UC Irvine UC Irvine Previously Published Works

Title Atmospheric sampling of Supertyphoon Mireille with NASA DC-8 aircraft on September 27,1991, during PEM-West A

Permalink https://escholarship.org/uc/item/3c03j813

Journal Journal of Geophysical Research: Atmospheres, 101(D1)

ISSN 2169-897X

Authors Newell, RE Hu, W Wu, ZX et al.

Publication Date 1996

DOI 10.1029/95JD01374

License https://creativecommons.org/licenses/by/4.0/ 4.0

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 101,NO. D1, PAGES 1853-1871,JANUARY 20, 1996

Atmospheric sampling of SupertyphoonMireille with NASA DC-8 aircraft on September 27, 1991, during PEM-West A

R. E. Newell,1 W. Hu,1 Z-X. Wu,1 Y. Zhu,1 H. Akimoto,2 B. E. Anderson,3 E. V. Browell,3 G. L. Gregory,3 G. W. Sachse,3 M. C. Shipham,3 A. S.Bachmeier, 4 A.R. Bandy, 5 D.C. Thornton,5 D. R. BlakefiF. S.Rowland, 6 J.D. Bradshaw,7 J. H. Crawford,7 D. D. Davis,7 S. T. Sandholm,7 W. Brockett,8 L. DeGreef,8 D. Lewis,8 D. McCormick,8 E. Monitz,8 J. E. CollinsJr., 9 B. G. Heikes,1ø J. T. Merrill,1ø K. K. Kelly,11 S.C. Liu,• 1 y. Kondo,12 M. Koike,12 C.-M. Liu, 13 F. Sakamaki,TM H. B. Singh,15 J. E. Dibb,16 and R. W. Talbot 16

Abstract. The DC-8 missionof September27, 1991,was designed to sampleair flowing intoTyphoon Mireille in theboundary layer, air in the uppertropospheric eye region, and air emergingfrom the and ahead of the system,also in the uppertroposphere. The objectivewas to find how a typhoonredistributes trace constituentsin the West Pacificregion and whether any suchredistribution is importanton the globalscale. The boundarylayer air (300 m), in a regionto the SE of the eye,contained low mixingratios of thetracer species 03, CO, C2H6, C2H2,C3H8, C6H6 and CS2 but high values of dimethylsulfide(DMS). The eye regionrelative to theboundary layer, showed somewhat elevatedlevels of CO, substantiallyincreased levels of 03, CS2 and all nonmethane hydrocarbons(NMHCs), andsomewhat reduced levels of DMS. Aheadof the eye,CO andthe NMHCs remainedunchanged, 03 and CS2 showeda modestdecrease, and DMS showed a substantial decrease. There was no evidence from lidar cross sections of ozone for the downwardentrainment of stratosphericair intothe eyeregion; these sections show thatlow ozonevalues were measured in thetroposphere. The DMS datasuggest sub- stantialentrainment of boundarylayer air intothe system,particularly into the eyewall region.Estimates of the DMS sulphurflux betweenthe boundarylayer and the freetro- posphere,based on computations of velocity potential and divergent winds, gave values ofabout 69 ggS m -2 d-•averaged over a 17.5øgrid square encompassing thetyphoon. A few hoursa•er samplingwith the DC-8, Mireille passedover Oki Island,just to the north of ,producing surface values of ozoneof 5.5 ppbv. These03 levelsare consistent withthe low troposphericvalues found by lidarand are more typical of equatorial regions.We suggestthat the central eye region may act like a Taylorcolumn which has movedpoleward from low latitudes.The high-altitudephotochemical environment withinTyphoon Mireille was found to be quiteactive as evidenced by significantlevels of measuredgas phase H202 andCH300H andmodel-computed levels of OH.

1Massachusetts Institute ofTechnology, Cambridge. 9Scienceand Technology Corp., Hampton, Virginia. 2UniversityofTokyo, Japan. løUniversityof Rhode Island, Kingston. 3NASALangley Research Center, Hampton, Virginia. I•NOAAAeronomy Laboratory, Boulder, Colorado. 4LockheedEngineering and Sciences, Co., Hampton, •2NagoyaUniversity, Nagoya, Japan. Virginia. •3NationalTaiwan University, Taipei. 5Drexel University, Philadelphia, Pennsylvania. 14NationalInstitute for EnvironmentalStudies, Tsukuba, 6UniversityofCalifomia, Irvine. Japan. 7GeorgiaInstitute of Technology,Atlanta. •SNASAAmes Research Center, Moffett Field, Califomia. 8AmesScience and Applications Aircraft Division, NASA 16UniversityofNew Hampshire, Durham. Ames ResearchCenter, Moffett Field, Califomia.

Copyright1996 by the AmericanGeophysical Union. 1. Introduction

Papernumber 95JD01374. The Global TroposphericExperiment (GTE), 0148-0227/96/95JD-01374505.00 undertakenover the past decadeby the National

1853 1854 NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE

ElI0 115 120 125 130 135 140 145 150 155 160 165 170 175E N45

LEGEND 6-tIR BEST 'IRACK POSITION BESTTRACKTC-21W !'' SPEED OF MOV 'EMENT (KT) 13SEP- 28SEP 91 INTENSITY (KT) 4O POSITION AT XXAXX)0Z MAXSFC WIND 130KT .... !...... +++• : 281...... •'...... TROPICAL DISTURBANCE MINIMUMSLP 910MB 'IROPICAL DEPRESSION TROPICAL STORM 0ki TYPI lOON

35 SUP[]*, TYPIIOON END '::,::::...... ' .....:...... i...... :•...... EX'IRA'IROPICAL SUPERSUBTROPICALTYPIIOON START 10•.ß 2• : DISSIPATING STAGE ....':..:.". : "•'5i ':'": Isla .4.•.:::%• .. FIRST WARNING ISSUED 24 39• 85! LAST WARNING ISSUF_.D 3O 110, 115 115 :!'11 25 115 •.. .:. . '11

...... •:'Ke•ting 1 24 65 2O

i:25 •7125 F- 16/00Z 15 ..•TC•A...... 20 19. 1817 16 *""7•øøo!ABpw t :.}':'':•'.. ':::"!25 15 /i øo ..•?.i ::;.::..:..:.i:130 10 • "'• ...... • : 130, i...... ::..::7 85 [ .'..:.:.:.'...'" :•": " ß.. : i !. .: :..•.. 1251 N5

Figure 1. Trackof TyphoonMireille from September13-28, 1991 [fromRudolph and Guard,1992]. Positionsof groundstations Kenting and Oki Islandare noted.

Aeronauticsand Space Administration,has carried out In the typhoon system,there is a substantiallateral atmosphericsampling in various regionsof the world. inflow of air in the lower troposphereto compensatefor Measurementsare used here from flights through a the risingconvective motions and outflow aloft, andthese typhoonduring the Pacific ExploratoryMission-West A quasi-horizontalflows are alsoimportant in redistributing (PEM-West A) NASA DC-8 deploymentto the western trace constituents. An opportunity to carry out an Pacific in September-October1991. The instrumentation experimentto checkthis redistributionwas presented when cardedon the DC-8 is describedby Hoe# et al. [thisissue] TyphoonMireille approachedan areato the southof Japan and the backgroundmeteorology by Bachmeieret al. [this which could be accessedfrom Tokyo. Air emeringthe issue]. Samplingwas plmmed in two separateseasons: one typhoonin the surfaceboundary layer, in the eye itself, toward the end of the summer monsoon and the other emergingfrom the eye and surroundingwall cloudregion toward the end of the winter monsoon. The summer in theupper troposphere, was sampled. monsoonof thetropical western Pacific always includes a TyphoonMireille was first notedon the weathermaps numberof ;five occurredduring the 6 weeksof as a significant system on September13, 1991, when the first PEM-West deployment. The large-scale positionedat 12øN,172øE moving westward (see Figure 1) manifestation of atmosphericconvection that these [from Rudolph and Guard, 1992] and qualified as a typhoonsrepresent could be importantin the redistribution typhoon,with 1-min sustained winds exceeding 32 ms'l, of atmospherictrace constituents.In someprevious on September16 while at 15øN,157øE. By September22 discussionsof typhoons, there was debate as to whythe air it had turned toward the northwest and winds had increased mass in the eye was apparently different from the to 67 ms-1, qualifying Mireille as a "supertyphoon." After surroundingair mass.Bergeron [1954], referring to the September23, Mireille beganto slowly weaken,although work of others, noted that "The two air masses are its size, as measuredby the diameter of its outermost separatedby a boundaryzone which sometimeshas been closedisobar, continued to increase. On September26, taken for the tmpopauseitself, suckeddown from its Mireille turned northward, then northeastward,and made normal height.... " In Bergeron'sview, "It is more land in westernKyushu at about 0600 UT on September plausiblethat this boundary zone has been formed by a 27. Duringthe eveningof September27 it passedover the similarprocess within the troposphericair itself." In the and gradually lost force, passing over presentexperiment, ozone, a goodstratospheric tracer, was northern and southern early on measured to test these ideas. September28. m ,. mm mmm m

• II - q

Plate 1. JapaneseGeostationary Meteorological satellite image for 0200 UT at time of DC-8 take-off (providedby first WeatherWing at YokotaAir Base).

