SECCION

RIESGOS VOLCANICOS

PELIGROS VOLCANICOS

Geroge P. L. Walker*

Resumen:

Los peligros volcánicos pueden ser evaluadosutlizando: 1) Registros históricos, 2) el registrogeológico, 3) a través del conocimiento del com-portamiento general de los volcanes. Esta. informa-ción es mejor presentada en mapas de zonificaciónde peligros volcánicos indicando peligros inmedia-tos y potenciales para flujos de lava y flujos piro-elásticos, y otros tipos de erupciones volcánicas.

Los volcanes de pendientes moderadas presen-tan el riesgo más elevado porque sus productos sonampliamente dispersados durante sus erupciones.Los valles de los ríos pueden actuar como canalespara lahares. permitiendo el transporte de los mate-riales fragmentarios más allá del área alrededor delvolcán. Las erupciones de ignimbntas son el tipode erupdones volcánicas más peligrosas debido aque causan ia devastación de grandes áreas. Losvo1can~s que hacen erupclón raramente ofrecen unproblema especial porque por lo general no son ob-servadas.

En Costa RIca hay muchas zonas de pehgros:los volcanes Poás e lrazú han producido numerososdepósitos freá ticos; San J osé y el Valle Central es-tán cubiertOs por ignimbritas y coladas de lava; de-pósitos de nubes ardientes se encuentran alrededordel Volcán Arenal. Existen numerosos volcanes jó-venes en Costa Rica. que como el Arenal, puedenreac tivarse.

VOLCANIC HAZARD

PART L VOLCANIC HAZARD IN GEÑERAL.

l. How is volcanic hazard assessed? There arethree information sources.

(a) The historical record of eruptions of a vo1ca-no, as a guide to possible future activity. Ge"nerally this record covers only the past fewhundred years.

(b) The record of eruptions of a vo1cano detemu-ned by stratigraphic mapping/volcanologicalinterpretation of the voIcanic products (eachyolcanic eruption leaves a layer which, whendeciphered, reveals the characteristics of theerupÜon l. Generally this record CQversa muchmore extended period and is eorrespondinglya more reliable guide to possible futurebehaviour than (a).

(e) Knowledge of the behaviour of volcanoes ingeneral.

To give an example of (a), the hlstOflcal re"cord of Irazu sine e 1700 is one of fairly frequentash-producing exploslve eruptions. The probabilityseems hlgh that eruptions of a similar type and ona similar scale will take place at a comparable fre-quency in future. To give an example of (b), the

. Hawaü lnstitute of Geophysics, 2525 Conea Road. Honolu.lu, Hawaii, %822.

41

record of the past ca. 40.000 yearS determined bystratígraphic study of pumice deposits from Taupoand Okataina rhyolitic vo1canoes (New Zealand)reveals that majar explosive eruptions occurroughly once per 1000 years, the latest being 700and 1800 years ago, and similar infrequent butlarge explosive outbreaks may be expected infuture. An example where general knowledge ofvolcanoes is helpful is Mt. St. Relens. In 1980 abulge developed on the northern side of this steepcone and this side of the cone becarne mechani-cally unstable. The consequent failure and collapseof part of the cone had devastating consequences.

2. How is informatíon on volcanic hazard pre-sented? The best method is by means 01'volcanichazard zone maps which delineate areas liable tospecific volcanic hazards, based in part on the pastrecord of the volcano and in part on general know-ledge of volcano behaviour. There is no unique ty-pe of hazard zone map. The simplest type delimitzones of greatest danger, requiring evacuation orrestricted entry should an eruption be impending,or alternatively it shows areas which are judge tobe mast liable to be affected for example by pyro-clastic flows, surges, lava flows, mudt1ows, or pyro-clastic faIls having a thickness greater than a speci-fied threshold. Another type takes into accountthe frequency or recurrence interval of specificevents ar (derived frorn that) the probability thatsuch events will happen within a given time periodoYet anather type delineates zones which shows thepro'oabUity, should <in eruption occur, that giveneffects will be experienced. Thus at Taupo there isa 0.3 probability that when an eruption occurs, py-roclastic flows wiH develop extendig to a radius of30 Km from the ven!.

3 Two broad types of hazard may be distingui-shed: inmediate, and potentlal. lnmediate hazardIS that which recurs relatively frequently (once ormore than once per century); the probabihty there-fore í5 hlgh thar people may experience a hazar-dúus ~vent in thelI lifetime. Potential hazard isth¡H which recurs relatively rarely (less than onceper century), the probability is therefore low thatpeúple will exp~f1enc'" a hazardoU5 event in theirlifetime.

