VOLCANIC
PROGRESS
HAZARDS
AND
THEIR
MITIGATION-
AN D PROBLEMS
RobertI. Tilling
U.S.GeologicalSurvey
Menlo Park, California
"Natural calamitystrikesat aboutthetimewhenoneforgetsits terror."
Japanese
proverb[seeShimozuru,1981]
Abstract. At the beginningof the twentiethcentury, disastersof the 1980s (Mount St. Helens, U.S.A. (1980),
E1 Chich6n, Mexico (1982); Galunggung, Indonesia
(1982); and Nevado del Rufz, Colombia (1985)) illustrates
the importanceof predisastergeosciencestudies,volcanic
hazards assessments,volcano monitoring, contingency
planning,and effectivecommunications
betweenscientists
and authorities. The death toll (>22,000) from the Rufz
catastrophe
probablycould have been greatlyreduced;the
reasonsfor the tragically ineffective implementationof
of volcanic disasters and crises.
A review of hazards
evacuationmeasuresare still unclearand puzzlingin view
mitigation approaches and techniques indicates that of the fact that sufficientwarningswere given. The most
significantadvanceshave been made in hazardsassess- pressing problem in the mitigation of volcanic and
ment, volcanomonitoring,and eruptionforecasting. For associatedhazardson a global scale is that most of the
example, the remarkableaccuracyof the predictionsof world's dangerousvolcanoesare in densely populated
dome-buildingeventsat Mount St. HelenssinceJune1980 countries that lack the economic and scientific resources or
is unprecedented. Yet a predictive capability for more the political will to adequatelystudy and monitor them.
voluminousand explosive eruptionsstill has not been This problem afflicts both developed and developing
achieved. Studies of magma-inducedseismicity and countries,but it is especially acute for the latter. The
ground deformation continue to provide the most sys- greatestadvancesin volcanic hazards mitigation in the
tematicand reliable data for early detectionof precursors near future are most likely to be achieved by wider
to eruptionsand shallow intrusions. In addition, some applicationof existing technologyto poorly understood
othergeophysicalmonitoringtechniquesand geochemical and studiedvolcanoes,rather than by refinementsor new
methodshave been refined and are being more widely discoveriesin technologyalone.
appliedand tested. Comparisonof the four major volcanic
volcanologybegan to emerge as a modem scienceas a
result of increasedinterestin eruptivephenomenafollowing some of the worst volcanic disastersin recorded
history: Krakatau (Indonesia)in 1883 and Mont Pe16e
(Martinique), Soufri•re (St. Vincent), and Santa Maria
(Guatemala)in 1902. Volcanologyis againexperiencinga
period of heightened public awarenessand scientific
growthin the 1980s,the worstperiodsince 1902 in terms
INTRODUCTION
AND
HISTORICAL
PERSPECTIVE
Most of the Earth'scrustis of magmaticorigin,attesting
to the enormousrole that volcanic and related magmatic
processes
haveplayedin formingthe outermostsolidrind
of our planet. In addition,the distributionof volcanoes,
pastand present,can be closelylinkedto the dynamicsof
the crust and mantle within a plate tectonicscontext.
Some foreign rock fragments (called "xenoliths")
contained in eruptive products represent the deepest
samplesof theEarth's interiorthathavebeenrecoveredto
datefrom any drill hole. Thus it is hardlysurprisingthat
manygeoscientists
work in terranesor on researchtopics
directlyor indirectlyassociatedwith volcanicrocks. Yet
withinthe geoscience
community,relativelyfew specialize
Thispaperis notsubjectto U.S.copyright.
in volcanology,the studyof the transportand eruptionof
magma [Sigurdsson,1987], with emphasison active or
potentiallyactivevolcanoes[Tilling, 1987a].
Of themorethan1300volcanoes
knownto haveerupted
in Holocenetime, abouthalf are classifiedas active (i.e.,
thosethathaveeruptedin recordedhistory). On average,
about50 of thesevolcanoeserupt eachyear, an eruption
frequencythat appearsto be obtainedfor all historicaltime
[Sirekinet al., 1981]. Individualvolcanoes,
however,may
remain in reposefor many centuriesor even millennia and
thus may be classified as dormant (i.e., could become
activeagain)or extinct(i.e., not expectedto eruptagain).
The shortcomings
of pigeonholeclassificationare evident;
in a generalway, the longer the period of intereruption
repose,the more energeticthe next eruption. Someof the
Reviews
of Geophysics,
27, 2 / May 1989
pages23 7-269
Publishedin 1989 by the AmericanGeophysicalUnion.
Papernumber89RG00436
'237'
238'Tilling: VOLCANICHAZARDSAND THEIRMITIGATION
27,2/REVIEWSOFGEOPHYSICS
Guatemala)killed more than 36,000 people. T. A. Jaggar,
Jr., a 31-year-oldassistantprofessorin geologyat Harvard
Universityat the time, was one of the geologistssentto
study the effects of the Mont Pe16eand Soufribreerupandthatgainedlaterfrom studiesof
eruptionof MountLamington,PapuaNew Guinea[Taylor, tions. Thisexperience
1958].
volcanoesand earthquakesin Alaska, Italy, Japan,and
Costa Rica convinced Jaggar that the expeditionary
VolcanicDisasters
andthe Emergence
of Volcanology methodof studywas inadequateand that to fully underworst volcanic catastrophesin history have occurred at
volcanoesbelieved to be "extinct," for example,the A.D.
79 eruption of Vesuvius that destroyedPompeii and
Herculaneum [Sigurdssonet al., 1985] and the 1951
it is necessary
to observeand measure
HistoriCally,
thestudy
oferuP•ve
phenomena
hasbeen standvolcanoes,
spurred by volcanic catastrophes. Indeed, the earliest
accuratedescriptionof an eruptionis containedin letters
theireruptive
andassociated
seismic
activity
ona continuousbasis
before,
during,
andaftereruptions.
Thedevasta-
from Pliny the Youngerto the RomanhistorianTacitus,
describingthe asphyxiationof his uncle, the famous
scholarPliny the Elder, who was observingthe A.D. 79
eruptionof Vesuviusas it destroyedPompeii. The 1815
eruption of Tambora (Sumbawa Island, Indonesia),
considered
the largesteruptionin recordedhistory,caused
morethan90,000 humandeathsandglobalclimaticimpact
[Self et al., 1984; Stothers,1984; Storereeland Storereel,
1983]. Yet becauseof the volcano'sremoteness,poor
global communications,and the immature statusof the
natural sciencesat the time, this huge eruptionattracted
little attentionbeyond the affectedpart of the Indonesian
archipelago. The first scientificexpeditionto study this
cataclysmicevent was not sent until 1847 [Zollinger,
1855]. That same year, a small primitive volcano observatorywasestablished
on the flank of Vesuvius.
tionJaggarwitnessedat Martinquealsoconvincedhim that
a better understandingof the processesthat could kill
"thousandsof personsby subterranean
machinerytotally
unknownto geologistsandthenunexplainable
wasworthy
of a life work" [Jaggar,1956,p. 62].
Jaggarwas a scientificvisionarywho recognizedthata
true understanding
of eruptivephenomenawould require
the establishmentof permanent observatoriesto study
active volcanoes [Macdonald, 1953].
His vision and
researchgoals were sharedenthusiastically
by the noted
Japanese seismologist, F. Omori, of the Imperial
EarthquakeInvestigationCommittee. In 1910, Omori
designedand deployeda seismometer
("tromometer") to
detect earthquakesbeneath Usu Volcano (Hokkaido),
marking the first instrumentalmonitoringof a volcano
[Shimozuru, 1981]. In 1911, Omori founded an obIn contrastto the feeble, belated scientificresponseto servatoryat AsamaVolcano(I-Ionshu).Meanwhile,Jaggar
and designed
the Tambora eruption, the 1883 eruption of Krakatau obtainedseveralOmori-type seismometers
(Sunda Strait, between Java and Sumatra, Indonesia) equipmentto measurethe temperatureof an active lava
promptedthe first well-organizedscientificinvestigation lake in Halemaumau Crater within Kilauea Volcano's
of a volcaniccatastrophe
andits aftermath[Verbeek,1885; summitcaldera[Jaggar, 1956]. A "technologystation"
Symons, 1888] (see also Sirekin and Fiske [1983]). was set up at the edge of Halemaumaulava lake in July
Scientific expeditions were quickly dispatched,and 1911, and, after many difficulties,F. A. Perret and E. S.
comprehensivestudies were made and publishedby Shepherd (Geophysical Laboratory of the Carnegie
scientists from the Netherlands and other countries within
Institution,Washington,D.C.) succeeded
in obtainingthe
a few yearsof theeruption. As emphasized
by Sirekinand first thermocouple
measurement
of the temperature(1000
Fiske [1984], who aptly termedKrakatau1883 "a classic + 25øC)of liquidbasaltic
lava(seeShepherd
[1912],cited
geophysicalevent," these studiesgreatly advancednot in the work of Bevenset al., [1988], andApple [1987]).
only volcanology,but also meteorology,oceanography, The Hawaiian Volcano Observatory (HVO) was
andbiology. From the 1883 Krakataueruption"the world founded "officially" in January 1912, upon Jaggar's
quickly learned that the impacts of large geophysical arrivalto assumethe directorship.Followingthe founding
events are global, and that they demonstratethe inter- of theobservatories
in Japanand Hawaii the youngscience
dependence
of land,sea,andair" [SimkinandFiske, 1984, of volcanologyhas continuedto mature through the
establishment of, and studies conducted at, additional
p. 48].
The study of the effects of the 1883 Krakatau volcanoobservatories
elsewherein the world. Jaggar's
catastrophe
providednew insightsinto the processes
and visionaryobjectiveof a global network of observatories
hazardsassociated
with a violentlyexplosiveeruption,but has only been partially achieved,becausemany of the
little was learned about how a volcano behaves before an
world's active and potentially active volcanoesstill are
eruption. Early in the twentiethcentury, volcanology little studiedand poorly understood.Very few are being
began to evolve from an expeditionary,after-the-fact adequately
monitored.
endeavorinto the modem,multidisciplinary
sciencethat it
is today. In the 6-monthperiodMay-October 1902, three Scopeand Purposeof ThisPaper
eruptionsin CentralAmericaandthe Caribbean(Soufri•re,
Many usefulbooksand review paperstreatingvarious
St. Vincent; Mont Pel6e, Martinique; Santa Mar/a, aspectsof volcanologyappeared in the 1970s [e.g.,
27,2/REVIEWSOFGEOPHYSICS
Tilling: VOLCANIC HAZARDSAND THEIR MITIGATION o239
UNESCO, 1972; Macdonald, 1972; Civetta et al., 1974;
interestedreader is referredto pertinentreview papers
Bullard, 1976; Williams and McBirney, 1979; Sheetsand
Grayson, 1979]; a new internationaljournal (Journal of
Volcanologyand GeothermalResearch)was launchedin
1976. Scientific and public interestin volcanoesand
volcanology
increased
dramaticallyin the 1980s,following
the well-documented reawakening and catastrophic
eruptionof Mount St. Helens in the spring of 1980
[LipmanandMullineaux,1981;Mansonet al., 1987]. It is
beyondthe scopeof thispaperto considerrecentadvances
in mineralogy,igneouspetrology,and geochemistrythat
have contributed to an improved understandingof
magmaticprocesses
operativein volcanicsystems. The
[Carlson, 1987; McCallurn, 1987; Marsh, 1987; Ghiorso,
1987]andthereferences
citedthereinandin thereportfor
1983-1986 compiledby Longhi [1987].
This paper primarily focuseson some geophysical
aspects
of volcanology,
andits purposeis threefold:(1) to
review the progressin volcano-monitoring
and hazards
mitigationstudiesmade since the early 1970s, (2) to
comparesomevolcanicdisastersand crisesin the 1980s
withina contextof hazardsmitigation,and(3) to highlight
problemsand challengesthat confrontvolcanologists
and
governmentofficialswho work to reducevolcanicrisk.
TABLE 1. SomeNotable Volcanic DisastersSincethe Year A.D. 1000 Involving 300 or More Fatalities
Primary Causeof Death
Country
Year
Pyroclastic
Flow
Merapi
Indonesia
1006
1,000'
Kelut
Indonesia
1586
Vesuvius
Etna
Merapi
Italy
Italy
Indonesia
1631
1669
1672
Awu
Indonesia
1711
3,200
Oshima
Cotopaxi
Japan
Ecuador
1741
1741
1,000
Makian
Indonesia
1760
Papadajan
Lakagigar
Asama
Unzen
Mayon
Indonesia
Iceland
Japan
Japan
Philippines
1772
1783
1783
1792
1814
2,960
Tambora
Indonesia
1815
12,000
Galunggung
Indonesia
1822
4,000
Nevado del Ruiz
Awu
Colombia
Indonesia
1845
1856
1,000
3,000
1,000
Volcano
Debris
Flow
Lava
Flow
Posteruption
Starvation
Tsunami
10,000
18,000-{lO,OOO-•
300*
1,480
9,340
1,150
15,190
1,200
Cotopaxi
Ecuador
1877
Krakatau
Indonesia
1883
Awu
Soufrigre
Indonesia
St. Vincent
1892
1902
1,560
Mont Pelie
Martinique
1902
29,000
80,000
36,420
1,530
Santa Maria
Guatemala
1902
6,000
Taal
Philippines
1911
1,330
Kelut
Indonesia
1919
Merapi
Lamington
Indonesia
PapuaNew
1951
1951
1,300
2,940
Hibok-Hibok
Agung
Philippines
Indonesia
1951
1963
500
1,900
Mount St. Helens
El Chich6n
Nevado del Ruiz
U.S.A.
Mexico
Colombia
1980
1982
1985
60$
>2,000
5,110
Guinea
Total
>22,000
65,140
53,900
28,000
89,340
53,090
An exceptionis madefor the May 1980 eruptionof Mount St. Helens,whichcausedthe worstvolcanic
disasterin thehistoryof the UnitedStates[fromTilling, 1989,Table 1.2]. Valuesareroundedoff to thenearest
tell.
*Includesdeathsfromassociated
mudflows;however,thevalidityof the 1006eruptionhasbeenquestioned
[Djumarmaet al., 1986].
'['Includes
deathsfromassociated
explosions
and/ormudflowactivity;estimates
areunreliableandprobablytoo
high.
$Principalcauses
of deathswerea laterallydirectedblastandasphyxiation.
27, 2 / REVIEWS OF GEOPHYSICS
240 ßTilling: VOLCANIC HAZARDS AND THEIR MITIGATION
TABLE 2. Principal Types of Volcanic Hazards and SelectedExamples
Example
Type
DIRECT
Fall processes
Tephrafalls
Ballisticprojectiles
Flowage processes
Pyroclasticflows, surges
Laterally directedblasts
Debris avalanches
Primarydebrisflows
(eruptiontriggered)
Floods(j6kulhlaups)
Lava flows
Other processes
Phreaticexplosions
Volcanicgasesandacidrains
Reference
HAZARDS
Vesuvius, 1906
Soufri&e (St. Vincent), 1812
Mount Pe16e,1902
Lacroix [ 1906]
Anderson and Flett [1903]
Bezymianny,1956
Fisher et al. [1980]
Gorshkov[ 1959]
Mount St. Helens, 1980
Nevado del Rufz, 1985
Herd and the Comitd de
Katla, 1918
Kilauea, 1960
Thorarinsson[1957]
Macdonald [ 1962]
Soufri&e(Guadeloupe),1976
DiengPlateau(Indonesia),1979
Feuillard et al. [ 1983]
Le Guern et al. [1982]
Voightet al. [1981]
EstudiosVulcanol6gicos[1986]
INDIRECT
HAZARDS
Sakurajima,1914
Shimozuru[ 1972]
Tsunami (seismicseawave)
Secondarydebris flows
Krakatau, 1883
Semeru, 1976
Simkin and Fiske [1983]
Posteruptionerosionand
sedimentationproblems
Atmospheric effects
Iraz6, 1963-1964
[1978]
Waldron [1967]
Mayon, 1814
Commission
on Volcanology
Earthquakes and ground
movements
Volcanological
Societyof Japan
Posteruptionfamine and disease Lakagigar(Laki), 1783
[19751
Thorarinsson[1979]
Principaldatasources:Blong[1984] andCrandellet al. [1984].
VOLCANIC
AND
RELATED
HAZARDS
Comparison
With Other Hazards
On a global basis and relative to most other hazards,
natural or man-made, volcanic and related hazards occur
Estimatessuggestthat the averageU.S. citizen is much
more likely to die from coronaryarrest while shoveling
snow or from a lightning strike than from either an
earthquakeor a volcaniceruption[White and Haas, 1975,
Figures3 and 4]. Nonetheless,sincethe year 1000, more
than300,000peoplehavebeenkilled directlyor indirectly
by volcaniceruptions(Table 1), and at present,about360
million people(about10% of the world'spopulation)live
on or near potentially dangerousvolcanoes[Peterson,
1986, Table 15.1]. The circum-Pacificregion,becauseit
containsmostof the world's activevolcanoesand densely
populatedcountries,hasfaced,and still faces,a disproportionatelyhigh risk, in termsof economiclossand human
deaths,posedby volcanicand relatedhazards[Simkinand
Siebert,1984;Blong, 1984].
infrequentlyand affect few people [Wijkmanand Timberlake,1984,Figures1 and 5]. Averageannualeconomic
loss and human casualty from volcanic hazards are
correspondingly
low if consideredglobally,but volcanic
disasters,like earthquakes,
can have significantshort-term
human and economicimpact. In the United States,for
example,theannualeconomiclossfromvolcaniceruptions
is at least an order of magnitudeless than the loss (0.6
billion dollars)causedby earthquakes,
which in turn is
nearlyan orderof magnitudelessthanthe lossassociated
with floodsandgroundfailures[Haysand Shearer,1981].
The deadliesteruption in history (Tambora, Indonesia, Direct and Indirect Hazards
The typesand natureof volcanicand associated
hazards
1815) killed 92,000 people (Table 1), comparedwith
500,000 killed in the worst hurricane (Ganges Delta, havebeenwell described[e.g., Macdonald, 1975;Blong,
Bangladesh,1970). Perhapsthe deadliestnaturaldisaster 1984; Crandell et al., 1984; Office of the UnitedNations
was the Huahsienearthquake(Shensi, China) in 1556, DisasterRelief Co-Ordinator(UNDRO)/UNESCO, 1985].
which killed more than 820,000 people [DeNevi, 1977]. A list of the principalhazards(Table 2) is presentedhere;
More recently, the 7.8-magnitudeTangshan (China) all examplesof hazardsgiven involved human fatalities
earthquakein July 1976,accordingto Chinesegovernment and/ordestructionof property. It shouldbe emphasized
figures,killed about240,000people[ShiDiguang,1987], thatan eruptioncommonlyproducesmultiplehazards.For
but some outside observers have claimed that the fatalities
example,the May 18, 1980, eruptionof Mount St. Helens
included a debris avalanche,a laterally directed blast,
mayhaveexceeded800,000.
27, 2 / REVIEWS OF GEOPHYSICS
Tilling: VOLCANIC HAZARDSAND THEIR MITIGATION o241
:
mudflows, pyroclastic flows and surges, steam blast within a given period of time; if sufficientdata exist, a
("phreatic") explosions,and tephrafall [Christiansenand probablilityshouldbe assigned
to a potentialhazard.
Peterson, 1981]. Of these, the avalanche,directedblast,
2. Risk is the probabilityof a loss (such as life,
and mudflows caused most of the deaths and devastation.
property,productive
capacity,etc.)withinan areasubject
Some terms in this paper may be unfamiliarto some to volcanic hazards. Assessment of risk involves the
readers. "Pyroclastic" ("fire-broken" in Greek) describes consideration
of therelationrisk= (value)x (vulnerability)
fragmented
moltenlavaand/orsolidrockexpelledduring x (hazard), wherevalue may includethe numberof lives
explosiveeruptions.Highly explosiveeruptionsare called threatened,
the economicworth of property,civil works,
"plinian," afterPliny theElder,whowaskilled duringthe andproductivecapacity,and vulnerabilityis a measureof
A.D. 79 eruption of Vesuvius. "Tephra" refers to thepercentage
(0 to 100%)of the valuelikely to be lostin
airbornefragmentalvolcanicejectaof any size; "ash" is a givenhazardousevent.
tephra with grain size less than 2 mm. "Lahar," an
3. Disasteris aneventthatis markedby thesignificant
Indonesian term, describes volcanic debris flows lossof value(as definedabove)resultingfrom volcanic
(including "mudflows"), which are slurriesof volcanic and related hazards.
debrisand water that can vary widely in proportion. A
4. Crisisis a situationduringwhich a volcanoshows
debrisflow is considered"primary" if it is triggeredby signsof instability
or unrest,interpreted
to augurimpenderuptive activity, most commonlyby melting of snow ingeruptive
activityandassociated
hazards.A crisismay
and/Or
icebyhotvolcanic
materials,
and"secondary"
if or may not culminatein a dangerouseruption,but it
causedby noneruptiveprocesses,most commonlyby alwayscausesanxietyand/orsocioeconomic
disruption
heavy sustainedrainfall in terrain underlainby uncon- amongthepopulaceaffected.
solidated
volcanic
deposits.
•'.•6kulhlaups"
(anIcelandic
term),alsocalled"glacierbursts,"are periodichighOF VOLCANIC
RISK
dischargefloods of subglacialwater, causedmost com- MITIGATION
monly by volcanicor geothermalactivitybut sometimes
by purelyglaciologicalprocesses.
As emphasizedby Tilling and Bailey [1985], the
Of thedirecthazards,lava flowsaretypicallyassociated effectivemitigationof volcanicrisk buildsfrom a foundawith nonexplosive
to mildly explosiveoutpourings
of fluid tion of long-termbasicresearchon volcanoes,inactiveas
lavas (e.g., basaltic to basalticandesite),such as those well as active(Figure1). An improvedknowledgeof
produced at Kilauea and Mauna Loa (Hawaii), Etna "how volcanoeswork," the focus of a recent international
(Sicily), and Piton de la Foumaise(ReunionIsland,Indian symposium[Tilling, 1987b, 1988a, b], is the common
Ocean). Most other direct hazards are linked with point of departure for all volcanic hazards studies.
explosiveeruptionsof steep-sidedcompositevolcanoes. Specifically,
five elements
areessential
to mitigatetherisk
Excellent descriptionsof pyroclastic processesand from volcanicand relatedhazards: (1) identificationof
products
characteristic
of explosiveeruptionsaregivenby high-riskvolcanoes;
(2) hazardidentification,assessment,
Fisher and Schmincke[1984]. Of the indirect hazards, and zonation;(3) volcano monitoringand eruption
secondarymudflows, atmosphericeffects (shock waves forecasting;
(4) engineering-oriented
measures;
and (5)
andelectricaldischarges),
andmagma-induced
earthquakes volcanicemergency
management.
and grounddislocationsare commonbut the leastsevere.
