Apr 20, 1994 - William B. Grant, ⢠Edward V. Browell, ⢠Jack Fishman, 1 Vincem G. Brackett,. Robert E. Veiga, 2 ...... DIAL systems [McGee et al., 1993].
JOURNAL OF GEOPHYSICAL
RESEARCH, VOL. 99, NO. D4, PAGES 8197-8211, APRIL 20, 1994
Aerosol-associatedchangesin tropical stratosphericozone following the eruption of Mount Pinatubo William B. Grant,• EdwardV. Browell,• JackFishman, 1 VincemG. Brackett,
RobertE. Veiga,2Dominique Nganga, 3A. Minga,3BernardCros,3 CarolynF. Butler, 2 MartaA. Fenn, 2 CraigS. Long, 4 andLarryL. Stowe • Abstract. The largeamountof sulfuricacidaerosolformedin the stratosphere by conversionof sulfurdioxideemittedby the eruptionof Mount Pinambo(15.14øN, 120.35øE)in the PhilippinesaroundJune 15, 1991, has had a pronouncedeffect on lower stratospheric ozonein the tropics. Measurements of stratospheric ozonein the tropics usingelectrochemical concentration cell (ECC) sondesbeforeand after the eruptionand the airborne[IV differentialabsorptionlidar (DIAL) systemafter the eruptionare comparedwith Stratospheric Aerosoland Gas ExperimentII (SAGE II) measurements from severalyearsbeforethe eruptionandECC sondemeasurements from the year prior to the eruptionto determinethe resultingchanges. Ozonedecreasesof up to 33 % comparedwith SAGE II climatologicalvalueswere foundto be directlycorrelatedwith altituderegionsof enhancedaerosolloadingin the 16- to 28-km range. A maximum partial-columndecreaseof 29+9 Dobsonunits (DU) was found over the 16- to 28-km rangein September1991 alongwith smallincreases(to 5.9+2 DU) from 28 to 31.5 km. A large decreaseof ozonewas alsofoundat 4øto 8øSfrom May to August1992, with a maximumdecreaseof 33+7 DU foundaboveBrazzavillein July. Aerosoldata from the visible channelof the advancedvery high resolutionradiometer(AVHRR) and the visible wavelengthof the UV DIAL systemwere usedto examinethe relationshipbetween aerosol(surfacearea) densitiesand ozonechanges. The tropical stratosphericozone changeswe observedin 1991 and 1992 are likely be explainedby a combinationof dynamical(verticaltransport)perturbations, radiativeperturbations on ozonephotochemistry, and heterogeneous chemistry. Introduction
Reductionsof midlatitudestratosphefic ozoneconcentration were observedfollowing the volcanic eruptionsof Mount Agungon March 17, 1963[GramsandFiocco,1967; Pittock, 1965, 1966], Fuego in October1974 [Jagerand Wege, 1990], andE1 Chich6non March 28 to April 4, 1982 [Datsch, 1985; Mantis et al., 1986; Adriani et al., 1987; Angell, 1988; Newell and $elkirk, 1988; Jager and Wege, 1990; Chandra and $tolarski, 1991]. While there has been some debate about whether the observed midlatitt•e
ozone
decreasesfollowingvolcaniceruptionspfior to 1982 could be attfibutedto the presenceof aerosolsin the stratosphere
from the eruptions,it is generallyagreedthatthe northern midlatitudeozonedecreases observedin early 1983 were at
least partially causedby the April 1982 eruptionof E1 Chich6n[Bojkov,1987;Newelland$elkirk,1988;Jagerand Wege, 1990]. It hasbeensuggested that thisdecreasewas a resultof heterogeneous chemistryoccurringin the midlatitude regions[e.g., Holmannand Solomon,1989]. This mechanism
has received
some confirmation
in the ozone
depletion observedabove Antarcticain 1991, which is attfibuted to heterogeneous processesoccurring on the aerosolsfrom CerroHudson[Hofmannet al., 1992], aswell as in the
increase
of
OC10
above
Antarctica
in
1992
attfibutedto the presenceof sulfufic acid aerosolsfrom
•Atmosphefic Sciences Division,NASALangleyResearch Mount Pinambo [Solomonet al., 1993]. However, mechanismswhichoperateto causeozonedecreases in themidlatiCenter, Hampton,Virginia. tude and polar regions, with different temperature andsolar 2Science Applications International Corporation,Hampradiation conditions, may not be active in the tropics. ton, Virginia. The eruptionof Mount Pinatubowas estimatedto have 3Universit6 Maien Ngouabi,Brazzaville,Congo. 4NOAA/NWS/NMC,ClimateAnalysisCenter,Washing- injected20 million tons (M0 of SO2into the stratosphere
[Bluth and Doiron, 1992], resultingin 20 to 30 Mt of sulfufie acid aerosol [McCormickand Veiga, 1992], the 5NOAA/NESDIS/ORA/Satellite ResearchLaboratory, largest amount for any volcaniceruptionof this century. Washington,DC. This event provideda uniqueopportunityto continuethe study of the effect of volcanicaerosolson stratosphefic Copyfight 1994 by the AmeficanGeophysicalUnion. ozone, especiallyin the tropics. High-resolution vertical proffieaerosolinformationin thetropicswasobtained using Papernumber93JD03314. 0148-0227/94/93JD-033 14505.00 the UV differentialabsorptionlidar (DIAL) systemfrom ton, DC.
