup of these exhaust products and their perturbation to stratospheric ozone is followed with two- and three-dimensional atmospheric chemical transport models. Chlorine enhancements due to the current rate of shuttle launches is small, on average less than 0.6% above the current background. Other gases emitted from the solid rockets appear to have even smaller global effects, although the impact of particulate alumina remains uncertain. stratosphere. made with two two-dimensional models below. Local of


INTRODUCTION
The launch of NASA's Space Shuttle and similar rockets injects chlorine compounds directly into the stratosphere, adding to the current burden of stratospheric chlorine. Depletion of the stratospheric ozone layer has been linked to increases in stratospheric chlorine compounds associated predominantly with chlorofluorocarbons (see recent assessments: World Meteorological Organization (WMO) [1986; 1990]; Watson et al., [1988]). The purpose of this study is to determine the magnitude of the chlorine increases that might be caused by the Space Shuttle and to assess the overall impact on the chemistry and composition of the global stratosphere.
The solid rocket motors on the Space Shuttle and Titan IV launch vehicles use a solid fuel composed of ammonium perchlorate, aluminum and a polymer binder (P. D. Evanoff, Thiokol Corp., private communication, 1989). The exhaust consists primarily of gaseous HC1, carbon monoxide, water vapor, molecular nitrogen and aluminum oxide. The potential damage to the ozone layer by the Space Shuttle's solid rocket motors was recognized during the Climatic Impact Assessment Program (CIAP focussed on supersonic aircraft but also studied rocket plumes, see Hoshizaki [1975]). The last assessment of the Shuttle in terms of stratospheric ozone was more than a decade ago [Potter, 1978], and our understanding of stratospheric chemistry and modelling has evolved much since then.
The launch scenario considered here consists of nine Shuttles and six Titans per year; it is typical of the current schedule, but it may fall short of the frequency needed for major space projects. The chlorine from these launches is used as a stratospheric source of Cly (total inorganic chlorine: In one numerical simulation, Cly enhancements in the middle stratosphere two days after a January launch of the Shuttle are still expected to be clumpy, but the exhaust plume is predicted to have spread over a region about 20 ø latitude by 30 ø longitude with an average increase of about 30 ppt (parts per 1012 ) or 2% above background. One month later the Shuttle plume is well mixed, and increases in Cly are less than 4 ppt throughout the stratosphere. The buildup of chlorine from these launches approaches a steady state limit after several years: on average, Cly would increase by about 10 ppt in the middle stratosphere of the northern hemisphere, less than 0.5% above current levels. Corresponding ozone depletions are predicted to be less than 0.2% locally, with smaller perturbations to the ozone column.
The profiles and amounts of chlorine injected from the solid rocket motors are summarized in section 2. The transient response to a single launch is shown in section 3, and model predictions of the steady state accumulation of Cly are described in section 4. The overall impact on stratospheric chemistry and ozone is discussed in section 5. as HC1, which we treat as Cly in the models.

CHLORINE FROM ROCKET EXHAUST
The total amount of chlorine released into the stratosphere (above 15 km) by the solid rocket motors is 725,000 kg (0.725 kilotons) per year, and can be compared with that associated with industrial halocarbons. The chemical industry's production of halocarbons exceeds 1250 kilotons of chlorine per year [WMO, 1990]. The release of chlorine during photochemical destruction of the chlorofluorocarbons (CFCs) occurs predominantly in the stratosphere, but happens slowly, on time scales of order 100 years, over the lifetime of the gas. The estimated annual source of stratospheric chlorine from the industrial halocarbons is about 300 kilotons of chlorine per year (AER model); the remainder of the annual emissions goes into the accumulating atmospheric burden of chlorinated halocarbons (about 600 kt/yr) or is destroyed in the troposphere. Thus, the launch schedule in Table 1 would add only about 0.25% to the current stratospheric source of Cly.
These results are qualitatively similar to a conclusion reached 16 years ago [Cicerone and Stealman, 1974].
The release of chlorine from the 15 launches summarized in Table 1 was averaged over the year and put into the models as a continuous source of Cly every time step. The vertical distribution specified in Table 1 was used by the models. The latitudinal location of the two launch sites was included in all three models; but the longitudinal location could only be specified in the three-dimensional simulation.