In terms of human fatalities, the indirect hazards of
Identificationof High-RiskVolcanoes
tsunamiand eruption-caused
famineare as significantas
Of the some600 active volcanoesknown in the world,
the direct hazards of pyroclasticflows and primary only a smallfractionof them havebeen,or are being,
mudfl0ws(Table 1).
studiedin detail. Economicallyand scientificallydevelopedcountrieslack sufficientresolveto studyandmonitor
all of the activeor potentiallyactivevolcanoeswithin their
Hazards,Risks,Disasters,and Crises
borders;
thesituation
is evenmoreacutefor thedeveloping
The distinction between the terms "hazards"
and
countries,which containmost of the world's explosive
"risks" and the terms "disasters" and "crises" is
volcanoes,many in denselypopulatedregions. Nonethecommonlyblurred; in particular,hazardsand risks are less,identificationof high-riskvolcanoesis requiredto
sometimesusedsynonymously.In this paperan attempt determinewhichonesshouldreceivethe mostattentionby
will be made to use theseterms consistentlywith the scientistsandpublic safetyofficials,within the limitations
following definitions, adapted in part from Fournier of whateverresourcesmay be available.
d'Albe [1979] andNewhall [1982]:
Attemptshavebeenmadeto compilelistsof high-risk
1. Hazard is the volcanicphenomenonthat posesa volcanoes[e.g., Shimozuru,1975; Matahelumual, 1982;
potentialthreat to personsor propertyin a given area Lowensteinand Talai, 1984; Yokoyamaet al., 1984]. All
242*Tilling: VOLCANICHAZARDSAND THEIRMITIGATION
GO VERNMENT
27,2/REVIEWSOF GEOPHYSICS
BODIES
DECISION
CONTINGENCY PLANS
MAKERS
DISASTER
POLICY MAKERS
WARNING
DISASTER PREPAREDNESS
RESPONDING
AGENCIES
LAND-USE PLANNING
ClENTIFIC
/
ERUPTION
•
HAZARD
WARNING
SCOMMUNITY
/ FORECASTING
•x EDUCATION
OF
THE
/AWARENESS-
•
PUBLIC,MEDIA,AND
RESPONSIBLE//PROMOTION
PROGRAM• DECISION
MAKERS
AGENCY /
//
/
HAZARDS-ZONATION
•
MAPS
'•x
VOLCANO
/ PAST
&PRESENT
BEHAVIOR
• MONITORING
/
/
I
\
\
\
/ • d / • • I i•o • • = t % e. \ •_• \ RESEARCH
/,?
/
I .s 5 I o
/
Figure 1.
Diagramshowingthat an effectiveprogramof
\
% \-%%
voc^.os
\ PROGRAMS
\
restof thet•iangleto indicatethedivisionof primaryrespon-
who must
volcanichazardsmitigationmustbe built on a strongfoundation sibilitybetweenthe scientistsand civil authorities,
and political factors in addition to
of basic studies,followed by successively
more specialized considersocioeconomic
(modifiedfromTilling
investigations
to preparehazardszonationmapsandto predict scientificinformationin makingdecisions
the future behaviorof the volcano. The apex is separatedfrom andBailey [ 1985,Figure1]).
suchlistsutilizeratingcriteriainvolvingsomeor all of the used to identify high-riskvolcanoes. Yokoyamaet al.
thatthe"low... ratingsfor certain
followingfactors: (1) frequency,sites,and natureof [1984,p. 21] cautioned
hencenot identifiedas 'high risk' by the criteria
recordedhistoricaleruptions;(2) informationon recent volcanoes,
prehistoric
eruptions
asinferredfrommapping
anddating used, may simply reflect incompleteand/or incorrect
studies;(3) known ground deformationand/or seismic information,not necessarilylow risk. In fact, the volevents("earthquakeswarms"); (4) natureof eruptive canoesnot listed(authors'italics)... shouldbe the focus
products
aspossible
indicators
of explosive
potential;
and of increasedscientific investigationsto establishtheir
(5) variousdemographic
determinants,
suchaspopulation actualpotentialfor eruptionsandrelatedhazards."
density,propertyat risk,andfatalitiesand/orevacuations
Hazard Identification,Assessment,
and Zonation
resulting
fromhistorical
volcanicdisasters
or crises.
In 1983, participantsin three UNESCO-sponsored Identification and assessment of volcanic and related
workshopsidentified89 high-riskvolcanoes: 42 in hazardswere not prime subjectsof scientificinquiryuntil
southeastAsia and the westernPacific, 40 in the Americas
the 1960s [e.g., Markhinin et al., 1962; Searle, 1964;
and the Caribbean, and seven in Europe and Africa
Crandell and Mullineaux, 1967].
However, the 1919
[Yokoyarna
et al., 1984]. However,this compilation
is eruptionof Kelut (Java),whichkilled about5000 people,
hardlydefinitive,because
geologicandgeophysical
data motivatedthe Dutch scientificcommunityto establisha
are inadequate
for soundevaluationof many volcanoes. "Volcano Watch" at the "most deadly East Indies
For example,had scientists,
usingthe criteriaadoptedat volcanoes." This "watch," which included a network of
at their "most dangerous"volcanoes(all
the 1983 UNESCO workshops,rated the E1 Chich6n six observatories
Volcano(Mexico)beforeitsviolenteruptionin 1982[Luhr in Java),was the forerunnerof the presentVolcanological
and Varekamp,1984], it would not havebeenscoredas Surveyof Indonesia(VSI). In additionto initiatingthe
beinghighrisk. Ironically,theNevadodelRu/z Volcano systematic observation of some volcanoes, another
(Colombia)was not identifiedas beinghigh-riskby the importantresult of the Volcano Watch was the first
attempt,beginningin the 1920s,to classifythe
1983 workshopparticipants.Its eruption2 yearslater systematic
caused more than 22,000 deaths, perhaps as many as Indonesianvolcanoesin termsof their hazards,including
of hazardszonationmaps. On thebasisof
27,000 [Podestaand Olson, 1988]. However,scientists thepreparation
only
topographic
considerations
and recordsof historical
are quiteawareof the deficiencies
in the availabledata
27,2/REVIEWSOFGEOPHYSICS
Tilling: VOLCANIC HAZARDS AND THEIR MITIGATION o243
complete. The eruptive behavior, vent locations, and
topographyof a volcano all may change with time.
Moreover, for many volcanoesthe period for which the
eruptiverecordandbehavioris known may be too shortto
to make useful volcanic hazards assessmentsinclude those
"capture" infrequent but potentially high magnitude
utilized in the identificationof high-riskvolcanoes. In events. The longer the period spannedby the data base
addition, further attention is paid to the geologic usedto reconstructpasteruptivebehavior,the more useful
eruptions,theserudimentarymaps delimitedzones of
"danger" to populationand "regions that may be
destroyed"[NeumannvanPadang,1960].
As detailedby Crandellet al. [1984], the essentialdata
(especiallystratigraphic),petrologic,and geochemical and reliable the hazards assessmentis, if the volcano has
information on the nature, distribution,and volume of the
not changedits "style."
Hazardszonationmaps at appropriatescalesshouldbe
eruptiveproducts. From such data it is possibleto
becausethey
reconstruct
a volcano'spasteventsand eruptivebehavior, an integral part of a hazardsassessment,
which in turn providesthe basisfor assessing
potential portraythe pertinentinformationin a summaryfashion
readily understoodby land use planners and decision
hazardsfrom futureeruptions.
Hazardsassessments
are generallypredicatedon the makers as well as by scientists. This cartographic
assumption
that the samegeneralareason a volcanowill representationof hazards zonation can vary widely,
most likely be affectedby the samekinds of eruptive dependingon the specific hazardsin question,and is
eventsat aboutthesameaveragefrequencyin thefutureas discussedin detail by Crandell et al. [1984]. Figure 2
in thepast. Obviously,thisassumption
maynotalwaysbe showsthe hazardszonationmap that wasprepareda month
750O0'
75015 '
75o30'
75o45 '
74045'
IManzanet'es
•
5o15 , __
/
-
Marquita
.,-
I::re843o
•.•"-'"'
'"'x.. Neira
Honda
Meode•
Ita
Palestina
5o00' -
,'
ß Santa Rosa
,••de Cabal
;rida
,•:•Dos
Quebradas
Pereira
4ø45, _
Venadillo
0
5
lO MILES
I
II 0I....5 I
I ....
1
10
1
15 KILOMETEFIS
EXPLANATION
Lava flow hazard
November
Pyroclasticflow hazard
Mudflowhazard
Ashfallhazard
Low angle blast hazard
•
'4•
"'i="'
' Hgh •------•
Hgh [,'•
....•:•Hgh
Moderate
.;,• Moderate
• Moderate
• Moderate
['•-•';'-";'1
Moderate
13, 1985 eruption
Mudflows
.'.-•...:**
•
Figure 2. The hazardszonationmap for Nevado del Ruiz
Volcano, Colombia (reprinted by Herd and the Comitd de
EstudiosVulcanol6gicos[1986, Figure 4]). Althoughthis map
accuratelyanticipatedthe nature and areal extent of potential
Ashfall
volcanic hazards and was available more than a month before the
catastrophic
eruptionon November13, 1985, its usefulness
was
negated by ineffective emergency management during the
disaster(seetext).
244'Tilling:
VOLCANIC HAZARDSAND THEIR MITIGATION
27,2/REVIEWSOFGEOPHYSICS
To date, seismicand ground deformationtechniques
accuratelythe areal extent of the mudflowsand ashfalls have been the most widely and routinely applied in
volcano monitoring [e.g., Shimozuru, 1972; Decker and
from theNovember13, 1985, eruption.
before the Ruiz volcanic disaster;this map anticipated
Volcanic hazards assessmentsand/or hazards zonation
Kinoshita, 1972; Kinoshita et al., 1974].
Several other
maps are now available for a numberof the world's geophysicalmonitoringmethodsshow promisebut at the
high-riskvolcanoesor volcanicregions(see,for example, moment must still be considered experimental:
the compilationof Scott [1989, Table 3.5]). However, microgravity[e.g.,Jachenset al., 1981;RymerandBrown,
hazards assessmentsare not available, or even forthcom1987], geomagnetic[e.g., Davis et al., 1984; Zlotnicki,
ing, at presentfor many potentiallydangerous
volcanoes 1986], geoelectdcal[e.g., Zablocki, 1975, 1978; Jackson
(e.g.,E1Chich6n(Mexico) andE1Misti (Peru)).
andKauahikaua,1987],remotesensing[e.g.,Malingreau,
1984; Francis et al., 1988], and thermalradiation[e.g.,
Moxham, 1972; Kieffer et al., 1981] methods. Another
promising but also still experimentalapproachis the
VolcanoMonitoringand EruptionForecasting
Systematicsurveillanceof volcanoesbeganearly this application of geochemicalmethods to monitor the
century. Experiencegainedat well-monitoredvolcanoes emissionandtemporalvariationof volcanicgases:general
indicatesthat most,andperhapsall, eruptionsarepreceded methodologyand total analysis [e.g., Le Guern, 1983;
and accompaniedby measurablegeophysicaland/or Greenland,1987], sulfurdioxide[e.g.,Stoiberet al., 1983;
geochemical
changes.Appropriately,volcanomonitoring Krueger, 1983], carbondioxide [e.g.,Harris et al., 1981;
providestheprimarydatafor short-termforecasts
(hoursto Casadevallet al., 1987], hydrogen[e.g., Sato and McGee,
months) of eruptions;longer-termforecasts(1 year or 1981; McGee et al., 1987], radon and/ormercury[e.g.,
longer) generally are premised on other data, most Chirkov,1975;Williams,1985],and helium[Friedman
and Reimer, 1987; Williams et al., 1987].
commonlythe long-termeruptiverecordof the volcano.
The rationaleand techniquesof volcanomonitoringand
Short-termeruptionforecastsare still most reliably
eruptionforecastinghave been summarizedin several made largely on the basis of seismicityand/or ground
review volumesandarticles[e.g., UNESCO, 1972;Civetta deformation alone. Experience worldwide, however,
et al., 1974; Decker, 1986; Newhall, 1984b]. Of these showsthat optimumvolcanomonitoringis best achieved
summariesthe compilationsof the 1970sstill remainthe by integratinga combinationof approaches,rather than
best and most comprehensivepresentations
of the basic relyingon anysinglemethodor precursory
indicator.
monitoring techniques, while the subsequentworks
emphasizerefinementsin, and applicationsof, volcano Engineering-Oriented
Countermeasures
Volcanic eruptions cannot be controlled, but some
monitoringin the 1980s. In addition to thesereview
papers,many studiesspecificto a volcanoor regionalso hazardousphenomenacan be temperedby engineering
have been conducted: Mount St. Helens [Lipman and measures
or structures
to lessenimpactor extent. To date,
Mullineaux, 1981], Mount Ema [Barberi and Villari,
most of the engineeringcountermeasures
have addressed
with flow processes
(lava flows, debris
1984], Campi Flegrei(Phlegraean
Fields) [Barberiet al., hazardsassociated
1984a], and Hawaii [Decker et al., 1987].
flows, and floods)and, to a lesserextent,tephraloading
Most monitoringtechniquesare designedto measure andeffectson buildings.Of these,lava diversion
perhaps
muchcoverage
by thenewsmediabecause
of
changesin the physicalor chemicalstateinducedby the hasattracted
movementof magmainto or within a volcanicsystem. In the dramatichumaninterestslant ("man fightsnature").
building toward an eruptionor intrusion,magma influx Attemptsto mitigatehazardsby divertinga lava flow from
into an uppercrustalmagmareservoircommonlyresultsin potentiallydestructivepaths have been summarizedby
the swelling(or inflation)of a volcano;suchinflationcan Macdonald[1972, p. 419]. The diversionmethodsinvolve
be measuredand trackedby seismicand geodeticstudies. the disruptionof active vent areasor flow channelsor the
With the onsetof an eruptionor intrusion,pressureon the erection of barriers or deflectors. The earliest known lava
reservoiris relieved, and the volcanotypically undergoes diversioneffort was duringthe 1669 eruptionat Etna; it
rapid shrinking(or deflation). An intrusioninvolvesthe was not successful. Rock diversion barriers were consubsurface movement/injectionof magma without stmctedby bulldozers
duringthe 1955and 1960eruptions
culminationin a surfaceoutbreak. For many volcanoes, of Kilauea Volcano,Hawaii, and attemptswere madeto
eruptiveor intrusiveeventsare precededand/oraccompa- disrupta lava flow channelby aerialbombingduringthe
nied by harmonicor volcanictremor, which is charac- 1935 and 1942 eruptionsof Mauna Loa. All of these
terized by nearly continuous narrow-band seismic attemptsin Hawaii were largely ad hoc and ultimately
vibration,predominantlyof a singlefrequency,believedto proved to be unsuccessful.However, more recent studies
be associatedwith motion of magma or of volcanically and effortsin lava diversionhavebeen more systematic
heatedfluids [e.g., Aki et al., 1977; Chouetet al., 1987; andperhapsmoresuccessful
(seethesectiononprogress
in
Koyanagiet al., 1987;Leet, 1988].
engineering-oriented
countermeasures).
27,2/REVIEWSOF
GEOPHYSICS
Tilling: VOLCANIC HAZARDS AND THEIR MITIGATION o245
The mitigationof the damagecausedby debrisflows
and floods commonly involves one or more of the
followingcountermeasures
[Sudradjatand Tilling, 1984;
Japan Sabo Association(JSA), 1988]: (1) dikes and
channelworksalongriverbanksto divertflow courses;
(2)
check and steel-slit dams to retard or prevent the
downstreammovementof large blocksand boulders;and
(3) sandpocketsto traploose,finer-sizedmaterials.These
structures,togetherwith seedingand soil stabilization
projectson steepvolcanicslopes,collectivelyareknownas
"sabo" in Japanand "saborem" in Indonesia,the two
countriesthat have been the most active in the development and construction
of suchsystems.In additionto the
physicalstructures
to containor deflectdebrisflows,some
society. The Japaneseproverbquotedat the beginningof
thispaper[Shimozuru,1981] is especiallyapt for volcanic
hazards.
That more effort in natural hazards studies should be
placed in emergency managementand related societal
impactshasbeenemphasized
in severalrecentworks[e.g.,
White and Haas, 1975; Sheets and Grayson, 1979;
UNDRO/UNESCO, 1985; Peterson, 1986, 1988], all of
which also attest to the complex interactionamong the
scientists,officials, land managers,news media, and the
generalpublic (see, for example,Peterson[1988, Figure
1]). The developmentof more effective emergency
managementof volcanic and other natural hazards
transcendsscientific issuesper se and must, in the final
"sabo works" also include various event detection devices
analysis, be addressedby decision makers within a
(seismicand thermoelectricdevices,"trip wires," etc.) frameworkof othersocietalconcerns(Figure 1). However,
could and must do much more in working
installedhigheron thevolcanoflankto registerthepassage geoscientists
of pyroclasticand/or debris flow to give warning of towardimprovedvolcanicemergencymanagement.
potentialdanger to the communitiesdownstream[e.g.,
Sumaryonoand Kondo, 1985; VolcanicSabo Technical
PROGRESS IN HAZARDS MITIGATION
STUDIES
Centre (VSTC), 1986].
Otherengineering-oriented
countermeasures
includethe
During recent years, considerableprogresshas been
lowering of the water level of reservoirsin the path of
made
in understandinghow volcanoes work, such as
potentialdebrisflows or mudflowsin orderto accommodate the debrisvolumeand to prevent/minimizeovertop- magmasupplyand delivery systems,eruptionfrequency
ping. This measurewas specificallyrecommended
by and dynamics,and eruptiveprocessesand products[Self
Crandell and Mullineaux [1978] in their volcanic hazards and Francis, 1987]. This improvedunderstanding
has in
assessment
of Mount St. Helensandimplementedin 1980, turn strengthenedthe basic underpinningsfor more
weeksbefore the May 18 eruption[Miller et al., 1981]. specializedhazardsmitigationstudies. While no major
Followingthe 1919 eruptionof Kelut (Java, Indonesia), breakthroughsin approach or methodology in hazard
duringwhichmorethan5000 peoplewerekilled by lahars, mitigationcan be claimed during the past decadeor so,
a systemof tunnelswasconstructed
to drainandlowerthe improvementsin instrumentation,data collection and
and dataanalysisand interpretationhaveled
level of waterin Kelut's craterlake. The drainagetunnel transmission,
systemproved its effectivenessin greatly reducingthe to correspondinglymore refined techniquesof volcano
lahar hazards during Kelut's next eruption in 1951 monitoringand eruptionforecasting. Below I highlight
somegeophysicalexamplesof suchadvances.
[NeumannVanPadang, 1960].
Engineering-orientedmethods to minimize tephra
hazardsare not consideredin this paper, becausethey VolcanoMonitoring
mainly require nongeoscienceapproaches,such as
Progresshas been greatest in the general area of
improvementsin the design of structuresand stricter geophysicalmethodsof volcanomonitoring. Reflecting
enforcementof building codes in the affected volcanic the advances in electronics and computerized data
regions. Blong [1981] providesa good review of the collection, seismic monitoring can now quickly and
precisely determine the hypocentral locations (the
impactof tephrafalls andvolcanicbombson buildings.
maximumerror is <2 km, but the averageis <1 km) of
magma-inducedevents [e.g., Klein, 1978; Fremont and
VolcanicEmergency
Management
Emergencymanagement
playsa pivotal,if not the most Malone, 1987]. The largerseismiceventscan be located
critical, role in coping with a volcanicdisasteror crisis. automaticallyby "real-time processing"(RTP) systems
Yet this importantelementin reducingvolcanicrisk draws [e.g., Hill, 1984]. The greatly enhancedaccuracyand
little attention from scientists and decision makers alike,
precisionin analyzingseismicdata have made possible
even in the developedcountrieswith activeor potentially detailedthree-dimensional
modelsof volcanicseismicity
activevolcanoes.This situationis lamentablebut perhaps [Kleinet al., 1987], which,togetherwith grounddeformaunderstandable,in view of the fact that volcanic hazards tion data,providethe primarybasisfor determiningthe
occur infrequentlyrelative to the human life span and configuration
of crustalvolcanicplumbingsystemsin a
comparedto mostothertypesof hazards,let alonerelative variety of plate tectonic settings: convergentplate
to the day-to-day demandsof an increasinglycomplex (subduction)
boundary(e.g., Mount St. Helens [Scandone
246'Tilling:
VOLCANIC HAZARDSAND THEIR MITIGATION
and Malone, 1985]), divergentplateboundary(e.g., Krafla
(Iceland)[Tryggvason,1986]),andintraplate(e.g.,Kilauea
[Ryan,1988]).
Grounddeformationmonitoringmethodsalsohavebeen
refinedwith improvements
in computerized
datacollection
and analysis, especially in connectionwith use of
telemetered, continuously recording tiltmeters [e.g.,
Westphalet al., 1983; Wyatt et al., 1984;Agnew, 1986].
The newest generationsof electronicdistancemeasurement (EDM) instrumentsused to monitor horizontal
grounddeformationare lighter weight,easierto use,and
haveself-contained
microprocessors
for rapidcomputation
of line distancesin the field. The two-colorgeodimeter
[Slater and Huggett, 1976; Langbeinet al., 1982], which
eliminatesthe need for atmosphericcorrections,has been
usedwith goodresultsin monitoringthe grounddeformation at Long Valley caldera [Hill, 1984; Linker et al.,
27,2/REVIEWSOFGEOPHYSICS
Helensand hasprovideda remarkablysimplebut highly
reliable method (Figure 3) to predict dome-building
eruptions
[Swanson
et al., 1983]. Thissimple,inexpensive
monitoringtechniqueshouldencourage
scientists
to devise
and test other "1ow-tech" monitoring techniques,
including
repeated
carefulfield observations
to detectany
visiblechangesin the stateof the volcanowith time.
PREDIC
T!ON
Predictive
ISSUED
--
8/26
• •window
100
I•u
1986].
Somepromisingresultshavebeenobtainedfrom recent
geoelectricalmonitoringstudies. For example,measure-
0
200-
ments of the Earth's weak natural electrical current, called
"self-potential" (SP), near an eruptivefissureabout 60
hoursafter the onsetof the September1977 eruptionat
Kilauea volcanoindicatedlocal increasesas great as 70
mV
[Dzurisin et al.,
1980].