37.00N

./

•27 00E 141.00E Fi•.r½ 2. Horizontal projectionof flight track, DC-8 measuredwinds, •nd •dditional commercial aircraft flight winds at 288 hPa (data from EuropeanCentre for Medium-RangeWeather Forecasts (ECMWF)files) for September27, 1991. Flag,25 ms-l;a full bar,5 ms-l;a halfbar, 2.5 ms-1. Time interval between DC-8 winds is 10 min. 1856 NEWELL ET AL' DC-8 SAMPLING OF TYPHOON MIREILLE

12

10

Plate 2. Three-dimensionalprojection of DC-8 flight.

27.90N I 3:15

3:10

2:50

2:42

26.70N 135.30E 136.80E

Figure3. Intensivesurvey DC-8 flight mutes and winds for September 27,1991. Flag, 25 ms'l; a full bar,5 ms'l;a halfbar, 2.5 ms'] . (a) Boundarylayer inflow region, 300 m. Timeinterval between DC-8 windsis 1 min. (b) Eye region,11360 m. Time intervalbetween DC-8 windsis 1 min. (c) Outflow region(circles are mdiosondewinds at 200 hPa), 11200m. Time intervalbetween DC-8 windsis 2 min. NEWELL ET AL' DC-8 SAMPLING OF TYPHOON MIREILLE ! 857

32.80N

5:: 13

5:

5:30

:1o

.:57

31.70N I 128.40E 129.70E

37.00N

/ : 7:10 7:14 •_•_•-•-- ---•- - .....i,•-•7:00 ......

5:32

30.00N 128.00E 138.00E

Figure 3. (continued)

The NASA DC-8 left Yokota Air Base, Tokyo, at section of Plate 2. The DC-8 entered the eye at about 0105 UT on September27, 1991,as Mireille approached 0457 UT througha gap in the highestclouds and remained Kyushufrom the southwest with a centralpressure of 935 in the eye, which by then was beginningto fill in with hPa. A boundarylayer mn at 300 rn was madeat about clouds, until about 0535 UT, then exited to the northeast. 28øN, 136øE,and the aircraftthen climbed to 11.3 km and At 0550 UT the aircraft headed toward the southeast at 12 penetratedthe cloudregion surrounding the eye (see Plate km and returnedalong the samepath, headingnorthwest, 1 satellite image) which, according to cloud top at 11 km. The purposewas to samplethe atmospheremore temperaturesand vertical temperature profiles, exceeded or less perpendicularto the outflow from Mireille. At 15.5km in height.The overallaircraft track is summarized 0638 UT, the end of this maneuver, the DC-8 turned in the horizontalprojection of Figure2, whichincludes the toward the east-northeastand returned toward Tokyo, aircraft-measuredwind velocities,and in the vertical cross where it landed at 0739 UT. 1858 NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE

540[ 520•- 500• 1700 E48o[ ,•sol- 1650 ,,,1 .... • .... i,, • • ,,t.,,,I .... I .... I, 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 Hour CUT) 440t 420•- 300f,,,,•.... •.... •.... •.... •.... •.... •.... •.... •.... •.... •.... •.... •,,, 400 •, 200[

100•0 a.... j.,,,L• I .... '"'•""'"'•,,I ...... 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 6 .. . Hour (UT) 100•.,,,•.... • .... • .... • .... • .... • .... • .... • .... • .... • .... • .... • .... •,,, r•4•- 75 .... , i., ,i .... i,,, i, ,i,,, i ,i i ,i, ,i .... i .... i so 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5

2 100 Ho• •) 90 == 80 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 Hour (U'D [ so70 = = 30001•,,,,• .... • .... • .... • .... • .... • .... • .... • .... • .... • .... • .... • .... •,,, 50 =soo 2000•- 20 = =

1000•- 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 soo•-0"' • • ,I ...... I 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 Hour (UT) 300•- Hour(UT) 14 250•- = = = 12 200[:- . . 10 •150• ". ." •O"100•. ======"======0 50•,, B E 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 2' ' •i,• 3 •.,• 4 •,I,• 515 6 6.5 7 7.5 Hour(UT) o "i'g ...... '" ...... ' .... ' Hour (LIT)

120

•-• 100 S0 ß , ** ..3000 ß , ßß ....? , *******,, • •oo •0 ß** , , * * *** * ** ** •2000 •0 ß • 1500 2 3 4 5 6 7 8 • 1000 240 ,-• 200 ß • 0 . , ,* ...... 4 5 6 7 160 , , • ***** 7• •120 ß* • 50• •0 ,•, •, • 4000 3000 2 3 4 5 6 7 8 • 2000 150

, -•- 12o 1 2 3 4 5 6 7 • 90 , ** *****% , , *** , ß , • •o ß , • 800 • 3o o 600

o • 400 1 2 3 4 5 6 7 8 000 • 200 800 ß* •* * * ß •**• • , ******•,•***• 0 2 3 • 5 6 7 8 600 400 200 B 0 1 2 3 4 5 6 7 8 HOUR (UT)

Figure4. Timeseries of selectedconsfituems for theentire flight for September27, 1991. B represemsboundary layer mn (043242UT-0315 UT) andE representstime spent in typhooneye region (4)457 UT-0533UT). (a) CH4, CO, 03, and H20 mixingratio from Lyman-otinstrument (plus flight altitude). Resolutionof dataused: CO and CH4, 5 s; 03, 10 s; H20, 10 s. Co)OCS, CS2,DMS, and SO2. (c) C6H6(benzene), C2H2 (ethyne[acetylene]), C3H8 (propane), and C2H6 (ethane).(d) CH3OOH(methyl-hydroperoxide), H202 (hydrogenperoxide), and OH frommodel by Daviset al. [this issue].(e) NO, NOy, peroxyacetyl nitrate (PAN), and ambient temperature (difference between GIT andKondo discussed by Crosley[this issue]). (f) Aerosolconcentrations. (g) Ratiosof acetylene/COand propane/ethane. (h) Detailed(10 s) temperature,water vapor, aircraft wind speed,and aircraft altitude from 0500 to 0700 UT. NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE 1859

200 C•rcle GA Inst Tech 180 Star Kondo oø o ß •.o[ 40 o

o 80 o o• •0 40 20

2 3 4 5 6 7 8 o.o...... 800 Qrele GA Inst Teeh HOUR •T) 700 Star Kondo

5• o 0.20[ 4•

2• 1• 1 2 3 4 5 6 7 8 0.00..... [ ...... I loo 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 90 HOURCUT) • 70 • 60 -25 "--" 50 ** 40 ** ** * Z 30 ** -30 ** * * ** **, • 20 ** x , , ** ** ** , ** ,x ** ** ** -35 0 1 2 3 4 5 6 7 8 -48

-45

lo -50 •-• o 4 5 50 55 60 65 70 7 5 o lO 3000 E-• 20 30 40 1580 50 2 3 4 5 6 7 8 1808 HOUR CUT) 580

4 5 50 55 60 65 70 7 5

48

38 50 AEROSOL 05 to 20 /•

40 lO

20 50 55 60 65 78 75 14

10 13

12 2 3 4 5 E 6 7 f I 11 HOUR (UT)

9

58 55 00 65 /0 75 HOUR (UT)

Figure 4. (continued)

