Detection of Volcanic Ash Clouds from Nimbus 7/Total Ozone Mapping Spectrometer

Michigan Technological University Digital Commons @ Michigan Tech

Department of Geological and Mining Department of Geological and Mining Engineering and Sciences Publications Engineering and Sciences

7-1-1997

Detection of volcanic ash clouds from Nimbus 7/total ozone mapping spectrometer

C. J. Seftor Hughes STX Corporation

N. C. Hsu Hughes STX Corporation

J. R. Herman NASA Goddard Space Flight Center

P. K. Bhartia NASA Goddard Space Flight Center

O. Torres Hughes STX Corporation

See next page for additional authors

Follow this and additional works at: https://digitalcommons.mtu.edu/geo-fp

Part of the Geology Commons, Mining Engineering Commons, and the Other Engineering Commons

Recommended Citation Seftor, C. J., Hsu, N. C., Herman, J. R., Bhartia, P. K., Torres, O., Rose, W. I., Schneider, D. J., & Krotkov, N. (1997). Detection of volcanic ash clouds from Nimbus 7/total ozone mapping spectrometer. Journal of Geophysical Research, 102(D14), 16749-16579. http://dx.doi.org/10.1029/97JD00925 Retrieved from: https://digitalcommons.mtu.edu/geo-fp/79

Follow this and additional works at: https://digitalcommons.mtu.edu/geo-fp Part of the Geology Commons, Mining Engineering Commons, and the Other Engineering Commons Authors C. J. Seftor, N. C. Hsu, J. R. Herman, P. K. Bhartia, O. Torres, William I. Rose, David J. Schneider, and N. Krotkov

This article is available at Digital Commons @ Michigan Tech: https://digitalcommons.mtu.edu/geo-fp/79 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 102, NO. D14, PAGES 16,749-16,759, JULY 27, 1997

Detection of volcanic ash clouds from Nimbus 7/total ozone mapping spectrometer

C. J. Seftor, • N. C. Hsu, • J. R. Herman, 2 P. K. Bhartia, 2 O. Torres, • W. I. Rose,3 D. J. Schneider,3 and N. Krotko½

Abstract. Measured radiancesfrom the Version 7 reprocessingof the Nimbus 7/total ozonemapping spectrometer (TOMS) 340- and 380-nm channelsare usedto detect absorbingparticulates injected into the atmosphereafter the E1 Chichoneruption on April 4, 1982. It is shownthat while the single-channelreflectivity determined from the 380-nm channelis able to detect cloudsand haze composedof nonabsorbingaerosols, the spectral contrastbetween the 340- and 380-nm channelsis sensitiveto absorbingparticulates such as volcanicash, desert dust, or smokefrom biomassburning. In this paper the spectral contrastbetween these two channelsis used to detect the volcanicash injection into the atmosphereand to track its evolution for severaldays. The movement of the ash cloudsis shownto be consistentwith the motionsexpected from the National Centers for EnvironmentalPrediction (NCEP)-derived balancedwind fields in the troposphereand lower stratosphere.The movementof the volcanic SO2 cloud detectedfrom TOMS data was also in agreementwith the NCEP wind at higher altitudesof up to 100-10 mbar. The verticalwind shearin the neighborhoodof the eruption site resultedin a clear separation of the ash and SO2 clouds.The location and movement of the ash cloud are consistent with informationobtained by the advancedvery high resolutionradiometer (AVHRR) instrumenton board the NOAA 7 satellite and to ground reports of ash fall.