A
controlled-source
electromagnetic (CSEM) technique for monitoring
resistivityhas been employedat Kilauea since 1979, and
correlations
wereobservedbetweenvariationsin resistivity
and some magma intrusions. Particularlysignificantwas
the finding that "some aseismicmagmamovementsthat
are not detectedby seismicor geodetictechniquesmay be
easilydetectedby CSEM monitors"[Jacksonet al., 1985,
p. 12,555]. Yukutakeet al. [1987] also observedsignificant magma-relatedtemporal variations in electrical
resistivityat OshimaVolcano,Japan. A combinationof
geoelectrical techniques (SP, CSEM, and very low
frequency (VLF)) is being used to monitor the
1983-presentPu'u 'O'o eruptionat Kilauea,the longestlived rift eruptionin Hawaii in historicaltime [Jackson,
i
1300-
1981
JUL I AUGI SEPi
Figure 3. An exampleof a usefulandinexpensive"low-tech"
monitoringtechniqueusedto successfully
predictdome-forming
eruptionsat Mount St. Helens. The plot showsthe cumulative
contractionof the distance(measuredby steeltapebetweentwo
benchmarks) across the toe of a thrust fault at base of the
growing lava dome. The arrow marks the issuancedate of the
eruptionprediction,the blackrectangleshowstheperiodduring
which the eruptionwas predictedto occur ("predictivewindow"), and the dashedverticalline indicatesthe eruptiononset.
Suchsimplemeasurements
arein goodaccordwith dataobtained
with more sophisticated
but more expensivemonitoringmethods
(seeFigure4) [fromSwansonet al., 1983,Figure3].
1988].
Recent advancesin geodeticapplicationsof satellite
positioningand otherforms of spacegeodesy,especially
the Global PositioningSystem (GPS) being tested in
Hawaii and elsewhere,hold increasingpromisethat the
few-parts-per-million resolution needed for volcano
monitoringmay be soonattained[e.g.,Prescottand Svarc,
1986; $chutz,1987]. The GPS, onceit is fully testedand
shown to be routinely preciseand accurate,can augment
and/orsupplantconventionalgeodeticvolcanomonitoring.
Not all recentprogressin volcanomonitoringhasbeen
"high tech" and expensive. For example,the manual
measurementof distancesbetweenbenchmarksby means
of a steel tape monitors movement acrosssmall thrust
faults at the baseof the growinglava dome at Mount St.
EruptionForecasting
and Prediction
Progressin volcanomonitoringhas not been matched
by commensurateadvancesin eruption forecasting.
Generallyspeaking,changesin the rate and amplitudeof
monitoredparametersprovideinsufficientinformationfor
preciseforecastsof the place, time, and characterof the
anticipatedactivity. Nevertheless,accordingto Decker
[1986,p. 269], precursory
seismicityhasguidedobservers
to arrive "within a few hundred meters of the outbreak
locations"beforetheonsetof "all theeruptions
of Kilauea
Volcano since 1979."
Strictlyspeaking,however,the successful
anticipation
of the likely sitesof eruptiveactivityat Kilaueadoesnot
constitute
a true "forecast"or "prediction,"becausethe
27,2/REVIEWSOFGEOPHYSICS
Tilling: VOLCANIC HAZARDSAND THEIR MITIGATION o247
specifictime of the outbreakcannotbe determinedin
advance.
The remarkable record of successful forecasts
and predictionsat Mount St. Helens since June 1980
[Swansonet al., 1983] provides a context for useful
distinctions between the terms "factual
predictivemethodsused at Mount St. Helens need to be
testedfor more voluminousand explosiveeruptionsthere
and at other volcanoes.
SUCCESSIVE
statement,"
PREDICTIVE
WINDOWS
"forecast,"and "prediction"asdefinedby Swansonet al.
[1985, p. 397]:
A factual statementdescribes
currentconditions
but doesnot
anticipatefuture events. A forecast is a comparatively
imprecisestatement
of the time,place,andideally,thenature
andsizeof impendingactivity. A predictionusuallycoversa
shortertime period than a forecastand is generallybased
dominantlyon interpretations
and measurements
of ongoing
processes
andsecondarily
on a projectionof pasthistory.
A classicexampleof a successful
long-termforecastis that
of Crandellet al. [1975, p. 441], who wrotethatMount St.
Helenswill "erupt again,perhapsbeforethe end of this
century." The volcanoerupted5 yearslater. An example
of one of the many successfulpredictionsat Mount St.
(1)
2
(2)3/15
SEISMIC
(3) 3/19
ENERGY
RELEASE
1
(x108ergs)1/2
0-
200
DOME
,,
-
o
EXPANSION
(cm)
0
3,000
-
TILT
-
(microradians)
Helens and associated factual statements is illustrated in
Figure4 and Table 3. About 10 hoursafter the updated
predictionat 0900 local time (LT) on March 19, 1982,
eruptionbegan with an explosion,precededby about 2
3/12
_
I
./-'
,
,
0
10, 000 SULFUR
DIOXIDE
EMISSION
(tons)
0
2/1
3/1
3/19
1/10/82
hoursof intenseseismicity. Extrusionof new lava began
about1 day later and continueduntil April 12 [Swansonet Figure 4. The accelerationof precursoryactivity before the
I
I
al., 1985]. In my opinion the near-perfectrecord of
predictionsof dome-buildingeruptionsat Mount St.
Helensis one of the mostsignificantadvancesin hazards
mitigation since the emergenceof volcanologyas a
I
onsetof the March-April 1982 eruptive activity at Mount St.
Helens(modifiedfrom Swansonet al. [1985, Figure3]). Dome
expansionis measuredby electronicdistancemeasurements
of
horizontaldisplacement,tilt changeby electronictiltmeter, and
sulfurdioxideemissionby correlationspectrometer
(COSPEC).
modem science. It must be remembered, however, that
Excerptsfrom predictions1, 2, and 3, issuedon March 12, 15,
theseeruptionsare relativelysmallevents,hazardousonly and 19, respectively,are givenin Table 3; solidbarsindicatethe
narrowingpredictivewindows.
to thoseworkingwithin the summitcrater. The successful successively
TABLE 3. Examplesof Factual Statementsand Predictionsof the Onsetof the March-April 1982 Eruptive
Episodeat Mount St. Helens
Time, LT
Date
Factual Statementor Prediction Issued*
0900
March
5
Factualstatement:"Seismicity... increasedaround21 Februaryandhasremainedat
a level somewhatabove backgroundsince that time.... Measurementsmade last
week (27 February) show only slow ground deformation... and no significant
increasein gasemissions."
(Measurementsafter 0900 on March 5 showincreasedratesof deformation.)
0800
March
12
Factualstatementand prediction1' "Seismicity... continuesat elevatedlevels....
Ratesof grounddeformationin the craterhave increasedduringthe last two weeks.
... Basedon rates of deformation,an eruptionis likely within the next 3 weeks.
Deformationis conf'medto the craterarea,suggestingthatreneweddomegrowth will
occur. ' '
(Measurements
on March 15 showedgreatlyaccelerated
deformation.)
1900
March
15
0900
March
19
Prediction3, updated: "An eruptionwill beginsoon,probablywithin 24 hours. The
characterof both the seismicityand deformationin the craterarea indicatesthat the
mostlikely typeof activityis domegrowth."
1927
March
19
Eruptionbegins(seetext).
Prediction2, updated: "An eruption,most likely of the dome-buildingtype, will
probablybeginwithin 1 to 5 days."
(Ratesof deformationandseismicenergyreleasecontinuedto increaserapidly.)
ModifiedfromSwanson
et al. [1985].
*QuotedfromSwanson
et al. [ 1985,pp.415-416].
248 o Tilling: VOLCANIC HAZARDSAND THEIR MITIGATION
Voight [1988a], drawing on data for dome-building
eruptionsat Mount St. Helens and Bezymianny(Kamchatka)in 1960,proposesa methodof predictionbasedon
a mathematicalanalogy between the terminal stagesof
materials failure and precursoryvolcanic behavior. A
novelfeatureof hispredictionmethodis theuseof inverse
rate-timeplots of monitoringdata (grounddeformation,
seismicenergyrelease,etc.), and Voight claimsthat his
approachcould improve predictionsby recognitionof
predictivewindows soonerthan other methodsand by
narrowingof theirwidths. However,the generalutility of
Voight's predictionmethodremainsto be testedagainst
explosiveeruptivestylesand a varietyof volcanotypes
[Tilling, 1988c]. To date,only a few explosiveeruptions
havebeensuccessfully
predicted: Tolbachik,Kamchatka,
[Tokarev, 1978]; Mount St. Helens [Swansonet al., 1985];
andSakumjima[Ishihara,1988].
Hazards Assessment and Zonation
27,2/REVIEWSOFGEOPHYSICS
Engineering-Oriented
Countermeasures
The massivedebrisavalanchefrom the May 18, 1980,
eruptionof Mount St. Helensblockedpartsof thepreexisting drainageof the north fork of the Toutle River to form
naturaldams,behindwhichlakeswereformedor enlarged
with rainfall and runoff [Meyer et al., 1986]. It was
recognizedthat the failure of these "landslide" dams,
which are composedof unconsolidated,
easily erodible
volcanic debris, could causecatastrophicmudflows and
floods, especiallyduring the winter, when rainfall and
snowpackare maximum. As an interim emergency
measure,the U.S. Army Corpsof Engineers,in the fall of
1982,beganto controlthe rise of the level of waterbehind
the largestof thesedebrisdams (Spirit Lake) by bargebasedpumpingand dischargeinto outlet channels. The
water level of and dischargefrom Spirit Lake are now
regulatedby a permanentsystemof pipe and runneloutlets
[Sagerand Chambers,1986].
The
diversion
of lava flows
to minimize
volcanic
hazardshas receivedmore attentionfrom the geoscience
community than other engineering-oriented
measures.
Althoughlava diversionprobablywill neverbe a major
meansof hazardsmitigationbecauseof its high costsand
enormouslegal implications[Decker, 1986], it may be a
viable option under special circumstances[Lockwood,
United States and other countries [see Scott, 1989, Table
1988]. Perhapsthe greatesteffort by man to controllava
3.5]. However, thesehazardsassessments
vary widely in flows occurredin 1973 at Heimaey, Iceland; massive
scaleand in the breadthand depthof data usedto prepare volumesof seawaterwere sprayedto cool the advancing
The reliabilityof a volcanichazardsassessment
depends
on the quality and abundanceof basicgeologicdata and
the time spanencompassed
by the data baseusedin the
assessment.Progressin hazardsassessments
in recent
decadesis measuredlargely by the fact that they are now
becoming increasinglyavailable for volcanoesin the
them.
flow front, and diversion barriers were constructed in an
Most assessments
focus on potential volcanic and attempt to protect a vital fishing harbor [Williams and
associated
hazards,commonlyrankedin termsof relative Moore, 1983]. Icelandicscientistsgenerallybelievethat
severity,but very few attemptto quantifythe probabilities the diversion effort was successful in the sense that the
of each type and magnitudeof hazard. Crandell et al. harborlikely would have been destroyedif no effortsat
[1984] briefly review the few attemptsmade in the diversionhadbeencarriedout. Duringthe 1983 eruption
quantificationof hazards,includingthe semiquantitative of Ema, explosivesand earthen barriers were used in
probabilistic assessmentof intermediate- (weeks to attemptsto divertmajor flows [Abersten,1984;Lockwood
months) and long-term (years to millennia) volcanic and Romano,1985]. This lava diversioneffort generated
hazards and risks at Mount St. Helens as proposedby considerable
controversyamongItalian scientists,
largely
Newhall [1982]. Subsequently,
Newhall [1984a] extended becauseof its high costs (3 million dollars), but most
his methodto considerthe probabilitiesof the following concededthat the efforts, especiallythe use of barriers,
five short-term(week-to-week or shorter)volcanic states: were successful.
quiet, slight unrest, severe unrest, dominantly nonexThe reawakening
of MaunaLoa in 1975,after25 years
plosivedome growth, and explosiveeruption. Newhall of quiet [Lockwoodet al., 1987a], generateda flurry of
cautionsthat uncertainties
in the estimatesof probabilities officialinterestin lavadiversionplansto protectthecity of
are at leastan orderof magnitude. More recently,Hoblitt Hilo from possiblefutureMaunaLoa flows [e.g.,Stateof
Pacific (CINCPAC),
et al. [1987], in a preliminarystatisticalevaluationof Hawaii, 1977; Commander-in-Chief,
tephrahazardsfor the entire CascadeRange,give calcu- 1978; United StatesArmy Corps of Engineers,1980].
lated minimum annualprobabilitiesof areasreceiving1, Someof the emergencyplanningincludedexperiments
on
10, and 100 cm of tephraduringfutureeruptions.Most of bomb-cratering tests [Torgerson and Bevins, 1976;
the uncertaintyin theircalculations
derivesfrom the large Lockwood
andTorgerson,1980]shouldaerialbombingbe
uncertaintyin the annual eruptionprobability;too few exercisedasan optionin lava diversion;bombinghadbeen
postglacial eruptions have been documentedfor most tried previouslyon Mauna Loa flows in 1935 and 1942
Cascade volcanoes.
with little planningor success.As it turnedout, no lava
27,2/REVIEWSOFGEOPHYSICS
Tilling: VOLCANIC HAZARDSAND THEIR MITIGATION o249
diversionefforts were necessaryduring the 1984 Mauna
Loa eruption, because flows advancing toward Hilo
"self-diverted" by naturalbranching[Lockwoodet al.,
disaster
in thehistoryof theUnitedStates,causingtheloss
of 57 lives, scoresof injuries, and more than 1 billion
dollars'worthof damage.Thefirstminuteof theMay 18
1985]. In 1986 a lava diversionbarrier was constructedto
eruption
produced
a 2.7-kin
3debris
avalanche,
a powerful
protect the Mauna Loa Observatory,an atmospheric laterallydirectedblast, and the start of a 9-hour-long
researchfacility (operatedby the National Center for plinianeruption
that ejectedmorethan 1.1 km3 of
AtmosphericResearch)situatedat 3425 rn elevationon the uncompacted
tephra(Figure5). Shortlythereafter,small
north flank of the volcano [Lockwood et al., 1987b]. The pyroclasticflows alsobeganto occur,andthe hot volcanic
effectiveness
of thisbarrieris yet to be tested.
materialsmeltedsnowandglacialice, causingdestructive
In recentdecades,Japanhasmade significantadvances mudflowsand floods. With the exceptionsof a few
in the general area of engineering-orientedcounter- smallerexplosiveepisodes,
the post-May 1980 eruptive
measuresto mitigate the hazardsfrom debrisflows and activity of Mount St. Helens has mainly involved the
from tephm. A vigorouslong-term national program endogenous
and exogenousgrowthof a compositelava
operates to upgrade existing "sabo works" and to dome (Figure6) within the craterformedon May 18
constructnewones,as well as to maintainandimprovethe [Swansonet al., 1987; Chadwicket al., 1988].
country'ssystemof tephra fallout shelters,dugoutsand
Becausethe eruptionof Mount St. Helens was thoropen spacesfor heliports, and other refuge facilities oughlydocumented
andreceivedcopiousmediacoverage
[DisasterPreventionBureau(DPB), 1988a]. As a specific worldwide, public and scientific awarenessof volcanic
exampleof suchimprovements,somesaboworksinclude phenomena
and hazardsincreaseddramatically. Almost
a warning system that automaticallyshuts down road overnight,"Mount St. Helens" becamehouseholdwords
traffic, in a manneranalogousto thatat railway crossings, synonymouswith volcanic catastrophe. The disaster
when the event detectionsensorhigh on the volcano's catalyzed,
andprovidedcompellingbudgetary
justification
flank registersthe occurrenceof a pyroclasticflow or for, an expandedprogramof volcanichazardsstudiesin
lahar.
theUnitedStates[TillingandBailey, 1985]. Before1980,
programsrepeatedlyproposedby the U.S. Geological
THE 1980s:
DECADE
OF VOLCANIC
DISASTERS
Survey(USGS)to increase
researchon andmonitoring
of
AND
CRISES
The catastrophic
eruptionof Mount St. Helenson May
18, 1980,usheredin the mostdisastrous
periodof volcanic
hazardssince the year 1902, when more than 36,000
peoplelosttheirlivesasa resultof volcanicactivity(Table
1). It thus seemsuseful to briefly review the volcanic
disasters
andcrisesin the 1980s[Tilling, 1986] within the
contextof hazardsmitigationstudiesand then to consider
the lessonslearned, or not learned, and the measuresthat
shouldbe takento reducelossesfrom futureeruptions.
Major Volcanic Disasters
U.S. volcanoes in addition to those in Hawaii failed to
receivefundingfromCongress.After May 1980,however,
increasedfunding permitted the USGS to establisha
permanentfacility at Vancouver,Washington,to monitor
the continuingeruptiveactivityat Mount St. Helensandto
acquirebaselinedata for other Cascadevolcanoes. On
May 18, 1982, this facility was formallydesignated
the
DavidA. Johnston
Cascades
VolcanoObservatory
(CVO)
in memoryof the USGS volcanologist
killed 2 years
earlier at Mount St. Helens. More recently,undera
cooperativeprogrambetweenthe USGS and the stateof
Alaska, an Alaska Volcano Observatory,considerably
smallerthan HVO or CVO, was formed in March 1988.
El Chich6n, Mexico, 1982 E1 Chich6nVolcano
(Chiapas,southeastern
Mexico), eruptedviolentlythree
timesbetweenMarch28 andApril 4, 1982. The eruptions
destroyed
a domeat the summitof E1Chich6n,leavinga
1-kin-wideand nearly 300-m-deepcrater in its place.
Pyroclastic
flowsandsurgeswipedoutall villageswithina
7-kin radiusof the volcano(Figure7), killing morethan
2000 people[Duffieldet al., 1984;Luhr and Varekamp,
1984]. In additionto wreakinglocalhavocthe eruptions
attractedworldwideattentionfrom atmospheric
scientists
magmaintrusioninto the volcaniccone[Endoet al., 1981; because
of theirpossibleimpacton globalclimate[e.g.,
Lipman et al., 1981]. Triggeredby a magnitude5.1 RampinoandSelf,1984;Galindoet al., 1984]. Although
earthquake
on the morningof May 18, the paroxysmal theE1Chich6nandMountSt. Helenseruptions
produced
Mount St. Helens, U.S.A., 1980 Before1980,Mount
St. Helenshadbeendormantsince1857,andpeopleliving
in the region did not think of the mountainas an active
volcano. Followinga week of intenseprecursoryseismicity, phreatic explosionsbegan on March 27, 1980.
During the ensuingintermittentphreaticactivitythrough
May 17, volcanomonitoringwas greatly expandedand
confirmedvisualobservations
that the volcano,especially
the north flank, was deformingat a high rate becauseof
eruptionof Mount St. Helens marked the worst volcanic
aboutthesameamount
of silicate
ejecta(0.3-0.5km3
Figure 5. The tephraandgasplume attaineda maximum altitude
of about24 km duringthe plinianphaseof the climacticeruption
of Mount St. Helens on May 18, 1980. The photographwas
takenby R. M. Krimmel (U.S. GeologicalSurvey) about5 hours
after the onsetof activity.
denserock equivalen0, El Chich6n injected much more
believedthat its last activity was during the early Pleis-
to be a high-riskvolcanoby
aerosols,
highlyenrichedin SO2, into the lower strato- tocene.It wasnotconsidered
sphere. Its stratospheric
volcaniccloud,consideredto be
among the largest ones in the twentieth century [e.g.,
Mitchell, 1982;Hofmann,1987],remaineddetectableuntil
late 1985.
The
1982
outburst of E1 Chich6n
caused the worst
volcanic disasterin Mexico's recordedhistory. Before
1982, El Chich6n was an obscure, little studied volcano
sinceits "discovery" in 1928 by Miillerried [1932], who
scientists,officials, or the local populace. The 1982
resumptionof activity came almost as a total surprise.
There were no recognizedimmediateprecursorsto the,
1982eruption,although(in hindsight)earthquake
activity,
somefelt by local inhabitants,increasedfor somemonths
before the first explosion [Havskov et al., 1983].
Moreover,during the period November 1980 to April
1981, two geologists,while doingfieldwork in connection
27,2/REVIEWSOFGEOPHYSICS
Tilling: VOLCANIC HAZARDS AND THEIR MITIGATION.251
with assessingthe geothermal potential of the area,
frequentlyheardloud noisesand felt earthquakes.In their
report [Canul and Rocha, 1981] they suggestedthe
possibility of "subsurface magmatic activity and/or
tectonicmovements." They specificallyconcludedthat
"in this area there is a high volcanic risk that must be
consideredif one wishesto developa geothermalfield"
(translation
quotedby Duffieldet al. [1984,p. 119]).
1986].
During the eruption'smost energeticphase
(mid-May throughAugus0, eruptioncolumnsrose20 km
into the atmosphere
and ash fell on the capitalcity of
Jakarta,350 km to the west. Severalpyroclasticflows
sweptdown the principal valleys, drainingthe volcano,
andmanyrainfall-triggeredlaharsoccurred.
Fortunately,the initial activity of the eruption was
relativelymild, and peopleliving on or near the volcano
Reconnaissance studies after the 1982 disaster indicated
wereableto evacuate
safely.Whiletherewerenoreported
thatE1Chich6nwasfrequentlyandviolentlyactiveduring humanfatalitiesdirectlyattributable
to the eruption,the
the past few thousandyears, including major eruptive 9-month-longeruptionof Galunggung,
unusuallyproepisodesoccurringon average about every 600 (+200) longedby Indonesianstandards,was a major volcanic
years[Tilling et al., 1984]. Might suchinformationbefore disaster
because
it causedmassivesocioeconomic
impact
1982 haveeliciteda differentresponse
from scientists
and andadverselyaffectedthedaily livesof morethan600,000
officials to E1 Chich6n'searly signsof unrestin 19817 peoplein West Java. More than 80,000 peoplewere
Unfortunately, we will never know the answer to this forced to evacuate;most were able to return home, but
qt•estion.
about35,000 were left permanently
homeless. Many
Galunggung,West lava, Indonesia,1982-1983
hundreds of homes, schools, and other structures were
Severalmillion peoplelive in the shadowof Galunggung
Volcano,which before 1982 had eruptedin 1822, 1894,
and 1918. Laharsof the 1822 eruptionkilled more than
4000 people,but the next two eruptionsdid little or no
harm. In early April of 1982, Galunggungcameback to
destroyed;transportation
and communications
systems
weredisrupted;
and someof the mostproductive
agricultural landsand fishponds(growingfish for food) were
ruined and/or threatenedby lahars. In all, the total
economicloss exceeded100 million dollars (U.S. curlife with virtuallyno precursory
activity,andtheeruption rency). While the long durationof the eruptionstrained
continuedintermittentlyuntil early 1983 [Katilit et al., the limited scientificresourcesof the Volcanological
ß
ß •_'•:• .:::_
_:'...
