Implications for Chlorine Chemistry - NASA

GEOPHYSICAL RESEARCH LETTERS, VOL. 22, NO. 7, PAGES 843-846, APRIL 1, 1995

Recent trends in ozone in the upper stratosphere: Implications for chlorine chemistry SushilChandra •, CharlesH. Jackman •, andEric L. Fleming 2 Abstract. We havestudiedtheimplications of recenttrends More recently, Chandra and McPeters (1994) have in theannualmeanandtheamplitude of theannualharmonic extendedthe ozone data baseby combiningthe Nimbus-7 of ozonein theupperstratosphere from the 15 yearsof the SBUV datawith the NOAA-11 SBUV/2 datafromJanuary combined data from the Nimbus-7 SBUV and the NOAA-11 1989 to December 1993. Though the combineddata set SBUV/2 instruments. This was done in the context of the coveringa 15 year period doesnot significantlyalter the GSFC 2D modelpredictionsof thesetrendswhichare based lengthof the time seriesfor studyinglongterm changesin on plausiblescenariosof anthropogenic Cly increasein the stratospheric ozone,it is thebestdatasetavailable to study atmosphere.The comparisonof the observedand model- the various implicationsof long term changesin the estimated annualmeanozonetrendsshowsomesimilarityin stratosphere.It hasbeenshownby Chandraet al.( 1993, to their latitude and altitude characteristics. Both the model and

datashowa maximumozonedecrease of-6 to -10 % per decadeat highlatitudesin theupperstratosphere. However, there are also significantdifferencesbetweenthe observed andcomputedtrendswhichmaybe relatedto boththemodel uncertaintyand the uncertaintyin correctingfor the long term instrumentdrift. The observations also suggesta decreaseof 10-25 % per decadein the annualamplitudeof ozoneat 2 mb between400-60øin bothhemispheres, with a relatively larger interannualvariability in the northern hemisphere.Thesevaluesare in generalagreementwith the model predictionsand thus provideadditionalsupportin favorof the chlorineinducedchanges in ozonein the upper stratosphere.

be referred to as C93) that both the annual mean and the

amplitude of theannualharmonic of ozoneareinfluenced by the chlorinecatalyticloss cycle. The annualharmonicat middle and high latitudesaccountsfor more than 80 % of the totalvarianceof the annualwave ( seealsoPerliskiand London,1989). A studyof longtermchangein theannual cycle of ozone thus provides an additionalconstraintin delineating theroleof oddchlorinein thestratosphere. The purposeof this paperis to studylongterm changesin the annualmeanandtheannualamplitude of ozonein theupper stratosphere at middleandhigh latitudesin the contextof the GSFC 2D model results as discussed in C93.

Data Description

Introduction

The datausedin this studyare basedon the monthly values of ozone time series from the Nimbus-7 SBUV and

Recently,Hoodet al. (1993)havereporteda decrease of the NOAA-11 SBUV/2 instruments from January1979 to 6-12 % in upper stratospheric ozoneat middle and high December 1993 as discussed in Chandra and McPeters latitudesover an elevenyear period from 1979 to 1990. (1994).Theyrespectively coverperiodsfromJanuary1979 Their study, basedon the version6 SBUV data from the to December1989 andJanuary1989to December1993 with Nimbus-7 satellite, supportsan earlier conclusionby an overlappingperiod of one year. To accountfor the Hilsenrathet al. (1993) that ozonetrendsin the upper differences in theabsolute calibrationof thetwoinstruments, stratosphere are statisticallysignificantand may havebeen the SBUV time series have been normalized in terms of the caused by an increasein thechlorinelevelof theatmosphere SBUV/2. The normalization is achievedby readjusting the relatedto anthropogenic perturbations.DeLuisiet al. (1994) SBUV time seriesby a factor calculatedfrom the mean of have shown that the trends derived from the SBUV data over

the ratio of the SBUV

and SBUV/2

values over the

thisperiodare in generalagreement withthosederivedfrom overlappingperiod. The normalizationfactor varies from the ozone profiles from the ground based Umkher about 4-6 % at and above 2 mb to about -1 to -5 % at 10 mb measurements. The SBUV trends also show general and below. The time series for each year are Fourier similaritywith thetrendsderivedfromthe2D photochemical analyzed to calculate the annual mean and the annual, models(WMO, 1991). These trends,however, are sig- semiannual,and terannualamplitudesand phases. Our nificantlylarger thanthe trendsreportedby McCormicket discussions will be limited to the annual mean and annual al. (1992) from their analysisof SAGE I andSAGE II ozone waveat middleandhighlatitudeswherechanges in levelsof profilesfor the periodFebruary1979to April 1991. atmospheric chlorinehavetheir largesteffects. We have also performeda Fourier analysison the analogous NMC temperature timeseries.The annualcycles •NASA/ GoddardSpaceFlightCenter,Greenbelt,MD of ozonein the upperstratosphere are drivenby the annual '-Applied Research Corporation, Landover,MD cyclesof temperaturethroughtemperaturedependent loss rates. The interannualvariabilityandlongtermchangesin ozoneare, therefore,expected to be closelycoupledwith the Copyright1995by theAmerican Geophysical Union. interannual variabilityandlongtermchanges in temperature. Unfommately, the long term changes in the NMC Papernumber94GL03395 temperaturein the upper stratospheremay have an 0094-8534/95/94GL-03395503.00 uncertaintyof 1-3 K becauseof the severaladjustments 843

made in the temperature time series using rocket measurements (Finger et al., 1993). Suchadjustments may alsohave affectedthe annualamplitudeof temperaturesince someof the adjustments were madeduringthe middleof the

SBUV-SBUV/2Trend (Z/decode) •

..........

1

x

•

year.

10

*"

"'--'

mean ozone time

series. The

model

assumes the total

chlorinein the sourcegasesat the groundto be about2.5 ppbv and 3.5 ppbv, respectively, for 1980 and 1990

-20

1 m ta

35

-40

20

Model

......

'"*• -60

-40

35•

_

30 •,

25