These two types of hazard should be treateddi1'ferently. Far \'olcanoes which present inmedla-te hazard, do;;:tauedhazard zone maps should ideal1ybe prepared delimiting the probable extent and fre-quency of the hazard, a volcano observatory be

established to monitor the volcano closely, contin-gency plans be drawn up should a volcanic emer-gency develop, and the people be educated regar-ding the vo1canic hazards they may have to face. Itmight moreover prove desirable to restrict cons~tructíon ar development within certain hazardzones. An example where safety features havebeen incorporated into the design of a constructionworks IS the provision of gatcs on intake tunnels atthe hydroelectnc power plants near Ruapehu vol-cano (New Zealand) which can be ciosed when la.haIS are approaching. Consideration might be gi-ven to no-cost/low-cost measures such as the cons.truction of buildings wíth a steep-pitched roof inareas that may be susceptible to pyroclastic falls,and the building of simple volqmo shelters (proofagainst nuees ardente or pyroclastic falls) as apossible alternatíve to evacuation in some volcanicareas. For vokanoes which present potentialhazard it is more appropriate merely to do quietresearch to establísh the behavioUI pa ttern of thevolcano, and to set up some kind of sImple moni-torin'g system so as to avoid the posslbilíty of aneruption breaking out without warning.

A few general cornments should be made.Fírst, it is a principie of volcanology that the morepowuerful or violent (and hence more dangerous) avolcanic event, the less impressive or conspicuous isthe resulting deposit when viewed on the scale of arock outcrop (the converse is not necessarily true).It is only when the deposit is mapped and its far-reaching nature recognised that the extent of thethreat may be assessed. An example is the 1902nuée ardente which is recorded today by a deposítonly 20 cm thick in the devastated town of Sto Pie-rre. Agaín the ash thickness is less than 10 cmthick over half of the forest blow-down area ofMay 18th. 1980 around Mt. St. Helens, and withina few years will probably be almost total]y remo-ved by erosion.

A second general comment is that steep volca-noes are steep b.::cause their eruptions are weak andtheir volcanic products are piled up inmediatdyaround the v.¿>nt.aod tlat volcanoes are f1at becausetheir eruptions are habitually powerfuL or viojentso that the volcanic products become widely dis~persed. Howevl'"r even small volcanic events onsteep volcanoes may be amplified by gravíty. ope-rating on the high and steep slope, into potential1ydangerous events.

A thlrd cornment is that lahars pose a especialhazard because. channelled as they are by river va-

42

lleys, they may present a hazard far outside thearea of the volcano. For example, some prehistoriclahars of Mt. Rainier are known to have travelledkm from the volcano. Note that the distance trave-lled is probably a rather simple function of thelahar volume.

A fourth cornment is that by far the greastestvolcanic hazard is that arising frem ignimbriteeruptioo. Such an event is prabably the only typewhich i5 capable effectively of tempararily "knoc-kíng out" a coun1;¡y the size of Costa Rica, Guate-mala, North Island, New Zealand, Kyusyu, ar NewBritam. Apart trom the devastation caused by theignimbrite itself, disruption wauld be caused ayera great area by the accompanying fall of co.ignirn-brite ash.

A final cornment 1S that the most insiduoustype of volcanic hazard is posed by valcanoeswhich erupt very rarely, at intervals say of 103 to105 years. Such volcanoes are commonly regardedas extinct ---an example is Mt. Lamington (NewGuinea) which erupted unexpectantly and withgreat violence in 1951 ---and will thus not bewatched. Eruptions which follow a long repose pe-nad tend moreover to be particulary large, power-fuI or violent events---examples are the plinianoutburst of Somma-Vesuvius in the yeaor 79, andthe ferocious eruption of Taupo in 186 A. D.which followed nearly 500 years of quiescence.

PART 2. VOLCANIC HAZARD IN COSTA RICA

The following discussion i5. based on impres-sions gained during a brief ten-aay visit to CostaRica and should not be regarded in any way a defi-nitive account.

1. Sa!l Jos~ and the Central Valley. Extensívelava tlCW'í dud igrÜmbrite are seen in many canyonexposures no::..JISan Jose 1 understand that thesear~ oí" the arder of a mIlhon years old. A pre-ig-nimbrite phnian pumice fall deposit underlying theignimbrite thickens and coarsens towards the northand east COilsistent with a SQurce underlying Barva.It i5 possli:Ú tilat Ihese extenslve lava fIows and ig-nim!:J.rite r~present a long-past period of more vigo-rous vc1canism that wiII not recur, but it seerns be-tter not au tomatlcally to assume that thís is so andto reg3.rd ;mÜlar eruptions in future as possible.posing a pot.::ntlaI hazard.

Resting on these older volcanoes are very pro-minent mudflows, and the reality of the mudflowhazard is shown by the 1964 mudflows that rava-ged Cartago. I visited the mudflow source area inthe Reventado Valley and saw how extremely uns-table it is. A high-intensity rainfan following apenod of wet weather would readly activate majormudflows perhaps much bigger than the 1963 one.It seems desirable to attempt to stabilize this mud-flow source area, although it would be very costlyto do so.

.