..
.:.-...:..:
:.¾•:.
• :.,•'!}
:..
.-::'
-.
- "• ....
•::•
<'.
. "'
'"%
:'•
:•-':•
"•:'
'• ':"':'
........
.i•'-•ß'::?:'.:.:.::....;:--.-',-'•-:•--.-:5'?"¾•
•:.•?•
•'•-.•":'.::--•"•'.
:"::"
.......
ß
........
:':...'..
_ ..- ....
.•
.:.};•j::•::•
....
'......::::......?
?.?.'
-:•.
...:.•.:
....
;•.:•_.
:•
-
•"•'•'•
':•.•
'.:/':'.'_..
•... -::•:,.,
'. •:'.•:.• ..•:.,•.:•:.---•-•.•
•••-•::7"•-..
•_....•..:.•:
•.."•,':•.?•...:•:..•...-._
...:.-......•....-•.
:•...:...•?.,•.
•_.....?
•'"'";..•/-• • •
¾..--. ...•••
....
:. .......
..........
.......
- ...,..?.,-.. -.:-::-•.••
•. ,...........
.-:.....•
.........
,...
;.:..
<
'
.
...
.:.?-"
..........•.......
...
:..• ••
• '"'
".::......
'.:
......
:?•
• .' •.• :.-.L:y'•-':..•:-'.
"-:z•••-.•:•"
•;* '-.-- " ' '•.... "• .... --'--:.'
..'" -'......"-.2-" ."
:...._.
, ..... ...............
.......
.....
.•,.•.... .-::......•.-..::
.
.
....
...,
....•.... ........-..•_
• .....'•.'.•:•
-.• '• .-•.'•.
.••.••:..•••.•..••::•...•
.•
....
:...
. .:....:..:•
Figure 6. Small eruption plume rises from the lava dome wastakenby L. Topinka(U.S. GeologicalSurvey)in May 1982,
growing inside the large, amphitheaterlikecrater formed by the when the dome measuredabout 850 m long, 800 m wide, and
May 18 climacticeruptionof Mount St. Helens. The photograph 230 m high.
252'Tilling:
VOLCANIC HAZARDS AND THEIR MITIGATION
27,2/REVIEWSOFGEOPHYSICS
Figure 7. Oblique aerial view showingthe site of Francisco (circled) remain of the village, locatedabout7 km southof the
Le6n, one of severalvillages that was obliteratedby pyroclastic volcano'ssummit.(Thephotograph
wastakenon June2, 1982,
flows and surges during the 1982 eruption of E1 Chich6n by R. I. Tilling (U.S. GeologicalSurvey)).
Volcano, southeasternMexico. Only the ruins of the church
Survey of Indonesia,it also afforded an unprecedented Nevado del Ruiz, Colombia, 1985 On the aftemoon
opportunityto apply a combinationof modem volcano- of November 13, 1985, a relatively small volume (0.029
eruption
beganat Arenas
Craterat the
monitoring techniques not attempted previously on km3) explosive
Indonesian volcanoes [Katili et al., 1986].
summit of Nevado del Rufz, a 5389-m-high, glacieractivevolcanoin
The Galunggungactivity also dramaticallyemphasized cappedmountainthat is the northernmost
the threat of volcanic ash clouds to international aviation.
the Andes. The hot ejecta melted and mixed with snow
On threeoccasions
in theperiodMay-July 1982,commer- and ice to form severalhighly destructivemudflowsthat
cial jetliners encounteredash plumes from Galunggung, raceddown the narrow,steepvalleys,drainingthe volcano
resultingin ash abrasionof cockpit windshieldsand/or andkilling morethan22,000 people,the vastmajorityin
engine failures. During two of these encountersthe thetownof Armeroin the valleyof Rio Lagunillas(Figure
ingestionof ash causedthe flameoutof two or more 8). In addition,the mudflowscausedseriousinjury to
enginesof Boeing 747s and emergencylandingsat the another 5000 people, left about 10,000 homeless,and
Jakartaairport [de Neve, 1986]. Fortunately,becauseof resultedin an economiclosstotalling212 million dollars
rapid and skillful pilot responses
to theseair emergencies, (U.S. currency). The Rufz catastrophewas the worst
disasterswere averted. However, theseincidentsprompted volcanicdisasterin the recordedhistoryof Colombiaand
the InternationalCivil Aviation Organization(ICAO) to the worst in the world sincethe 1902 eruptionof Mont
create, in September,a Study Group on Volcanic Ash P½16e
[Herd and the Cornitdde EstudiosVulcanol6gicos,
Warnings(VAW), composedof volcanologists,
pilots,and 1986; Naranjo et al., 1986; Lowe et al., 1986; Voight,
civil aviationspecialists[Scarone,1985]. In 1987, acting 1988b].
on recommendations
of this studygroup,the International
In contrastto E1Chich6nin 1982 the 1985 eruptionand
Air TransportAssociation(IATA) institutedthe use of associated mudflows at Rufz should not have come as a
"experimental" forms (SpecialAir Reports)to be carried surprise. Accordingto Herd and the Comitdde Estudios
on all international flights for reporting of volcanic Vulcanol6gicos[1986, Figure 2], at least 10 major
activity, as well as other measuresto facilitate rapid eruptionsand severalsmallereventshaveoccurredat Rufz
communications between volcanologists, air traffic Volcanoduringthepast4000 years.Its lastmajoreruption
controllers,andpilots.
was in A.D. 1595, which apparentlyproducedhuge
27,2/REVIEWSOFGEOPHYSICS
Tilling: VOLCANIC HAZARDS AND THEIR MITIGATION'253
Figure 8. Remainsof the village of Armero, devastatedby
lahars triggeredby the relatively small magmaticeruptionof
Lagunillas(upperright comer),killing morethan22,000people.
The photographwas taken by D. G. Herd (U.S. Geological
November 13, 1985, at Nevado del Ru•z Volcano, Colombia.
Survey).
The deadly lahars surged from the canyon mouth of Rfo
mudflowsin the Rfo Lagunillasvalley, upon which the
town of Armero later developed [Voight, 1988b]. In
February1845a lahar,generated
by a largeearthquake
or a
phreaficeruption,again sweptdown the Rfo Lagunillas
and killed more than 1000 people; this mudflow also
reached the modem site of Armero [Herd and the Comit•
de EstudiosVulcanol6gicos,
1986].
The story of the 1985 Ru•z tragedybegan in late
November1984,whenprecursorysignals(felt earthquakes,
increasedfumarolicactivity,phreaticexplosions,
andeven
burstsof harmonictremor) were first noticed [Tomblin,
1985; Herd and the Comit• de EstudiosVulcanol6gicos,
1986]. Anomalous behavior continued into 1985, and,
largelythroughthe effortsof J. Tomblin,Office of the
United Nations Disaster Relief Co-Ordinator (UNDRO,
Geneva),effortsweremadethroughout
theyearto increase
volcanosurveillance(seeHerd and the Comit• de Estudios
Vulcanol6gicos
[1986]for detailedaccoun0.
By summer,rudimentary
seismicmonitoringhadbegun,
and the Comit6 de Vigilancia del Riesgo Volcfinicodel
Ru•z (Ru•z Volcanic Risk Committee) was formed.
However, it was not until September11 (when a strong
phreatic eruption produced measurable ash fall at
Manizales, the capital of Caldas Province (population
230,000), andsomesizeabledebrisflows in valleysof R•o
Azufrado and R•o Gua10 that the level of concernwas
heightenedand additionalinternationalscientific assistancewasrequested.By October7 a preliminarydraft of
the Rufz volcanic hazard zonation map (Figure 2) was
completedand released. By mid-Octobera monitoring
networkof five smokedrum seismometers
and four "dry
flit" arrays was established. Meanwhile, intermittent
phreaticactivitycontinuedeven as seismicitywaned,and
Colombianscientists
discussed
the hazardsmap with local
officialsandbriefedthemon potentialhazards.
The November13 eruptionbeganat 1505 (local time);
explosionsoccurredintermittently,but the strongestones
happenedshorfiyafter 2100 LT (Herd and the Comit• de
Estudios Vulcanol6gicos,1986; Tomblin, 1988]. The
destructivemudflows did not begin striking populated
27, 2 / REVIEWS OF GEOPHYSICS
254 øTilling: VOLCANIC HAZARDS AND THEIR MITIGATION
areas in the affected river valleys until about 2230 LT.
Despite various warnings and alerts given beginning
around 1600 LT, "there apparentlywas little or no
responseto calls to evacuate. It is unclearif the general
populaceof Armero receivedan order to evacuate"(Herd
and the Comitdde EstudiosVulcanol6gicos,1986, p. 459].
However, Tomblin [1988, p. 10] remarks that a partial
evacuation,promptedby the first eruptionsin the afternoon, "two hours later appearedto be unnecessary
and
then became unpleasantbecause of heavy rain and
nightfall. Consequently,the report of new explosionsat
9.00 [sic] that evening, which were not adequately
describedas significantlylarger,met with scepticism
from
local authorities and populations over the need to
historical volcanic event (Tambora, 1815).
Mount
Mazama
and Tambora
events
are
Both the
orders of
magnitude smaller than the three enormous caldera-
forming eruptionsthat took placeat Yellowstonewithin
thepast2 millionyears[Christiansen,
1984]. Fortunately,
within recordedhistory, mankindhas not experienced
caldera-forming
eruptionson the scaleof Yellowstone,but
thereare geologicreasonsto supposethat sucheruptions
surelywill happenagain.
From seismicand othergeophysical
evidencewe know
that a magma reservoir still exists beneathYellowstone
[e.g., Eaton et al., 1975; Iyer, 1978; Smith and Braile,
1984]. Alsorecentgeodeticstudiesindicatethat the floor
of thepresentYellowstonecalderahasundergone
significantverticaldisplacements
(tensof centimeters)
duringthe
Another contributoryfactor was that throughoutthe past60 years[e.g., Pelton and Smith, 1982; Dzurisin and
year-long volcanic crisis, the local populacereceived Yamashita,1987]. Yellowstonecalderadoesnot posean
evacuate."
contradictorymessagesfrom governmentofficials and
other influentialpeople. For example,on September28
the Archbishopof Armeroanda localmedicaldoctor,who
was a prominentcivic leader,pronouncedthat "the alarm
over the potentialhazardsfrom Rufz was anti-socialand
irresponsible"(asquotedby WilliamsandMeyer [1988,p.
1554]). On November 13 the citizens of Armero were
advised"eruptiondetected,but 'stay calm'" a few hours
before
the destructive
mudflows
struck.
In
the most
detailedaccountto date, Voight [1986b, p. 30] concludes
that the Rufz tragedy"was not producedby technological
ineffectiveness
or defectiveness,
nor by an overwhelming
eruption of unprecedentedcharacter," but was caused
"purely and simply,by cumulativehumanerrorrubymisjudgment,indecision,andbureaucratic
shortsightedness."
It remainsunclearto this day what actuallytranspired
during the critical hoursbetweeneruptiononsetand the
arrival of the first mudflowsin populatedareas. In a
systematicattemptto reconstructwhat went wrongat Rufz
a group of Latin American geoscientistsand social
scientistsis currentlystudyingthe scientificand public
response
aspectsof the Rufz catastrophe.In retrospect
it is
soberingto rememberthat the Rufz eruptionejectedonly
about3% of the volumeof ash producedduringthe May
18, 1980,eruptionof Mount St. Helens. The Ru•z disaster
emphaticallydemonstrates
thatin denselypopulatedareas,
even very small eruptionscan causewidespreaddevastation andkill manypeople.
Volcanic Crises at Calderas
The mostpowerfulbut leastfrequentvolcaniceventsare
caldera-formingeruptionsof ignimbrite(ash flow tuffs)
from large-volumemagmaticsystems,suchas the Mount
Mazama eruption6850 yearsago that formedCraterLake,
Oregon [Bacon, 1983]. The Mount Mazama eruption
immediatevolcanic threat. However, unrestat three other
calderasduring the 1980s has causedvolcanic crisesof
seriousconcernto scientists,
officials,and the populace
affected.
Campi Flegrei,Italy, 1982-1984 CampiFlegrei
(PhlegraeanFields) is a 12-km-diametercalderacentered
about 15 km west of Naples, Italy. Calderaformation
occurredduringa majoreruptionabout34,000 yearsago
and was followedby severalother successively
smaller
eruptions,the most recent being that of Monte Nuovo in
A.D. 1538 [Lirer et al., 1987]. Campi Flegrei has
undergonelarge vertical movementssinceat leastRoman
times. Subsidence
thatprevailedsincethe 1538 eruption
wasreversedin theperiod1969-1972,duringwhichabout
1.7 m of uplift occurred,accompaniedby a modest
increasein seismicity[Corradoet al., 1976/1977]. For the
next 10 years,grounddisplacements
showedminorannual
fluctuations
of the order of 10-15 cm, and seismicity
oscillated at low levels.
In the summerof 1982,rapiduplift beganandcontinued
for twoyears,accompanied
by intenseearthquake
activity,
resultingin another1.8 m of uplift centeredwithin the
coastalcity of Pozzuoliinsidethe caldera(Figure 9).
Geophysical
andgeodetic
measurements
(gravity,leveling,
EDM) as well as tide gaugedata all indicatedmaximum
uplift centeredat Pozzuoli[Berrinoet al., 1984]. Several
particularlyenergeticearthquakeswarms, including
magnitude4 events,in October1983 causedsomeof the
older brick buildingsin Pozzuoli to collapseand led
officialsto evacuate
andresettlenearly40,000people.In
responseto the growingcrisisthe seismicand geodetic
monitoring
wasaugmented
by geochemical
monitoring
of
thermal springsand fumarolic gases [Barberi et al.,
1984b].
The uplift andseismicitypeakedin early 1984 andthen
vented
about
50 km3 of magma,
about
thesame
volume beganto decline. By January1985, therewaslittle or no
(denserock equivalent)eruptedduringthe largestknown detectableseismicity,and a weak subsidencetrend had
27,2/REVIEWSOFGEOPHYSICS
Tilling: VOLCANIC HAZARDS AND THEIR MITIGATION o255
Quarto
52'
25
\
\
\
\
\
\
Na
50'
Solfatara
Po:
"',Baia
N
.o
ß
ß
o
.
..
ß
..
ß
ß
.
ß
ß
ß
.
..
.
ß
ß
ß
ß"..... .
ß
Miseno 0
1
2 KILOMETERS
I
47'
ß
ß
ß
.
.
10'
14•05 '
Figure 9. Grounddeformationassociated
with the calderaunrest
at Campi Flegrei, Italy. Contour lines show the vertical
displacements
(in centimeters)for the period 1980-1983. The
city of Pozzuolilies within the centerof maximumuplift [from
displacements
of benchmarks
(in centimeters)
duringthe period Berrinoet al., 1984,Figure12;Decker,1986,Figure10].
January1982 to January1985; the vectorsindicatehorizontal
begun. Italian scientists(G. Luongoet al., in the work of
ScientificEvent Alert Network (SEAN) [1985, p. 5])
considerthe net cumulative uplift (350 cm) since 1969
"the major geologicalevent at Campi Flegrei in the last
180 years" and "comparablewith that which occurred
before the eruptionof Monte Nuovo in 1538, and . . .
considerthisphaseasa possibleprecursor(on a time scale
of years) to new volcanic activity." Becauseof the
populationdensity of the Campi Flegrei region, about
200,000 people could be affected by even a minor
eruption,say comparableto Monte Nuovo in 1538. With
the 1982-1984 crisis abated, the people who were
evacuated"have slowly beenmovingback into the town.
The new settlement on the NW
side of the caldera has
almostbeencompleted,but it is not certainthat all of the
people evacuated. . . will move in. The fears of an
impendingeruptionare alreadygone,and the interestof
the mediahasvanished"(G. LuongoandR. Scandone,in
the work of SEAN [1986, p. 8]). In fact, many people
returnedto their original homesrather than moving into
the new settlement.
Rabaul, Papua New Guinea, 1983-1985
The
calderaat Rabaul (New Britain Island) is about the same
size(14 km by 9 km) as CampiFlegrei. The volcanohas
producedthree caldera-formingeruptionsbetween3500
and 1400 yearsago as well as severalsmall- to moderatesizedexplosiveeruptionsin historicaltime. Its mostrecent
eruptionswere in 1937-1943, during which about 500
peoplewerekilled [Johnson
and Threlfall,1985]. Today,
nearly105,000peoplecouldbe affectedby a catastrophic
eruption,andabout70,000 by a moderate-sized
explosive
eruption.
Following two magnitude8.0 regionalearthquakes
in
1971 in the nearbyNew BritainTrench,Rabaulbeganto
exhibitsignsof seismicityandgrounddeformationsimilar
to thoseprecedingpreviouseruptions[McKeeet al., 1985].
In August1971 the monthlycountof shallowearthquakes
doubledto about 100. During the next 10 years,several
seismicswarmsoccurred,eachsuccessively
more intense
(Figure 10). Leveling data and measurableuplift of
beachesduringthis sameperiod indicatedthat the central
partof the calderawaselevated1 m. In August1983 the
seismicity and rate of ground deformation increased
abruptly. The monthly count of shallow (<4 km)
earthquakes
in Augustwas about330, in Septemberwas
2135,andby early 1984exceeded10,000(Figure10). The
deformation
rate accelerated
correspondingly.Maximum
tilt ratesincreasedto between50 and 100 grad/month,and
256 o Tilling: VOLCANIC HAZARDSAND THEIR MITIGATION
27,2/REVIEWSOFGEOPHYSICS
14000
Number
of
earthquakes
per month
'Stage-2'
12000
10/29/83-
Alert
•
11/22/84
10000
8000
6000
4000
2000
1972
1974
1976
1978
1980
I 982
I 984
Figure 10. Seismicity beneath Rabaul caldera, Papua New
irregularlybut sharply,returningto 1982 levels by May 1985.
Guinea,from July 1971 throughlate 1985 (modifiedfrom Mori The hatched area indicates the period during which a
and McKee [1987, Figure 1]). The seismicityand ground government-imposed
"stage2" alertwasin effect(seetext).
deformationpeaked in April 1984 and then began to decline
the uplift rate in the area of maximum deformation
increasedto 5-10 cm/month. The RabaulVolcanological
Observatory(RVO) expandedits volcanomonitoringby
initiatingtilt andhorizontaldeformationmeasurements.
On October 29, 1983, after consideringthe technical
information
from
RVO
and socioeconomic factors,
governmentofficialsdeclareda "stage2" volcanicalert,
which implied that an eruptionwould occurwithin a few
months.The seismicityandrate of deformationcontinued
to increasethroughMay 1984andthendeclinedirregularly
for the remainderof the year. On November22, 1984,
officials downgradedthe volcanic alert to "stage 1,"
which implied that the "anticipatederuptionis not now
expectedto occurbefore severalmonthsto a few years"
(P. L. Lowenstein, in the work of SEAN [1984]).
However, at the end of 1984 the seismicityand ground
deformation rate were still 1-2 times pre-1983 rates
begunwhen public concernwas high remain unfinished,
and progressalso has stalled on initiatives to establish
alternative monitoring capability should RVO be incapacitatedor destroyedduring an eruption [Lowenstein
and Mori, 1987].
LongValley, California,19110-19114kongValley
caldera, 17 x 30 km in size,is locatedon theeasternfront
of the SierraNevada. It was formedduringthe cataclysmic eruptionof the BishopTuff about0.7 million years
ago;ashof the BishopTuff hasbeenidentifiedthroughout
much of the western United States, as far east as Nebraska
andTexas[Baileyet al., 1976;Izett, 1981]. The youngest
of the postcalderaeruptions from the Long Valley
magmaticsystemoccurred100,000 yearsago, and other
nearby magmatic sources(e.g., Inyo-Mono craters) fed
smallereruptionsas recently as 550 years ago [Miller,
1985; Achauer et al., 1986]. The resort town of Mammoth
Lakes(population4000) lies within the southwestern
part
In the period from 1985 to the present,activity at of the caldera,and during the height of the winter ski
Rabaul has remained at a low level, with periods of seasonits populationsnowballsto 25,000.
Beginning on May 25, 1980, as Mount St. Helens
standstilland localized deflation followed by a gradual
resumptionof inflationsinceMarch 1988. With the lifting producedan explosiveeruption much smaller than the
of the stage2 alert in November 1984 and the continuing catastrophic
eruptionof a week before, the Long Valley
relativecalm at Rabaul,emergencypreparedness
measures area was rockedby four magnitude6 earthquakeswithin
[McKee et al., 1984].
27,2/REVIEWSOFGEOPHYSICS
Tilling:VOLCANICHAZARDS
AND THEIRMITIGATIONo257
48 hours,accompanied
and followedby myriadsmaller- hazardnoticeat Long Valley, simplifiedthe three-tiered
magnitude
aftershocks
[Bailey,1982]. Not surprisingly,hazardsnotificationsysteminto a single"hazardwarnofficialsandpeoplein theLongValleyareawereanxious ing" procedure.The criteriafor a "geologichazard
and concerned. On May 27 the USGS issued an warning"in thenewsystem
are [FederalRegister,1984,
earthquake "hazard watch," the second level in the
p. 3938]
three-tierednotificationsystemthen in use,becausethe
a) a degreeof risk greaterthannormalfor the area;or a
unusualseismicitywasthoughtat thetimeto be of tectonic
hazardous
condition
thathasrecently
developed
or hasonly
been recentlyrecognized;and b) a threat that warrants
origin. The hazardwatchdesignation
was usedwhen
consideration
of a near-term
publicresponse.
information
indicated"a potentiallycatastrophic
eventof
generallypredictablemagnitudemay be imminentin a TheLongValley situation
in early1984failedto meetthe
general
areaor regionandwithinanindefinite
timeperiod newcriteria. Accordingly,
boththe noticeof potential
(possibly
monthsor years)"[FederalRegister,1977,p. earthquake
hazardand the noticeof potentialvolcanic
19,292].
hazardwere de facto discontinued;the volcanichazard
After May 1980 and continuinginto 1983, several noticewasformallyterminated
onJuly11, 1984[Goriand
seismicswarmsof varying durationand magnitude Shearer, 1987].
occurredsporadicallywithin the southernpart of the
Asof mid-1988,continued
slowinflationof thecaldera,
caldera.In late 1980,SavageandClark [1982]foundthat ata ratemuchlessthanthatduring
the1980-1982period,
the earthquakes
were accompanied
by uplift of the is still indicated
by grounddeformation
monitoring
data,1988). With thecurrentreduced
caldera's
resurgent
dome. Thisfinding,coupledwiththe (USGS,unpublished
possibility
thatmagmamay existat shallowdepth(7-8 levelof activity
thecrisisatLongValleyisrapidly
fading
km)beneath
thedome[Hill, 1976;RyallandRyall,1981], frompublic
memory.However,
theintensive
monitoring
suggested
thatrenewedinflationof the magmareservoir continues
because
the USGSstill considers
the Long
mightaccount,
at leastin part,for theincreased
seismicityValleyareato havepotential
for volcanicactivity. A
andmeasured
uplift. If so,thenthepotentialfor reactiva- response
planforvolcanic
hazards,
prepared
in early1984
tionof thevolcanic
system
wouldbeincreased.