Intensivesampling was carried out in threeregions: the also containsthe flight altitude,and Figure4e includes boundarylayer inflow region at about 300 m between ambientair temperature.Time resolutionof the dataused 0241 UT and0315 UT; theeye and surrounding wall cloud in Figure4a is givenwith theFigure 4a caption.For other region, samplingat about 11.3 and 12 km between0457 constituentsthe data are collected at variable intervals and and 0535 UT; and the outflow region(a term usedhere to are plottedat the midpointof the collectioninterval. Each denotethe northeastquadrant) between 0550 and 0714 UT. time seriesset hasthe boundarylayer and eye regions Theaircraft-reported winds are plotted in Figure3a-3c for markedon the time axis. Constituentshaving their lowest thesethree regions. Where available, radiosondewinds valuesin theboundary layer included CO, 03, SO2,C2H2, (see,for example,Figure 3c) andwinds from commercial C2H6,C3H8, C6H6, NO, NOy, and PAN. PANis actually aircraft flights (AIREPs) are also included (see, for belowthe detection limit; its concentration isvery sensitive example,Figure 2). Mostof thebody of thispaper deals to temperature,decreasing at highertemperatures [Singh, with the atmosphericchemistry findings in thesethree 1987]. Thus indicationsare that the air enteringthe regions and their possible interpretationin terms of typhoonfrom the southin the southeasternsector is clean atmosphericmotions. marine air. The marine origin of the inflow was also indicatedby the fact that the highestDMS valuesof the missionoccurred in the boundarylayer. The higher aerosolconcentrations in the boundary layer (Figure 40 are 2. Resultsof Boundary Layer Sampling thoughtto be associatedwith seasalt. In additionto thetime series of constituentsthemselves, time seriesof ratiosof two pairsof constituentswhich can Time series of selected constituents are examined first be interpretedin termsof air massage are presentedin to place the boundary layer inflow region in proper Figure4g. As discussedelsewhere [Gregory et al., this perspective. Figure 4 includestime seriesfor the entire issue],an acetylene/COratio of <0.5 correspondsto well- mission of CH4, CO, 03, H20, OCS, C82, CH3SCH3 agedair, whilea ratio>1 suggestsemissions only a few (DMS), 802, C2H2,C2H6, C3H8, C6H6, NO, NOy, .daysold. Thesevalues apply when acetylene isexpressed CH3C(O)OONO2 (peroxyacetylnitrate (PAN)), H202, in pptv and CO is in ppbv. The lifetimeof acetylene CH3OOH,OH (modelcalculated), and aerosol. Figure 4a (C2H2)is about11 days,while thatof carbonmonoxide is 1860 NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE

Rodriguezet aL, unpublishedmaterial, 1995].) The same generalview is givenby the propane/ethaneratio; these two gaseshave lifetimesof 8 and 34 days (see above 27 commenton OH levels). Using all of the abovecriteria, the boundarylayer air enteringfrom the southcan be considered well aged. Further comments about the hydrocarbondata are given below. The watervapor mixing ratios and aircraft temperatures 2•7 28 29 '30 31 '32 33 for the tropical boundary layer are shown togetherin Figure 5, from which it can be seen that higher temperaturesare associatedwith drier air; correlation coefficients between these two variables are about -0.5. 17 The spatialseparation of the descendingdry plumesis tens of kilometers,as may be seen from a map of specific humidityduring the boundarylayer sampling(not shown). The higher temperaturesat 300 m are produced by subsidence,implying a downwardheat transportby these 14 2 7 2.8 2.9 • 3.0 3.1 3.2 3.3 motions. Becausesinking motions are relativelydry and, HOUR OJT) as will be seenlater, there is overall convergenceinto the typhoonregion in the lower levels,rising motions in the Figure 5. Time series of DC-8 1-s samplingof samearea will be relativelymoist, and the net effect is an temperatureand specific humidity during boundary layer upwardtransport of moisture. •'here is a largevertical mn at about300 m. Units, degreesCelsius and gramsper gradientof H20 mixing ratio in erielowest 2 kin, as seen kilogram. from verticalprofiles made on descentto and ascentfrom theboundary layer (an exampleis shownin Figure6, to be discussedbelow). This is maintainedby the strongwinds about34 days;by contrast,under the sameconditions, (>15 ms-l) and consequemhigh evaporationfrom the DMS would have a lifetime of approximately2 days. surface. The temperaturegradient between 850 hPa (1.5 (These lifetime estimatesuse a diurnally averagedOH km) and 300 m is about6.5øC km 'l, close to moist level of 1.4 x 106cm '3 which was basedon averaged adiabatic.The subsidingplumes are an importantfactor in atmosphericconditions during PEM-West A centered ventilating the boundary layer with air from the free around a latitude of 30øN [see Davis et al., this issue; troposphere. Other speciesmeasured in the boundarylayer, but not illustratedhere, include 7Be, which has a concentrationof O3 (ppbv) 46 fCi m'3, relativelylow comparedwith values of over •.o...... !,9 ...... •,o, ...... ,•,o, ...... ,•,o, ...... ,•,o, ...... ,•,o, ...... •o, ...... ,¾, ...... ¾, ...... •.o 200fCi m-3 in theupper troposphere, and 2]øpb, which has 7.5 a value of 3.6 fCi m'3, quite similarto two'other measurementsin the upper troposphere made before 7.0 7.0 penetrationof the eye and not far from the value measured in the eye. Well ahead of the eye, a final fifth 6.5 6.5 measurementyielded-•0.5 fCi m'3. These data are

6.0 03 6.0 obtained from aerosol collection and typically involve H20 samplingtimes of 30-40 min. 5.5

5.0 5.0

4.5 5:33 - 4.0

3.5 3.5 -

3.0

2.5 2.5

2.0

1.5 1.5 -

1.0 1.0

0.5 0.5

.

0.0 ).0

31 70N t I I C I L 128 40E 129 70E Figure6. Verticalprofiles of specifichumidity and ozone from DC-8 measurements.This profile was madeduring Figure 7. Map of DC-8 air temperaturein vicinityof eye the descentto the boundarylayer between 0223 UT and at about11.3 km on September27, 1991. Pointsare 60 s 0242 UT on September27, 1991. apart. Units,degrees Celsius. NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE 1861

32 80N 32 80N , • • • , • , , , , , • 5:33 5:33 _ 1•3. /.•.o

1 .

_

.,

_ - 7. ' 3' 5:10 i5.7 - 3 •1:57 _ O. 1. ß -

_

3170Nt 129 70E 128 40E 129 70E 128 40E Figure 8. Map of DC-8 water vapor mixing ratio in Figure 9. Map of ozonein vicinityof the eye at about vicinity of eye at about 11.3 km on September27, 1991. 11.3km September27, 1991. Pointsare 60 s apart. Units, Pointsare 60 s apart. Units,parts per millionby volume. partsper billion by volme.