2. The slopes of Poas and Irazu. Repeated cen-tral vent eruptions of Poas have produced manymud~rich phreatic deposits which are up to 0.5 mthick at 8 km radius, besides strombolian depositsand a pumice falI (probably of sub-plinian type)which is 1 m thick at 12 km radius. Irazu has hadmany strombolian and like eruptions; similar cen-tral vent eruptions may be expected in future.They pose a hazard to life only very near the erup.tive vent, but have a senous if temporary effect onthe farming community, exemplified by the 1963-65 ash-falls of Irazu. There is olso a mudflow ha-zard. Poas and Irazu voIcanoes have a broad shield~like form, and my irnpression was that lava flowsare a major component of their broad edifices, andare now mostly covered by a thick voIcanic1asticveneer. One lava flow (Formación Cervantes) cave-ring 38 km1 and extending 13 km from vent hashowever been mapped on the south side of Irazuand is reatively young.

3. Arenal. The 1968 and 1975 nuées ardent ori-"

.

ginated, as do al! nuees arden te, on a steep coneand may be expected to develop..~:fnfuture' erup-tions. The stratigraphic study of the earhe pro-ducts of the volcano is weIl in hand and should

.reveal the fuIl capabihty of the volcano.

4. Guanacaste Province. 1 did not reach the an-desite cones of OroSl, RIncón de la Vieja, and Mira-valles, so 1 have no cornments to make On the ha-zards they presento 1 did however visit (he extensi-ve ignimbrite field on the south.wesI slde oI' thesecones. A number of ignimbrite eruptions (at least5) have occurred there, and it is cIear that if onesuch eruption were to occur today it wou!d havevery serious effects. Stratigraphic mapping of theignimbrite field and datíng of the individual ignim-brites is a necessary prelude to assessing hazardfrom possible future eruptions.

S. Other voIcanoes. There are many volcanoesin Costa Rica which are currently not c1assified as

43

active. The outburst of ATP.nalin 1968 serves as areminder of howimportantit is to investigate "dor-

mant" and "extinct" volcanoes as well as the "ac.tive" ones when assessing volcanic hazards.

SECCION

QUIMICA DE GASES

BALANCE DEL S Y CL PARA CUERPOS\f AGMA TICOS CALCALKALIN OS POCO

PROFUNDOS

WiUiam l. R ose.

Resumen

El presupuesto volátil de un volcán puede serd~t~rminado de la siguiente forma:

( 1) Midiendo volátiles en inclusiones de vidrio-va-por en minerales para detenninar el contenido ini-cIal de volátiles en el magma; (2) usando eqUIpo desensares remotos y aparatos de muestreo de gasespara determinar la cantidad de volátiles liberadosdurante las emisiones de bajo nivel entre las erup-,'iones; O) el muestreo y análisis de los materialeseruptados para determinar la concentración de vo-látiles retenidos en los materiales eruptados; y (4)midiendo la cantidad de volátiles liberados duranteuna erupcIón. que es lo más difícil de estimar exac-tamente y tiene muchas fuentes de error.

Resultados midales muestran para erupcionesr~Cl'~,1tesdel volcán Fuego, Guatemala (1974) yelMonte Santa Ekna ( 1980). qUe la gran mayoría de!azufre y doro fueron liberados del magma antc:s ydurante !a enlpción. a pesar de esto una porciónsignifIcatIva de ambos gases se libera durante losperíodos J~ emIsión baja entre empclOnes.

Los presupuestos gaseosos pueden ser usadospara ( l) estimar el tamaño de cuer:Jos magmáticos

superficlales; l2) detectar el arrivc de nu~vo mag-ma: (3) detenn-inar el grado de dt gasificación delmagma '5uperficial: y (4) detennin..,

"cambias en la

temperatura y fugacidad de] oxíge; o en e! magma

S and CI Budgets for Shallow Calc-AIkaHc MagmaBodies.

ABSTRACT

Particulariy in the case of explosive magmas,it is almost never possible to dlrectlj' obtam repre-sentative samples of magma tic gases: As a result wehave limited understanding 01' the degasslI1g 01'sha-110wmagma bodies. lf a program of selected geo-chemical and petrograph.1c observatlons can be ma-de on volcanic samples, howevtr. it is passibJe {Oinfer much abour the magma inmedlatdy belowvolcanic vents.

To make an interpretatian of the yolatilebudget we need to know at least: 1) The in¡tialvolatBe concentrations of magma when Lt reachesthe shallow reservoir, 2) the amounts of vo!atilesreleased during eruptive actlvity, 3) the amountsreleased during low~level ernissions betw~en erup.t1005. 4) the concentrations of vola tiles remainingin the lava and ash samples after eruption and soli-dlfica tian. Recen t anal)' tical methods al101,1,llS tomake estlmltes of l by dekrrnmatlOn and intuvre-tation 01' the chemical composition of glass mdu-sions trapped within phenocrysts by rapid skdetalgrowth, 3 can be measured by using remote sen-sing equipment and gas sampling. 4 can oe estima-ted by samphng and analysis of materials erupted.2 LSthe hardest to accurately estimate and thesources of error iD its determinatíon ar¿ numerous.Nevertheless because of atmospheric samp!mg of

*Michigan Technological Um'i.