8197
8198
GRANT
ET AL.'
TROPICAL
OZONE
CHANGES
AFTER
MOUNT
PINATUBO
ERUPTION
that 4-5 % of the decreasecouldhavebeendue to the QBO. This analysis,based on the earlier analysisreportedin eruptionwasobtainedusingthe advanced very highresolu- Chandra and Stolarski [1991], assumedthat the 30-mbar tion radiometerinstrumenton NOAA 11. High-resolution winddirectionaboveSingapore couldbe usedasa proxyfor ozoneprofileswereobtainedon nearlya weeklybasisbefore the effect of the QBO. However,a closereadingof Chanand after the eruptionusing electrochemical concentration dra and Stolarski indicates that the correlation coefficient of cell (ECC) sendesin the tropicsandon the 1992UV DIAL zonalwind andtotalozoneat the equatoris only0.6 to 0.8. flights. Stratospheric aerosoland gasexperimentII (SAGE Indeed, in 1989 the correlationhad broken down, with the I• dataobtainedbeforethe eruptionwereusedto providean zonalwindbeingof a phaseandmagnitude thatshouldhave ozone climatology. Data from thesevarious sourcesare corresponded to 3-4% lesscolumnozonethanwasobserved combined to examine the effect of the volcanic aerosols on by TOMS. In addition, Rind et al. [1992] show that changes in theozonedistribution in thetropicalstratosphere. volcanicaerosolsin the tropicsaffectthe circulationin the Decreased tropical stratosphericozone values were stratosphere.With the eruptionof MountPinambo,a nonreportedfollowingthe eruptionof Mount Pinamboon June historical factor is added which could also affect the relation 15, 1991 [Grant et al., 1992; Waterset al., 1992; $choeberl betweenthe total ozoneand the zonal wind. Thus, we have et al., 1993; Chandra, 1993; Grant et al., 1993; Hermann seriousreservationsaboutthe conclusions of Chandra[1993] et al., 1993]. While there may well have been similar regardingthe magnitude of theeffectof theQBOontropical decreases followingvolcaniceruptions priorto thatof Mount total ozoneafter the eruptionof MountPinatubo. Pinatubo, such as that of El Chich6n [Hermannet al., 1993], measurements followingthoseeruptionswere either Instrument Descriptions
flights in October1991 and in JanuaryandFebruary1992. Global-scale
aerosol
information
before
and
after
the
very sparseor difficult to interpret,due in part to natural variationsfrom the quasibiennial oscillation(QBO) [e.g., Chandra and $tolarski, 1991; Herman et al., 1993]. Verdecchiaet al. [1992] suggested thatdecreases in tropical stratospheric ozonefollowingthe eruptionof El Chich6n shouldhave occurredat up to 3-4% of the columnamount due to an aerosolradiativeperturbation. In Grant et al. [1992], the decrease dueto the presence of volcanicaerosolsin the 16- to 28-kmaltituderegionwas estimated to be 13 to 20 Dobsonunits(DU-). In Grantet al. [1993], madebetteruseof SAGE-II climatological data,and the estimate was refined to 27 DU decrease below 28 km and an increase of about 5.4 D U from 28 to 32 kin. The
The airborneUV DIAL systemusesa pair of Nd:YAG lasersto pumpdye laserswhichare frequencydoubledto generatethe ultraviolet wavelengthsused for the ozone measurements [Browell, 1989]. For stratospherie measurements,wavelengths of 301.5 and310.85 nm are employed, and the temperature-dependent absorptioncoefficientsof Molina and Molina [1986] are usedin the analysis. The residualundoubleddye laserandundoubledNd:YAG laser pulseenergiesare usedto measureatmospheric backscatter profiles in the visible (603 nm) and near-IR (1.064 /an) spectralregionsin order to determineaerosoldistributions. The UV DIAL systemhas been shownto make measurementsof stratospheric ozonewith an accuracyof •