3. Resultsof Samplingin Eye and Wall Cloud the eye, comparedwith typical values of about 2000 fCi m-3 in thestratosphere. Again the implication is that Region stratosphericair is not being drawn down into the eye region. The lidar crosssections (Plates 3a and 3b) show very low ozoneinside the eye, generallycharacteristic of The aircraft reachedthe eye at about 0500 UT at 11.3 equatorialboundary layer air, from 5 km to a heightof km and ascended to about 11.9 km between 0528 and about 16 km. The gap in coveragein the 0511-0513 UT 0530 UT, just beforeleaving the eye. The DC-8 1-nfinair periodin theseimages occurs when the aircraft was turning temperaturesare mappedin Figure 7; the eye centeris into the wall cloud from the north (see Figure 8 water about 6øC higher than its surroundings(compare 0457, vapor values). It is also worth noting that there is a 0517, and 0520 UT values). Temperaturevalues every negativecorrelation between H20 and03 in the time series 10 s are in Figure4h andshow variation of up to 6øCin the andgenerally a positivecorrelation between H20 andwind eye. Sinking motion in the eye with concomitant speed(cf. Figure 4h). This indicatesthat moist, high- compressionand adiabatic warming is suggestedand velocityair hasless ozone than dry slower-movingair; it would account for the lack of cloud there in the GMS can be inferred from the limited radiosonde data available imagesat and before0000 UT. A relevantschematic of thatwind speedsare decreasing above -12 km, sothe latter this type of circulationhas been given by IVilloughby air is subsiding.Carbon monoxide is characteristicallylow [1988]. The water vapor mixing ratio measuredby the in the stratosphere(-50 ppbv) and can also be usedas a Lyman-a instrumentshowed quite dry patches,270-500 tracerof stratosphericair; but in thiscase (Figure 4a), the partsper million by volume(ppmv), in someregions of the boundarylayer air gavethe lowestvalue (-70 ppbv),and eye, againsuggesting subsidence, mixed with valuesup to theair aheadof thetyphoon showed 115-120 ppbv (Figure 2580 ppmv, and it appearsthat cloudparticles could also 4a), so the fact that the eye containeda few valuesin the be involved in the latter (Figure 8). Patchy cloudswere 103-110ppbv (Figure 10) rangecould not be interpretedas quite abundantwithin the eye region,and thesemay be indicatingstratospheric air; in fact, it couldbe claimedthat responsiblefor the high aerosol concentrationafter someof the air originatedfrom the boundary layer. It was 0518 UT. The ozone concentrationsin the eye region noted earlier that DMS had its highestvalues in the (Figures9 and 4a) are relatively low (-25-30 pans per boundarylayer (-90 partsper trillion by volume(pptv); see billionby volume(ppbv)), although them are a few spikes Figure4b). After theDC-8 left theboundary layer, DMS up to 40-50 ppbvin the centraldrier region, but themare valuesreturned to their limit of detectionuntil 0454 UT, no high values (>100 ppbv) as might be expectedif just beforeentering the wall cloudconvection region; at stratosphericair waspresent. Suggestions have been made thatpoint they increased to 80 pptv. In theeye, there were that stratosphericair couldbe entrainedinto the eye, but threeobservations: 25, 42, and48 pptv(see Figure 4b). On theseozone measurements suggest that Bergeron[1954] leavingthe eye at 0537 UT, highvalues (78 pptv) were provideda more attractiveexplanation of the higher againassociated with the wall cloudregion; values are temperaturesin the eye by arguing,as notedearlier, that considerablylower in the outflowregion, as maybe seen the subsidence responsible occurred within the fromFigure 4b, addingweight to theinterpretation that the troposphere. Nevertheless,there may be small-scale highestvalues are transported from the boundary layer into intrusionsof stratosphericair into the upper troposphere the uppertroposphere in the wall cloudregion. Someof well above the aircraft, and one possible example is this wall-cloudair sinksinto the eye, while the remainder included in the ozone lidar cross section of Plate 3a. The movesaway from the typhoon in theupper troposphere. concomitant aerosol lidar cross section shows cirrus cloud In general,tracer species showed a rangeof mixing almostcompletely covering the eye at about16 kin, thus ratiopatterns when viewed in theboundary layer, the free indicatingno large-scalesinking motion at thatlevel. It is troposphereahead of the eye, and the eye itself. For alsoworth noting that the highest 7Be concentration is200 example,the mixing ratios of 7Be,03, C2H2,C2H8, C3H8, fCi m'3 in theupper troposphere and about 80 fCi m-3 in C6H6,C82, and CO werelowest in theboundary layer, 1862 NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE

TYPHOON PEM WEST-A FLIGHT 9 27 SEP 91

RELATIVE AEROSOL SCATTERING (IR) 0 400 800 1200 1600 2000 I I I I I I

5:08 5'12 5:16 UT , I • I • I

18- -18

.J17-

16-

- u.J15_ - 14- - 13--

32.34 32.01 32.25 N LAT

128.94 129.34 128.92 E LON

OZONE (PPBV) 60 120 180 240 300 I I I I I

5:08 5'12 5:16 UT ,. ! • I • I

19-

<18- ,,,½ - u,j 17- • - • 16- I-- - J <15-

14- -14

32.34 32.01 32.25 N LAT

128.94 129.34 128.92 E LON

Plate 3. Airbornelidar aerosoland ozonecross sections of the eye region. (a) Lookingup. (b) Lookingdown. NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE 1863

TYPHOON PEM WEST-A FLIGHT 9 27 SEP 91

RELATIVE AEROSOL SCATTERING (VS) 0 500 1000 1500 2000 2500 I I I I I I

5'08 5:12 5'16 UT , I • I • I

32.34 32.01 32.25 N LAT , i , t , ! 128.94 129.34 128.92 E LON

OZONE (PPBV) 20 40 60 80 100 I I I I I

5:08 5:12 5'16 UT , I • I i I

10 9 _ 8 7 6 5 5 4- 4 3 3 2- 2 1- 1 - 0- 0 32.34 32.01 32.25 N LAT i , I I 128.94 129.34 128.92 E LON

Plate 3. (continued) 1864 NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE

32 80N contactwith the wall cloudregion, may havehad higher 5:33 _ -,poe concentrations of some of these constituents than were measuredin the boundarylayer to the southeastof the storm.In consideringthe latter hypothesis, it is important to keepin mindthat while the sampled boundary layer was overthe open ocean, the air whichfed directlyinto the wall cloudwas over the southernportion of Kyushuat thetime of sampling,and the shallowbays and coastal waters and land areascould have had much higher concentrations of severalof thetracer species. The observed relatively high values of SO2 may also have originatedfrom surface :57 anthropogenicsources in Japan.However, it is alsoto be •9 notedthat them were two activevolcanoes in the area, Sakura-Jima(31.58øN, 130.67øE, 1360-m elevation) and Unzen(32.75øN, 130.03øE, 1118-m elevation), and as may be seenfrom Plate 2, air fromtheir vicinity could easily havebeen entrained into the eye region. 31 70N ,t 128 40E 129 70E eye:Aerosol5-10 pam,'clesvalues cm-(Fi.•ure from4f) 0457show totwo 0518 ranges UT andnear 15-35 the Figure 10. Map of carbonmonoxide in eye regionat particlescm -a until the end of the flight. As for the hydrocarbons,these aerosols may have originated from the about 11.3 km on September27, 1991. The time interval continent. betweenclosely spaced points is about60 s. There are a numberof missingdata points. Units, parts per billion by volume. 4. Upper TroposphereOutflow RegionAhead of Mireille whileDMS washighest. Approximate mixing ratio values for severalof thesespecies (e.g., 7Be, 03, CS2,and CO) Wind velocitiesin this regionare shownin Figure3c for all threeregions are shown in Table1, togetherwith the and temperature,H20, DMS, and 03 in Figures 11-14. deducedfraction of boundarylayer air that, when mixed The closenessof the two legs to the southeastand with air from the free troposphere,accounts for the northwest(0550-0635 UT) enablesa verticallapse rate to observedmixing milos in the eye. Problemsarise from the be foundas 7.4øCkm -1 at 0610UT (-48.5øCat 11.95km assumptionthat the eye air is drawn only from two versus-39.5øC at 10.73 km). High watervapor mixing sources,from the fact thatthe boundary layer air sampled ratiosare encounteredbetween the eye and 34øN, 13IøE was far removedfrom the boundarylayer contributing (Figure 12), and thesecorrespond to intenseclouds on the mostdirectly to the wall cloud regionand from the fact GMS image(Plate 1). DMS declinesafter passage through thatthe free troposphere contribution came from a variety thiscloudy, moist region (Figure 13), againsuggesting that of air massesthat were entrainedhorizontally by the the active convectionis bringingup materialfrom the typhoonsystem. Nevertheless, these results suggest that a boundarylayer that has been processed by passagethrough significantfraction of the air in the eye wall regioncould the typhoonsystem. What is termedoutflow is thereforea haveoriginated from the boundary layer. mixtureof air thathas been advected into the typhoon area The four hydrocarbonsshown in Figure4c all increased in theupper troposphere and air whichhas passed through with altitudeand were found to be higherin andahead of the rising motion region of the typhoon. An aerosol the eye thanin the boundarylayer. The hydrocarbonsare samplefrom 0540 to 0638UT gavea 7Bevalue of 63 only slightly diluted,if at all, in the eye wall region, fCi m-3, which compares with 46 fCi m-3 in theboundary suggestingthey may have originated (see comments layer,and therefore gives further evidence for this point. below)from the continent.Carbon disulfide (CS2) may The ozonevalues along the sortieto the southeastshow a have had a similar path as the NMHCs. Still another generaldecrease with altitude(Figure 14), whereascarbon possibilityis that the boundarylayer which is in direct monoxideshows little change(Figure 4a).