Of special [USGS, 1984], remainsin effect.
concern was the possibility that a major tectonic
earthquake
in theregion[Hill et al., 1985]mighttriggeran Comparisonof the Disasters
eruption
fromtheLongValley'srestless
magmareservoir.
The four majorvolcanicdisasters
that havealready
On thebasisof continuing
caldera
unrestanda prelimi- occurred
in the 1980saffordillustrativeexamplesof the
naryvolcanichazardsassessment
[Miller et al., 1982] the currentstateof theart in volcanichazardsmitigationin a
USGS issued,on May 25, 1982, a "Notice of Potential globalcontext.In Table4 I haveattempted
to summarize
VolcanicHazard"in additionto the existingearthquakemy knowledge and perceptionsof the circumstances
hazards
watch,andsystematic
geophysical
andgeochemi- surrounding
therecentvolcanicdisasters;
othersmaynot
calmonitoring
studies
werebegunby theUSGSandother fully sharemy views.
organizations.
Thetypesof studies
usedto monitorLong
The response
to, andhandlingof, theMountSt. Helens
Valleyaredescribed
in detailby Hill [1984]. Theissuance disaster
appearto beadequate
by thecriterialistedin Table
of the volcanic hazards notice created considerable 4, in largemeasure
because
of the substantial
bodyof
consternation
amongthelocalofficialsandresidents
in the knowledge about the volcano that was available to
Mammoth
Lakesarea. In a comprehensive
studyof the scientists
andofficialsbeforetheeruption.A long-term
publicresponse
duringthecrisis,Maderetal. [1987,p. 38] forecast
thatMountSt. Helenswouldbe themostlikely
concluded,
"Initially people... reactedto the noticewith
volcanoin the conterminousUnited Statesto reawaken and
feelingsof fear, confusionand denial. Local officials and
erupt,"possibly
beforetheendof thiscentury,"[Crandell
businesses
didnotwelcome
theUSGSnotice.Communityet al., 1975, p. 441] was made in 1975. A detailed
concernfocusedon the adverseeconomic
impactsof the volcanic
hazards
assessment
followedin 3 years[Crandell
noticeratherthanonpublicsafety."Eventually,
however, andMullineaux,1978],delineating
themosthazardous
and
"peopleseemedto believetheywouldbe warnedbefore vulnerable
areasin theeventof renewed
eruptive
activity.
aneruption
andhavetimeto escape
thearea.Theyrelied Littlereadat thetimeof its publication,
thisstudylater
ontheUSGS,theagency
responsible
fortheobjectionablebecamethereportmostwidelyreadby localofficialsand
notice,to providethatwarning"[Maderet al., 1987,p. scientists
between
theonsetof activity(March27) andthe
39].
climacticeruptiononMay 18, 1980.
Following
a briefearthquake
swarmin January
1983the
TheCrandell-Mullineaux
reportwasdulysubmitted
to
seismicactivity at Long Valley largely returnedto the Department
of Emergency
Services
of Washington
pre-1980levels. In January1984theUSGS,perhaps
in State and other involvedagencies. Saarinenand Sell
partresponding
to the "publicoutcry" over the volcanic [1985,p. 4] succinctly
summarize
thepre-1980
response:
258'Tilling:
VOLCANIC HAZARDSAND THEIRMITIGATION
Mount St. Helens was the eighth notice of the new USGS
hazardwarningsystem. On December20, 1978, a letter was
sent to the WashingtonState governor'srepresentative
from
the USGS notifying federal, state, and local officials of the
potential hazard.
The governor's representative
misinterpretedthe notification, thinking an eruption was
imminent,and a specialmeetinginvolvingrepresentatives
of
many stateof Washingtongovernmentdepartments
and the
USGS officials was called in January,1979, to clarify the
situation. Althoughat the time this actionwas regardedan
overreaction,the meeting might, in retrospect,have been
usefulin alertingstateofficialsto thepotentialproblem.
Despitethe initial flurry of interestand concern,public
officials were not convinced of the need to prepare
long-rangecontingencyor land useplans.
Ironically,the USGS itself failed to respondinstitutionally to a long-termforecastmadeby its own scientists.
Becauseof budgetaryconstraintsand other (at the time)
higher-priority programmatic commitments, USGS
officials (myself included) made no decision to begin
baselinemonitoringstudiesat Mount St. Helens. Earlier
however,on his own initiative a USGS scientist(D. A.
Swanson)had proposedin the winter of 1971-1972 a
project titled "Geodimeter Studies of Cascade Volcanoes." In the summer of 1972, after consultationwith
D. R. Crandell and C. Hopson(Universityof California,
Santa Barbara) regardingwhich volcanoesshould have
first priority, Swansonestablished
five electronicdistance
measurement(EDM) lines at Mount St. Helens and one at
Mount Hood (Oregon). Accordingto Swanson(written
communication,1989), "clearly expediencyplayed some
role in decidingto do St. Helens first, but the overriding
reasonwas the perceivedhigh hazardand particularlythe
perceivedlikelihood that St. Helens would be the first
Cascadevolcanoto erupt." The measurements
were made
usingborrowedequipmentanda loanedfield assistant,
and
they constitutedthe only pre-1980 volcanomonitoring
27,2/REVIEWSOFGEOPHYSICS
conducted at Mount St. Helens.
Because of the lack of
funding and the subsequentunavailability of loaned
equipment
theprojectterminated
by 1973.
The intensivepredisaster
volcanomonitoringduringthe
phreaticeruptivephase(March 20 to May 17) failed to
provide diagnosticinformationfor a precise short-term
forecast
of theMay 18 magmaticeruption,eventhoughthe
data clearly indicatedthat the north flank of the volcano
was becomingdangerouslyunstable. In hindsight,some
scientists
speculatethat the rate of growthof the "bulge"
[e.g.,Lipmanet al., 1981, Figure 97] might haveincreased
prior to failure, if the deformationprocesshad not been
"short-circuited"by the magnitude5.1 earthquakethat
triggerederuption. In any case,the possibilityof a major
avalanche,and a large magmaticeruptionwiggeredby it,
wasrecognizedby the scientificteam at Mount St. Helens
and explainedto officials before May 1 [Miller et al.,
1981;Decker, 1986].
A detailed account of hazards assessments at Mount St.
Helens is given by Miller et al. [1981], who show the
importanceof updatesof potentialhazardsas conditions
changeand of daily meetingswith land use managers,in
this casethe U.S. ForestService(USFS), to brief them on
the hazards. The scientistsprovidedthe geotechnicaland
hazardsinformationusedin the preparationof the "Mount
St. HelensContingency
Plan," issuedby USFSon April 9,
as well as in the establishmentof restrictedzones by
WashingtonStateand USFS officials. Had thesezonesof
restrictedaccessnot been in effect on May 18, greater
humanlossandinjury surelywouldhaveresulted.
In contrastto Mount St. Helensthe E1 Chich6nexperiencerepresentsthe worstpossiblecasebecauseof the total
lackof predisaster
volcanichazardsstudiesandemergency
preparedness.
Before 1982, virtually nothingwasknown
about its past history of frequentand violent eruptions
[Tilling et al., 1984], and no volcano monitoringwas
conductedbefore or during the brief eruption. The
TABLE 4. Comparison
of FourVolcanicDisasters
Since1980in Termsof VolcanicHazardsMitigation
"DecisionMaking"
Government Bodies'
ScientificCommunity'
s Responsibility
Responsibility
Knowledge
Volcanic
Disaster
Mount St. Helens,
Long-Term
Precursors
Short-Term
Precursors
(Weeks
to
Months)
(Hoursto
Days
Predisaster
Monitoring
yes
none
much
good
good
?
none
none
no
probably(but
unrecognized)
nonereported
none
little or
none
none
fair
no
yes
?
limited
poor
poor
Predisaster
ofPast
Behavior
Hazards
Assessment
Long-Term
Forecast
Short-Term
Forecast
good
detailed
yes
no
virtually
none
fair*
none
no
no
rudimentary
no
fai•
preliminary
no
Predisaster
Contingency
Planning
Emergency
Management
U.S.A. (1980)
El Chieh6n,
Mexico (1982)
Galunggung,
Indonesia(1982)
Nevadodel Ruiz,
Colombia(1985)
Theratingscaleis qualitative,
fromnoneto excellent.Figure1
illustrates
the divisionof primaryresponsibility
betweenthescientific communityand"decisiÜn-making"
bodiesshownin thistable.
*Historicaleruptiverecordknownonly.
•Both historicalandprehistoric
eruptiverecordknown.
27,2/REVIEWSOF
GEOPHYSICS
Tilling: VOLCANIC HAZARDS AND THEIR MITIGATION'259
propheticobservations
of CanulandRocha[1981] about
thepossible
reawakening
of thevolcanowerecontained
in
an internalreport submittedin September1981 to the
GeothermalDepartmentof the Comisi6n Federal de
Electricidad(CFE), whose mission is to develop and
managethe energy resourcesof Mexico, not to worry
aboutpotentialvolcanichazards.
Fortunately,E1 Chich6n is located in a relatively
sparsely
populated
ruralarea;otherwise,
its 1982eruption
authoritiesin the United Statesonly have had to copewith,
at most, a few hundredevacueesin responseto volcanic
crisesin the pastcentury. A NationalCoordinationBody
of Natural Disasters (BAKORNAS in Indonesian) is
responsible
for thecoordination
of government
andprivate
efforts in hazards mitigation in Indonesia. The VolcanologicalSurvey of Indonesia is the governmentdesignatedorganizationresponsiblefor volcanichazards
studiesand directly interacts with BAKORNAS in the
would have causedconsiderablymore deathsand destruc- event of a volcanic disaster or crisis.
In contrastto Indonesiathe countryof Colombia,before
tion. Unfortunately,the presentstateof knowledgeabout
manyvolcanoes
in far moredenselypopulated
regionsis 1985, hadno nationalprogramfor volcanichazardsstudies
not much better than that available for E1 Chich6n before
nor a single institutiondesignatedto be responsiblefor
itsdeadlysurprise.Immediately
followingtheE1Chich6n suchstudies. In responseto the Rufz disasterthe Colomdisaster, there were some efforts made in Mexico to biangovernment,
with substantial
supportof the Office of
developa nationalvolcanichazardsprogram. To date, ForeignDisasterAssistance(OFDA) of the U.S. Agency
however,the Mexican govemmenthas neitherdeveloped for InternationalDevelopment (USAID), establishedthe
sucha programnordesignated
a principalinstitution
to be ObservatorioVolcano16gicoNacional at Manizales,under
responsiblefor volcanomonitoringand hazardsassess- the managementof the Instituto Nacional de InvesGeo16gico-Mineras
(INGEOMINAS). This new
ments. In the United States the USGS is the congres- figaciones
sionallymandated
organization
(by theDisasterReliefAct observatory,one of the bestequippedand staffedvolcano
in Latin America, is monitoringthe continuof 1974) responsible
for providingtimely warningsof observatories
volcanic and related hazards.
ing intermittentbut weak eruptiveactivity at Nevado del
The Galunggungand Nevado del Ru/z casesin some Rufz and initiating preliminary studiesof nearby potenvolcanoesas well.
respectsare intermediatebetweenthosefor Mount St. tially dangerous
Helens and E1 Chich6n (Table 4). Prior to the disasters,
someknowledgeexistedabouttheirrespective
historiesof
past eruptionsand associatedhazards. In addition, PROBLEMS AND CHALLENGES
volcanichazardszonationmaps were available for both
In recentdecades,significantadvanceshavebeenmade
volcanoes
prior to thedisasters
(Figure2) [Sudradjatand
Tilling, 1984] to guide authoritiesin the preparationof in understandingvolcanic phenomena;these advances
measurecontingency
or evacuationplansand in takingemergency reflect progressin geophysicalinstrumentation,
measures.
And because of the abundance of active
menttechniques,
observational
capabilities,datacollection
volcanoesin Indonesiathepeopleandthe officialsarevery and analysis (computerization),modeling studies, and
accustomed
to eruptiveactivity. Thuspeopleneededlittle theoreticalformulations. Certainly,our basicknowledge
promptingfrom officials to evacuatewhen Galunggung of eruptiveprocessesand productsis increasingrapidly
beganto eruptin April 1982 with virtuallyno warning. comparedwith a century ago. Few would disputethat
Luckily, the initial eruptionswere sufficientlyweak to volcanologyhas arrivedas a modemscience;yet for me,
permitsafeevacuation.As mentioned
earlier,the reasons naggingquestionspersistabout successfulapplicationof
why evacuationmeasuresfailed at Nevadodel Ru/z are the scienceto societalproblems. Have the advancesin
unclear and still under study. However, from what volcanologybeenfully translatedinto advancesin hazards
information is available it seems reasonable to conclude
mitigation? Are we winninga few battlesbut losingthe
do to reduce
that "the scientificteam performedremarkablywell and war? What more can or shouldgeoscientists
volcanic
risk?
providedtimely warnings
....
But despitetheir efforts,
time ran out beforeeff•five planscouldbe implemented
Problem
by the emergency-response
authorities"[Tilling, 1985,p. TheMost Pressing
230].
Despiteits limited scientificand economicresources,
Indonesia,which containsmore active volcanoesthan any
other country,has an effectivenaturalhazardsprogram
thatdeservesgreaterrecognition.During the pastdecade,
Indonesiahas successfullyevacuatedmore than 140,000
personsthreatenedby volcanicactivity in the successful
mitigation of volcanic hazards. In comparison,civil
The incidence of human fatalities from volcanic hazards
(Table 5) providesone, and perhapsthe only, reasonably
well documentedmeasurein trying to addressthe above
questions.Still, it mustbe rememberedthat the figures
givenin Table5 represent
an incomplete
sampling,largely
reflectinga few infrequentbut highly destructiveeventsin
only a few countries(seeTable 1). For example,the three
27, 2 / REVIEWS OF GEOPHYSICS
260 ßTilling: VOLCANIC HAZARDSAND THEIRMITIGATION
TABLE 5. Human Fatalities From Volcanic Activity, 1600-1986
Primary Causeof Fatalities
1600-1899
1900-1986
Pyroclasticflows and debris avalanches 18,200
(9.8%)
36,800
(48.4%)
Mudflows (lahars) and floods
8,300
(4.5%)
28,400
(37.4%)
Tephra falls and ballisticprojectiles
8,000
(4.3%)
3,000
(4.0%)
(23.4%)
(49.4%)
(0.5%)
400
3,200
100
1,900'
2,200
76,000
(0.5%)
(4.2%)
(0.1%)
(2.5%)
(2.5%)
(100%)
Tsunami
Disease,starvation, etc.
Lava flows
Gasesand acid rains
Other or unknown
Total
Fatalitiesper year(average)
43,600
92,100
900
......
15,100
186,200
(8.1%)
(100%)
620
880
Modified fromBlong [ 1984,Table 3.2]. Valuesin parentheses
refer to percentages
relative to total fatalitiesfor eachtime period.
*Includes the deathscausedby lethal gas burstsat two volcaniccrater lakes in
Cameroon,Africa: 37, Lake Monoun, August 1984 [Sigurdssonet al., 1987]; and
>1700, Lake Nyos, August 1986 [Kling et al., 1987]. While investigatorsagreethat
the lethalgas(carbondioxide)in boththesecasesis of magmaticorigin,theydisagree
on the causativemechanismof gasrelease.
"predict virtually any explosiveeruption" (D. A. Swanson, written communication, 1988). However, as disalone would account for about 75% of the total deaths
cussedlater, volcanologistsfirst must develop refined
from volcanicactivityin the twentiethcentury.
techniquesand criteria to distinguishunambiguously,
if
At face value the tabulation shows that the average possible,the diagnosticprecursorypatternsfor explosive
numberof fatalitiesper year for the 1900-1986 period andnonexplosiveeruptions.
2. Gains in hazards mitigation from scientific and
(880) is higherthanthat for the 1600-1899 period(620).
advances(almostexclusivelyconfinedto the
Table 5, however,doessuggestsignificantimprovementin technological
the twentiethcenturyin reducingthe incidenceof deaths developednations)are offset by the word population
causedby eruption-induced
starvationand tsunami. The explosionin the twentiethcentury(largelyin the developreductionin fatalitiescausedby posteruption
starvationis ing nations).In anygivenareaan eruptionnow, with no or
real and reflects the existenceof modem, rapid com- ineffectivehazardsmitigationmeasures,
wouldkill many
municationsand disaster relief delivery systems. The morepeoplethanthe sameeruptionin 1600 becauseof the
apparentreductionin volcano-relatedtsunamicasualties, increasedpopulation density with time. The world's
however, stems from the fortunate fact that no large humanpopulationwas about0.5 billion in 1600, became5
eruption-triggered
tsunamihas occurredin this century, billion in 1987, and is expectedto grow to about6 billion
even though a highly effective internationalwarning peopleby the year 2000; greatestgrowth is projectedto
systemnow existsfor tsunamigenerated
by distantevents. take place in the developing countries [Gwatkin and
No improvementis observedin the twentiethcenturyfor Brandel, 1982; Haub, 1988].
the numberof fatalitiesassociatedwith pyroclasticflows
3. Volcanic hazardsmitigation is inadequatein the
destructiveeruptionsin 1902 (Mont Pe16e,Soufri•re (St.
Vincent), and SantaMaria) and the 1985 eruptionof Rufz
and mudflows.
Given the advancesin volcanologyin the twentieth
century, the above noted observationsare somewhat
unexpected.Whetheror not thedifferencesin fatalitydata
(Table 5) are statisticallysignificant,they still reflectsome
or all of the followingfactors:
1. Virtually all the fatalitiesarecausedby volcanicand
relatedhazardsassociatedwith infrequentmajorexplosive
eruptions,for which reliable short-termforecastsare still
not possiblefor most volcanoes. This situationmainly
stems from insufficient funds and trained personnelto
conduct the needed volcano monitoring and hazards
assessment
studies.Somevolcanologists
areconfidentthat
given adequatemonitoringnetworksand prompt interpretationof the acquireddata, it shouldbe possibleto
denselypopulateddevelopingregions.While thedeveloping countriessuffer the most loss of life, the developed
countriesincurthe greatesteconomiclossfrom damageto
property[Wijkmanand Timberlake,1984]. For example,
more than 99% of the eruption-caused
deathssince 1900
occurred
in thedeveloping
countries
or regions(Table1).
4. Advancesin volcanomonitoring,eruptionforecasting, and hazardsassessment
have not been fully factored
into effectivevolcanicemergencyplanningand management. This failure in part derives from the fact the
scientificinformationgenerallyis still not specificenough
aboutthe time, place, and magnitudeof an anticipated
hazardouseventeruptionto compelofficialsto takeaction.
Even
from
the limited
information
available
it is
abundantlyclear that on a global basis,futureadvancesin
Till
ing: VOLCANIC HAZARDSAND THEIR MITIGATION'261
volcanichazardsmitigationshouldbe in the developing
countries,especiallythosein the circum-Pacific"Ring of
Fire" [Latter, 1987]. While thesecountriescontain most
of the world'shigh-riskvolcanoes,
theylack the scientific
and monetaryresourcesto undertakethe neededfundamentalgeoscience
andvolcanichazardsstudies.
Simply stated,the mostpressingproblemin reducing
volcanicrisk in a global contextis that the world's most
dangerous
volcanoes
are the leastunderstood
and studied.
This problemconfrontsboth developedand developing
countries,but it is especiallyacutefor the latter. For the
developedcountries,tacklingthe problemwould involve
shiftsin nationalprioritiesand the greaterallocationof
resourcesto volcanicrisk mitigationprograms. For the
affected developingcountriesthe ideal solutionto this
problemis to achieveself-sufficiency
in volcanologyand
related socioeconomic
capability in hazardsmitigation.
Such a solutioninvolvesa long-termprocessrequiting
decadesin any realisticscenario.In the interim,geoscientistsmustwork closelywith internationalorganizationsto
createworkable,stableinternationalprogramsof mutual
assistanceand rapid responseduring volcanic crises.
These programsmust also include preparationand/or
upgradingof hazardsassessments,
acquisitionof baseline
monitoring data, and training/educationfor scientists,
emergencymanagementofficials, and the public. Such
programscostrelativelylittle but mustbe stablyfundedfor
long-termcontinuityto be effective. For example,the
annualbudgetof a programproposedfollowinga seriesof
UNESCO-sponsored
workshopsin 1983 [Yokoyamaet al.,
1984] would be a small fraction of the daily helicopter
expensesincurredduringthe disasterrelief effortsat Rufz
in 1985, or (I. G. Gass, written communication,1986)
"about a thirdof what is paid for a goodsoccerplayer!"
Tilling and Newhall [1987, p. 61] concludethateffective
volcanic hazards mitigation "will not be achievedby
technologyalone,but also requiresovercomingnational
capabilityto dispatchon short notice a Volcanic Crisis
AssistanceTeam (VCAT) to a volcanictrouble spot, in
responseto a requestfrom the host country government
throughdiplomaticchannels.The World Organizationof
Volcano Observatories(WOVO), with funding from
UNESCO, has coordinatedvolcano-monitoring
assistance
at Apoyo Volcano (Nicaragua) and Tacanfi (MexicoGuatemala). Also, a programjoinfly fundedby OFDA/
USAID andUNDRO is now in progressto increasestudies
of high-riskvolcanoesin Ecuador. While theseisolated
smallprojectsare helpfulandillustratesomeaspectsof the
work needed,they do not begin to addressthe problemon
theglobalscale.
27, 2 / REVIEWS OF GEOPHYSICS
An International
Decade of Natural Hazard Reduction
(IDNHR) has been proposedas a global program for the
1990s [Housner, 1987; DPB, 1988b]. The IDNHR is a
componentof the proposed international Decade for
NaturalDisasterReduction,endorsedby a resolutionof the
United Nations passed in December 1987 and by a
resolutionof the U.S. Houseof Representatives
(H. Con.