1. Introduction On the basisof radiative transfer calculationsin a Rayleigh scatteringatmosphere, the presenceof aerosolschanges the Atmosphericradiance information from the Version 7 Nim- spectraldependence in the radiances.Depending on the char- bus 7 and Meteor 3 total ozone mapping spectrometer acteristicsof the aerosols,they can be either absorbingor (TOMS) data setshas been usedto generatea 16-yearrecord nonabsorbingin the UV spectral region [DMtmeida, 1987; of ozonechanges. Of the sixwavelength channels available on Patterson,1981; Pattersonet at., 1983; Pattersonand McMahon, these instruments, the three that are insensitive to the amount 1984]. The spectraldependence is most pronouncedfor UV- of ozonein the atmosphere(340, 360, and 380 rim) canbe used absorbingaerosols, which cause R x to increase with wave- to determine the effectivereflectivity of the lower boundary, length. While the ozone retrieval algorithm is designedto R x. R x can be determined by comparing the measured up- minimize the spectraldependence from clouds,other typesof wellingradiance to the atmosphericbackscattering from a pure nonabsorbingaerosols, under certain conditions,can causeR x Rayleighatmosphere over a Lambertian surface.If cloudsare to decreasewith wavelength. TOMS data can therefore be present, two different Lambertian surfacesare assumed,one usedto clearlydistinguish between absorbing particulates (e.g., representingthe ground and one representingclouds. The smokefrom biomassburning, desert dust, and volcanicash) calculatedreflectivity using the 380-rim channel radiance is a and nonabsorbingparticulates (e.g., water clouds,haze, and measureof the presenceof clouds,haze, or a reflectiveground volcanicH2SO 4 aerosols). surface such as snow or ice. The recalibration of Nimbus In practice, the spectral contrast is measured and tracked 7/TOMS in the Version 7 data set permits the wavelength through a quantity known as the aerosol index: dependenceof the effectivereflectivities to be calculatedwith an accuracyof about 0.1% and the absolutereflectivities to be calculatedto better than 1%. Details of the algorithmused to A= -100log10 I3807 ....q- 100 log10 I38ø/] calc' generatethe Version 7 data setsare given by McPeterset at. where (I34o/I38o).... is the measuredspectral contrast be- [1996]. Details of the Version 7 calibration of the Nimbus-7 tweenthe 340- and 380-nmradiances and (1340/1380)calcis the TOMS instrumentare givenby Wettemeyeret at. [1996] and of spectralcontrast between the 340- and 380-nm radiancescal- the Meteor 3 TOMS instrument are given by Seftor et at. culatedusing a Rayleigh scatteringatmosphere and reflectivity [1997]. determined from the 380-nm channel. Since the effective reflectivity is determined by requiring 1HughesSTX Corporation,Greenbelt, Maryland. 2NASAGoddard Space Flight Center, Greenbelt, Maryland. (138o).... : (138o)calc, 3Departmentof GeologicalEngineering and Sciences,Michigan A = - 100 log10[ (I340).... /(I340)calc]' (2) TechnologicalUniversity, Houghton. Copyright1997 by the American GeophysicalUnion. Without additional information about the absorbingparticu- Paper number 97JD00925. latesobserved in the TOMS data (e.g.,the refractiveindex and 0148-0227/97/97 JD-00925 $09.00 particle size distribution),the aerosol index can be used to