47

eruphvc douds in recent years and because os stu-Jies of fresh pyroclastics collected before and du-ring faIlout we have some primitive estimaks.

Results so far have focused on S and C1 bud-gets for the recent eruptions of Fuego. Guatemala(1974) and Mount St. Helens. USA (1980). Themagmas lose the vast majonty of S and most CI be-fore and dunng eruptlon. There is considerable gasfrom mtruslve magma whlch participates 1fl explo-SlVI:'eruptions. Mafic magma ma}' have hlgher Scontenl. A significam portion of both gases is lostduring low level emlssions between eruptions. The

composltlOn:) or emisslOns vark" conslderdbly 'Wlrh

time.

In terpretatlon af the gas budgets af shallawmagma bodies offers informatian of obvious valuein forecasting of activity. 1) The sizes of shJJlowmagma bodies may be estimateJ. 2) The arrival ofnew magma fram depth may be detected. 3) Thedegree of degassing of [he shaIlow magma may bemferred. 4) Changlng temperature and oxyg:en fu-gacity wndl1lOnS in the magma body may be tn-ferred.

EMISION DE DIOXIDO DE AZUFRE Y OTROSGASES EN EL VOLCAN MASAYA,

NICARAGUA

Stoiber, Richard E., *

El volcán Masaya se encuentra en un estadode alta emisión gaseosa, produciendo 1000 a 2000toneladas de S02 por día. Estos estados ocurrenaproximadamente a intervalos de 25 afios y se ex-tienden por un lapso de 5 años. Lluvia ácida conun PH de 2.5-3.5, se precipita de la pluma de gasdel Masaya. Filtros especiales se han colocado enel cráter y en la dirección del viento para colectarS02, HCI, HF, HBr, S04' Cl y F (ver tablas). Estosdatos combinados con los datos suministrados porun espectrógrafo de correlación permiten hacer es-tL'11aciones de la emisión de gases diaria. Un anali-zador de gas Intersca electroquímlco de S02 fueusado para medir las concentraciones de S02 y HCIen la pluma de gases. Estos datos permiten evaluarlos peligros ambientales derivados de la degasifica-ción del volcán Masaya.

,. Departmcm of E.uth SciencesDartmoutb Collcges, Honovcr, N. H. 03755. USA.

Sulfur Dioxíde and Other Gas Emission FromMasaya Volcano, Nicaragua

A group of scientfsts- from Dartmouth Colk-ge (Stoiber, R. E., S. N. Wilhams. and N. M. Jolm"son), Umversity of Virginia (R. A. Parnell). Stan-ford University (William Winner). and ColoradoCollege (B. Huebert) are studying the geology ofthe Masaya Caldera, Nicaragua wlth special att~n-tion to the gas emissíon fram its crater, Santiago.We are now in the midst of a penad of high emis-SlOn, 1000 to 2000 tons per day 502 and oftenmore. These periods have recurred at aproximatcly25 year intervals and last about 5 years.

Gas is almost a!ways at ground leve! as It pJS-ses over a ridge WhlCh is traverse to the prevajlmgwind direction, 15 km downwind frcm ~he Cr.IkfConcentrations of up to 1 ppm saz are fOllnd hefl:'.MeJ.surements of 502 flux passing over the ridgewere made in 93 m5tances In January to March1981. At this time an average rate of 1300 tooswas indicated.

There 15 a wide vana tia os in emlSSl0n, greaterthan W~ hav.; measured at other vokanoes,

""i.:h

quite :jurel)' ret1ects a changing output at the crat::r.Our collaborator lohnson finds strongly acid rainf.llls tl1ru plume, pH 2.5-3.5. Tht soil reactioosare und~r study. Winner is descnbing the dekte-nous effects of the gas on the coffee plan ts.

Special filters appropriate for the determinJ-tlOn of certam gases have been placed in the gas plu.me both at the era ter and downwind at Llano Pal'J.-ya. The 3 filters m each se! allow the weight of

49

Sampie Loation Gases AelOSOIs Gas to aerosol

SO4 ~/0 - la - IF- 502/504" HCL/CL-

Santiago Crater SOZ/HO Ha/HF HO/HBr HF/F

Masaya VolcaDO 3.4 >370 840 1.3 2.3 90.5 29.5 H.5 <.5

Uan? Pacaya, 15 km 6.7 > 10 Not 10 3.4 10.8 1.77 H .68 < .68

downwind foom Santiago detennined

Rate 01 Emission 5°2 HO HF }Gr.

Tons/day 1300 400 5 0.5

TonslYeaI 500000 150000 3300 180

'" based on very few data

S02, HCI, HF (and infrequently HBr) which passesthrough them to be measured. The filters also ca-Hect aerosols: 804 ", CI -, F -. From these data

Table 1

gas ratios, aerosol ratios and gas to aerosol ratios'have been cakulated (Tabk 1).

