Oct 15, 1998 - merry the following BrO mixing ratios were measured; .... of simultaneously performed ground-based, balloon-borne and ERS-2 satellite (GO-.
GEOPHYSICAL RESEARCH LETTERS, VOL. 25, NO. 20, PAGES 3843-3846, OCTOBER 15, 1998
StratosphericBrO profiles measuredat different latitudes and seasons:Atmospheric observations
H. Harder 3'1,C. Camy-Peyret 2, F. Ferlemann 1, R. Fitzenberger 1, T. Hawat2 H. Osterkamp 1, M. Schneider I D Perner 3 U PlattI P Vradelis I andK •
ß
•
'
•
'
9
'
Pfeilsticker I
to spaceand •ime coverage,and restrictedto low ambient pressures[Brune et al., 1988, 1989; Avallone et al., 1995;Mc Kinney et al., 1997]. Remotestratosphe-
Abstract.
Stratospheric BrO profiles were measured at differentlatitudes and in differentseasons in 1996/97 during three flights of the LPMA/DOAS balloon gondola (LPMA/Laboratoire PhysiqueMo16culaire et Application and DOAS/Differential Optical Absorption Spectrometry). Using direct sunlight DOAS spectromerry the following BrO mixing ratios were measured; (1) 9 to 14 ppt in the heightrangefrom 20 to 30 km (at solarzenithangles,SZA < 88ø) duringascent,(2) about (14-t-2) ppt for altitudesabovethe balloonfloat altitu-
ric BrO observations usingthe DOAS (differentialoptical absorption spectroscopy . ) techniquewith scattered sunlight or direct moonlight were performedmore widely. Presumablybecauseof the still limited accuracy
(+ 30%)of previousmeasurements the BrO observed by
both techniquesapparentlyagreewith expectationsbasedon photochemical modelcalculations.Somestudies, however,found a disagreementwith our presentunderstandingof the stratosphericBrO chemistry[Wahner
de at 30.6km, 30.0km, and39.8km, and(3) 5 to l0 ppt in the 20-30 km region during sunset. The lower BrO concentrationsduring sunsetthan those observedprior at daytime indicate a conversionof BrO into nightti-
et al., 1990;Butler et al., 1995;Oftenet al., 1998]. For example, Wahner et al. [1990]reportednightti-
me stratospheric BrO in the wintertime activated arc-
me reservoirspecies(BrONO2, HOBr, and BrC1). The
overall agreementof our UV spectroscopicBrO profiles tic vortex(whereC1activationis expected).They inwith recent measurementsusing the chemicalconversi- terpreted this observation as an indication for different on/resonance.fluorescence techniqueis good. Our BrO vertical concentrationprofilesof C10 and BrO, since profiles are also in reasonableagreementwith the pre- the rapid conversionof BrO by reaction with C10 into
sentstratosphericBry burdenand chemistry.Conversi- BrC1wouldotherwiseremoveBrO from the nighttime ly collocatedground-basedand satellite columnmeasu- stratosphere. More recently,stratosphericBrO measurements, however, show significantlymore total atmos- red on the ER-2 [Butler et al., 1995]indicateda total phericBrO (50- 100%)than the integratedstratosphe- stratosphericbromine burden closerto 25 ppt than to arison of round-based mearic BrO balloon profiles can accountfor. This indicates 20 ppt. In addition, a comp a global troposphericBrO background,estimated at 1 with photochemical calculations of the SLIMCAT mo2 ppt.
surements ofBrO atKiruna/Swedengin winter 1994/95
1.
del [M. Chipperfield, priv. comm.]showed a significant deficit (30%) in the modelledcomparedto the observedslantcolumndensityof BrO (SCD) at highstratos-
Introduction
duringtwilight lOtten et al., In the lower stratosphere bromine is photochemically phericNO2 concentrations
releasedfrom sourcegases(CHaBr, CH2Br2,CBrC1F9., CBrFa,..) which presentlyamountto 19+1 ppt in total Bry [Wamsleyet al., 1998;$chauffieret al., 1998]. After releaseBr is presentin inorganicform (BrO, Br, Br2, BrC1, BrONO2, HOBr, and HBr known collecti-
vely as Bry). Due to the lowerbindingenergyof HBr
1998]. BrO profilesmeasuredby balloon-borneDOAS (SAOZ) over northern Swedenin winter 1997 showed goodagreementwith photochemical modelcalculations, exceptfor 20 March 1997whenin the heightrangefrom 17 km to 25 km up to 30% more BrO wasmeasuredthan
modelled[Pundt, 1997].
