Impacts of climate change on stratospheric ozone recoveryD. W. Waugh,1L. Oman,1S. R. Kawa,2R. S. Stolarski,2S. Pawson,2A. R. Douglass,2P. A. Newman,2and J. E. Nielsen3Received 4 October 2008; accepted 5 January 2009; published 5 February 2009.[1] The impact of increasing greenhouse gases (GHGs) onthe ‘‘recovery’’ of stratospheric ozone is examined usingsimulations of the Goddard Earth Observing SystemChemistry-Climate Model. In this model, GHG-inducedclimate change has a large impact on the ozone evolutionand when O3recovery milestones are reached. The twodistinct milestones of ‘‘O3returning to historical values’’and ‘‘O3being no longer significantly influenced by ozonedepleting substances (ODSs)’’ can be reached at verydifferent dates, and which occurs first varies betweenregions. GHG-induced cooling in the upper stratospherecauses O3to increase, and O3returns to 1980 or 1960values several decades before O3is no longer significantlyinfluenced by ODSs. In contrast, transport changes in thetropical and southern mid-latitude lower stratosphere causeO3to decrease. Here O3never returns to 1980 values, evenwhen anthropogenic ODSs have been removed from theatmosphere. O3returning to 1960 (or 1980) values shouldnot necessarily be interpreted as O3recovery from theeffects of ODSs.Citation: Waugh, D. W., L. Oman, S. R.Kawa, R. S. Stolarski, S. Pawson, A. R. Douglass, P. A. Newman,and J. E. Nielsen (2009), Impacts of climate change onstratospheric ozone recovery, Geophys. Res. Lett., 36, L03805,doi:10.1029/2008GL036223.1. Introduction[2] There is considerable interest in how stratosphericozone will evolve through the 21st century. Due to regu-lations imposed by the Montreal Protocol and its amend-ments, the concentration of the sum of all ozone-depletingsubstances (ODSs) peaked in the 1990s and is expected todecrease back to 1960 levels around the end of this cen-tury [ World Meteorological Organization (WMO)/UnitedNations Enviroment Programme (UNEP), 2007], with acorresponding ‘‘recovery’’ of stratospheric ozone. However,changes in temperature, transport, and nitrogen and hydro-gen ozone-loss cycles caused by the anticipated continuedincrease i n well-mixed greenhouse gases (GHGs) arelikely to affect this ozone ‘‘recovery’’ [e.g., Haigh andPyle, 1979; Brasseur and Hitchman, 1988; Rosenfield et al.,2002; Chipperfield and Feng, 2003; Austin and Wilson,2006; Eyring et al., 2007]. The probable influence ofclimate change on the evolution of stratospheric ozonecomplicates quantification of the recovery process, includ-ing the interpretation of reaching different milestones thathave been used to define ozone recovery.[3] Traditionally the date that ozone returns to a specifiedhistorical value (e.g., pre-1980 values) has been used todescribe full ozone recovery [e.g., Rosenfield et al., 2002;Shindell and Grewe, 2002; Austin and Wilson, 2006].However, this milestone does not require attribution toODSs, and may be reached because of natural variabilityor other climate changes, or may never be reached evenwhen all anthropogenic ODSs are removed from theatmosphere. This lack of attribution can lead to misinter-pretation of observations [e.g., Kane, 2008]. To demon-strate recovery from halogen-induced destruction it isnecessary t o attribute changes in ozone to changes in ODSsand to changes in clima te [Shindell and Grewe, 2002]. Analternative, attributed, milestone for ‘‘full ozone recoveryfrom ODSs’’ introduced by WMO/UNEP [2007] is thedate when ozone is no longer significantly affected byODSs. Recent studies of the simulated ozone evolutionthrough the 21st century [e.g., Rosenfield et al., 2002;Chipperfield and Feng, 2003; Austin and Wilson , 2006;Eyring et al., 2007; Shepherd and Jonsson, 2008] havenot examined the milestone proposed in WMO/UNEP[2007].[4] This study uses simulations of NASA’s GoddardEarth Observing System Chemistry-Climate Model (GEOSCCM) [Pawson et al., 2008] to examine the impacts ofstratospheric climate change on ozone recovery, focusing ondifferences in the dates when the above ‘‘full recovery’’milestones are reached.2. Model and Simulations[5] The GEOS CCM includes representations of atmo-spheric dynamics, radiation, and stratospheric chemistryand their coupling through transport and radiative processes.Pawson et al. [2008] show that the climate structureand ozone in GEOS CCM agree quite well with observa-tions. Two deficiencies are a high bias in total O3whenchlorine loading is low (in the 1960s) and the anomalousnature of the Antarctic vortex [Pawson et al., 2008], butthese should have only a minor effect on the presentanalysis, which focuses on relative ozone changes. Addi-tional evaluations of GEOS CCM [Eyring et al., 2006;Perlwitz et al.,2008;Oman et al., 2008] reveal goodcomparisons with observations. Further, Waugh and Eyring[2008] demonstrate that GEOS CCM is one of the better-performing models from the CCMs included by Eyring etal. [2006].[6] This study uses a pair of GEOS CCM simulationswith identical surface concentrations of CO2,N2O, andGEOPHYSICAL RESEARCH LETTERS, VOL. 36, L03805, doi:10.1029/2008GL036223, 2009ClickHereforFullArticle1Department of Earth and Planetary Sciences, Johns HopkinsUniversity, Baltimore, Maryland, USA.2Atmospheric Chemistry and Dynamics Branch, NASA Goddard SpaceFlight Center, Greenbelt, Maryland, USA.3Global Modeling and Assimilation Office, NASA Goddard SpaceFlight Center, Greenbelt, Maryland, USA.Copyright 2009 by the American Geophysical Union.0094-8276/09/2008GL036223$05.00L03805 1of6CH4, and sea surface temperatures (SST) but different ODSconcentrations. The first simulation has ODSs fixed at their1960 values, so that changes in O3are due to climatechange (this is referred to as the climate-only simulation).In the second simulation the surface ODS concentrationsvary with time according to the WMO/UNEP [2003] Abscenario, and O3changes are due to changes in both climateand ODSs (this is referred to as the climate + ODSsimulation). The difference in O3between the two simula-tions is then the ODS-induced change. These ODS-inducedchanges include both the direct ODS chemical impact and‘indirect’ feedbacks, such as that in the upper stratospherewith respect to temper ature decreases caused by ODS-induced ozone loss.[7] The pair of simulations include separate runs for thepast (1950–2004) and the future (2000 –2099) [Pawson etal., 2008; Perlwitz et al., 2008]. The past