Average ratios of weights of gases and aerosols fram filter samples

Our present data indicate that ratios from 3fIlter collections. a11 on Llano Pacaya, a11on thesame day. are very similar. The majar differencesin these ratios are the contrast between ratios fromf1lter eollections at the era ter and downwind asshown in Table l. The aerosol component increasesdownwind. This is marked in the gas to aerosol ra-tios for S02/S04 ..

. and especially HCl/CI -. Thegas aerosol ratio is greatest for S°.1.' less for HCI

Table 2

and least for HF.

We have calculated the probable S02, HCI.and HF flux at the era ter using the ratios and thedownwind correlation speetrometer average value.The ratios allow one to take account of the S0210ss due to aerosol produetion downwind from theera ter and to include the arooums in aerosols at thecrater (Table 2).

Tons per day and per year of selected gases emitted fraro Santiago Crater

Concentrations of 502 and HCl in the plum ewere measured using an Interscan electrochernlcalSOz gas analyzer at the era ter and downwind.Because of the difficulty of sampling the most con-centrated part of the plume, values (Table 3) mustnecessarily be viewed as mínimum only.

The data relative to Masaya volcano degassingare being used in a continuing study of the enviran-mental impact downwind fraro the Santiago era ter.The study emphasis the importance of volcanic gasplumes as a volcanic hazard under eertain eondi.tions of winds and local topography.

50

Method LocadoR SO: HO

Filter data Crater 96 30

Calculated) Llano Pacaya 2.8 i) ;

. ..- . . . ..-.. .. -- . . .. - . ~-.

Ilnterscan data Crater » 10

llano Pacaya 1

I

Aircraf traverse 0.5Over Llano Pacaya20 km downwind 0.4

ground level

Table 3

Concentrations (ppm by weight)

SECCION

SISMOLOGlA VOLCANICA

ANAL YSIS OF VOLCANIC TRE~fOR FROMPAVLOF, FUEGO, PACAYA, SAN CRISTOBAL

ANO MASA YA VOLCANOES,

Steven McNutt*,

RESUMEN

Tremor volcámco estrechamente relacionadoen el tiempo con la actividad eruptiva de cinco es-tratovo1canes de los arcos circumpacíficos fue ana-lizado y comparado con señales de baja amplitudprodu~idas por agua fluyendo a través de túnelesde descarga de la r~presa Tarbela en Pakistán. Aná-hSIS de frecuencia fueron hechos para cada volcánusando los métodos de la discreta transformada rá-pida de Fourier y el análisis Burg de máxima entro-pí::. La represa y los volcanes tienen resonanciasde tubo de órgano. exhibIendo picos angostos eigualmente espaciados.

Usando la frecuencia de la vibración. una velo-cidad de 2 Km/sec para las ondas P en el magma, lalongItud de onda puede ser determinada. De la 10n-gÜud de onda, s,," puede calcular la longitud delconducto magmático (tabla 1). que es positivamen-te correlacíonable con las alturas de[ volcán.

Altanatlvamente, si los conductos están abier-tos en ambo:> ~xtrt.'mos. en vez de ser solo en la par-te superior como en el modelo anterior, ]OS con-ductos serían más largos por un factor de 2. Si elmagma contIene burbuJas, entonces la velocida.d delas ondas "P" dc.::recería reduciendo ]a longitud delconducto calculado por un factor hasta de 5.

Todos los casos estudiados aquí involucrantr¿mores superfk [ales. que put.'den representar tinfenómeno de resonancia asociada con los conduc-

Análisis de tremor volcánico de los volcanesPavlof. Fuego, Pacaya, San Cristóbal y Masaya.

tos magmáticos en un edificio o maCIZOvokánküque pueden considerarse como llna zona de la bajaresistencia mecánica cuyos esfu~rzo5 mtcrno..; L'sedngobernados por fuerzas de cuerpo (gravedad) envez de ser por fuerzas tectónicas ..;omo en la pro-fundidad.

Vo1canic tremor 15 generalJ}' detlned as contl-nuous seismic ground oscillations J,>so iatl'd \1,'Jthvolcanic activity (Tokarev, 1981). Tremor ha~been classified in the literaturc in gt'vcrdl \\a~..., b:J-sed upon' 1) auptIve vs non-erupti\t' actl'llty ofthe volcano: 11) frequency cantent and ""ismlc wa.ve type: UI) time dl.lration (canunuOl.~ or mtcr-mittent on a time seale 01' minutes lO month<; ¡; :.JI1UIV) source mechanism, including dos.:-d \" IH m e-chanisms (jerky crack growth. contll1uous hlockmotian, magma chamber oscHlatlOns). and openvent mechanisms (exptosions. continuous gas anlllava ejection. magma flow in ca ndu lts. gJS 05..:11!J-tlOns) (Tokarev, 1981: Kubotera. 197-+: Shimozuru.¡ 9fj 1. 1971 t This srudy 15concern~J w¡th tr --'mor

which is c1ose1y related in time to surfac.:- t'ruptivL'activityat five circum-Pacific lsland are \trutovolca-noes; Pavlof (Alaska), Fuego and PacJya (Guatema-la). San Cristóbal and Masaya (Nicaragua), BnddescriptlOns of tremor episodes at th~sL' volcano\.'sfollow.