Res. 290) introduced in May 1988 [Oaks, 1988;
Sigurdsson,
1988]. In May 1988 the U.S. Subcommittee
of the InternationalAssociationof Volcanologyand the
Chemistryof theEarth'sInterior(IAVCEI) for the IDNHR
presentedin its report [Heiken et al., 1988] to the U.S.
Geodynamics
Committeethe followingrecommendations,
in addition to continued fundamental research on volcanic
phenomena, pertinent to global volcanic hazards
mitigation:
1. Monitoring networks should be establishedon
volcanoesdeemedto be of "highest-risk" on the basisof
preliminarygeoscience,demographic,and socioeconomic
studies.
2. After the GlobalPositioningSystem(GPS) is fully
testedand shownto be routinelypreciseand accurate,it
canbe usedin placeof conventional,
morelabor-intensive,
geodeticmonitoring. To assurewide applicationof GPS
and international bureaucratic inertia to create needed
data for eruptionprediction,its costs,includingmaintenanceexpenses,must be reducedto be affordablefor use
long-termprograms."
To date,internationalor bilateralprogramshavebeenad in developingnations.
3. Presently, experimentalgeophysicalmethods of
hoc, ephemeral,and utterlyinadequate,consistingmainly
of "too little too late" responsesto volcanic disasters. eruptionprediction,includinggravity, geomagnetic,and
However,since1984, a few smallstepshavebeentakento geoelectrical
techniques,
mustbe improvedandcontinueto
initiate predisasterstudiesat a few high-riskvolcanoes, betestedon volcanoes
well monitored
by othertechniques.
4. The equipmentpresentlyusedfor volcanomonitormostlyin Latin America. With the supportof the OFDA/
USAID the USGS has providedtechnicalassistance
and ing must be simplified, miniaturized,and made more
consultationto the VolcanologicalSurveyof Indonesiain rugged. Lightweight,sturdy,and easyto usegeophysical
the period 1979-1988. In addition,the USGS, again in instruments
will simplifyvolcanomonitoringin developcooperationwith OFDA/USAID, has begun a modest ing countries,in field situations
remotefrom the luxuryof
programof rapidresponse
to volcaniccrises(or of baseline experienced
electronictechnicians
andspareparts.
5. "High-tech" volcano-monitoring
methodscan be
acquisition
andtrainingexercises
if nocrises
occur)called
the InteragencyVolcano Early Warning DisasterAssis- too expensivefor developingcountries. An emphasis
tanceProgram(VDAP) [Banks,1986]. A specialfeature should be made on research leading to "1ow-tech"
of thisprogram,whichis focusedon Latin America,is the monitoringtechniques,
which,thoughpossiblylessprecise
262 oTilling: VOLCANIC HAZARDSAND THEIRMITIGATION
27, 2 / REVIEWS OF GEOPHYSICS
technologicaladvancesalone,but ratherby wider application of existingtechnologyto poorly studiedvolcanoesin
denselypopulatedregions,thenthe majorchallengesto the
geoscience
communityare as follows:
1. Apart from their "regular" researchon volcanic
phenomenaand/or volcanichazardsstudies,scientistsin
bothdevelopedanddevelopingcountriesmustplay a much
moreactiveandvisiblepart in increasingpublicawareness
of volcanoesand their potentialhazards. Peterson[1986,
1988] discussesthe role and responsibilitiesof volChallenges
to the GeoscienceCommunity
Much remains to be learned in understandinghow canologistsin societyand believes [Peterson,1988, p.
have"an ethicalobligationto convey
volcanoeswork, especiallythose that characteristically 4161] volcanologists
produce explosive eruptionsand the most destructive effectivelytheirknowledgeto benefitall of society."
volcanic hazards. Even for the well-studied volcanoes in
2. For potentiallydangerousvolcanoesin regionsstill
the developednations,reliable short-termpredictionsof not heavily populatedor developed,geoscientistsmust
explosiveeruptionsare still not routinely possible. A redouble efforts to prepare the most detailed hazards
as availabledata permit. Then they mustbe
majorscientificchallengeconfronting
volcanologists
is the assessments
needto developa reliablemethod,usingpatternrecogni- willing to work closelywith decisionmakers,to encourage
tion in conjunctionwith newerapproaches,
to identifythe and persuade them to consider the volcanic hazard
zonationmapsin thedevelopmentof local or regionalland
diagnostic
precursors
of explosiveeruptions.
Anotherseriousproblemis the lack of reliablecriteria useplans.
3. For volcanicallyactiveareasthatalreadyaredensely
for distinguishingbetweenthe precursorypatternof an
eruptionand that of an intrusion. Magma intrusionsare populatedand have land use patternsfixed by demand,
often,but incorrectly,considered"false alarms,"but they culture,or tradition,the only availableoptionsin hazards
are better termed "aborted eruptions" [Walker, 1982]. mitigation are to develop improved monitoring and
Decker [1988] suggeststhat intrusionsmay "outnumber predictivecapabilitiesto enablescientiststo give timely
eruptionsby at least2 to 1." Currently,the besthopefor warningsto officials. In addition, the scientistsmust
recognizingthe differences,if any,betweenthe precursory interact closely with civil authoritiesto devise and test
plansbeforeany volcaniccrisisstrikes.
indicatorsof an eruptionversusthoseof an intrusionlies in contingency
the careful, continuousmonitoringof a volcanothrough
4. Geoscientists
must try harderto convincedecision
many eruptive,intrusive,and dormantcyclesto learn its makers that the conveningof more internationalworkfull range of characteristicpatterns. However, this is a shopsand symposiato furtherstudythe topicof volcanic
processof trial and error, in which the errors can be hazardsmitigationwould yield diminishingreturns.
minimizedbutprobablynevereliminated.
submitthat the scopeof the problemis sufficientlywell
The technologyto developa more reliable predictive definedto lake decisiveactionto immediatelyimplement
capability already exists, but the societalresolve and programsrepeatedlyproposed. Monies earmarkedfor
commitment to achieve the objective are lacking. more"talking about" the problemwouldbe betterspent
Nonetheless,we mustcontinueto refine existingmonitor- to make a hazardszonationmap or baselinemeasurements
ing techniques,
to devisenew ones,andto betterrecognize at a high-riskvolcanoin a developingcountry.
the onset and the nature of the premonitorysignalsof
5. Higherprioritymustbe placedon thepreparation
of
renewederuptionat explosivevolcanoes.In particular,we generalinterestpublications,movie f'llms, videotapes,
must improve the instrumentationand systems to trainingmanuals,and otheraudiovisualaidsin a concerted
recognize,and to transmitreal-timewarningsof, danger- programto educatethe emergencyresponseofficialsand
ous eventsand products(e.g., lahars)as they occur. An the generalpublic aboutthe typesand natureof volcanic
importantadjunct to hazardsmitigationresearchis the andassociated
hazardsand theirdestructive
potential. As
better characterization of unrest at dormant calderas and its
a small step in this direction, under the auspicesof
bearing on the probability of renewed caldera-forming IAVCEI, effortsare currentlyunderway to raisefundsto
eruptions.More thanhalf of the 200 Quaternarycalderas produce two 25-min video source tapes illustrating
consideredby Newhall and Dzurisin [1988] in a literature dramaticexamplesof volcanichazardsand their adverse
surveyapparentlyexhibitedsomeformof unrestwithinthe impact.Thesevideotapes
will be usedby volcanologists
to
past 100 years;suchunrestdoesnot necessarily
culminate assistin educatingthe decisionmakersand the affected
in an eruption,nor doesit providemuchinformationabout public about the natureof the hazardsand what can be
done about them.
the typeandsizeof theeruptionthatmightoccur.
If, as I contend,the most pressingproblemin interna6. Scientistsin the developedcountriesmustovercome
tional volcanic hazardsmitigation cannot be solved by theirreluctance
to becomeinvolvedin effortsto promote
and sensitive, can be applied locally, easily, and
inexpensively.
6. Data from monitoringequipmenthave limited use
and may even give false impressionsunless adequate
meansare availableto interpretthem on a scientificand
timelybasis.Properlyinterpreted
volcano-monitoring
data
mustbe presentedimmediatelyto the local civil defense
authorities
in a way readilyunderstood.
Tilling: VOLCANICHAZARDSAND THEIRMITIGATION o263
27, 2 / REVIEWS OF GEOPHYSICS
term by wider application of existing technology to
high-risk but poorly understoodvolcanoesin densely
populatedregions. The effective mitigationof volcanic
hazardson a globalscalereally boilsdown
stemsfrom the perceptionthat theseefforts,which andassociated
generallydo not lead to publicationsin refereedjournals, to a simple "have-have not" issue. The developing
are little appreciatedand rarely rewarded within the countrieshavemostof the world'shigh-riskvolcanoesbut
scientificcommunity. Anotherlarge obstacleis that too do not have the economic and scientific means to admany scientists
believethat their own work is too impor- quatelystudyandmonitorthem. Clearly,in the shortterm
tant to be interruptedfor suchnonresearch
activities,even the developed countriesneed to boost their technical
throughbilateralor internationalprogramsuntil
though for some, their researchis funded by hazards assistance
reductionprograms. Worse still, somescientists
are the the developingcountriesattain self-sufficiencyin volvolcanological
equivalentsof "ambulancechasers."They canology. The primary challengeto both the scientific
are eager to experience the scientific excitement in community and the "decision-making" bodies is to
observingan eruptionor studyingits impact and fresh prevent volcanic crises from turning into volcanic
eruptiveproductsbut unwilling to assumeany respon- disasters.
sibilityin assistingthe officialsand publicaffectedor to
sharetheir informationwith otherscientists
who are doing
ACKNOWLEDGMENTS. My views on volcanic hazards
thereal work in hazardsmitigation.
7. Last but not least, to gain full cooperationfrom mitigationon a globalbasishavebeenshapedby my experience
in directly studying,or in helping to coordinateinternational
decision makers and the populace, and to maintain
scientificresponses
to, eruptionsin Hawaii, at Mount St. Helens,
credibility in the eyes of the public, geoscientists
must E1Chich6n,Galunggung,Nevado del Rufz, and elsewhere.Over
better handlethe problemof "false alarms" or "crying theyearsI havelearnedmuchfromcolleagues
withinandoutside
wolf." As already noted, the occurrenceof precursory the U.S. Geological Survey who have been at the forefront of
activity often does not culminate in eruptions,thus volcanic hazards mitigation. I have been influenced by
compounding
this vexingproblem,which is not likely to stimulating interactions and discussionswith them during
and developthe neededinternationalprograms,to help
train scientistsfrom developingcountries,and to try to
educateofficialsand the public. This reluctancein part
emergencyas well as calm circumstances,
andnot all of them,I
am certain, would fully share the opinions expressedin this
betweenprecursoryeruptiveand intrusivebehaviorcan be paper. My deepthanksgo to WendellA. Duffield andWilliam
made routinelyand unambiguously,
Bankset al. [1989] E. Scott,who thoroughlyreviewedearlierdraftsof thepaperand
offered detailed constructivecriticisms and suggestionsthat
makethe followingrecommendation:
materially improved it. I also appreciatereceiving helpful
It seemsprudentto treat every occurrenceof unmistakable general comments and reactions to earlier versions of this
precursoryactivityashavingthepotentialfor eruptionandto manuscriptfrom the following colleagues: Roy A. Bailey,
adviseemergency-response
officials accordingly. With this Thomas J. Casadevall,Dwight R. Crandell, Robert W. Decker,
prudentapproach,"false alarms"(actuallyabortederuptions) Richard S. Fiske, L. J. Patrick Muffler, Donal R. Mullineaux,
will be unavoidable. A current,and much needed,partial ChristopherG. Newhall, Donald W. Peterson,and Donald A.
solution to the false-alarm problem is to educate the Swanson.In addition,I thankjournaleditorDavid W. Schollfor
governmentofficials and generalpublic aboutthe probabil- his helpful editorial suggestionsand efficient handling of the
istic natureof forecastsandpredictionsand of the limitations manuscript.However,any errorsof commissionor omission,or
inherentin the scientificinformationupon which they are of misplacedemphasis,restsolelywith me.
Finally, I wish to dedicatethis paperto DonaldB. McIntyre,
based. However, an equally seriouschallenge...
is to
minimize false alarms through more reliable pattern an inspiring teacherwho introducedme to the wondersof
recognition
....
If societywishesto maximizeeffective geologywhen I was a studentat PomonaCollege,California,
response
to warningsof volcanichazards,it mustbe prepared many, many years ago.
to accept the unavoidable false alarms. False alarms
themselvescan provide,throughobjectiveassessment
of the
scientificandpublic responseto a volcaniccrisisthat ended
REFERENCES
without eruption, valuable lessonsuseful in making or
improvingcontingencyplansfor the next crisis,which could Abetsten, L., Diversion of a lava flow from its natural bed to an
culminatein an eruption.
artificialchannelwith the aid of explosives:Etna, 1983,Bull.
be solved within the foreseeable future.
Until distinction
Volcanol., 47, 1165-1177, 1984.
In summary, the geosciencecommunityitself must Achauer,U., L. Greene,J. R. Evans,and H. M. Iyer, Nature of
the magma chamber underlying the Mono Craters area,
becomemore aware of the urgentneedto developstable
eastern California, as determined from teleseismic travel time
long-terminternationalprogramsneededto addressthe
residuals,
J. Geophys.Res.,91, 13,873-13,891, 1986.
global problem of volcanic hazards mitigation on a Agnew, D.C., Strainmetersand tiltmeters,Rev. Geophys.,24,
579-624, 1986.
systematic
ratherthanan ad hocbasis. While refinements
Aki, K., M. Fehler, and S. Das, Source mechanism of volcanic
in methodologyand new technologieswill be neededto
tremor: Fluid-drivencrackmodelsandtheir applicationto the
improve hazards assessmentsand eruption forecasts,
1963 Kilauea eruption, J. Volcanol. Geotherm. Res., 2,
259-287, 1977.
significantgainsare morelikely to be obtainedin the near
264'Tilling:
VOLCANIC HAZARDS AND THEIR MITIGATION
27,2/REVIEWSOF
GEOPHYSICS
Christiansen,
R. L., Yellowstonemagmaticevolution:Its bearing
on understandinglarge-volume explosive volcanism, in
ExplosiveVolcanism: Inception, Evolution, and Hazards,
Martinque,Philos. Trans. R. Soc. London,Ser. A, 200,
353-553• 1903.
edited by F. R. Boyd, Jr., pp. 84-95, G•physics Study
Committee,National ResearchCouncil, National Academy
Apple, R. A., T. A. Jaggat,Jr., and the Hawaiian Volcano
Press,Washington,D.C., 1984.
Observatory,
Volcanismin Hawaii, editedby R. W. Decker,
R. L., and D. W. Peterson,Chronologyof the 1980
T. L. Wright, andP. H. Stauffer,U.S. Geol.Surv.Prof. Pap., Christiansen,
1350, 1619-1644, 1987.
eruptiveactivity,The 1980 Eruptionsof Mount St. Helens,
Washington,editedby P. W. Lipmanand D. R. Mullineaux,
Bacon, C. R., Eruptivehistoryof Mount Mazama and Crater
U.S. Geol.Surv.Prof. Pap., 1250, 17-30, 1981.
Lake caldera,CascadeRange,U.S.A., J. Volcanol.Geotherm.
Res., 18, 57-118, 1983.
Civetta, L., P. Gasparini,G. Luongo, and A. Rapolla (Eds.),
PhysicalVolcanology,
Dev. in SolidEarth Geophys.Ser.,vol.
Bailey,R. A., MammothLakesearthquakes
andgrounduplift:
6, 333 pp., ElsevierScientific,New York, 1974.
Precursors
to possiblevolcanicactivity?,in Yearbook,
Fiscal
Pacific, Lava flow controlto protectthe
Year 1982, pp. 4-13, U.S. GeologicalSurvey,Reston,Va., Commander-in-Chief,
1982.
city of Hilo, Hawaii, CINCPAC CONPI.AN 5107, 10 pp.,
United StatesPacific Command, Honolulu, Hawaii, 1978.
Bailey,R. A., G. B. Dalrymple,andM. A. Lanphere,Volcanism,
on Volcanology,
MayonVolcano,pp. 6-9, Manila,
structure,
and geochronology
of Long Valley caldera,Mono Commission
Philippines,1975.
Coumy,California,J. Geophys.
Res.,81,725-744, 1976.
Banks,N. G., Hazardmitigationandrapidresponse
to criseswith Corrado, G., I. Guerra, A. Lo Basico, G. Luongo, and R.
mobilevolcanoobservatories
(abstract),Eos Trans.AGU, 76,
Rampoldi, Inflation and microearthquakeactivity of
398, 1986.
Phlegraean Fields, Italy, Bull. Volcanol., 40, 169-188,
1976/1977.
Banks,N. G., R. I. Tilling, D. H. Harlow,andJ. Ewert,Volcano
monitoringand short-termforecasts,in Short Course in Crandell, D. R., and D. R. Mullineaux, Volcanic hazards at
Mount Rainier,U.S. Geol.Surv.Bull., 1238, 26 pp., 1967.
Geology,vol. 1, VolcanicHazards,editedby R. I. Tilling, pp.
Crandell, D. R., and D. R. Mullineaux, Potential hazards from
51-80, AGU, Washington,D.C., 1989.
Barberi,F., andL. Villari (Eds.),Mt. Ema andIts 1983 Eruption,
futureeruptionsof Mount St. HelensVolcano,Washington,
Bull. Volcanol., 47, 877-1177, 1984.
U.S. Geol.Surv.Bull., 1383-C, 26 pp., 1978.
Barberi, F., D. Hill, F. Innocenti,G. Luongo, and M. Treuil Crandell, D. R., D. R. Mullineaux, and M. Rubin, Mount St.
Helens volcano: Recent and future behavior, Science, 187,
(Eds.), The 1982-1984 BradyseismicCrisis at Phlegraean
438-441, 1975.
Fields(Italy), Bull. Volcanol.,47, 173-412, 1984a.
Barberi, F., G. Corrado,R. Innocenti,and G. Luongo,Phlegraean Crandell, D. R., B. Booth, K. Kusumadinata, D. Shimozuru, G.
Fields 1982-1984: Brief chronicleof a volcanoemergencyin
P. L. Walker, and D. Westercamp, Source Book for
Volcanic-HazardZonation,97 pp., UNESCO, Paris,1984.
a denselypopulated
area,Bull Volcanol.,
47, 175-185,1984b.
Berrino, G., C. Corrado, G. Luongo, and B. Toro, Ground Davis, P.M., D. R. Pierce, R. L. McPherron, D. Dzurisin, T.
deformationand gravity changesaccompanyingthe 1982
Murray,M. J. S. Johnston,
andR. Mueller, A volcanomagneticobservation
on Mount St. Helens,Washington,
Geophys.
Pozzuoliuplift,Bull. Volcanol.,47, 187-200, 1984.
Res. Lett., 11,233-236, 1984.
Bevens,D., T. J. Takahashi,andT. L. Wright (Eds.),The Early
volcaniceruptions,Annu.Rev.Earth
Serial Publicationsof the Hawaiian VolcanoObservatory, Decker,R. W., Forecasting
Planet, Sci., 14, 267-291, 1986.
vol. 1,565 pp., Hawaii NaturalHistoryAssociation,
Hawaii
National Park, 1988.
Decker, R. W., Reducingvolcanicrisk, Geol. Soc. Am. Abstr.
Programs,20, A81, 1988.
Blong,R. J., Someeffectsof tephrafallson buildings,in Tephra
in
studies,NATO Adv. StudyInst. Ser., editedby S. Self and R. Decker, R. W., and W. T. Kinoshita,Geodeticmeasurements,
The Surveillanceand Predictionof VolcanicActivity,pp.
S. J. Sparks,pp.405-420, D. Reidel,Hingham,Mass.,1981.
47-74, UNESCO, Paris, 1972.
Blong,R. J., VolcanicHazards: A Sourcebook
on theEffectsof
Decker, R. W., T. L. Wright, and P. H. Stauffer (Eds.), VolEruptions,424 pp., Academic,SanDiego,Calif., 1984.
canism in Hawaii, U.S. Geol. Surv. Prof. Pap., 1350, 1667
Bullard, F. M., Volcanoesof the Earth, 579 pp., University of
Texas Press, Austin, 1976.
pp., 1987.
Canul, R. F., and V. S. Rocha, Informe geo16gicode la zona de Neve, G. A., Effects in the atmosphere
of the Galunggung
eruption,in LetusanGalunggung1983-1983, editedby J. A.
geot6rmicade "El Chichonal," Chiapas,report, 38 pp.,
Katili, A. Sudradjat,and K. Kusumadinata,pp. 367-372,
Geotherm. Dep. of the Com. Fed. de Electr., Morelia,
VolcanologicalSurveyof Indonesia,Bandung,1986.
Michoacfin,Mexico, Sept. 1981.
Carlson, R. W., Geochemical evolution of the crust and mantle, DeNevi, D., Earthquakes,230 pp., Celestial Arts, Millbrae,
Calif., 1977.
Rev.Geophys.,25, 1011-1020, 1987.
Anderson,T., and J. S. Flett, Report on the eruptionsof the
S0ufri•rein St,..Vincent,andona visitto Montagne
Pe16e
in
Casadevall, T. J., J. B. Stokes, L. P. Greenland, L. L. Malin-
conico,J. R. Casadevall,
andB. T. Furukawa,
SO2 andCO2
emission rates at Kilauea Volcano, 1979-1984, Volcanism in
Hawaii, edited by R. W. Decker, T. L. Wright, and P. H.
Stauffer,U.S. Geol.Surv.Prof. Pap., 1350, 771-780, 1987.
Chadwick, W. W., Jr., R. J. Archuleta, and D. A. Swanson,The
mechanics of ground deformation precursory to dome-
Disaster Prevention Bureau, Volcanic Disaster Countermeasures
in Japan, 23 pp., National Land Agency,Prime Minister's
Office, Tokyo, 1988a.
Disaster Prevention Bureau, Natural Disasters and International
Cooperation: The Abstract of White Paper on Disaster
Countermeasures
in 1988 Toward the IDNDR, 32 pp.,
National Land Agency, Prime Minister's Office, Tokyo,
1988b.
building extrusionsat Mount St. Helens 1981-1982, J.
Geophys.Res.,93, 4351-4366, 1988.
