next up previous
Next: Mid-Winter Chemistry Up: Polar Ozone Depletion Previous: Initial Vortex Composition

Chlorine Activation and Polar Stratospheric Clouds

Observations of OClO, a proxy for ClO, from McMurdo in 1991 [ Sanders et al., 1993] indicate that conversion of reservoir to reactive chlorine can occur in sunlit regions of the antarctic vortex early in the winter, just after the appearance of PSCs. The following year, Waters et al. [1993] observed large enhancements of ClO from UARS in the outer edge of the antarctic vortex in regions coincident with PSC temperatures. Similar abundances of ClO were observed by satellite and in situ instruments onboard the NASA high-altitude ER-2 aircraft within the lower arctic polar vortex in mid-December [ Waters et al., 1993 and Toohey et al., 1993b]. Using back trajectory analyses, Newman et al., [1993], Toohey et al. [1993b], and Schoeberl et al. [1993b] showed that these enhancements correlated well with the onset of temperatures low enough (approximately 196 K at 20 km) for PSCs composed of nitric acid trihydrate, or NAT (also called Type I PSCs). Meteorological analyses [ Newman et al. 1993] for the '91/'92 winter indicated that nearly all of the air within the lower vortex had experienced temperatures below the NAT threshold by January, but that very little of the air had dropped below ice temperatures (189 K at 20 km) at any time. ER-2-based measurements of HCl by Webster et al. [1993a] showed extensive depletions that were highly correlated with enhancements of ClO and with PSC threshold temperatures. Abundances of reactive chlorine were about twice as large as the assumed loss of HCl. The results are consistent with the laboratory findings of Abbatt and Molina [1992a,b], Hanson and Ravishankara [1992,1994], and Chu et al. [1994], that ClONO reacts rapidly with HCl on PSC-like materials to form compounds that photolyze readily to produce reactive chlorine.

Where studies of chlorine activation have answered many questions, those of aerosols have raised others about the phases and reactivities of PSCs. For example, based on laboratory work of Middlebrook et al. [1992] and Worsnop et al. [1993], it is not clear whether Type I PSCs form initially as dihydrates (nitric acid dihydrate, or NAD) or trihydrates. Particle and NO measurements from the ER-2 reported by Kawa et al. [1992a] indicate that these PSCs do form above the frost point, but at temperatures lower than NAT saturation. These conclusions seem to be at odds with the trajectory studies of ClO and HCl that implied significant enhancements at NAT threshold temperatures. Laboratory findings of Molina et al. [1993] and Beyer et al. [1994] might resolve this discrepancy. They show that cold liquid sulfuric acid absorbs nitric acid efficiently below 200 K, and that HCl and ClONO react as readily on these ternary solutions as with frozen NAT. Therefore, the rapid chlorine activation observed in the stratosphere may not require the formation of ``pure'' NAT crystals. Indeed, Pueschel et al. [1992] find the signature of nitric acid within aerosols near the NAT saturation point. Another explanation is that NAT or NAD forms on a small fraction of sulfate aerosols at saturation temperatures by selective nucleation, and that chlorine activation occurs rapidly on these few particles. As discussed by Dye et al. [1992], aerosol spectrometers might not detect significant changes in particle numbers in this case, but they might detect changes in size distributions.

These two cases imply differences for the rate of chlorine activation: in the former, activation would be a continuous function of temperature, increasing with decreasing temperature, and would occur throughout the winter for temperatures below 200 K. In the latter case, activation would be discontinuous, occurring suddenly at the NAT saturation temperature. In both cases, chlorine activation would be rapid at temperatures at or below the NAT saturation threshold. Models that use a parameterization that is a simple function of temperature seem to realistically simulate the distribution of reactive chlorine [ Douglass et al., 1993].This would also imply that chlorine activation could occur in most years so long as temperatures drop to the NAT saturation at some point during the winter [ Toohey et al., 1993b]. On the other hand, if ternary solutions of nitric acid, water, and sulfuric acid effectively process chlorine near 200 K, slow, but continuous, activation might occur in years when temperatures do not fall below the NAT threshold. In either case, it is the growth, sedimentation, and evaporation of solid particles that will determine the extent of denitrification, which in turn determines the rate of recovery of chlorine [ Schoeberl et al., 1993b] and the magnitude of subsequent ozone losses [ Salawitch et al. 1993].

There is an alternate explanation for the apparent discrepancies between the chlorine activation observations and the particle measurements. The former measurements were obtained during the winter following the eruption of Mount Pinatubo, while the latter were made before. Therefore, it is possible that the large burden of stratospheric sulfate aerosols (SSAs) helped to activate chlorine. Since extensive NAT clouds were not encountered by the ER-2 aircraft during AASE II, it is impossible to prove what aerosols were responsible for the observed activation. Based on data from the previous AASE mission, Dye et al. [1992] argued that formation of Type I PSCs might be catalyzed by the freezing of supercooled SSAs a few degees below the NAT saturation point. This freezing point may have differed for background and Pinatubo aerosols. Subsequent evaporation of NAT would leave behind frozen sulfate (sulfuric acid tetrahydrate, or SAT) that could remain a catalyst for chlorine activation well above 200 K [ Hanson and Ravishankara, 1993]. Processing could then occur continuously throughout the winter until such particles melted, at temperatures that are, at present, not well defined [ Middlebrook et al., 1993 and Zhang et al., 1993]. As suggested by Dye et al. [1992], frozen SAT might be a more effective nucleus for condensation of NAT the next time air cools, such that high supersaturation ratios would not be necessary. Thus, PSC formation might exhibit a memory effect with later events occurring at higher temperatures, such that observations at different times might yield different results. Or, the additional surface area of the Pinatubo aerosols, some of which entered the vortex in the fall [ Wilson et al., 1993], could have enhanced the rate of chlorine activation by a purely liquid sulfate process [ Hanson et al., 1994]. Finally, an important, but as yet unknown, chemistry may occur on SSAs [ O. Toon et al., 1993]. Results from 1994 ER-2 aircraft observations of PSCs over Antarctica may help to resolve this issue.



next up previous
Next: Mid-Winter Chemistry Up: Polar Ozone Depletion Previous: Initial Vortex Composition



U.S. National Report to IUGG, 1991-1994
Rev. Geophys. Vol. 33 Suppl., © 1995 American Geophysical Union