16,749 16,750 SEFTOR ET AL.: DETECTION OF VOLCANIC ASH CLOUDS FROM TOMS determinetheir locationand the relative amountof particulate a scanninginstrument, and the size of the footprint and cor- matter [Hermanet al., 1997]. respondingsize of the rectangleincrease with the scanangle. This paper is one of a seriesof papers reporting on the Figures la-ld show the derived SO2 cloud for four days, ability of TOMS instrumentsto detect troposphericaerosols. April 4, 5, 6, and 8 (data were missingon April 7) as TOMS Previouspapers have describedusing TOMS data to detect passedoverhead at localnoon (1800 UT). Figurela showsthat biomassburning [Hsu et al., 1996] and map the global distri- the initial SO2 cloud coveredan area extendingeast past the bution of UV-absorbingaerosols [Herman et al., 1997].Subse- Yucatan Peninsula and west to the Atlantic Ocean. Winds quent paperswill detail the theoreticalradiative transfer as- derived from NCEP data indicate predominantly westerly pectsof detectingtropospheric and volcanic aerosols at TOMS winds in the troposphereand easterlywinds in the strato- wavelengths. sphere.The zonal spreadingof the cloudindicates shearing of In this paper the detectionof volcanicash is demonstrated thevertical column of SO2 by these winds. On April 5 and6 the by looking at the April 4, 1982, eruptionsof E1 Chichon.The displacementof the SO2cloud was mainly toward the westand developmentof the ash cloud is mappedand its motion is northwest,following the winds at roughly 10-30 mbar. These followed. Since the TOMS wavelengthchannels can also be motionsare comparableto those seen in other observations usedto derive the total column amount and geographicaldis- [Barthet al., 1983]and in previousversions of the TOMS data tribution of SO2 in the atmosphere[Krueger et al., 1995],com- [Krueger,1983]. Krueger alsoindicated that the amountof SO2 parisonsare madebetween the motion of the ashcloud and the releasedinto the atmospherereached values higher than 700 SO2cloud. The ashcloud motions are confirmedfrom infrared DU; the upper limit of 200 placed on the SO1 precludesa satelliteobservations obtained by the NOAA 7 advancedvery comparisonwith values obtained from the Version 7 algorithm. high resolutionradiometer (AVHRR). Finally, in order to demonstratethe abilityto determineoptical depths from such data, a samplecalculation is performedfor April 6, 1982,from 3. The Ash Cloud a radiative transfermodel usingaerosol parameters consistent with volcanic ash. Ash from the eruption of E1 Chichonwas seen to circum- navigatethe globe [Robockand Matson, 1983] and was ob- servedto spreadvertically between 15 and 35 km over Mauna 2. The El Chichon Eruption and Resulting Loa [De Luisi, 1982].TOMS was able to track the formation SO2 Cloud and dispersionof the ash for the first few days after each Two of the E1 Chichoneruptions occurred on April 4, 1982, eruption.The distributionof the TOMS detectedash cloud at 0135 and 1122 UT, respectively[Sigurdsson et al., 1984]. afterthe April 4 eruptions,as represented by positivevalues of the aerosolindex, is shownin Figures 2a-2d. In these figures, These two eruptions (which, along with a smaller one on March28) injected7 (___2)x 109kg of SO2[Bluth et al., 1992] measurementartifacts such as sea glint, which can be easily determined from geometry, have been removed. The areal to levels as high as 20-26 km in the atmosphere[Carey and coverageof the ashcloud on April 4 (Figure 2a) agreeswell Sigurdsson,1986]. The SO2 effectswere significantfor a period with the coverageseen by the GOES satellite,using both vis- of a few weeks and then decreasedas the SO2 became con- ible (0.55-0.75 /xm) and thermal infrared (10.5-12.5 /xm) verted into H2SO4. The H2SO4 aerosolpersisted in the strato- bands,at approximatelythe sametime [Matson,1982]. Exten- sphereand tropospherefor severalyears before gradually fad- sive field studies of ash fall were coducted at 99 locations ing backto pre-emptionamounts. This samephenomenon was coveringan areaof about45,000 km 2 [Varekampet al., 1982]. alsoobserved by TOMS after the Mount Pinatuboeruption in These studiesindicate that ash from the April 4 eruptionwas June 1991 [Torreset al., 1995]. depositedprimarily in an eastwarddirection, with the 1-mm The ability to determine the presenceof SO2 from TOMS isopachat approximately210 km. Theseresults are consistent radiance measurementshas long been known, and an algo- with the area of extent determinedby TOMS. rithm designedto detect SO2 was containedin previousver- While the displacementsof both the ashcloud and the SO2 sionsof the TOMS processingsystem [Krueger, 1983]. For the cloudwere roughlythe sameon April 4, there wasa significant Version7 processingof TOMS data an improvedalgorithm to geographicalseparation of the ash cloudfrom the SO2 cloud detect SO2 was implemented.This algorithmroutinely pro- by April 5. The ashcloud itself separatedinto two clouds,one ducesan index,called the sulfurdioxide index (SOI), whichis movingnorth and one moving south. related to the amount of SO2 in the atmosphere.Radiative The movementsof the ash cloud were comparedwith tra- transfer studies indicate that the Version 7 SOI is an accurate jectoriescalculated from NCEP balancedwind data; the winds measure of the amount of SO2 for ash-free plumes in the were calculatedon a 2ø latitude by 5ø longitude grid. Figures3a atmosphereup to 100 Dobsonunits (DU) (with an SOI of 1 and 3b show the NCEP balanced winds at various altitudes for being equivalentto 1 DU). The SOI beginsto lose some April 4, 5, and 6. The latitude and longitude of the winds accuracybetween 100 and 200 DU and becomesunreliable shownare the grid pointsclosest to the centerof the two ash above200 DU. The presenceof ashcauses the SO1to under- clouds.The windsspeed and directionsindicated in Figure 3a estimatethe amount of SO2; the underestimationis propor- coincide with the movement of the northern cloud, while those tional to the ashoptical thickness. Details of the algorithmare in Figure 3b coincidewith the southerncloud. Since there was givenby McPeterset al. [1996].For thiswork an upperlimit of only one ash cloud on April 4, the wind speedand directions 200 was placed on the SO1, and it was usedto determinethe for this day are duplicatedin Figures3a and 3b. geographicallocation and extent of the SO2 cloud. As shownin Figure2a, a stronglypositive index was detected The SO2 cloudrepresented by the SO1 for April 4 is shown by TOMS overthe E1Chichoneruption site (17.33øN, 93.2øW) in Figure la. The areal coverage("footprint") of the TOMS on April 4. Becausethe TOMS measurementswere made measurementshas been approximatedby rectangles.TOMS is more than 6 hoursafter the eruption,a considerablespread of SEFTORET AL.:DETECTION OF VOLCANICASH CLOUDS FROM TOMS 16,751