... Columbia UniversltyLamont Doherty GeoL Obs.. Pahsadcs. NY. 10%4, USA.

55

Tremor at Pavlof volcano is observed as shortbursts of 2-26 minutes time duration and also ascontinuous tremor of 1 1/2 days duration aecom-panying major strombohan eruptions. At Fuegovolcano, tremor episodes often begin with or areaccompanied by large explosions from the surnmitera ter: these episodes last fram several hours to ma-ny days duration. Tremor froro Pacaya volcanooccurs during strong strombelian eruptions and isof several hours duration. At San Cristóbal volca-no, tremor increased in amplitude gradually over aperlOd of severa! days prior to a strong ash andsteam eruption on March 10, 1976. FinaUy, con-tinuous tremor of low amplitude is observed nearthe active lava Jake of Masaya volcano.

Signals resembJing volcanic tremor are alsoobserved on seismograrns from stations near Tarbe-la dam and reservoir in Pakistán. Tms signal is ob-served when water is flowing through the dam'soutilow tunnels. Frequency analysis have beenperfomed on the tremor fram each volcano and onthe Tarbela signals using two spectral analysis me-thods: Discrete Fast Fourier Transfonn (FFT) andBurg maximun entropy (MEM). The Tarbela signaJsare shown to be organ pipe resonanees in the out-fIow tunnels, with both even and odd hannonicspresent O. e., i\" 1/4,Q,1/2,Q, 3/4,Q 2, 5/49., ete.).The observed spectra change with location aroundthe reservoir and with the size of the outllow ope-ning, which is gate controlled.

We infer that the volcanic tremor signals arealso organ pipe resonances, presumably of a mag-ma~f1l1ed conduir. The tremor spectra exhibit na-rrow, evenly~spaced speaks, similar to the Tarbeladaro signals (Figure 1). We measure the frequeneyof the fundamental mode of vibration, and by assig-ning a velocity of 2.0 km/sec to P~waves in magma(Murase and McBirney, 1973), we are able to deter-mine the wavelenght from the relation:

v = XC or i\ = v/r

where V ~ velocity in km/sec, A. wavelength inkm, and f'" frequency in Hz. Based on our resultsfrom Tarbela daro, we assume tha! both odd andeven harmonics are present. We then ealculate thelengh of the magma conduit based on the relation

for the fundamental mode of vibra tlon for an Of-gan pipe wlth one open and one closed cnd-

A2 =4

where Q '" length in km, A= wavelength m km. Re.sults are shown in Table 1. We -note a rough poslti-ve correlatlOn between the length determined bythis method and the heigh of the volcano above itssurroundigs.

We also note two possible sources of error Inthis type of analysis. It is p.ossible that the funda-mental mode of vibratíon 18 that for a plpe wlthtWQ open ends; we cannat determine this unambi-guously. If true, our lengths gíven in Table 1 shourdbe multiplied by a factor of 2. There is also a largeuncertainty in the value to assign for the velocityof P-waves in magma. Addltion of even a few per-eent of bubbles to the magma drastically lowersthe P-wave velocity, so our value of 2.0 km/sec isprobably a maximum. Values as low as 0.3 kmísechave been measured in the field (Aki d a1., 1978).Using this value reduces the lengths given in Table1 by a factor of 6.

All the examples of tremor examined in thisstudy appear to be of shallow erigin. ~ e speculatethat tremors originatmg at shallow depths may re-present a resonance phenomena associated withmagma conduits within the volcanic pile. whichcan be regarded as a zone of relatively low mecha-nical strength whose internal stress field i<;largelygoverned by body forces (gravity) rather than tec-tonic forces at depth.

T ABLE 1

Lengths of resonating structureswith v: 2.0 km/sec

PavlofFuegoPacayaSan CristóbalMasaya

1.6:t 0.3 kmla:!: O.l kmL 1 :!:0.2 km0.7 :!: O. I km0.5 =: 0.1 km

56

REFERENCES

AKl, K., B. CHOUET, M. FEHLER, G. ZANDT, R. KOYA.NAGI, J, COLP, and R. G. HAY, 1978. Seismic pro-perties of a shallow magma reservoir in Kilauea 1ki byActive and passl'le experiments, J. Geophys. Res.. 83(B5), 2273.2282.

KUBOTERA, A., 1974. Volcanic tremors at Aso Volcano,in Pbysical Volcanology. edited by Civetta el al.,Developmenu in Soljd Earth Geophysics, 6. Elsevier,Amsterdam, 29-47.

MURASE, T., and A. R. McBIRNEY, 1973 Properties ofsome common igneous rack:>and their melts at hightemperatures, GeoI. SocoAmer. BulI., 84,3563.3592.