Djumarma,A., S. Bronto,I. Bahar,F. X. Suparban,
R. Sukhyar,
Chirkov, A.M., Radon as a possiblecriterionfor predicting
C. G. Newhall,R. T. Holcomb,N. G. Banks,T. Torley,J.P.
eruptionat KarymskyVolcano,Bull. Volcanol.,39, 125-131,
Lockwood,R. I. Tilling, M. Rubin,andM. A. del Marmol,
1975.
Did MerapiVolcano(centralJava)eruptcatastrophically
in
1006 A.D.? (abstract), in Abstracts, International_VolChouet, B., R. Y. Koyanagi, and K. Aki, Origin of volcanic
canological Congress, Auckland-Hamilton-Rotorua,New
tremor in Hawaii, II, Theory and discussion,Volcanismin
Hawaii, edited by R. W. Decker, T. L. Wright, and P. H.
Zealand,p. 236, Royal Societyof New Zealand,Auckland,N.
Z., 1986.
Stauffer,U.S. Geol.Surv.Prof. Pap., 1350, 1259-1280, 1987.
27,2/REVIEWSOF
GEOPHYSICS
Tilling: VOLCANIC HAZARDS AND THEIR MITIGATION o265
Duffield, W. A., R. I. Tilling, and R. Canul, Geology of E1
Chich6n Volcano, Chiapas,Mexico, J. Volcanol. Geotherm.
Res., 20, 117-132, 1984.
Dzurisin, D., and K. M. Yarnashita,Vertical surface displacements at Yellowstone caldera, Wyoming, 1976-1986, J.
populationgrowth in the Third World, Sci. Am., 246, 57-64,
1982.
Harris, D. M., M. Sato, T. J. Casadevall, W. I. Rose, Jr., and T. J.
Bornhorst,
Emission
ratesof CO2 fromplumemeasurements,
The 1980 eruptionsof Mount St. Helens,Washington,edited
by P. W. Lipman and D. R. Mullineaux, U.S. Geol. Surv.
Geophys.Res.,92, 13,753-13,766,1987.
Dzurisin, D., L. A. Anderson,G. P. Eaton,R. Y. Koyanagi,P.
Prof. Pap., 1250, 201-207, 1981.
W. Lipman, J.P. Lockwood,R. T. Okamura,G. S. Puniwai, Haub,C., A population
nightmare,SanJoseMercuryNews,pp.
M. K. Sako,and K. M. Yamashita,Geophysicalobservations
1C and4C, SanJose,Calif., July31, 1988.
of Kilauea Volcano, Hawaii, 2, Constraintson the magma Havskov,J., S. De la Cruz-Reyna,S. Singh,P. Medina, and C.
supplyduringNovember1975-September1977,J. Volcanol.
Guti6rrez,Seismicactivityrelatedto the March-April 1982
Geotherm. Res., 7, 241-269, 1980.
eruptions
of E1Chich6nVolcano,Chiapas,Mexico,Geophys.
Res. Lett., 10, 293-296, 1983.
Eaton, G. P., R. L. Christiansen,H. M. Iyer, A. N. Pitt, H. R.
Blank, Jr., I. Zietz, D. R. Mabey, andM. E. Gettings,Magma Hays, W. W., and C. F. Shearer,Suggestions
for improving
beneath Yellowstone National Park, Science, 188, 787-796,
decisionmakingto face geologic and hydrologichazards,
1975.
Facing Geologic and Hydrologic Hazards: Earth-Science
Endo, E. T., S. D. Malone, L. L. Noson, and C. S. Weaver,
Considerations,
editedby W. W. Hays,U.S.Geol.Surv.Prof.
Locations,magnitudes,and statisticsof the March 20-May
Pap., 1240-B, B103-B108, 1981.
18 earthquakesequence,The 1980 Eruptionsof Mount St. Heiken,G., C. G. Newhall,D. Nielson,P. W. Lipman,P. Sheets,
Helens, Washington,edited by P. W. Lipman and D. R.
T. Simkin, and R. I. Tilling, Volcanic hazardsMullineaux, U.S. Geol. Surv.Prof. Pap., 1250, 93-107, 1981.
recommendationsfor research during the International
Federal Register, Warnings and preparednessfor geologicDecadefor NaturalHazardsReduction,report,20 pp., U.S.
related hazards: Proposedprocedures,
42, 19,292-19,296,
Geodyn.Comm.,Natl. Res.Counc.,Washington,D.C., 1988.
1977.
Herd, D. G., and the Comit6 de EstudiosVulcano16gicos,
The
Federal Register,Revision of terminologyfor geologichazard
1985 Rufz Volcano disaster,Eos Trans. AGU, 67, 457-460,
1986.
warning,49, 3938, 1984.
Feuillard, M., C. J. Allegre, G. Brandeis,R. Gaulon, J. L. Le Hill, D. P., Structureof Long Valley caldera,California, from a
Mouel, J. C. Mercier, J.P. Pozzo, and M.P. Semet, The
seismicrefractionexperiment,J. Geophys.Res.,81, 745-753,
1976.
1975-1977 crisisof la Soufri•re de Guadeloupe(F.W.I.): A
still-bornmagmaticeruption,J. Volcanol.Geotherm.Res.,16, Hill, D. P., Monitoringunrestin a largesiliciccaldera,theLong
317-334, 1983.
Valley-Inyo Craters Volcanic Complex in east-central
California, Bull. Volcanol., 47, 371-396, 1984.
Fisher,R. V., and H.-U. Schmincke,PyroclasticRocks,472 pp.,
Hill, D. P., R. E. Wallace, and R. S. Cockerham, Review of
Springer-Verlag,
New York, 1984.
Fisher, R. V., A. L. Smith, and M. J. Roobol, Destruction of St.
evidenceon the potential for major earthquakesand volPierre, Marfinqueby ash cloud surges,May 8 and 20, 1902,
canismin the Long Valley-Mono Carters-White Mountains
Geology,8, 472-476, 1980.
region of easternCalifornia, EarthquakePredict. Res., 3,
571-594, 1985.
Fournierd'Albe, E. M., Objectivesof volcanicmonitoringand
Hoblitt, R. P., C. D. Miller, and W. E. Scott, Volcanic hazards
prediction,J. Geol. Soc.London,136, 321-326, 1979.
Francis,P. W., S. L. de Silva, andD. A. Rothery,Satelliteremote
with regard to siting nuclear-powerplants in the Pacific
sensingappliedto volcanoes(abstract),in AbstractsVolume,
Northwest,U.S.Geol.Surv.OpenFile Rep.,87-297, 196pp.,
1987.
KagoshimaInternational Conferenceon Volcanoes,p. 50,
OrganizingCommittee,KagoshimaInternationalConferences Hofmann, D. J., Perturbationsto the global atmosphere
on Volcanoes,Kagoshima,Japan,1988.
associated
with the E1 Chich6nvolcaniceruptionof 1982,
Fremont, M.-J., and S. D. Malone, High-precisionrelative
Rev. Geophys.,25, 743-759, 1987.
locationsof earthquakes
at Mount St. Helens,Washington,J. Housner, G. W. (Ed.), Confronting Natural Disasters: An
Geophys.Res.,92, 10,233-10,236, 1987.
International
Decadefor NationalHazardReduction,
60 pp.,
Friedman, I., and M. Reimer, Helium at Kilauea Volcano, I,
U.S. NationalAcademyof Sciences,
Washington,D.C., 1987.
Spatial and temporalvariationsat SulphurBank, Volcanism Ishihara,K., Predictionof summiteruptionby tilt andstraindata
in Hawaii, edited by R. W. Decker, T. L. Wright, and P. H.
at SakurajimaVolcano,Japan(abstract),in AbstractsVolume,
Stauffer,U.S. Geol.Surv.Prof. Pap., 1350, 809-813, 1987.
Kagoshirna
InternationalConferenceon Volcanoes,p. 44,
Galindo, I., D. J. Holmann, and M.P. McCormick (Eds.),
OrganizingCommittee,KagoshimaInternationalConferences
Proceedings
of the Symposiumon Atmosphericeffectsof the
on Volcanoes,Kagoshima,Japan,1988.
1982 Eruptions of E1 Chich6n Volcano, XVIII General Iyer, H. M., Deep structureunder YellowstoneNational Park,
Assembly of the International Union of Geodesy and
U.S.A.: A continental"hot spot," Tectonophysics,
56,
165-197, 1978.
Geophysics(IUGG), August 1983, Hamburg, Germany,
Geofis.lnt., 23, 113-304, 1984.
Izett, G. A., Volcanicashbeds: Recordersof upperCenozoic
Ghiorso,M. S., Thermodynamics
of mineralsand melts, Rev.
silicic pyroclasticvolcanismin the westernUnited States,J.
Geophys.,25, 1054-1064, 1987.
Geophys.Res.,86, 10,200-10,222, 1981.
Gori, P. L., and C. Shearer,A chronologyof hazardswamings: Jachens,R. C., D. R. Spydell,G. S. Pitts,D. Dzurisin,andC. W.
1976-1986, Proceedings
of Conference
XL: A Workshopon
Roberts,Temporal gravity variationsat Mount St. Helens,
the U.S. Geological Survey's Role in Hazards Warnings,
March-May 1980, The 1980 Eruptionsof Mount St. Helens,
editedby P. L. Gori andW. W. Hays, U.S. Geol. Surv. Open
Washington,editedby P. W. Lipman and D. R. Mullineaux,
File Rep.,87-269, 76-85, 1987.
U.S. Geol.Surv.Prof. Pap., 1250, 175-189, 1981.
Gorshkov,G. S., Giganticeruptionof the volcanoBezymianny, Jackson,D. B., Geoelectricobservations
(includingthe SepBull. Volcanol., 20, 76-109, 1959.
tember1982 summiteruption),The Pu'u 'O'o Eruptionof
Greenland,L. P., Hawaiianeruptivegases,Volcanismin Hawaii,
KilaueaVolcano,Hawaii, Episodes1-20, January3, 1983
editedby R. W. Decker,T. L. Wright, andP. H. Stauffer,U.S.
ThroughJune8, 1984, U.S. Geol. Surv. Prof. Pap., 1463,
Geol. Surv.Prof. Pap., 1350, 759-770, 1987.
Gwatkin, D. R., and S. K. Brandel, Life expectancyand
237-251, 1988.
Jackson,D. B., and J. Kauahikaua,regional self-potential
266'Tilling:
VOLCANIC HAZARDSAND THEIR MITIGATION
27,2/REVIEWSOFGEOPHYSICS
anomalies at Kilauea Volcano, Volcanism in Hawaii, edited
Valley caldera,Science,232, 213-216, 1986.
by R. W. Decker,T. L. Wright andP. H. Stauffer,U.S. Geol. Lipman, P. W., and D. M. Mullineaux (Eds.), The 1980
Surv.Prof. Pap., 1350, 947-959, 1987.
Eruptionsof Mount St. Helens,Washington,U.S. Geol. Surv.
Jackson,D. B., J. Kauahikaua,and C. J. Zablocki,Resistivity
Prof. Pap., 1250, 844 pp., 1981.
monitoringof an active volcanousing the controlled-source Lipman,P. W., J. G. Moore,andD. A. Swanson,
Bulgingof the
electromagnetictechnique: Kilauea, Hawaii, J. Geophys.
north flank beforethe May 18 eruption--Geodeticdata,The
Res., 90, 12,545-12,555, 1985.
1980 Eruptionsof Mount St. Helens,Washington,editedby
Jaggat, T. A., My ExperimentsWith Volcanoes, 198 pp.,
P. W. LipmanandD. M. Mullineaux,U.S. Geol.Surv.Prof.
Hawaiian Volcano ResearchAssociation,Honolulu, 1956.
Pap., 1250, 143-155, 1981.
JapanSabo Association,Sabo Worksfor Active Volcanoesin Liter, L., G. Luongo,and R. Scandone,
On the volcanological
Japan,44 pp.,Tokyo, 1988.
evolutionof CampiFlegrei,Eos Trans.AGU, 68, 226-234,
Johnson, R. W., and N. A. Threlfall, Volcano Town:
The
1987.
1937-43 Eruptionsat Rabaul, 151 pp., Robert Brown &
Lockwood,J.P., Lava diversion--techniques
and strategy
Associates,Bathurst, Australia, 1985.
(abstract),in Abstracts Volume, KagoshimaInternational
Katili, J. A., A. Sudradjat,and K. Kusumadinata
(Eds.),Letusan
Conferenceon Volcanoes,p. 492, OrganizingCommittee,
Kagoshima Intemational Conferences on Volcanoes,
Galunggung1982-1983 (sevenpapersin English,21 papers
in Indonesian),693 pp., VolcanologicalSurveyof Indonesia,
Kagoshima,
Japan,1988.
Bandung,1986.
Lockwood,J.P., andR. Romano,Diversionof lava duringthe
Kieffer, H. H., D. Frank, and J. D. Friedman, Thermal infrared
1983 eruptionof Mount Etna, Earthquakelnf. Bull., 17,
124-133, 1985.
surveys at Mount St. Helens--Observationsprior to the
eruptionof May 18, The 1980Eruptionsof MountSt. Helens, Lockwood,J.P., andF. A. Torgerson,
Diversionof lavaflowsby
aerialbombing--Lessonsfrom Mauna Loa Volcano,Hawaii,
Washington,editedby P. W. Lipman and D. R. Mullineaux,
Bull. Volcanol., 43, 727-741, 1980.
U.S. Geol.Surv.Prof. Pap., 1250, 257-277, 1981.
Kinoshita, W. T., D. A. Swanson, and D. B. Jackson, The
Lockwood,J.P., N. G. Banks,T. T. English,L. P. Greenland,D.
measurement of crustal deformation
related to volcanic
B. Jackson,D. J. Johnson,
R. Y. Koyanagi,K. A. McGee,A.
activityat KilaueaVolcano,Hawaii, in PhysicalVolcanalogy,
T. Okamura,andJ. M. Rhodes,The 1984 eruptionof Mauna
Loa Volcano, Hawaii, Eos Trans. AGU, 66, 169-171, 1985.
editedby L. Civettaet al., pp. 87-115, ElsevierScientific,
New York, 1974.
Lockwood,J.P., J. J. Dvorak,T. T. English,R. Y. Koyanagi,A.
Klein, F. W., HypocenterlocationprogramHYPOINVERSE,
T. Okamura,M. L. Summers,andW. R. Tanigawa,Mauna
U.S. Geol.Surv.OpenFile Rep., 78-694, 113 pp., 1978.
LOa1974-1984: A decadeof intrusiveandextrusiveactivity,
Klein, F. W., R. Y. Koyanagi,I. S. Nakata,andW. R. Tanigawa,
Volcanismin Hawaii,editedby R. W. Decker,T. L. Wright,
The seismicityof Kilauea's magma system,Volcanism in
and P. H. Stauffer, U.S. Geol. Surv. Prof. Pap., 1350,
537-570, 1987a.
Hawaii, edited by R. W. Decker, T. L. Wright, and P. H.
Stauffer,U.S. Geol.Surv.Prof. Pap., 1350, 1019-1185, 1987. Lockwood, J.P., H. J. Moore, E. Robinson, and S. Wu, Lava
Kling, G. W., M. A. Clark, H. R. Compton,J. D. Devine,W. C.
diversionstructures
to protectthe Mauna Loa Observatory,
Evans,A.M. Humphrey,E. J. Koenigsberg,
J.P. Lockwood,
Hawaii (abstract),83rd CordilleranSectionMeeting,Geol.
M. L. Tuttle, and G. N. Wagner,The 1968 Lake Nyos gas
Soc.Am. Abstr.Programs,19, 399, 1987b.
disaster in Cameroon, west Africa, Science, 236, 169-175,
1987.
Koyanagi,R. Y., B. Chouet, and K. Aki, Origin of volcanic
tremor in Hawaii, I, Data from the Hawaiian Volcano
Longhi,J. (Ed.), Volcanology,Geochemistry,
and Petrology,
U.S. NationalReportto the International
Unionof Geodesy
and Geophysics1983-1986, Rev. Geophys.,25, 1007-1130,
1987.
Observatory1969-1985, Volcanismin Hawaii, editedby R. Lowe, D. R., S. N. Williams, H. H. Leigh, C. B. Connor,J. B.
W. Decker,T. L. Wright, and P. H. Stauffer,U.S. Geol. Surv.
Gemell,and R. E. Stoiber,Laharsinitiatedby the November
Prof. Pap., 1350, 1221-1257, 1987.
13, 1895 eruptionof Nevado del Rufz, Colombia,Nature,
324, 51-53, 1986.
Krueger,A. J., Sightingof E1Chich6nsulfurdioxidecloudswith
the Nimbus 7 total ozone mapping spectrometer,Science, Lowenstein, P. L., and J. Mori, The Rabaul seismo-deformation
220, 1377-1379, 1983.
Lacroix, A., The eruptionof Vesuviusin April, 1906, Smithson.
Inst.Annu.Rep.,1906, 223-248, 1906.
Langbein,J. O., M. F. Linker, A. McGarr, and L. E. Slater,
crisisof 1983-1985: Monitoring,emergency
planningand
interactionwith the authorities,the media, and the public
(abstract),
Abstracts
of theXIX GeneralAssembly,
IUGG, vol.
2, p. 429, IntemationalUnionof GeodesyandGeophysics,
Observations
of strain accumulation across the San Andreas
Washington,D.C., 1987.
Fault near Palmdale, California, using a two-color Lowenstein, P. L., and B. Talai, Volcanoes and volcanic hazards
geodimeter,
Science,218, 1217-1219, 1982.
in Papua New Guinea, ChishitsuChosashoHokoku, 263,
Latter, J. H., Volcanoes and volcanic risk in the circum-Pacific,
in Proceedingsof the Pacific Rim Congress,pp. 745-752,
AustralasianInstitute of Mining and Metallurgy, Parkville,
Australia, 1987.
315-331, 1984.
Luhr, J. F., and J. C. Varekamp(Eds.), E1 Chich6nVolcano,
Chiapas,Mexico, J. Volcanol. Geotherrn.Res., 23, 1-191,
1984.
Leet, R. C., Saturatedand subcooledhydrothermalboiling in Macdonald,G. A., Memorial to Thomas AugustusJaggat,
VolcanoLett. Hawaiian VolcanoRes.Assoc.,519, 1-4, 1953.
groundwaterflow channelsas a sourceof harmonictremor,J.
Geophys.Res.,93, 4835-4859, 1988.
Macdonald,G. A., The 1959 and 1960 eruptionsof Kilauea
Volcano, Hawaii, and the constructionof walls to restrict the
Le Guem, F., Magmaticgasmonitoring,in ForecastingVolcanic
Events,editedby H. Tazieff andJ. C. Sabroux,pp. 293-310,
spreadof the lava flows,Bull. Volcanal.,24, 249-294, 1962.
Elsevier Scientific, New York, 1983.
Macdonald,
G. A., Volcanoes,
510 pp.,Prentice-Hall,
Englewood
Cliffs, N.J., 1972.
Le Guern,F., H. Tazieff, and R. Faivre-Pierret,An exampleof
health hazards, people killed by gas during a phreatic Macdonald,G. A., Hazardsfrom volcanoes,in Geological
eruption:Dieng Plateau(Java,Indonesia),February19, 1979,
Hazards,editedby A. A. Bolt et al., pp. 63-131, SpringerBull. Volcanol., 45, 153-156, 1982.
Verlag,New York, 1975.
Linker, M. F., J. O. Langbein, and A. McGarr, Decreasein Mader, G. G., M. L. Blair, and R. A. Olson,Living with a
deformationrate observedby two-colorlaserrangingin Long
volcanicthreat: Response
to volcanichazards,LongValley,
27, 2 / REVIEWS OF GEOPHYSICS
Tilling: VOLCANIC HAZARDSAND THEIR MITIGATION o267
Calfiomia, report, 105 pp., William Spangleand Assoc.,
the consequences
of volcanicoutbursts,Bull. Volcanol.,Ser.
PortolaValley, Calif., 1987.
2, 23, 181-192, 1960.
Malingreau,J.-P.,Remotesensinganddisaster
monitoring:A Newhall, C. G., A methodfor estimatingintermediateand
review of applications
in Indonesia,paperpresentedat 18th
long-term
risksfromvolcanicactivity,withanexamplefrom
International
Symposium
on RemoteSensingof the EnvironMountSt. Helens,Washington,
U.S.Geol.Surv.OpenFile
ment, Environ. Res. Inst. of Mich., Paris,Oct. 1-5, 1984.
Manson, C. J., C. H. Messick, and G. M. Sinnott, Mount St.
Helens--A bibliography
of geoscience
literature,1882-1986,
U.S.Geol.Surv.OpenFile Rep.,87-292,205pp.,1987.
Markhinin, E. K., A. N. Sh'in, K. M. Timerbaeva,and P. I.
Rep.,82-396,59 pp., 1982.
Newhall,C. G., Semiquantitative
assessment
of changing
volcanicrisk at Mount St. Helens,Washington,
U•. Geol.
Surv.OpenFile Report84-272,30 pp., 1984a.
Newhall,C. G., Short-termforecasting
of volcanichazards,
Tokarev, Experienceof volcanic-geographic
zoning of
Proceedings
of the Geologicand HydrologicHazards
Kamchatkaand Commander
Islands,Bull. 32, pp. 52-70,
TrainingProgram,
U.S.Geol.Surv.OpenFile Rep.,84-760,
507-592, 1984b.
Volcanol.Sta.,Petropavlousk,
Kamchatskiy,
USSR,1962.
Marsh,B., Magmaticprocesses,
Rev.Geophys.,
25, 1043-1053, Newhall,C. G., and D. Dzurisin,Historical
unrestat large
1987.
Quaternarycalderasof the world, U.S. Geol. Surv. Bull.,
Matahelumual,
J., Gunungapi
danBahayanya
di Indonesia,
Bull.
1855, 1108pp., 1988.
105,65 pp.,Volcanol.Surv.of Indones.,
Bandung,1982.
Oaks, S. D., Decade for Natural Disaster Reduction: An
McCallum,I. S., Petrology
of theigneous
rocks,Rev.Geophys.,
opportunityfor internationalscientific cooperation,Eos
25, 1021-1042, 1987.
McGee, K. A., A. J. Sutton, and M. Sato, Use of satellite
telemetry
for monitoring
activevolcanoes,
with a casestudy
of a gas-emission
eventat KilaUeaVolcano,December1982,
Trans. AGU, 69, 753-757, 1988.
Office of the United Nations DisasterRelief Co-Ordinator/
UNESCO,Volcanic
Emergency
Management,
86 pp.,United
Nations, New York, 1985.