compared to HC1 and to the more rapid photolysis of Here we report on UV spectroscopicBrO profile meaBr2, BrC1, BrONO2, and HOBr comparedto their C1 surementsperformedwith the LPMA/DOAS balloon
analogsabout half of the Br• consistsof the activespe- gondola during flights conductedat mid-latitude duringfall (Le6n/Spain,Nov. 23, 1996;flight1) andsumciesBr and BrO during daytime [Lary el al., 1996a,b], mer (Gap/France,June20, 1997;flight3) and at high whereastypically only a few percent of inorganicchlorine is present as C10. Therefore, although much less northernlatitudesin winter (Kiruna/Sweden,Feb. 14, abundant in theatmosphere (Bru/Cly• 0.6%),Brycan 1997;flight 2). remove O3 at a rate comparable to that of chlorine.
2.
Observations
In-situmeasurements of stratospheric BrO (mostof these made by the chemicalconversion/resonance fluore- For the measurementof stratosphericprofiles of O3, scencetechnique) are sparse,quite limited with respect NO2, 04, BrO, NO3, H20, and possiblyCH20, and IO, a novelDOAS-instrument(and two NO2 photolysis sensors)were installedon the LPMA gondola. For • Institut f•r Umweltphysik, Universityof Heidelberg, detailsof the gondolasee Camy-Peyretet al. [1993], Heidelberg/Germany 2 Max-Planck-Institut ffir Chemie, Saarstr. 23, and for the DOAS-instrumentFerlemannet al. [this issue,and 1998].The combination of a UV/vis spectroM ainz/ Germany 3 LPMA, CNRS, Universit6Pierre et Marie Curie, Pa- meter with a Fourier transform infrared IR-instrument both performing solar occultation measurementsoffers
ris/France
severaladvantages:(1) Both instrumentsusethe same Copyright 1998bytheAmerican Geophysical Union.
infrastructure
and the number of measured constituents
PapernumberGRL-1998900026. 0094-8276/98/GRL-1998900026505.00
F11 ...), a comprehensive characterization of the chemi.... cal composition of the stratosphere is obtained. In the
is muchlarger(extendingto HNO3, HC1,C1NOa,CH4,
3843
3844
HARDER ET AL.,' STRATOSPHERIC PROFILES OF BRO
Table 1. Overview of simultaneously performed ground-based, balloon-borne andERS-2satellite(GOME) BrO measurements. Date and
Location
Time(UT) Nov. 23, 96 14:55-16:52
Feb. 14, 97
12:19-14:49
June20, 97 3:30-7:42
Le6n
42.6ø N,5.7ø W
Kiruna
67.9ø N,21.1ø E
Gap 44.0øN,6.1øE
Range
Height
of SZA
Range
Integrated BrO
(degree) (km) 74-86.4 5.1-30.6 31.0-22.0
2.31-t- 0.45
82.5-88.8
3.99 4- 0.6
86.5-93.0
64.4
5.8-28.8
30.0-13.0
80.6 80.7
-
90.0-55.9 94.8-90.0
39.8-31.4 17.3-397
Comment
BrO above
ballonfloat (i)
(1013/cm2) -
-
-
88 9-94.1
Total atmos-
abovethetropopause phericBrO-VCD
-
(4p p_•t )2.5)
Ascent
(14.