SHIMOZURU, D., 1961. Volcanic micro-seisms, discussionon the origm, BuII. Vole. SocoJapan, 2, no. S, 154-162.

SHIMOZURU, D., 1971. A seismological approach to theprediction af 'lolcanic eruptians, in The Surveillanceand Prediction of Volcanic Activity. UNESCO EarthScience Monograph No. 8, París, 19-45.

TOKAREV, P. I., 1981. Volcanic tremor, 1981 IAVCEISymposium on Are Volcanism, Abstracts Volume,Volcanol. Soc. Japan and IAVCEI, 386.

57

A

B

e

lLJo:::)~-~:E«

I

D)(~

1'~M~'--

~,G

H

o 2 3 4

Figure i

58

5 hz

SECCION VOLCAN ARENAL

PETROLOGY OF ARENAL VOLCANO

M. Se. Eduardo Malavasst R. '"

ABSTRACf

Petrochemestry of Arenal yolcano is similar tovolcanic rocks of other circum-Pacific Continentalmargins, characterized by the presence of orogenicandesite and related rocks. The MgO, CaO, TiO:zand total Fe conrents decrease, and aIkalis increasein these rocks with increasing Si02 (see figure 1).

Arenal volcano magmas show typical cal-alka-¡ic trends in the AFM diagram with neither enrich.:11ent non empovenshment in the total Fe relativeto MgO and total alkalies (see tigure 2). Composi-!ians of rocks from Arenal volcano plat in the calc-alkalic field in the total Fe versus total Fe/MgO,which differentiate between rhe calc-alkalic andtholeitic series of island ares and actlve continentalmargins (see figure 3)-

Bilsed O!1 chemical analyses most vo1canicrocks from Arenal VoIcano are bas;11tic-andesites orandesltes. The exceptions are these daeitic !apillilayers. a basa!t and several gabbroic accidentalblocks.

Mmeralogy and mineral proportlOns of Arenaltephra layers show the most contrasting composi-tion variatlOns. Juvenile ejects from these layersrange t'rom two pyroxene basalts, to two pyroxene

* Program... de Investigación Vu1canológica, Escuela de CienciasGe.ográf"I;Cas. UniverSidad Nacional Heredia, Costa R.1ca.

basaltic-andesites, to one pyroxene hornblende an-desite, to hornblende dacites. Texture ranges fromhyalopilitic to trachytic.

Arenal volcano lavas are two pyroxene basal-tic-andesites and andesites, gray to dark gr('y in co-lor and hyalopilitic to pilotaxitic tex turco HistoncArenal lavas (1968-present) are two pyroxenebasaltie-andesites. porphyritlc hyalopil ilie texture:with exception of the first lava flow erupted in1968 (Sáenz, 1977) which had homblende pheno-crysts.

Chemica! trend of lavas erupkd bl'twe;;::nd1968-1973 (Melson and Sáenz, 1973) rdlo::cts theextrusion 01' Iess differentiated lava w¡t~, time Sewchemical analyses of lavas erupted bet\~ l:t'n 1974-1978 do not support a simple evolution toward lessdifferentiated lavas with tune smcC' they nave mi-nor chemical variations, but generally are very SI-milar lnd fairly basic for calcalkaline volcanicrocks (see figure 4).

::PETROLOGIA DEL VOLCAN ARENAL

La petroquímica de las rocas del volcán Are-nal es similar a la de las rocas volcánicas de otrasmárgenes continentales de la región clrcumpacífica,caracterizados por la presencia de andesita orogé-niea y rocas relacionadas. En las rocas del volcánArenal el contenido de MgO, CaO. TiO:! y Fe totaldecrece y el contenido de alcalis crece wn el incre-mento en silice, ver figura l.

61

Los magmas del volcán Arenal muestran típi-cas tendencias calcoalcalínas en el diagrama AFMsin incremento o decremento de Fe total relativo alMgO y a1calis totales, ver figura 2. Las rocas delvolcán Arenal se ubican en el campo calco-alcalinoen el gráfico Fe total contra Fe total/MgO, que per-mite diferenciar entre las series volcánicas calco-al-calinas y toleiticas de los arcos de islas y márgenescontinentales activas, ver figura 3.

La petroquímica sugiere que la mayoría de lasrocas del volcán Arenal S~)flandesitas basálticas oandesitas. Las excepciones conocidas.son tres eyec-tas juveniles daciticas de las capas de tefra, un ba-salto y muchos bloques accidentales gabroides.

La mineralogía y proporciones minerales delas capas de tefra evidencian variaciones químicascontrastantes en composición. Las eyectas juveni-les de estas capas gradan de basalto con dos piroxe-nos, a andesitas basálticas con dos piroxenos, a an-desita con homblenda y un piroxeno, a dacitas con

......-',t- .... ...

MtO

.~-1 '.5

LI1.0

o.. "lO.o.1t.-.(

o..., ~0.5-41. ,,)¡ ..-

eI."8,O.