Table 1. Mixing Ratiosin BoundaryLayer, Free 5. Air Mass SourcesFrom Meteorological Troposphere,and Eye Regionat 11-12 km Data TogetherWith ResultingFraction of Air in theEye DerivedFrom the Bounda .ry Three approachescan be used to examine possible Layer sources:trajectory analysis, streamline analysis, and considerationof the rotationaland irrotational components Boundary Free % of Eye of the wind. In trajectoryanalysis, one tries to follow the Layer(BL) Troposphere Eye FromBL motion of an individual parcel of air; the procedure, assumptions,caveats, and applicationsto PEM-West A casestudies are given by Merrill [thisissue]. The problem CO, ppbv 70.0 116.0 105.0 24.0 with applicationof trajectoryanalysis techniques to a 7Be,fCi m-3 46.0 205.0 81.0 78.0 typhooncase is that large verticaldisplacements occur in DMS, pptv 94.0 6.0 38.0 36.0 regions of strong convection, which invalidate the 0 3, ppbv 15.0 55.0 30.0 62.0 assumptionthat motionfollows an isentropicsurface at the C2H2,pptv 30.0 185.0 160.0 16.0 large scale. Other problems arise because machine CS2, pptv 1.0 5.8 4.5 27.0 analysesoften lack the fidelity necessaryto reproduce reality [cf. Danielsen, 1974], and they are basedon an NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE 1865

37 OON 37 00N ,, , ,

-_ -4••6:•6 ø•a•'• ::

30 00N 30 00N • I • ] I • I i I 128 00E 128 00E 138 00E

Figure 11. Map of DC-8 air temperaturein the outflow Figure 12. Map of DC-8 watervapor mixing ratio in the regionon September27, 1991. Circles are radiosonde outflowregion at about 11.2 km on September27, 1991. temperaturesat 195 hPa. Time interval betweendata Time intervalbetween data pointsis 5 min. Units, parts pointsis 5 min. Units,degrees Celsius. per millionby volume. observational network that cannot catch the spatial v = vir+ vr =-V Z + k x V• (1) variationsat the typhoonscale. Streamlineanalysis too is limitedby the horizontalspacing between radiosondes that Takingthe divergence and curl in mmyields go into it, but streamlineanalysis can be usedqualitatively to identify the broad-scaleair massesin the region. V. v =-V2z (2) Streamlinesand wind vectorsare shownin Figure 15 for a level of 1000 hPabased on grid poim data suppliedby the EuropeanCentre for Medium-RangeWeather Forecasts cud v = V:• (3) (ECMWF). There is convergenceof a marine air mass from the east and a continental air mass from the west in the vicinity of 130øE. Streamlineshave been used to Fromthe observed or analyzedwind field, Z andW may be characterizetropical storms in the southernhemisphere and obtainedby solutionof the Laplaciansinvolved, and givepatterns which are almost a mirrorimage of Figure15 thence,v i, andv, maybe found. It wasfortunate that the [Holland, 1984]. The DC-8 samplingof the boundary ECMWF'analysigwas chang. edon September 17, 1991, 10 layer at about27øN, 136øEwas madein the marineair. daysbefore the DC-8 missionto TyphoonMireille, to a Accordingto the traceconstituent measurements reported higher resolution,T213, which gives the smallest in section2, this air is cleanand aged. wavelengthresolved at theequator to be 190km. While To estimatethe overall massflux in the typhoonregion, thisanalysis cannot resolve the eye region, it is sufficient the ECMWF wind velocitieswere decomposedinto two sothat estimatesof the irrotationalflow over the largearea components:an irrotationalpart which hasno curl and a coveredby thetyphoon may be made; it is thislarger-scale rotationalpart whichhas no divergence.While the first is flowthat represents the response to therising motion in the usuallysmaller than the second,it is themost important for eye wall region. Examplesof the divergentflow the massflux throughthe system. These parts may be componentsare given in Figures16 and17 for 1000and written in terms of a velocity potential Z and a stream 700 hPa. There is a significantdrift inwardtoward the function centerof up to-8 ms-l, andMireille clearly dominates this

37 00N 37 00N

30 00N 30 00N t I I I • I I I I 128 00E 138 00E 128 00E 138 00E

Figure 13. Map of DMS in the outflow regionat about Figure 14. Map of ozonein the outflow regionat about 11.2 km on September27, 1991. Time interval between 11.2 km on September27, 1991. Time interval between datapoints is 5 min. Units,parts per trillion by volume. datapoints is 5 min. Units,parts per billion by volume. 1866 NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE

60N

,

50N

40N

30N,

20N ... :.,,c--•ee---•- -. -,':,"' ' , :• I , z", -ill-- , , ',k..•--?,.- ff-:-?.... •------k--

i; c ' 'r•,' - --',b - -,'

EQ ;',," ! ,,',,: 100E 110E 120E 130E 140E 150E 160E 100E 110E 140E 150E 160E

Figure 15. Streamlinesand wind vectorsfor 0000 UT, Figure 17. Velocity potentialand divergentwind September27, 1991,based on ECMWF analysisfor 1000 componentat 700 hPafor 0000 UT, September27, 1991. hPa. Units, metersper second. Maximum vector (near Windvectors in metersper second.Velocity potential in eye) is 21 metersper second. 104m 2 s']. Maximumvector (near eye) is 6 metersper second. componentof the flow over a large area &110ø-150øEat 30øN). At 700 •a it appearsthat a substantialfraction of Asiancoast from the west. Theoverall picture is thatthe air nearthe typhoon-influenced region has originated from typhoonboundary layer from a very large area is the continent;more air entersfrom the west than from the convergentinto the central region, rises in the cloud east, whereas at 1000 hPa, contributions from east and systemssurrounding the eye, andthen moves out from the westare almost the same.By 300 hPa(not shown) there is eye regionaloft. Tracegases in the boundarylayer are a divergentregion in andahead of thetyphoon-influenced thus redistributedinto the upper troposphere;the region;the streamlinesshow a troughapproaching the magnitudeof this processwill be examinedfurther below. In addition,there is substantialinflow in themiddle levels, muchof it fromthe continent. From the cloud pattern and the higher temperaturesin the eye it is clear that subsidenceoccurs there in the uppertroposphere [cf. l,Villoughby, 1988], but the total massinvolved would 50N

H 40N 199 150100 I/ •/• • .

30N '•' ,' ,',, -- •d'-'. -.. . /,.4' •,: • 300 ' ;/"'""'*• '•[••'•" '"•' " 20N i[!:'•]]i]•!•i?J:L-'...... ' ...b...•..l._..k ..•,...•: ...... ,__ !...... j._]-:_ r-

• 500 '

I ....::::::::::::::::::::::::::::::::::::::::• .....[ • • [,-, .... .::•::• ...... :• [ ,_,,,.,i : ', , ,_:__• ton:::::::::::::::::::::: ....L•. Zz•z=:L..'•:::•;':•.:[.._, ...... •.... a--g---•o• ...... :...... ,..... , .... 'ZOO , ,

1000 ' ' :'.' . i . EE.0N E6.0N :30.0 34 0 :38.0 IOOE 1 fOE IEOE 130E 140E 150E 160E 127 5E 127 5E 127 5 127 5 127 5 127 5

Figure 16. Velocity potentialand divergentwind Figure 18. Cross sectionof vertical motion through componentat 1000hPa for 0000 UT, September27, 1991. Mirefile (located at about 30øN, 127.5øE)from ECMWF Windvectors in metersper second.Velocity potential in analysis at 0000 UT, September27, 1991 based on 10n m2 s']. Maximumvector (near eye) is 9 metersper divergent wind calculated from ECMWF data. Units, second. Pascalsper second.Contour interval is 0.2. NEWELL ET AL' DC-8 SAMPLING OF TYPHOON MIREILLE 1867

net income flux net income flux -0.137 -0.338 -0.201 5.576 1.451 -4.124 lflO hPa lot hPa II -5.032 - - -5.309 -1.202 - = -2.253 150 hPa + + 15fl hPa + -7.424 - - -10.224 -1.726 - -•- -5.329 2fit hPa + + 200 hPa + -6.155 - - -4.167 -0.679 - -•- -5.928 258 hPa. + + 25fl hPa + -4.096 - - 0.471 0.230 - -•- -3.377 + 3Off hPa + + 300 hPa + -3.894 - - 0.299 -0.685 - -•- -3.187 q- 4tO hPa + + 4fir hPa + 0.710 - - 0.872 -0.578 - -•- -0.360 5fit hPa + + 5fit hPa q- 7.730 - - 4.894 3.296 - -+- 4.703 7fit hPa + + •' 7fit hPa 7.722 - - 4.372 3.149 - •-- 5.231 850 hPa + + 850 hPa + 10.302 - - 8.591 3.771 - -.- 6.375 1 tilt hPa 1 tilt hPa W E S N

