GEOPHYSICAL RESEARCH LETTERS, VOL. 17, NO. 4, PAGES 561-564, MARCH SUPPLEMENT 1990 LOSS OF OZONE IN THE ARCTIC ¾ORTEX FOR THEWINTE• OF 1989 Ross J. Salawitch 1 ,Michael B. McElroy 1, John H. Yatteau I,Steven C. Wofsy I Mark R.Schoeberl 2, Leslie R. Lair 2, Paul A.Newman 2, K. Roland Chan 3 Max Loewenstein s, James R. Podolske 3, Susan E. Strahan 3and Michael H.Proffitt 4 Abstract. Measurements of C10 (Brune et al., 1990) acquired during the Airborne ArcticStratospheric Expedition are used to 'utter concentrations of reactive chlorine (C10+2xC1202). Observed fields of potential temperature and potential vorticity are •,ed to extrapolate in situ data to largerregions of the vortex. Cal- • •a• values of the lossrateof Os, based on estimates of reactive •o.fine and measurements of BrO (Toohey et al., 1990), suggest th• theloss of O3 wasabout 12% for levels of the atmosphere •th potential temperatures between 440 and470 K overthe 39 day duration of the ER-2 flights into the polarvortex. Calculated loss •s agree withobserved rates of removal of 03, although sig- •ficant uncertainties exist for each. introduction The Airborne Arctic Stratospheric Expedition (AASE) offered dramatic evidence that concentrations of reactive chlorine (CtO+2x C1202 --CI*)and bromine (BrO) were surprisingly high /n thepolar vortexduring winter,comparable to values measured forthe antarctic vortexduring the spring of 1987 (Brune et al., 1990;, Toohey et al., 1990). The enhancement of reactive chlorine is believed to result from reactions onthesurfaces of polar stratos- •c clouds (PSCs) 0VIcElroy et al., 1986; Solomon et al., 1986), and /s maintained, in part, by permanent removal of nitrogen o•des fromthe polar stratosphere associated with sedimentation ofPSC particles (Toon et al., 1986). The boundary of the polar vortex is deemed by a steep gradient • !•otential vordcity (q)ona surface of constant potential tempera- ture (0), with a concurrent drop in the concentration of N20 (Loewenstein et al., 1990).Thepattern is depicted in Figure 1, w•ch summarizes datafrom 10 February !989. The ER-2 was wiedn the polar vortex from approximately 1!50to 1320 GMT. 'The solar zenith angle (Z) was less than 90 øfor theentire flight. Variations in C10, for constant 0, arestrongly correlated with e ,--han•s in q. Inthis, paper, we determine the dependence ofthe concentra- t• ofC1 , denoted by[CI*], on 0 and q using in situ measure- mints of CIO (Bmne eta!., !990) and a model incorporating eff•ects of diurnal changes in insolation. Concentrations of 121' for t'•arger regions of the vortex are inferred using the functional rela- tiorrship with (0,q)(Schoebefl etal.,1989) and maps ofq on isen- tropic surfaces (Newman etal.,!989). Estimates are developed for rates of 03 loss in the vortex on6 and 16January and 10February us•tr•g [C-'l*] determined in this fashion, observations of [BrO] w•thin thevortex (Toohey et al., 1990), and a photochemical model, Uncertainties in the kinetics of CIOand C!202 introduce DePamnent of Earth and Planetary Sciences and Division of pplied Sciences, Harvard University NASA Goddard Space Flight Center NASA Ames Research Center NOAA Aeronomy Laboratory and Cooperative institute for Re•..,arch inEnvironmental Sciences, University ofColorado Copyright 1990 by the American Geophysical Union. Paper number 90GL00184. 00,94-8276 / 90/ 90GL-0Ot 84 $03.00 •'•10 1.0- o 0.8 _• 0.• • 0.2 • ' ' I' l' ' i• [, , - 400 • 2.0 200 o !.0 ! i 9 10 11 12 !3 14 15 !6 TIME (G•T hrs) •i•. 1. •c•men• of QO, N•O, e, c• Un• indicate •e •d•es of •e •1• vopex •d •e w• m•on. Boodles were fo•d from exmina•on of •e nora of ß e g•iem of q on imn•pic ••es. large uncertainties in the determination of [CI'] forair sampled in thedark (Z> 90ø), aswas generally the case for theJanuary flights. Two estimates of the time integrated loss of 03 aredeveloped: one is based ondailyloss rates estimated for 6 and16 January and10 February; the other uses the best determination of [C-'l*], derived from sunlit(Z <90ø) observations on !0 February, and assumes that comparable values of CI* applied at earlier times. Time integrated loss rates arecompared with empirical estimates based on measurements of 03. The DiumaI Variation of QO The presence of elevated levels of QO in the vorteximplies that the abundance of NO2 is very small; otherwise, reaction of NO2 with C10 wouldresult in rapidproduction of C1NOs, which would persist throughout thedayin thelow lightlevels of thepolar regions during winter. The diurnal variation of CIO, under these conditions, is determined by interchange with C1202. For a given value of[O*], [C10] may be obtained by solving the equation: d[C!O] =2 (J +ka) [C!202] - 2 kF M [ClO] 2 , (1) dt where kF, ka, andY define rates for the self-reaction of C10 pro- ducing C1202, thermal decomposiQ .on of C1202, 8Ild phot01ysis of C1202, res•ctively. Values reported by Sander et al. (!989)are 561