...le'. M... . tt08IA

510.

hornblenda. Su textura grada desde hJaloplJíticahasra rraquítica.

Las lavas del volcán Arenal son andesltas ba-sálticas y andesitas de dos piroxenos. color gns os-curo a gns, y textura pllotaxítica a hialopilítlca.Las lavas históricas (1968-presente) son andesitasbasáltlcas con dos piroxenos y textura porf¡ríticahIalopilítica; a excepción de la primera lava erup-rada en 1968 (Sáenz, 1977) que contenía fenocris-tales de hornblenda.

Las tendencias petroquímicas de lavas erupta-das entre 1968-1973 (Melson y Sáenz, 1973) sugIe-ren la extrusión de lavas menos diferencIadas con eltiempo. Nuevos análisis químicos de las lavas erup-tadas entre 1974-1978 no respaldan una simpleevolución hacia lavas menos diferenciadas, ya queellas tienen sólo variaciones menores pero general-mente son muy similares y básicas para volcanescalcoalcalinos, ver figura 4.

,.c;5.0

No! O.4

e e. '5.

~

.. 'A...~.

.'5..2

U

2.00.'0

"10 . . 0.1'8

"

..." :,:,,. ... 0.420.3011.0...

OM

0.1M

e

.'.. t;.

CIJO

.... ... 74

.. .. a

, 1.0~ ~ 88 60 el ..

SiC.

ni: I .-~ 01 ~ HAMI[ItI"MA~ aoCASDI1. ~

"t.ll.f¡~ ~IIS "OR ARPtAL ~CII'S.

62

Fe O'

FeO . ..Hiena total expresado como hOTotaltron as FeO

NOZO + ~OM90

:~: DELvou:.ut .r.aeKAL ""EHAL VOLCANO ROC8CS.

~8HIP8iA rOPIJ.t.,,--- T'OLEtTlCA. lI8C8. IIt.At, COHTI-~ ¡llCTH...,

TOFUAT"UO. 101..11'.,,10 ..-Ia O,ISLAMO AltCS AMO ""TIVI: ~-TAL JMMJU blYIffOMMUff.

'

.~IICY WMIUSWIt ~RDtU.-o.) IIttOVUM'" TOLDTICA CONT1-

---,

SKAIt D IMSTMISIOM(LtOUiD)'1.1110 COM'1ltI8IToIiLTMO\.III'tIC~"OYMtCI:

t f""",,, VOLCU"""', MIItE"~~'o:C*.J" .. -.coa ~ lILA8.. COIII~M.T"'l.u AC'1IYM

""'MAYOLCoUtO,TMNO, ~IC

SERIES Of' I8L"MD ARC AMO~TIVECowrUtt:M'1 AL IHS E"V 1M" tlEMT.

'i~ 2aIbP:I~" A"' HoJIIA.ocA, .. IUIt4L

"". .rWA l"OfItMaU.l. 1tOC88.

s. . ..S!!tIE~ CAliNACAi.C-ALICAO."IULM 2.37 2." 5.39

Ft<f/M9Q

seal! TOLEITICATifOt.ElITlC KII:IES""...

-''''''...

"'. '.~"'"~

'",

""""""...

ac.

!I.90P." !l.80

FIG.3

-""Ia) De Mt8tftO TOT"L IXJIItESADOCOMOf.O'lERSUSI.A ULACK»I F, 0.1 MOO. LIME-'S'U~1II1'tA",".0, A. (1,'7'4). IL 81111"1<:;0III!UMEU$ ,VOLCAIIICA. CIoaAOT""I~U DI U)8 AACOS 1)1:ISLAS '( .A~ ~ALIS ACT1VAS

~ O" TOTAl.lItCIIIII8.' Fle USTK€..,t,TlO

~ IIHMI AS F.~ lit O. L'" oV'TIlt ..,....-~IRO, A. n. "+), 8itA1181It!\IIIW V'OLCANIC-SUIES OP "LMO .ute. AIID ACTlvl COIIITIM!:NTAL1ILAR.11d IMVIa<¡MIIINT.

63

:7 TIOZ O

..1073~ ,.,.lII71"pg 68 SI l'On T4~ 76 7'li7718

coa

o OO O O 00A

O

s.MgO O O

00 00 O

O4.

.

O

g OO

OO

o

"tFeO

o o o o o o oo

ooo

K20 g oo o 0<1

o o00o

.

oo .o.

M

Oo o00

o

SiOto000 Alz03

10

o

oo oo

o'. o

o

NazO

. "70~",., Jni .8170'1'374757'tIJlTT7IAÑ08""UltS AMos-YENtS

FIG.4UlCOI DI LA C8íIfIoOSIC/ON\fUI.US TIDIf'O

MIL\ L.U UIIIQ MTOIUCAS DEL VOL.CANARENAL.

VA'''ATIOIt . ni -~n'IOM VERSUS TI lilE INAIIUAL W)U.IN& M.TOfIIIC LAVAS.