Figure 19. Exampleof massflux componentsacross the walls of thebox in Figure16 for September 27, 1991,based on divergent wind calculated from ECMWF data. Units, 10 9 kg s-1 . appearto be only a smallfraction of that cardedupward in The ECMWF crosssection of vertical motion through the cloud regions. In an effort to estimatethe vertical the storm(Figure 18) is not inconsistentwith theselow motion in central core, it is assumedthat the heat balance values. In fact, the subsidencein the eye is ratherlimited, within the core is between adiabatic compressionand andthe cylindricalvolume shown could be regardedas a radiativecooling. The relationshipis Taylor column, as outlined further below. These conclusionsdo not agreewith the relativelylarge sinking RT Q motion at 5 km in the core suggestedby Willoughby to 3p pCp=•p = W[Fd- F] (4) [19881. 6. Large-Scale Mass Flux whereF 4 is the dry adiabaticlapse rate (= g/Cv,about 10øCkm 4) and F = - 3T/3z. Q was calculatedfrom the nearestradiosonde to the eye with good data. Station A 17 1/2ø x 17 1/2ø latitude-longitudebox was drawn 47807 at 33.5øN, 130.4øE at 0600 UT was used. To surroundingthe typhoon(Figure 16), andflow into and out simulate conditions in the eye, air temperature was of the box was evaluatedeach day from the divergemwind increasedby 6øCat 250 hPabetween the environmentand component.An exampleappears in Figure 19 where it is the eye, taperingto IøC at 100 hPa and 3øC at 700 hPa. evident that the majority of the i•fflow occursbelow 500 Specifichumidity was taken directlyfrom the sounding. hPa. We are interested here in the mass flux passing Cooling rates calculatedfor the upper tropospherewere betweenthe boundarylayer and the free troposphereand about3.3øC d 'l. Thelocal lapse rate was 7.7øC km 'l. The havetherefore calculated the verticalflux through850 hPa. valuefor w wasthus -1.7 cm s-l, a smallsubsidence. Values in Table 2 show that Mireille was sampledon the dayof maximummass flux, 29 x 109kg s-1 Thesevalues are comparedwith TyphoonOrchid sampledon October8, Table 2. Mass Flux From Lower to Upper Troposphere in Table 2; Mireille has about doublethe flux by Orchid but clearly a larger sample needs to be studied and Flux perUnit Area TotalFlux comparedwith the flux by middle-latitudecyclones before 1991 10-3 kg m '2 s'1 109 kg s4 the overall significanceof thesevalues from Mireille can be assessed. It would also be valuable to make an estimate SupertyphoonMirei lle of the actualarea involvedin the risingmotion, perhaps by usingweather radar. It is alsoworth noting from Figure 19 September24 5 20 that as well as the boundary layer, the next two layers September25 6 21 abovealso contribute substantially to the massflux through September26 6 27 the system. For examp•le,the 700- to 500-hPa layer September27 9 29 contributesabout 20 x 10v kg s-• withat leastone third of September28 5 19 this materialcoming in from the adjacentAsian continent September29 3 13 to the west,as may be seenfrom Figure 17. Is the typhoonimportant in the generalcirculation in TyphoonOrchid carrying trace constituentsinto the upper troposphere? October 5 3 14 This questionis examinedbelow for the caseof DMS. October 6 4 17 The PEM-West missions showed that DMS only October 7 5 20 reachedthe free tropospherefrom the boundarylayer on October 8 5 18 threeoccasions: on September16 overthe North Pacificon October 9 4 15 a day of very strongconvection as seenfrom the cloud October 10 5 16 videos; on September27 in associationwith typhoon Mireille; and on October 8 in associationwith typhoon 1868 NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE

100 lOO 90 OKI 90 KENTING 80 80 70 70 60 60 50 50 40 40 30 30 z 20 o 20 10 lO 0 o

DATE DATE

Figure 20a. Surface ozone variation at Oki Island Figure 20b. Surfaceozone variation at Kenting,Taiwan (36.3øN, 133.2øE)in the period September17 to October (22øN, 120øE),in the period September5 to October3, 23, 1991 (provided by H. Akimoto). Units, parts per 1991 (providedby C.M. Liu). Units,parts per billion by billion by volume. volume.

Orchid (Table 3). For typhoon Mireille the mass of Where30½/3z > 0 andthe systemis stable,the air spreads sulphurcarried as DMS to the free tropospherefrom the out horizontally,forming the layersnoted. A summaryof boundarylayer can be estimatedfrom the valuesfor total the propertiesof theselayers is given elsewherein this mass flux derived above in conjunctionwith the mass issue [Newell et al., this issue],where it can be seenthat mixing ratio of DMS. The molecularweight of DMS is the characteristicsof Figure6 arefairly typical. 62, and the conversionto equivalentmass of sulphurfrom volumemixing ratio of DMS is therefore(ppv) x (62/28.9) 8. Ozone From Surface Stations x (32/62)= 1.11. For a boundarylayer mixing ratio of 80 pptvand a massflux fromTable 1 of 9 x 10-*kgm -2 s'l, The only surfacestation sampling trace constituems the flux of sulphurto the free troposphereis 80 x 1.11 x which was close to the path of Mireille was Oki Island 10-12 x9x 10-3kgS m'2 s'l, or69 •gSm '2d'l. This (36.3øN, 133.2øE),which was in operationas part of the compareswith recent estimatesof surfaceoceanic DMS experimem"Perturbation by East Asian ContinentalAir sulphurflux which have yielded 96 •g S m'2 d'l at 30øNin Mass to Pacific OceanicTroposphere" (PEACAMPOT). It thecentral Pacific [Quinn et al., 1990],227 •g S m-2 d 'l in is noteworthythat ozone mixing ratiosas the eye passed theequatorial Pacific [Bates et al., 1993],and 27 gg S m'2 over Oki at-4200 UT were downto 5.5 ppbv(Figure 20a). d-I in theNortheast Pacific [Bates et al., 1994].Typhoon The stationat Kenting,Taiwan (22øN, 120.9øE)reported a Mireille thereforegenerated a DMS sulphurflux into the value of about3 ppbv on September27 (Figure 20b), when upper free troposphere comparable to that normally the surfacepressure in the Kenting area was about 1007 associatedwith the flux from the oceaninto the boundary hPa. At 0600 UT on September26, Mireille was about layer. 550 km to the northeastof Kerning,and there was a trough The use of machineanalysis may underestimatethe massflux circulatingthrough the typhoonsystem, as the radiosondewind fields may be smoothedexcessively. Table 3. Mixing Ratiosof Dimethylsulfide(DMS) in the There are relatively few estimatesof the massflux in the BoundaryLayer and the Upper Troposphere literature. McBride [ 1981] usedabout 800 rawinsondesto make a compositeof Pacific typhoonsand computed •,,.,,,.:•oencefrom the surface to 350hPa as 28 x 10-3 kg Mission 1991 BoundaryLayer >8 km m-2 s-I evaluatedover an 8ø diametercircle, while our value from Figure l0 for the surface to 400 hPa is 21 4 September16 320,260 0 x 10-3 kgm -2 s-I evaluatedover an areaabout 6 timesas 5 September18 170 90 large. 6 September22 30, 40 3-6 7 September24 35, 12 4 8 September25 90, 110,80 2 7. Small-Scale Features 9 September27 95 80 10 October 1 120 2 During the descentto the boundarylayer at 0228 UT 11 October 2 near28øN, 136øEthe aircraftmeasured the vertical profiles 12 October4 11, 8 2 of ozoneand water vapor shownin Figure 6. The layered 13 October6 45, 48 2 structure between about 5 and 6 km shows that water- 14 October 8 34 45 vapor-richair is associatedwith lower ozonethan water- 15 October 12 65 5 vapor-poorair. Theselayers seem to be about400 m thick; 16 October 13 30 2 similarlayers show in the lidar echoesnear to the centerof 17 October 15 22, 27, 30/38 6 18 October 18 35 6 the eye (Plate 3b). They appearto provideevidence of air from the boundarylayer folded into higherlevels, probably 19 October 19 23 3 by convectiveprocesses. It is noteworthythat for this to 20 October20 24, 30 2 occur,there must be substantialinstability; that is, 30e/3Z < 0. (0½is equivalentpotemial temperature and z is altitude.) Units:Parts Per Trillion by Volume NEWELL ET AL: DC-8 SAMPLING OF TYPHOON MIREILLE 1869