a SOI for April 4, 1982

...... Loo...... • .• .. : gg , , . , 5 ...... 1...... •... :•j:;•..•......

ß ß ß

SOI

b SOI for April 5, 1982

--,

i : .•.•.-..•..+-• •: •::,: 7": -•.----".:f...... ,?•:::'":::: '?:•:•.• ...... :•':•'-'•!?•:,:i:

: ...... -:...... •t ...... •,-•......

. , , •iZO .....',•lO0...... 80 , .

• ...... : ?:;•te'•......

SO1 Figure1. Thederived SOIcloud, asseen byTOMS, for (a) April 4, (b) April 5, (c) April 6,and (d) April 8,1982. April 7is not shown because adata gap severely limited areal coverage. Theareal coverage ofeach TOMSscan-element measurement isapproximated byrectangles; thesize of the field of view (and the correspondingrectangle) increases with scan angle. theash cloud from the source was observed. The ashcloud was shownin Figure3a, in theneighborhood of the eruption site transportedin the east-west directions covering most of the andat approximately 70-200 mbar (near the tropopause). areain southernMexico and northern Guatemala. These mo- OnApril 5 (Figure2b), one section ofthe ash cloud moved tionsare consistent with the 1200UT NCEPbalanced wind, northinto the Gulf of Mexico,while the bulk of the ashcloud 16,752 SEFTORET AL.' DETECTION OF VOLCANIC ASH CLOUDS FROM TOMS

½ SOI for April 6, 1982 ...... :.... •...-:•;•iii...... •'-

25. ' ...... •.•(- - , .

0 . ,

, :..::,:..- •.. . :: .... • ...... • ...... , :•..--.--;:.--• :..z..... ,

...... ß ...... ,

}120 __ ..!.-100 ...... ½80

. 5 '

......

C;• tO-1 • tO-1 C;•

ß $o!

d SOI for April 8, 1982

• C,,-i

. ß SOl

Figure 1. (continued)

drifted slightlyto the south.According to the NCEP wind radiosondedata from Veracruz [Matson,1982; Matson and patternin Figure3a, the northwardmovement was only pos- Robock,1982]. Reports of a low-altitude(1.5-2 km) cloudwere siblein the lower troposphereat an altitude at or below 850 reportedblowing over the southern Texas coast on April 5, and mbar.The troposphericnature of thiscloud is alsoindicated in a lightashfall occurred in Houstonduring the nightof April 7 SEFTOR ET AL.: DETECTION OF VOLCANIC ASH CLOUDS FROM TOMS 16,753

a AerosolIndex for April 4, 1982 f ...... : ...... -: -' :•?'""'•:• .....m!•i' ...... :...... -• ......

: ...... :...... •...... •...... '" ...... ":'-:.-,.d-Z':;;•L,.•,•: ....•:":;•' ""• ......

......

Aerosol Index

b AerosolIndex for April 5, 1982 ... / ...... i•....; ...... •...... -...... ,..•:•...... •'-",• '" •:25.:' ...."' ...... : '...... :":-"--'""-•""-'"•'•:':•••--- "-••.:-.•••,!•i......

Figur• •. Tho ashcloud, as soonby TO•S throughtho calculationof tho aorosolindox, for (a) April 4, April 5, (c) April 6, and (d) April 8.