Occultation
5.2•1-2.2
above 31.6 km
GOME overpass totalatmosphere
-
Ascent
(15.5 4- 2.8)
-
-
Occultation
above 30.0 km
-
7.5 4-2.0 6.5 4-1.8
GOME (13 Feb. 97) ground-based
-
'h -
-
Descent Occultation
(15.3 4- 2.8) above39.6 km
(i) Averagemixingr•tio (•s determined from• linearregression of measured SCD of BrO versustotM Mr slant columnin the line of sightfor SZA < 90ø, seetext) •boveb•lloon flo•t •ltitude.
caseof 03 and NO2 a crossvalidationis alsopossible. ratios up to 18 km. This is expected,sinceinorganic Balloon Flights: Bry releasedfrom its organicprecursors[Wamsleyet Three successfulballoon flights were performed in al., 1998]and most part beingpresentin BrO during 1996/97(Table1). daytime[Lary et al., 1996a,b]. During fi•ght 1 on Nov. 23, 1996, over northern Spain For both afternoonascents(flightsi and 2), BrO voluat • 39øN, direct Sun spectra were collectedduring balloonascent(5 km to 30.3 km altitude) and during me mixing ratios were in the range of 9 ppt to 15 ppt in the lower-stratosphere (20 km- 30 km). Duringsosunset to a tangent height of • 21.7 km. Flight 2 on Feb. 14, 1997, over northern Scandinavia lar occultation(Fig. lb) only about50%of the BrO is (68øN) followeda similar pattern and spectrain air- foundat thecorresponding altitudes (Fig.la). Thismamassesjust outside the arctic vortex were collected. A
nifeststhe BrO conversion (throughreactionwith NO2
triple blind test of 03 and NO2 amongthe DOAS, LP- andHO2) into the brominenighttimereservoir species MA, and ILAS instrument, the latter operated on the BrONO2andHOBr (andlessimportantintoBrC1,sinJapaneseADEOS satellite, showedan overallagreement
of better than 10% (mixingratio). Flight 3 on June20, 1997oversouthernFrance(42ø N)
40
waslaunchedat night and aimedat the investigationof the mid-latitude summer 03 chemistry at dawn. The sun-trackercaught the solar disk at a tangent height of 2 km located over northern Switzerland, while the balloon was flying at an altitude of 40 km over the Rhone valley (420 N). Direct Sun measurements werecon-
30
I
....
I
(a) Flight1
....
I
....
I
....
I
'
'
'
'
Ascent
Flight2
20
tinued during the slow balloon descent until 31.4 km
(SZA of 55.½) (Table 1). The tropospheric part of the
measurements are reported elsewhere. During each flight about 1000 spectra were recorded and analyzed. 3.
lO
Tropopause • !
Results
•- 40
Vertical BrO profiles were derived from two observation modes, namely from balloon ascent and from solar
30
titudes aboveballoonfloat (> 39.8 km) [Ferlemannet al., this issue].
20
In the retrieved profiles photochemical processescausing spatial and temporal variations of the stratospheric BrO during the measurementswere not corrected for, mainly becauseof uncertainties in the stratospheric BrO chemistry. Sincethe onion peeli_ngand the matrix inversion/)rofile retrieval techniques[Ferlemann et al., this issue]react differently on photochemicalchangesthe overall consistency in the retrieved BrO profilesusingboth inversion
i
i
I
i
i
i
I
I
i
i
i
i
OcaJltat/on
(+10.106 )
occultation atfloat (Fig. la,and Fig.lb)•Note, that
for flight 3 due to the incomplete descent day profile could not be derived, but accurate BrO-SCD for the al-
i
(+30.106 )
10
Flight1
Flight2
: Flight ,
I .... o
i , , • , I , , , , I ....
10
20
30
40
50
BrOConcentration [106molecule/cm 3]
Figure 1. Verticalprofileof BrO measured during
techniques(Fig. la) suggests only smallerrorsof the flights1, 2, and3. (a) BrOconcentration (1/cm3) verascentprofiles due to photochemistry. In addition, from a comparison of measured and modelled ground-based
susheightfor the balloonascents. Notethat for flight
profilewasmeasured (seetext);(b) same BrO-SCDs,Friefi et al. [1998]concluded to a negligible 3 nodescent
occultation measurements (for photochemicalchangein BrO for SZA