in the directionof Kenting associatedwith the imerplay observationswere not taken in the maximum velocity that hadbeen taking place between Typhoon Mireille and regiongives a reasonabletime of about4 days,well within TyphoonNat whichhad previously passed over Kenting. the C2H6 lifetime. It is clear that convectionanywhere The low ozonevalues at Kenting (Figure 20b) may have along the back trajectory could have introduced the been associatedwith air which had beenbrought up from material into the upper troposphereand that the vertical the lower latitudesby the typhoons. Unfortunately,the shear could have produceda vertical gradient of C2H6, station power was turned off during the passageof such as was observedlanding at Tokyo, where values at TyphoonNat on September23. In the caseof Oki the 7.5 km were aboutdouble those in the boundarylayer. transferis more easilyunderstood, as the eye passedover the station(see Figure 1). The low (5 ppbv) ozonevalues hereare in generalagreement with thosereported by lidar 11. Photochemical Environment in the troposphericregion of the eye (see Plate 3). We speculatethat the eye region is a cylindricalsample of Althougha detaileddescription of the photochemical tropical air "packaged"by the circulationwhich has a environmentof typhoonMireille is not the main focusof continuityrather like that of a classicalTaylor column. this paper,for purposesof completenessthe authorshave Taylor [1923] performedsome experiments to checkan summarizedsome of the generalphotochemical features of analysisby Proudman[1916] which showed that all slow this systemand have then comrastedthese to air parcels steady motions of a rotating liquid must be two sampled in the same geographical region under dimensional. Taylor showed that when a sphere or nontyphoonconditions. For example,in Figure 4d the cylinderis pulled acrossa rotatingtank of water, it is observedmixing ratios for the photochemicalproducts accompaniedby a cylinderof water aboveit havingthe H202 andCH3OOH are reportedas a time series,together samediameter. He did thisby ejectingdye into the water with model-calculatedprofiles of the centrallyimportant abovethe sphereor cylinderand notingthat it diverged oxidizingspecies, the hydroxyl radical. If themidday time aroundan imaginarycylinder above the sphereor cylinder. period of 0330 UT to 0540 UT is examined,it can be If dye was ejectedinto the imaginarycylinder, it stayed shownthat the medianlevel of H20 2 (basedon 3-min insidethe cylinder and moved across the tank with it. It is averageddata) is 1032pptv, whereas that for CH3OOHis this imaginarycylinder which is calleda Taylor column. considerablylower, being 193 pptv. Model-estimatedOH Yih [1977]reports that Taylor also experimemed with fish concentrationsfor thissame time period ranged from 1.6 x in thetank, and they would try to avoidthe cylinder as if 106to 8.1x 106molecules cm -3, giving a medianvalue of sensingthe presenceof a solidobstacle. Low mixing 3.6x 106molecules cm '3. Someof thehighest estimated ratiosof tropicalozone found in both the laser images OH valuesare seenfor thosetimes when the aircraftwas (Plate3) and in the Oki surfacestation data (Figure20a) approachingor passingthrough the eye wall region;it was are thoughtto correspondto the dye in the experiment. at thesetimes when highly elevatedlevels of NO were The equivalentpotential temperature computed from the observedby the GeorgiaTech instrument. By way of ECMWF data(Figure 21) showsneutral stabili .ty near the contrastto the typhoonenvironment, clear air observations typhooncenter, but againthe linfitedhorizontal resolution over a similar altitude rangeand geographicallocation as must be borne in mind. typhoonMireille gave midday median mixing ratios for H202and CH3OOH of 412and 125 pptv, respectively; and theav, erage midday median OH levelwas estimated at 2.1 9. Total Ozone Relationships x 10ø molecules cm'3. Thus one might tentatively concludethat the typhoon environment on averagehad an There have been several studies of the association enhancedphotochemical environment relative to the betweentotal ozoneand tropical cyclones, including those backgroundupper troposphere. For further details on the in the western North Pacific. Some rather complex photochemistryof this system,the readeris referredto relationshipswere foundwhich depended on whetherthe otherpapers in thisissue [e.g., Davis et al. andHeikes et systemswere intensifyingor not intensifying[e.g., Stout andRodgers, 1992]. We wereprovided with Total Ozone MeasurementsSatellite (TOMS) dataeach day in the field, and after the missionwe obtaineddigital data. The maps 12. Conclusions for September26-27 showed no special relationship between Mireille and total ozone in accord with the lidar As notedin the introduction,typhoons were sometimes and cirrus cross sections discussed earlier. These thoughtto draw stmtosphericair down into the upper commentsapply to the regionwe couldmeasure directly. troposphere.Ostlund [1968] found experimemal evidence The questionconcerning stratosphere to troposphere for such transfer from aircraft measurements of tritium exchange outside the eye cannot be examined which, like ozone,is higherin the stratospherethan in the experimentallywith our measurements.However, it was troposphere.The DC-8 explorationof Mireille foundno noted that high potential vorticity (PV) values were experimental evidence for this process in the lidar indicatedon the ECMWF 315K analysisdrawn up for this measurementsof ozone; in fact, lidar measurementsin the experiment. middle tropospherewere very low. Them was evidence, particularlyfrom DMS, thatboundary layer air wascarded into the upper troposphere. The maximum vertical 10. Possible Remote Sources transportseems to occurin the wall cloudregion; some of the DMS-rich air which reachesthe upper troposphere A crosssection along the Asiancoast of the zonalflow movesinto the eye and subsides,while the majorityis in (not shown;see Bachmeier et al. [this issue]for related the divergent flow that spreads out in the upper data)illustrates that materialmay haveentered the region troposphere.There is thereforea net transportof sulphur from the west. This material could have reached the Asian from the boundary layer to the upper free troposphere. westcoast from Europein a few days. Even allowingfor Computationsof the irrotationalcomponent of the flow meridional meandering and the fact that aircraft suggestthat the boundarylayer from a largearea is drawn 1870 NEWELL ET AL' DC-8 SAMPLING OF TYPHOON MIREILLE

lOO 30 N lOO 30 N 100 30 N 126.5 E 127.5 E 128.5 E

15o 15o 150

200 200 200

250 250 a• 250 300 300 • 300

400 400 • 400

5OO 500 500

600 600 60O

700 70O 700

850 850 850

lOOO ß , J 1000 1000 , ,t, , , ,•• '•1'• ! i , • 320 360 400 280 328 360 488 280 320 360 400 K K

lOO 29 N lOO 31 N 127.5 E 127.5 E

150 150

200 2OO

250 a• 250 300 • 300

400 • 400

5OO 500

600 60O

700 70O

850 850

1000 , a , , , , \,•'• , , , i 1000 ß ' 288 320 360 400 288 328 368 488 K K

Figure 21. Potentialtemperature (solid lines) and equivalent potemial temperature (dashed lines) for pointsnear the eye center on September27, 1995at 0000UT, basedon ECMWF analysis. into the centralregion. The massflux from the boundaN also observed. These enhancementsmay have arisen if layerto thefree troposphere by Mireillereaches 3 x 10lø someof the air involvedentered the systemin the lower or kg s-1, peaking on the day the typhoon was sampled by the middle tropospherefrom the continent. The tracegas NO DC-8. This is about 50% greater than the flux from was in some casesnearly 2 ordersof magnitudehigher typhoon Orchid, which the DC-8 passed through on than in the boundarylayer, and them is evidencethat one October 8, and almost triple the flux from an active of the importantsources of this NO was in situ production cyclonein the Noah Pacific on September6, 1991 (values from lightning [see Davis et al., this issue]. The DMS not shown). The flux of sulphuras DMS transportedinto appearsto haveoriginated from the marineboundary layer, theupper free troposphere is approximately 69 gg S m'2 althoughsome of this may have enteredthe systemfrom d'l, whichis estimatedto be nearlyhalf the meanflux the coastalregions, as suggestedby Plate2. The tracegas from the ocean into the boundary layer. To assessthe SO2 may havebeen entrained from the vicinity of Japanas importanceof typhoonsin the globalsulphur budget, more a resultof volcanicactivity. samplesare required and these, in turn, will need to be In sampling Typhoon Mireille, it is important to incorporatedinto a globalmodel to form an assessment. recognizethat both continental and marine boundary layers Boundarylayer marineair was found to be low in the were mixedtogether by the typhoon'scirculation patterns. tracerspecies 03, CO, CS2,C2H2, C2H6 C3H8,and C2H6, Thusin view of thefact that the continentalboundary layer andrich in DMS. In theupper levels of iheeye region and was not sampleddirectly, it is difficult to distinguish in the outflowregion, NMHCs, 03, andCS2 were foundto contributionsfrom the continentalboundary layer and the be far more abundant, and modest increasesin CO were continentalfree troposphere. NEWELL ET AL' DC-8 SAMPLING OF TYPHOON M! REILLE 1871