[SmithsonianInstitution Scientific Event Alert Network(SEAN), tors at 14øNand 90øW on April 5. The weak south-southeast 1989].These reports also indicate that the northerncloud seen movementof the ash cloud is in agreementwith NCEP winds by TOMS is containedin the low troposphere. at approximately50-70 mbar, and Veracruz radiosondedata Figure 3b showsthe verticalprofile of horizontalwind vec- also indicate that this cloud is stratospheric[Matson, 1982]; 16,754 SEFTOR ET AL.' DETECTION OF VOLCANIC ASH CLOUDS FROM TOMS

½ AerosolIndex for April 6, 1982

/...... i••.••'...... •...... •!•...... •11.... •'...... •...... "

. • ,

. ,

......

.!-80 .

...... :•

Aerosol Index

d AerosolIndex for April 8, 1982 ...... • ...... -.•-,•,:•...... : ...... ,:• ......

: ...... : ...... -•--•i•g•!i'...... •:...... •.:.::•:..:•,:• -:•i•5•: '"-'--'."•'.-•;•,, "'-?:•':

: ...... : '...... •....• ...... •..•, ...... :i ...... iI.

•::•.•'•2•i•:•::•:•'•"•':•i•i•i• •:.•:.:•:.::•:..:..:.::.:...... : .-::.,:.5•-•.:.• :•::::•:::•'•

Aerosol Index

Figure 2. (continued) the SO2 cloud follows the NCEP wind pattern at about the lowertroposphere winds for April 6 depictedin Figure3a. 10-30 mbar. The bulk of the ash cloud observedin the coastalregion near On April 6 (Figure2c) the bulk of the ashcloud detected the PacificOcean on April 5 movedto the southwest,as seen previouslyin the Gulf of Mexicoon April 5 wasstill visible and by TOMS on April 6, again under the influenceof strato- only driftedslightly to the west.Again, this is consistentwith sphericwinds. This ashcloud continued to travelto the west, SEFTOR ET AL.: DETECTION OF VOLCANIC ASH CLOUDS FROM TOMS 16,755

N 70mbar • 500 mbar 500 mbar ',

, ,

, 100mbar • 700mbar • 700mbar • _ '.

, 200mbar • 850 mbar 850 mbar •

500 mbar 1000 mbar 1000mbar ""-•,,'...•..

18N, 95W 22N, 95W 22N, 95W 4/4/82 4/5/82 4/6/82

Northern Ash Cloud

b N 30 mbar 5mbar • •E ,: I 50 mbar 10mbar • 10 mbar :

70mbar • 30 mbar 30 mbar ' 7

100mbar xx.•.;..•' 50mbar • 50mbar •"---•N

200mbar x,...... b// 70 mbar 70mbar •

100mbar 100mbar • 500mbar• , 18N, 95W 14N, 90W 12N, 95W 4/4/82 4/5/82 4/6/82

Southern Ash Cloud

Figure 3. NCEP balancedwinds at locationsand altitudesconsistent with the movementof the (a) northern ash cloud and (b) southernash cloud. following the wind circulation in the lower stratosphereon and the surfaceis large enough,brightness temperatures de- April 7; it then was split by the northerly and southerlyflow in terminedfrom band 4 (10.3-11.3 txm) and band 5 (11.5-12.5 the region near 12øN and 98øW, as shownin Figure 2d. txm) are used to discriminatevolcanic ash cloudsfrom mete- The separationof the ashcloud from the SO2 cloudfor the orologicalones [Wen and Rose, 1994]. These data can be used E1 Chichoneruption appears to have resultedfrom the vertical to retrieve optical depth and effectiveradius of volcanic ash wind shearin the altitude regionwhere the two cloudsresided. particles,and ash burdens and mass retrievals can also be This separationis consistentwith lab experimentsperformed performed(D. J. Schneideret al., Observationsof sulfurdiox- to studygas-ash separation processes [Holasek et al., 1996]. ide and volcanicash in the April 4-7, 1982, E1 Chichonvolca- Data from AVHRR instruments can also be used to detect nic cloud as seenwith total ozone mappingspectrometer and volcanicash. If the temperaturecontrast between the ashcloud advancedvery high resolutionradiometer, submittedto Jour- 16,756 SEFTOR ET AL.: DETECTION OF VOLCANIC ASH CLOUDS FROM TOMS