Volcanism
in Hawaii,editedby R. W. Decker,T. L. Wright, Pelton,J. R., andR. B. Smith,Contemporary
verticalsurface
and P. H. Stauffer,U.S. Geol. Surv. Prof. Pap., 1350,
displacements
in Yellowstone
NationalPark,J. Geophys.
821-825, 1987.
Res., 87, 2745-2761, 1982.
McKee, C. O., P. L. Lowenstein,P. de St. Ours, B. Talai, I,
Peterson,
D. W., Volcanoes:Tectonicsettingandimpacton
Itikarai,andJ. J. Mori, Seismicandgrounddeformation
crises
society,
in ActiveTectonics,
pp.231-246,Geophysics
Study
at Rabaulcaldera: preludeto an eruption?,
Bull. Volcanol.,
Committee,
NationalResearch
Council,NationalAcademy
47, 397-411, 1984.
Press,Washington,D.C., 1986.
McKee,C. O., R. W. Johnson,
P. L. Lowenstein,
S.J.Riley,R. J. Peterson,
D. W., Volcanichazardsand publicresponse,
J.
Blong,P. de St. Ours,andB. Talai, RabaulCaldera,Papua
Geophys.Res.,93, 4161-4170, 1988.
New Guinea: Volcanichazards,surveillance,
anderuption Podesta, B., and R. S. Olson, Science and the state in Latin
contingencyplanning, J. Volcanol. Geotherm. Res., 23,
195-237, 1985.
Meyer, W., M. A. Sobol,andR. L. Schuster,Landslidedammed
lakes at Mount St. Helens, Washington, Landslide
Dams--Processes,
Risk, and Mitigation,editedby R. L.
Schuster,
Geotech.Spec.Publ. 3, pp. 21-41, Am. Soc.Civ.
Eng., New York, 1986.
America--Decision
makingin uncertainty,
in Managing
Disasters,
Strategies
andPolicyPerspectives,
editedby L. K.
Comfort,pp.296-311, DukeUniversityPress,Durham,N.C.,
1988.
Prescott,
W. H., andJ. Svarc,Precision
of GlobalPositioning
Systemmeasurements
for crustaldeformationstudies: Initial
results,Proc.4th Int. Geod.Symp.Satell.Positioning,
2,
993-1002, 1986.
Miller, C. D., Holoceneeruptions
of the Inyo volcanicchain,
California--Implications
for possibleeruptionsin Long Rampino,M. R., and S. Self, The atmospheric
effectsof E1
Valleycaldera,Geology,33, 14-17, 1985.
Chich6n,Sci. Am., 253, 48-57, 1984.
Miller, C. D., D. R. Mullineaux, and D. R. Crandell,Hazards Ryall, A., and F. Ryall, Attenuationof P and S wavesin a
assessments
at Mount St. Helens,The 1980 Eruptionsof
magma
chamber
in LongValleycaldera,
California,
Geophys.
MountSt. Helens,Washington,
editedby P. W. Lipmanand
Res.Lett., 8, 557-560, 1981.
D. M. Mullineaux,U.S. Geol. Surv. Prof. Pap., 1250, Ryan, M.P., The mechanicsand three-dimensionalinternal
789-802, 1981.
Miller, C. D., D. R. Mullineaux, D. R. Crandell, and R. A.
Bailey,Potential
hazards
fromfuturevolcanic
eruptions
in the
Long Valley-Mono Lake area, east-centralCalifornia and
southwestNevada--A Preliminaryassessment,
U.S. Geol.
Surv.Circ., 877, 10 pp., 1982.
structureof active magmaticsystems: Kilauea Volcano,
Hawaii,J. Geophys.
Res.,93, 4213-4248, 1988.
Rymer,H., andG. C. Brown,Causes
of microgravity
change
at
Po•isVolcano,CostaRica: An activebut non-erupting
system,Bull. Volcanol.,49, 389-398, 1987.
Saarinen,
T. F., andJ. L. Sell,WarningandResponse
to the
Mitchell,M., E1 Chich6n: Weather-maker
of the century?, MountSt.Helen'sEruption,
240pp.,StateUniversity
of New
Weatherwise,35, 252-259, 1982.
York Press,Albany,1985.
Mori, J., andC. McKee,Outward-dipping
ring-faultstructure
at Sager,J.W., andD. R. Chambers,
Designandconstruction
of the
Rabaulcalderaas shownby earthquake
locations,Science,
SpiritLakeoutlettunnel,MountSt. Helens,Washington,
235, 193-195, 1987.
Moxham, R. M., Thermal surveillanceof volcanoes,in The
Surveillance
and Predictionof VolcanicActivity,pp.
103-124, UNESCO, Paris, 1972.
Miillerried,
F. K. G., DerChich6n,
einbisher
unbekannter
t•itiger
Vulkan im nordlichenChiapas,Mexiko, Z. Vulkanol.,14,
191-209, 1932.
Naranjo,J. L., H. Sigurdsson,
S. N. Carey,andW. G. Fritz,
Eruption of Nevado del Rufz volcano, Colombia, 13
November 1985: Tephra fall and lahars,Science,233,
961-963, 1986.
Landslide
Dams--Processes,
Risk,andMitigation,
edited
by
R. L. Schuster,
Geotech.
Spec.Publ.3, pp.42-58, Am. Soc.
Civ. Eng.,New York, 1986.
SatoM., andK. A. McGee,Continuous
monitoring
of hydrogen
onthesouthflankof MountSt.Helens,
The 1980Eruptions
of MountSt.Helens,
Washington,
editedby P. W. Lipman
andD. M. Mullineaux,
U.S.Geol.Surv.Prof.Pap.,1250,
209-219, 1981.
Savage,
J. C.,andM. M. Clark,Magmatic
resurgence
in Long
Valley caldera, California: Possible cause of the 1980
MammothLakesearthquakes,
Science,
217,531-533, 1982.
Neumann
VanPadang,
M., Measures
takenby theauthorities
of Scandone,
R., and S. D. Malone,Magmasupply,magma
theVulcanological
Surveyto safeguard
thepopulation
from
discharge
andreadjustment
of the feedingsystem
during
268 o Tilling: VOLCANICHAZARDSAND THEIRMITIGATION
1980, J. Volcanol. Geotherm.Res., 23, 239-262, 1985.
Scarone, H., Volcanic ash clouds: A continuing threat to
internationalaviation, UNDRO News, pp. 17-21, Office of
the United Nations Disaster Relief Co-Ordinator, Geneva,
Nov./Dec. 1985.
Schutz,B. E., Satellitepositioning,
Rev.Geophys.,
25, 883-887,
1987.
27,2/REVIEWSOFGEOPHYSICS
Evolution,and Hazards, pp. 110-121, GeophysicsStudy
Committee,National ResearchCouncil, National Academy
Press,Washington,D.C., 1984.
Simkin, T., L. Siebert,L. McClelland, D. Bridge, C. Newhall,
and J. H. Latter, Volcanoesof the WorM: A Regional
Directory, Gazetteer,and Chronologyof VolcanismDuring
the Last 10,000 Years, 232 pp., Hutchinson Ross,
Stroudsburg,
Pa., 1981.
Slater, L. E., and G. R. Huggett, A multiwavelengthdistancemeasuring instrument for geophysical experiments, J.
Geophys.Res.,81, 6299-6306, 1976.
Scientific Event Alert Network, Rabaul caldera,Bull. 9, p. 9,
SmithsonianInst.,Washington,D.C., 1984.
ScientificEvent Alert Network, CampiFlegrei,Bull. 10, pp. 3-6,
Smithsonian
Inst.,Washington,D.C., 1985.
Smith, R. B., and L. W. Braile, Crustal structureand evolution of
ScientificEventAlert Network,CampiFlegrei,Bull. 11, pp. 7-8,
an explosivesilicic volcanicsystemat YellowstoneNational
Smithsonian
Inst.,Washington,D.C., 1986.
Scott,W. E., Volcanic-hazardzonationand long-termforecasts,
Park, in ExplosiveVolcanism: Inception, Evolution, and
Hazards,pp. 96-109, GeophysicsStudyCommittee,National
in ShortCoursein Geology,vol. 1, VolcanicHazards,edited
ResearchCouncil,NationalAcademyPress,Washington,D.
by R. I. Tilling, pp. 25-49, AGU, Washington,
D.C., 1989.
C., 1984.
Searle,E. J., Volcanicrisk in the AucklandMetropolitanDistrict,
Stateof Hawaii, EmergencyPlan to ProtectHilo From a Lava
N. Z. J. Geol. Geophys.,7, 94-100, 1964.
Self, S., and P. W. Francis,Volcanology,U.S. NationalReport
Flow, document,6 pp. and7 annexes,Civ. Def. Div., Dep. of
Def., Honolulu, 1977.
to the International Union of Geodesy and Geophysics
1983-1986, edited by J. Longhi, Rev. Geophys.,25, Stoiber, R. E., L. L. Malinconico, Jr., and S. N. Williams, Use of
1065-1078, 1987.
the correlationspectrometerat volcanoes,in Forecasting
Self, S., M. R. Rmapino,M. S. Newton, and J. A. Wolff,
VolcanicEvents,editedby H. Tazieff andJ.-C. Sabroux,pp.
425-444, Elsevier Scientific, New York, 1983.
Volcanological
studyof the greatTamboraeruptionof 1815,
Stommel,H., and E. Stommel,VolcanoWeather: The Storyof
Geology,12, 659-663, 1984.
the Year Without a Summer, 177 pp., Seven Seas Press,
Sheets,P. D., and D. K. Grayson(Eds.), VolcanicActivity and
Newport,R. I., 1983.
HumanEcology,644 pp., Academic,SanDiego,Calif., 1979.
Shepherd,E. S., Temperature
of the fluid lava of Halemaumau, Stothers,R. B., The great Tamboraeruptionin 1815 and its
aftermath, Science, 224, 1191-1198, 1984.
July, 1911, specialreport, pp. 47-51, Hawaiian Volcano
Observatory, Hawaii Volcanoes National Park, 1912. Sudradjat,A., and R. I. Tilling, Volcanic hazardsin Indonesia:
The 1982-83 eruptionof Galunggung,Episodes,7, 13-19,
(Reprintedin The Early Serial Publicationsof the Hawaiian
1984.
VolcanoObservatory,editedby D. Bevens,T. J. Takahashi,
and T. L. Wright, pp. 51-55, Hawaii Natural History Sumaryono,A., and K. Kondo, Developmentprogrammeof
Association,Hawaii National Park, 1988.)
volcanic mud-flow forecasting and warning system in
Indonesia,in Proceedingsof the InternationalSymposium
on
Shi Diguang, Tangshanearthquakeand its damageto civil
Erosion,DebrisFlow andDisasterPrevention,pp. 481-486,
engineering,in Proceedingsof Expert Working Group
ToshindoPrinters,Tokyo, 1985.
Meeting cum Workshopon the Urban Geologyof Coastal
Areas (First Draft), pp. 113-117, Economic and Social Swanson,D. A., T. J. Casaderail, D. Dzurisin, S. D. Malone, C.
Commission for Asia and the Pacific, United Nations,
G. Newhall, and C. S. Weaver,Predictingeruptionsat Mount
St. Helens,June 1980 throughDecember1982, Science,221,
Bangkok,Thailand,1987.
1369-1376, 1983.
Shimozuru,D., A seismologicalapproachto the predictionof
volcanic eruptions,in The Surveillanceand Prediction of Swanson,D. A., T. J. Casaderail, D. Dzurisin, R. T. Holcomb,
C. G. Newhall, S. D. Malone, and C. S. Weaver, Forecasts
VolcanicActivity,pp. 19-45, UNESCO,Paris,1972.
and predictionsof eruptive activity at Mount St. Helens,
Shimozuru, D., list of dangerousvolcanoes and associated
USA: 1975-1984, J. Geodyn.,3, 397-423, 1985.
information,Bull. Volcanol.,38, suppl.,81 pp., 1975.
Shimozuru, D., Volcano surveillance in Japan: An intricate Swanson,D. A., D. Dzurisin, R. T. Holcomb, E. Y. Iwasubo, W.
W. Chadwick, Jr., T. J. Casaderail, J. W. Ewert, and C. C.
cooperative
program,Episodes,1, 23-27, 1981.
Heliker, Growth of the lava dome at Mount St. Helens,
Sigurdsson,H., Volcanology, in Encyclopediaof Physical
Washington,(USA), 1981-1983, The Emplacementof Silicic
Scienceand Technology,vol. 14, pp. 406-435, Academic,
SanDiego,Calif., 1987.
Domesand Lava Flows, editedby J. H. Fink, Spec.Pap.
Geol. Soc. Am., 212, 1-16, 1987.
Sigurdsson,
H., Volcanichazardsmitigation: A componentof
the U.N. International Decade of Natural Disaster Reduction,
Eos Trans. AGU, 69, 771 and 776, 1988.
Sigurdsson,H., S. Carey, W. Cornell, and T. Pescatore,The
eruption of Vesuvius in A.D. 79, Nat. Geogr. Res., 1,
332-387, 1985.
Sigurdsson,
H., J. D. Devine,F. M. Tchoua,T. S. Presser,M. K.
W. Pringle, and W. C. Evans,Origin of the lethal gasburst
from Lake Monoun, Cameroun, J. Volcanol. Geotherm. Res.,
31, 1-16, 1987.
Simkin, T., and R. S. Fiske, Krakatau 1883: The Volcanic
Symons,G. J. (Ed.), The Eruptionof Krakatoa,andSubsequent
Phenomena,
Reportof the KrakatoaCommitteeof the Royal
Societyof London,494 pp., TrtibnerandCo., London,1888.
Taylor, G. A.M., The 1951 eruptionof Mount Lamington,
Papua,Bull. Bur. Miner. Resour.Geol. Geophys.Aust.,38,
117 pp., 1958.
Thorarinsson,S., The j6kulhlaup from the Katla area in 1955
compared with other j6kulhlaups in Iceland, Joekull, 7,
21-25, 1957.
Thorarinsson,
S., On the damagecausedby volcaniceruptions
with special reference to tephra and gases, in Volcanic
Eruption and Its Effects, 464 pp., SmithsonianInstitution
Press,Washington,D.C., 1983.
ActivityandHuman ecology,editedby P. D. SheetsandD. K.
Simkin,T., andR. S. Fiske,Krakatau1883: A classicgeophysiGrayson,pp. 125-159, Academic,SanDiego,Calif., 1979.
cal event in History of Geophysics,
1, vol. editedby C. S. Tilling, R. I., A lessonfrom the 1985 Ruiz volcanic disaster
(guesteditorial),Episodes,8, 230, 1985.
Gillmot, pp. 46-48, AGU, Washington,D.C., 1984.
Simkin, T., and L. Siebert, Explosive eruptionsin space and Tilling, R. I., Major volcaniccrisesin the 1980's (abstract),Eos
Trans. AGU, 67, 396, 1986.
time: Duration, intervals, and a comparisonof the world's
active volcanic belts, in ExplosiveVolcanism: Inception, Tilling, R. I., Volcanology,in Encyclopediaof Science and
27,2/REVIEWSOFGEOPHYSICS
Tilling: VOLCANIC HAZARDS AND THEIR MITIGATION o269
Technology,vol. 19, pp. 275-281, McGraw-Hill, New York,
1987a.
Volcanological
Societyof Japan,Bull. Volcanol.,16, suppl.32,
1978.
Tilling, R. I., Introductionto specialsectionon how volcanoes Waldron,H. H., Debrisflow anderosioncontrolproblemsby the
asheruptionsof Irazd Volcano, Costa Rica, U.S. Geol. Surv.
work: Part1, J. Geophys.Res.,92, 13,685-13,686, 1987b.
Bull., 1241-1,37 pp., 1967.
Tilling, R. I., Introductionto specialsectionon how volcanoes
work: Part2, J. Geophys.Res.,93, 4159, 1988a.
Walker, G. P. L., Volcanichazards,ISR Interdiscip.Sci. Rev., 7,
148-157, 1982.
Tilling, R. I. Introductionto specialsectionon how volcanoes
Westphal,J. A., M. A. Cart, W. R. Miller, and D. Dzurisin,
work: Part3, J. Geophys.Res.,93, 14757-14,758, 1988b.
Tilling, R. I., Volcano prediction: Lessonsfrom materials
Expendable
bubbletiltmeterfor geophysical
monitoring,
Rev.
science,Nature, 332, 108-109, 1988c.
Tilling, R. I., Introductionand overview, in Short Course in
Geology,vol. 1, VolcanicHazards,editedby R. I. Tilling, pp.
1-8, AGU, Washington,D.C., 1989.
Tilling, R. I., andR. A. Bailey,VolcanoHazardsProgramin the
United States,J. Geodyn.,3, 425-446, 1985.
Tilling, R. I., and C. G. Newhall, Volcanology,Geotimes,32,
Sci. lnstrum., 54, 415-418, 1983.
White, G. F., and J. E. Haas,Assessment
of Researchon Natural
Hazards,487 pp., MIT Press,Cambridge,Mass.,1975.
Wijkman,A., andL. Timberlake,NaturalDisasters:Actsof God
or Acts of Man?, 145 pp., Earthscan,Washington,D.C.,
1984.
Williams, H., and A. R. McBirney, Volcanology,397 pp.,
59-61, 1987.
Freeman,Cooper,SanFrancisco,Calif., 1979.
Tilling, R. I., M. Rubin,H. Sigurdsson,
S. Carey,W. A. Duffield, Williams,R. S., Jr., andJ. G. Moore,Man AgainstVolcano: The
and W. I. Rose, Holoceneeruptive activity of E1 Chich6n
Eruptionon Heimaey,Vestmannaeyjar,
Iceland,Gen. Interest
Volcano,Chiapas,Mexico,Science,224, 747-749, 1984.
Publ. Ser., 28 pp., U.S. GeologicalSurvey,Reston,Va., 1983.
Tokarev,P. I., Predictionandcharacteristics
of the 1975 eruption Williams, S. N., Soil Rn and Hgø distributionand relation to
of Tolbachik Volcano, Kamchatka, Bull. Volcanol., 41,
magmatic resurgenceat Long Valley caldera, California,
251-258, 1978.
Tomblin, J., Ruiz Volcano, Colombia: Two mission reports,
Office of the U.N. Disaster Relief Co-Ord., Geneva, 1985.
Tomblin, J., UNDRO's role in respondingto volcanic emergenciesin UNDRO News, pp. 7-10, Office of the United
Nations Disaster Relief Co-Ordinator,Geneva, March/April
1988.
Torgerson,F. A., and J. C. Bevins, Bomb crateringtests in
Hawaiian lava, Oper. Anal. Tech. Note 76-2, 74 pp.,
Headquarters
Pac.Air Forces,Honolulu,Hawaii, 1976
Tryggvason,E., Multiple magma reservoirsin a rift zone
volcano: Ground deformationand magmatransportduring
the September1984 eruptionof Krafla, Iceland,J. Volcanol.
Geotherm. Res., 28, 1-44, 1986.
Science, 229, 551-553, 1985.
Williams, S. N., and H. Meyer, A model of Nevado del Rufz
Volcano,Colombia,Eos Trans. AGU, 69, 1554-1556, 1988.
Williams,S. N., Y. Sano,andH. Wakita,Helium-3emissionby
Nevadodel Rufz Volcano,Colombia,Geophys.Res.Letters,
14, 1035-1038, 1987.
Wyatt, F., R. Bilham, J. Beavan,A. G. Sylvester,T. Owne, A.
Harvey, C. Macdonald,D. D. Jackson,and D.C. Agnew,
Comparingtiltmetersfor crustaldeformationmeasurement-A
preliminaryreport,Geophys.Res.Lett., 11,963-966, 1984.
Yokoyama,I., R. I. Tilling, and R. Scarpa,Internationalmobile
early-warningsystem(s)for volcanic eruptionsand related
seismic activities, Rep. FP/2106-82-01(2286), 102 pp.,
UNESCO, Paris, 1984.
UNESCO, The Surveillanceand Predictionof VolcanicActivity:
A Reviewof Methodsand Techniques,166 pp., Paris,1972.
United States Army Corps of Engineers,Lava flow control,
Island of Hawaii: Review report and environmentalimpact
statement,116 pp. and5 appendices,
Honolulu,Hawaii, 1980.
U.S. Geological Survey, USGS responseplan for volcanic
hazards in the Long Valley caldera-Mono Craters Area,
Califomia, U.S. Geol. Surv. Open File Rep., 84-500, 12 pp.,
Yukutake,T., T. Yoshino, H. Utada, H. Watanabe,Y. Hamano,
Y. Sasai, and T. Shimomura,Changesn the electrical
resistivityof the central cone, Miharayama,of Izu-Oshima
Volcano, associatedwith its eruptionin November, 1986,
Proc. Jpn.Acad.,Ser. B., 63, 55-58, 1987.
Zablocki,C. J., Mappingthermalanomalieson an activevolcano
by the self-potentialmethod, Kilauea, Hawaii, Proc. U31.
Syrup.Dev. UseGeotherm.Resour.2nd, 3, 1299-1309, 1975.
1984.
Zablocki, C. J., Applicationof the VLF inductionmethodfor
Verbeek,R. D. M., Krakatau (in Dutch),495 pp., Imprimeriede
studying some volcanic processesof Kilauea Volcano,
l'Etat, Batavia, Indonesia, 1885.
Hawaii, J. Volcanol. Geotherm.Res., 3, 155-195, 1978.
Voight, B., A method for predictionof volcanic eruptions, Zlotnicki, J., Magneticmeasurements
on La Soufri[re Volcano,
Nature, 332, 125-130, 1988a.
Guadeloupe(LesserAntilles), 1976-1984: A re-examination
Voight, B., Countdownto catastrophe,
Earth Miner. Sci., 57,
of the volcanomagnetic
effectsobservedduringthe volcanic
17-30, 1988b.
crisis of 1976-1977, J. Volcanol. Geotherm. Res., 30,
83-116, 1986.
Voight, B., H. Glicken, R. J. Janda, and P.M. Douglass,
Catastrophicrockslide avalancheof May 18, The 1980 Zollinger,H., Besteigung
des VulkanesTamboraauf der Insel
Eruptionsof Mount St. Helens,Washington,editedby P. W.
Sumbawaund Schilderungder Erupziondesselbenim Jahr
Lipman and D. M. Mullineaux, U•. Geol. Surv. Prof. Pap.,
1815, pp. 1-20, Winterthur,Wurster,Switzerland,1855.
1250, 347-377, 1981.
Volcanic SaboTechnical Centre,Debris Flow Forecastingand
WarningSystemat Mt. Merapi, 6 pp., Yogyakarta,Indonesia,
1986.
R. I. Tilling, U.S. Geological Survey, MS 910, 345 Middlefield Road, Menlo Park, CA 94025.