The low ozone in the eye region, shownby the lidar Quinn,P. K., T. S. Bates,and J. E. Johnson,Interactions between measurementsand by Oki Island station,may have been the sulphurand the reducednitrogen cycles over the central transporteddirectly from low latitudesin a processakin to PacificOcean, J. Geophys.Res., 95, 16,405-16,416,1990. a Taylor column. Low ozonevalues aloft would also fit Rudolph,D. K., and C. P. Guard,1991 AnnualTropical Cyclone with an ultimate low-latitudesource for the air in the eye Report, pp. 108-111, Joint Typhoon Warning Cent., region.Ostlund (1968) also mentioned that some of his COMNAVMARIANAS, , Mariana Islands,1992. tritium resultscould be explainedif the eye containedold Singh,H. B., Reactivenitrogen in the troposphere,Environ. $ci. air which had been trapped since the formation of the Technol.,21,320-327, 1987. circulation. Finally, the NMHC data suggest the Stout,J., and E. B. Rodgers,Nimbus-7 total ozoneobservations possibilitythat distantsources to the west may contribute of westernNorth Pacifictropical cyclones, J. Appl.Meteorol., to materialfound in the typhoon. 31,758-783, 1992. Taylor, G.I., Experimentson the motion of solid bodies in Acknowledgments. The work was supportedunder the rotatingfluids, Proc. R. $oc.London, A, 104, 213-218, 1923. NASA GTE programgrant NAG-l-1252. We acknowledgehelp Willoughby,H. E., The dynamicsof the tropicalcyclone core, in the field with real-timetrajectory analyses and afterwardwith Aust.Meteorol. Mag., 36, 183-191, 1988. datafrom ECMWF. The meteorologicalstaff at YokotaAir Base Yih, C. S., Fluid Mechanics,p. 82, West River Press,Ann Arbor, kindly helped throughoutthe Mireille flight and provided Mich., 1977. analyzedsynoptic maps afterward.

References H. Akimoto,University of Tokyo,4-6-1 KomabaMeguro-ku, Tokyo 153 305, Japan. (e-mail: [email protected] Bachmeier,A. S., R. E. Newell, M. C. Shiphain,Y. Zhu, D. R. tokyo.ac.jp) Blake, and E. V. Browell, PEM-West A meteorological B. E. Anderson,E. V. Browell, G. L. Gregory,G. W. Sachse overview,d. Geophys.Res., this issue. and M. C. Shipham,NASA LangleyResearch Center, Hampton, Bates,T. S., K. C. Kelly, and J. E. Johnson,Concentrations and VA 23681. (e-mail' [email protected]' fluxesof dissolvedbiogenic gas (DMS, CH4, CO, CO2) in e.v.browell•larc.nasa.gov; [email protected]' the equatorialPacific during the SAGA 3 experiment,d. g.w.sachse•larc.nasa. gov; shipham•weathennan. larc.nasa.gov) Geophys.Res., 98, 16,969-16,977,1993. A. S. Bachmeier,Lockheed Engineering and Sciences,Co., Bates,T. S., R. P. Kiene,G. V. Wolfe, P. A. Matrai, F. P. Chavez, Hampton,VA 23681. (e-mail:[email protected]) K.R. Buck, B.W. Blomquist,and L. Cuhel, The cyclingof A. R. Bandy and D.C. Thornton, Drexel University, sulphur in surface seawaterof the Northeast Pacific, d. Philadelphia,PA 19104. (e-lnail:arb@ac 1.chemistry.drexel.edu; Geophys.Res., 99, 7835-7843, 1994. dct•ac 2.chemistry. drexel.edu) Bergeron,T., Reviexvsof ModemMeteorology, 12, The problem D. R. Blake and F. S. Rowland,University of California at of tropicalhurricanes, Q. d. R. Meteorol. Soc., 80, 131-164, Irvine, Irvine, CA 92717. (e-mail: dblake•orion.oac.uci.edu; 1954. rowland•uci.edu) Crosley,D. R.,NOy BlueRibbon panel, d. Geophys.Res., this J. D. Bradshaw, J. H. Crawford, D. D. Davis and S. T. issue. Sandholm,Georgia Institute of Technology,Atlanta, GA 30332. Danielsen,E. F., Reviewof trajectorymethods, Adv. Geophys., (e-mail' jb57•prism.gatech.edu; [email protected]. 18(B), 73-94, 1974. gatech.edu;dd [email protected]; ss27@ prism.gatech.edu) Davis,D. D., et al., Assessmentof ozonephotochemistry in the W. Brockett, L. DeGreef, D. Lewis, D. McComfick, and E. western North Pacific as inferred from PEM-West A Monitz, AmesScience and Applications Aircraft Division, NASA observationsduring fall 1991, d. Geophys.Res., this issue. Ames ResearchCenter, Moffett Field, CA 94035. Gregory,G. L., A. S. Bachmeier,D. R. Blake, B. G. Heikes, D. J. E. Collins Jr., Scienceand TechnologyCorp., Hampton, C. Thornton,A. R. Bandy, J. D. Bradshaw,and Y. Kondo, V^ 23681. (e-mail:j.e. collins•larc.nasa.gov) Chemicalsignatures of agedPacific •narine air: Mixed layer J. E. Dibb and R. W. Talbot,University of New Hampshire, and free troposphereas measuredduring PEM-West A, J. Durham,NH 03824. (e-mail:[email protected]) Geophys.Res., this issue. B. G. Heikes and J. T. Merrill, University of Rhode Island, Heikes, B. G., et al., Hydrogen peroxide and peroxide Kingston, RI 02881. (e-mail: zagar•notos.gso.uri.edu; distributionsrelated to ozoneand odd hydrogenover the jmerrill•boreas.gso. uri. edu) NorthPacific in thefall of 1991,d. Geophys.Res., this issue. R. E. Newell (correspondingauthor), W. Hu, Z-X. Wu, and Y. Hoell, J. M., D. D. Davis, S.C. Liu, R. E. Newell, M. C. Zhu, MassachusettsInstitute of Technology, 54-1824, 77 Shipham,H. Akimoto,R. J. McNeal, R. J. Bendura,and J. W. MassachusettsAvenue, Cambridge, N• 02 • •o r.... •. Drewry,Pacific Exploratory Mission-West A (PEM-WestA): newell(•}newell1.mit. edu) September-October,1991, d. Geophys.Res., this issue. K. K. Kelly and S.C. Liu, NOAA Aeronomy Laboratory, Holland, G. J., On the climatologyand structureof tropical R/E/AL-4, 325 Broadway, Boulder, CO 80303. (e-mail: cyclonesin the Australian/southwestPacific region, I, Data liu•al.noaa.gov) andtropical stonns, Aust. Meteorol. Mag., 32, 10-15, 1984. Y. Kondo and M. Koike, NagoyaUniversity, Japan. (e-mail: McBride, J. L., Observationalanalysis of tropical cyclone kondo•stelab.nagoya-u.ac.jp;koike•stelab.nagoya-u.ac.jp) formation,1, Basicdescription of datasets, J. Atmos.Sci., 38, C.-M. Liu, Departmentof AtmosphericScience, National 1117-1131, 1981. Taiwan University, Taipei, Taiwan. (e-mail' Merrill, J. T., Trajectoryresults and interpretation for PEM-West [email protected]) A, J. Geophys.Res., this issue. F. Sakamaki, National Institute for Environmental Studies, Newell, R. E., et al., Vertical fine-scaleatmospheric structure Tsukuba,Japan. measuredfrom the NASA DC-8 during PEM-West A, d. H. B. Singh, NASA Ames ResearchCenter, Moffett Field, Geophys.Res., this issue. CA 94035. (e-mail:hanwant_singh•qmgate.arc.nasa.gov) Ostlund,H. G., Hurricanetritium, II, Air-seaexchange of water in Betsy1965, Tellus,20, 577-594, 1968. Proudman,J., On the motionof solidsin a liquid possessing (ReceivedJune 23, 1994; revisedMarch 11, 1995; vorticity,Proc. R. Soc.London, A, 100, 114-121, 1916. acceptedMarch 30, 1995.)