AVHRR Temp Difference, April 4, 1982 a ,'11,,..•.::.•:.',:i:,•,i,.•i.i:.,•.•.:•,•::.L:,i:i...... ,.ij•!½-•.•-•i•i"•'•'•"•"••"•••'••••••••!'•' •" ' •'

Degrees C

' AVHRR Temp Difference,April 5, 1982 b '."::.':'.:.:....:....":'• ...... '•::•'i•..y..'-•:i ....'•------•-.-----•:•••-•••.•••-.--•:, & • • • 6

Degrees C Figure 4. The ashcloud, as seen by AVHRR throughthe differencein brightnesstemperature, for (a) April 4, (b) April 5, and (c) April 6. Note the shiftin scalein Figure4c. AVHRR Temp Difference,April 6, 1982 ; ..,';.'..,...'••..,•,...... •.•.•.•.••, ...... • :..::.:,..-.-•..t•i•.:D•i•.•....,.•••••••..,•...••,--,---•.,,••--•--••-•.,-.•••••••!•...... •.•?....:...•...... ,•

Degrees C Figure 4. (continued)

OpticalThickness for April 6, 1982

1-80 ,

Figure5. Thecalculated optical depth calculated forApril 6, 1982. The aerosol model used in the radiative transfercalculation is givenin the text. 16,758 SEFTOR ET AL.' DETECTION OF VOLCANIC ASH CLOUDS FROM TOMS nal of GeophysicalResearch, 1997) (hereinafterreferred to as maps derived from AVHRR. For this volcanicevent, TOMS Schneideret at., submittedmanuscript, 1997). For this paper, data showed that the ash cloud motions were different from the only the original brightnesstemperature differencesare pre- SO2 cloud motions due to vertical wind shear.When the data sented.Figures 4a-4c showthe movementof the E1 Chichon are usedin conjunctionwith informationon the opticalprop- ash cloud as seenby AVHRR on board the NOAA 7 satellite ertiesof absorbingparticulates (layer altitude,size distribution between 2030 and 2230 UT. of the ash,and refractiveindex), radiative transfer calculations The developmentof the cloudand its movementgenerally can be used to estimatethe ash optical thickness. agreewith that observedby TOMS. The ashsignal detected by Although the current Earth Probe and ADEOS TOMS in- the AVHRR in the Gulf of Mexico is very weak and is cut off strumentsdo not have a 340- or 380-nm wavelengthchannel, when convertedto massfraction. This might be due to the fact an aerosol index constructedfrom the spectral contrast be- that the cloud is very low in elevation, leading to a very small tween the 331- and 360-nm channels has also been shown to be temperaturecontrast between it and the underlyingsurface. effective in detecting and tracking absorbingtropospheric Many of the differencesbetween the two data setscan be aerosols such as volcanic ash. explainedby the fact that the AVHRR observationsare 2-3 The use of TOMS radiance measurements to detect both hours later than thosefrom TOMS. An area of disagreement SO2 and ash cloudsnot only providesa way of comparingthe not explainedby the time differenceexists south of Baja Cal- distributionof gas and ashwithin an eruption cloud usingthe ifornia on April 5. AVHRR detectsa sizablearea of ashin this samesatellite sensor, but the resultsobtained can be compared area, yet there is no indicationof sucha cloudfrom the TOMS to laboratorystudies designed to examinethe processesin such measurement. Since the ash burden of this cloud, as deter- gas-ashseparations [Holasek et al., 1996]. Furthermore, the mined from the AVHRR measurements(Schneider et at., detection of ash cloudsfrom TOMS complementsAVHRR submittedmanuscript, 1997), is an order of magnitudelower measurementsin areas where the temperature contrast be- than the main cloud to the southeast, the amount of ash is tween the cloud and the surface is small and AVHRR cannot possiblytoo low for TOMS to detect. detect suchclouds. Finally, the ability of TOMS to detect the injectionof volcanicash into the atmosphereand to track the motion of the resulting ash cloud during the days following 4. Optical Depth for April 6, 1982 future eruptionswill provideuseful information for emergency The aerosolindex observedby TOMS may be convertedto planningas well as for air flight operations[Heffter, 1993]. optical thicknessusing radiative transfer calculations provided that an adequateaerosol optical propertiesmodel is