TRANSIENT RESPONSE TO A SINGLE LAUNCH
One launch of the Space Shuttle injects a single, very large pulse of 68,000 kg of chlorine into the stratosphere. Although this amount of chlorine is inconsequential on a globally averaged scale, the greatly enhanced levels of Cly in the vicinity of the exhaust plume may lead to large ozone depletions over a spatially limited region. We examined the transient response of stratospheric Cly to a single Shuttle launch using the three-dimensional GISS model. The chlorine is released over Cape Canaveral (29øN, 80øW) by using oneninth of the annual source given in Table 1 as an instantaneous source.
Simulations were initiated on January 1 and July 1, and continued for one month. Immediately following the launch, the exhaust plume will not be completely mixed over the scales resolved by the model grid, and the calculations are intended to represent the average concentration of the resulting nonuniform distribution. Figures 1 and 2  One month after the Shuttle launch (Figures ld and 2d), the added chlorine is predicted to have spread over most of the upper stratosphere in the northern hemisphere, and the plume is expected to have mixed thoroughly. The winter stratosphere is dispersive and Cly perturbations are less than 1 ppt everywhere. In summer the exhaust products remain predominantly over mid-latitudes (20øN-50øN) with perturbations still as large as 3 ppt. The globally averaged depletion of ozone associated with a single launch should be less than that caused by the steady state buildup of chlorine, as discussed below. Local destruction of ozone in the immediate vicinity of the rocket plume could be significantly larger and is not explicitly resolved in these global models. The launch scenario assumes nine Shuttles (A) and four Titans (B) from Cape Canaveral, Florida (29øN 80øW) and two Titans (C) from Vandenberg AFB, California (34øN 121øW).

STEADY STATE ACCUMULATION
The three models used the continuous source of stratospheric chlorine from the rocket launches as described in section 2. The ultimate removal for the injected stratospheric Cly is transport into the lower atmosphere (troposphere) where most inorganic chlorine species are soluble and therefore removed rapidly by rainfall and other processes. In these models this sink was applied by imposing either rapid loss for Cly below 10 km (two-dimensional models) or a negligibly small concentration of Cly below 1 km (three-dimensional  Figures 3b-5b) peak strongly at 30øN between 30 and 50 km altitude in all three models, with monthly and zonally averaged maxima of more than 12 ppt in the GSFC and GISS models, and 9 ppt in the AER model. In order to determine the relative perturbation to Cly from the rocket launches, the two-dimensional models compared the calculated Cly enhancements (Figures 3-4)  1000 ppb. However, a 100 km 2 area comprises less than 1/1,000,000 of the mid-latitude stratosphere, and the global or chlorine into the stratosphere. This amount is small compared even regional effects of complete ozone destruction within this to the current background of stratospheric chlorine (3 ppb), corridor would be inconsequential. Furthermore, the chlorine which is generated from the photochemical destruction of is released predominantly as HC1 and would need some time industrial and natural halocarbons within the stratosphere at a to be chemically processed into more catalytically active forms This early stage following the launch is not adequately modeled in these calculations, but we believe that strict limits can be placed on the potential for global ozone destruction because the amount of chlorine injected is small compared with that contained in a 1000 km by 1000 km region (less than 0.2 ppb out of more than 2 ppb Cly). In order to have a significant impact on ozone globally, the plume must mix with the stratospheric environment and lead to significant perturbations over scales of at least 1000 x 1000 km. In the few days following a launch, stratospheric winds will have stretched and dispersed the exhaust plume to scales greater than 1000 km. The average increase in stratospheric chlorine levels over such an area is modest, at most +5% within a 20 ø latitude by 20 ø longitude area. By the end of the month, these perturbations decrease rapidly to less than 0.2% above background levels as the chlorine is mixed laterally throughout the stratosphere.
A continuous series of rocket launches will lead to a buildup of chlorine in the stratosphere whose magnitude is governed by the frequency of launches and the rate of the stratosphere-century.
The Shuttle and Titan IV solid rocket motors comprise the largest source of stratospheric chlorine that is expected from the current space fleet. The major launch vehicles from the United States and other space agencies use nonchlorinated (e.g., liquid) fuels or employ much smaller rockets. The U.S. strategic nuclear arsenal uses solid fuel containing chlorine, but in such launches most of the chlorine would be released below 15km.
The exhaust from the solid rocket motors also contains other possible stratospheric pollutants. The major gaseous components are CO (24% by wt), HC1 (21%), H20 (10%), N 2 (9%), CO 2 (4%) and H 2 (2%). The perturbation to CO should be about the same in mixing ratio as that to Cly, and would not significantly affect the background levels of order 100 ppb. Clearly, the HC1 represents the largest fractional perturbation to the background stratosphere; the remaining effluents should have negligible impact on the stratosphere.
Another principal exhaust product, particulate A1203 (30%), has the potential to perturb stratospheric chemistry. Most of the alumina is reported by Thiokol Corporation to form troposphere circulation. For a scenario of nine Shuttle and six particles of radii greater than 1 micron; these fall out of the Titan IV launches per year, the accumulation of chlorine in the stratosphere more rapidly than the gaseous products which are stratosphere is still modest, ranging from 0.2 to 0.6% over removed by the circulation exchanging air between the northern mid-latitudes and much less in the tropics and stratosphere and troposphere (modeled here). The stratospheric southern hemisphere. Corresponding ozone depletions are abundance of larger alumina particles appears to be increasing even smaller, less than 0.25% locally and less than 0.1% in the and has been attributed to solid rocket motors as well as space