available. Acknowledgments. The authors would like to thank the other The mostcritical parameters are the effectiveradius of the size membersof the Ozone ProcessingTeam for their supportin generat- distributionand the complexindex of refraction.Knowledge of ing the Version 7 TOMS data set.We would alsolike to thank Charlie the ash vertical distribution and the reflectivityof the under- Schnetzleras well as the reviewerof this manuscriptfor their helpful lying surfaceis alsorequired. comments. To illustratethe feasibilityof derivingash optical depth from the TOMS measured aerosol index, radiative transfer calcula- References tions were carried out using an assumedaerosol model con- sistingof a potydispersionof silicateash composed of spherical Barth, C. A., R. W. Sanders, R. J. Thomas, G. E. Thomas, B. M. particleswith a lognormal particle size distributionof mode Jakosky,and R. A. West, Formation of the E1 Chichon aerosol cloud, Geophys.Res. Lett., 10, 993-996, 1983. radius (to) 1.5/x and width (or) 1.5. The complexrefractive Bluth, G. J. S., S. D. Doiron, C. C. Schnetzler,A. J. Krueger, and L. S. index at 340 and 380 nm was taken to be 1.5-0.0023i and Walter, Global tracking of the SO2 clouds from the June, 1991 1.5-0.0018i, respectively,based on reported measurements Mount Pinatuboeruptions, Geophys. Res. Lett., 19, 151-154,1992. [Patterson,1981; Patterson et al., 1983]. The ashvertical distri- Carey, S. N., and H. Sigurdsson,The 1982 eruption of E1 Chichon volcano,Mexico, 2, Observationsand numericalmodeling of tephra- bution is representedby a Gaussianfunction with a peak at 15 fall distribution, Bull. Volcanol., 48, 127-141, 1986. km above sea level (approximately700 mbar). The derived D'Almeida, G. A., On the variabilityof desertaerosol radiative char- optical depth is shownin Figure 5. It shouldbe emphasized acteristics,J. Geophys.Res., 92, 30t7-3026, 1987. that the validity of the optical depth derivedfrom this calcu- De Luisi, J., Measurements of the E1 Chichon dust cloud from Mauna lation is closelytied to the assumptionsof the input aerosol loa observatory,in Proceedingsof the SeventhAnnual ClimateDiag- nosticWorkshop, pp. 383-385, Natl. Oceanic and Atmos. Admin., model; any deviationsbetween the actual aerosolproperties of U.S. Dep. of Commer., Washington,D.C., 1982. the E1 Chichon ash cloud and those of the aerosol model would Heffter, J. L., Volcanic ash forecast transport and dispersion lead to differencesin the actual optical depth and the one (VAFTAD) model, WeatherForecasting, 8, 533-541, 1993. calculated. Herman, J. R., P. K. Bhartia, O. Torres, N. C. Hsu, C. J. Seftor, and E. Celerier, Global distributionof absorbingaerosols from Nimbus 7/totalozone mapping spectrometer, data, J. Geophys.Res., in press, 1997. 5. Conclusion Holasek, R. E., A. W. Woods, and S. Self, Experimentson gas-ash The Version 7 reprocessingof the Nimbus 7/TOMS data set separationprocesses in volcanicumbrella plumes, J. Volcanol.Geo- containsaccurate information about the spectralcontrast be- therm. Res., 70, 169-181, 1996. Hsu, N. C., J. R. Herman, P. K. Bhartia, C. J. Seftor, O. Torres, A.M. tween the 380- and 340-nm channels which can be used to Thompson,J. F. Gleason,T. F. Eck, and B. N. Holben,Detection of detect absorbingatmospheric particulates. In this work an biomassburning smokefrom TOMS measurements,Geophys. Res. aerosolindex has been definedand usedto map and track the Lett., 23, 745-748, 1996. motion of the volcanicash cloud from the April 4, 1982,vol- Krueger,A. J., Sightingof E1 Chichonsulphur dioxide clouds with the Nimbus-7 total ozone mapping spectrometer,Science, 220, 1377- caniceruption of E1 Chichon.For this eruption the evolution 1379, 1983. of the ash cloud (until April 8) is presentedand shownto be Krueger, A. J., L. S. Walter, P. K. Bhartia, C. C. Schnetzler,N. A. largelyconsistent with NCEP balancedwind fieldsand with ash Krotkov, I. Sprod,and G. J. S. Bluth, Volcanic sulfurdioxide mea- SEFTOR ET AL.: DETECTION OF VOLCANIC ASH CLOUDS FROM TOMS 16,759

surementsfrom the total ozonemapping spectrometer instruments, Global Volcanism1975-1985, edited by L. McClelland et al., Pren- J. Geophys,Res., 100, 14,057-14,076,1995. tice-Hall, EnglewoodCliffs, N.J., 1989. Matson,M., The 1982 E1 Chichonvolcano eruptions--A satelliteper- Torres,O., J. R. Herman, P. K. Bhartia,and Z. Ahmad, Propertiesof spective,J. Volcanol.Geotherm. Res., 23, 1-10, 1982. Mount Pinatubo aerosols as derived from Nimbus 7 total ozone Matson, M., and A. Robock, Satellite detection of the 1982 E1Chichon mappingspectrometer measurements, J. Geophys.Res., 100, 14,043- eruptionsand stratosphericcloud, in Proceedingsof the SeventhAn- 14,055, 1995. nual ClimateDiagnostic Workshop, pp. 359-370, Natl. Oceanicand Varekamp,J. C., J. F. Luhr, and K. L. Prestegaard,The 1982eruptions Atmos. Admin., U.S. Dep. of Commer., Washington,D.C., 1982. of El ChichonVolcano (Chiapas,Mexico): Characterof the erup- McPeters, R. D., et al., Nimbus-7 total ozone mapping spectrometer tions, ash-fall deposits,and gasphase,J. Volcanol.Geotherm. Res., (TOMS) data productsuser's guide, NASA Ref. Publ.,1384, 1996. 23, 39-68, 1982. Patterson,E. M., Measurementsof the imaginarypart of the refractive Wellemeyer, C. G., S. L. Taylor, G. Jaross,M. T. Deland, C. J. Seftor, index between 300 and 700 nanometers for Mt. St. Helens ash, G. Labow,T. J. Swissler,and R. P. Cebula,Final report on Nimbus-7 Science,211,836-838, 1981. TOMS Version 7 calibration,NASA Contract.Rep., 4717, 1996. Patterson,E. M., and C. K. McMahon, Absorptioncharacteristics of Wen, S., and W. I. Rose, Retrieval of sizesand total massesof particles forestfire particulatematter, Atmos. Environ., 18, 2541-2551, 1984. in volcanicclouds using AVHRR bands4 and 5, J. Geophys.Res., 99, Patterson,E. M., C. O. Pollard, and I. Galindo, Optical propertiesof 5421-5431, 1994. the ashfrom E1 Chichonvolcano, Geophys. Res. Lett., 10, 317-320, 1983. P. K. Bhartia and J. R. Herman, NASA Goddard Space Flight Robock,A., and M. Matson,Circumglobal transport of the E1 Chichon Center, Code 916, Greenbelt, MD 20771. volcanic dust cloud, Science,221, 195-197, !983. N. C. Hsu, N. Krotkov, C. J. Seftor, and O. Torres, Hughes STX Seftor, C. J., G. Jaross,J. R. Herman, X. Gu, L. Moy, S. L. Taylor, and Corporation,7701 Greenbelt Road, Greenbelt, MD 20770. (e-mail: C. G. Wellemeyer,The Meteor 3/totalozone mapping spectrometer [email protected]) Version 7 data set: Calibration and analysis,J. Geophys.Res., in W. I. Rose and D. J. Schneider,Department of GeologicalEngi- press,1997. neering and Sciences, Michigan Technological University, 1400 Sigurdsson,H., S. N. Carey, and J. M. Espindola,The 1982 eruptions TownsendDrive, Houghton, MI 49931. of E1 Chichon,Mexico: Stratigraphy of pyroclasticdeposits, J. Vol- canol. Geotherm. Res., 23, 11-37, 1984. (ReceivedSeptember 18, 1996;revised March 13, 1997; SmithsonianInstitution Scientific Event Alert Network (SEAN), acceptedMarch 26, 1997.)