Volcano Hazards in the San Salvador Region, El Salvador · By J.J. Major, S.P. Schilling, D.J. Sofield, C.D. Escobar, and C.R. Pullinger U.S. GEOLOGICAL SURVEY Open-File Report 01-366

VVVVVolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvador 11111

Open-File Report 01–366Open-File Report 01–366Open-File Report 01–366Open-File Report 01–366Open-File Report 01–366

U.S. Department of the InteriorU.S. Department of the InteriorU.S. Department of the InteriorU.S. Department of the InteriorU.S. Department of the InteriorU.S. Geological SurveyU.S. Geological SurveyU.S. Geological SurveyU.S. Geological SurveyU.S. Geological Survey

VVVVVolcano Hazards in the San Salvadorolcano Hazards in the San Salvadorolcano Hazards in the San Salvadorolcano Hazards in the San Salvadorolcano Hazards in the San SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El Salvador

22222 VVVVVolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvador

Cover photographCover photographCover photographCover photographCover photograph

San Salvador volcano viewed from the southeast. The broad edifice of El Boquerón is on the left,and steep-sided El Picacho is on the right. San Salvador city is in the foreground. (Photograph byJ.J. Major, U.S. Geological Survey).


Volcano Hazards in the San SalvadorVolcano Hazards in the San SalvadorVolcano Hazards in the San SalvadorVolcano Hazards in the San SalvadorVolcano Hazards in the San SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El SalvadorByByByByBy J.J. Major, S.P. Schilling, D.J. Sofield, C.D. Escobar, and C.R. Pullinger J.J. Major, S.P. Schilling, D.J. Sofield, C.D. Escobar, and C.R. Pullinger J.J. Major, S.P. Schilling, D.J. Sofield, C.D. Escobar, and C.R. Pullinger J.J. Major, S.P. Schilling, D.J. Sofield, C.D. Escobar, and C.R. Pullinger J.J. Major, S.P. Schilling, D.J. Sofield, C.D. Escobar, and C.R. Pullinger

U.S. GEOLOGICAL SURVEYU.S. GEOLOGICAL SURVEYU.S. GEOLOGICAL SURVEYU.S. GEOLOGICAL SURVEYU.S. GEOLOGICAL SURVEYOpen-File Report 01-366Open-File Report 01-366Open-File Report 01-366Open-File Report 01-366Open-File Report 01-366

Vancouver, Washington U.S.A.2001


U.S. DEPARTMENT OF THE INTERIORGale Norton, Secretary

U.S. GEOLOGICAL SURVEYCharles G. Groat, Director

This report is preliminary and has not been reviewed for conformity with U.S. Geological Survey editorial standards. Any use oftrade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

:morf desahcrup eb nac troper siht fo seipoC:ot etirw noitamrofni lanoitidda roF

yevruS lacigoloeG .S.UegrahC-ni-tsitneicSsecivreS noitamrofnIyevruS lacigoloeG .S.U

68252 xoB .O.P01 .gdlB ,truoC lanidraC ES 003152208 OC ,revneD38689 AW ,revuocnaV

0124-202 )303(0098-399 )063(FAX: (360) 993-8980


CONTENTSCONTENTSCONTENTSCONTENTSCONTENTS

Introduction..................................................................................................................................... 1

Volcanic Phenomena ....................................................................................................................... 2

Hazardous phenomena at composite volcanoes ......................................................................... 4

Tephra ................................................................................................................................... 4

Pyroclastic flow and pyroclastic surge .................................................................................. 5

Lava flows and lava domes ................................................................................................... 5

Volcanic gases ....................................................................................................................... 6

Debris avalanche, landslide, and lahar .................................................................................. 6

Hazardous phenomena at monogenetic volcanoes ..................................................................... 7

Past Events at San Salvador Volcano .............................................................................................. 7

Future Activity at San Salvador Volcano ........................................................................................10

Events at Other Volcanoes Can Affect the San Salvador Region ................................................... 11

Volcano-Hazard-Zonation Maps ..................................................................................................... 11

Proximal volcanic hazard zone ..................................................................................................12

Lahar hazard zones ..................................................................................................................... 12

Regional volcanic hazard zone ................................................................................................... 13

Hazard Forecasts and Warnings ...................................................................................................... 13

Protecting Communities and Citizens from Volcano-Related Hazards .......................................... 14

References ....................................................................................................................................... 15

Additional Suggested Reading ........................................................................................................ 15

End Notes ........................................................................................................................................ 15

PLATES PLATES PLATES PLATES PLATES [In pocket]

1. Volcano hazards in the San Salvador region, El Salvador. Proximal-volcanic and lahar hazardzones from an event at San Salvador volcano.

2. Volcano hazards in the San Salvador region, El Salvador. Regional volcanic hazard zone thatcould be affected by eruptions of monogenetic volcanoes.

FIGURESFIGURESFIGURESFIGURESFIGURES

1. Location of major cities and significant Quaternary volcanoes in El Salvador ......................... 2

2. Simplified sketch showing hazardous events associated with volcanoes like San Salvador .... 3

3. Summary of eruptive history of San Salvador volcano, based largely on Sofield (1998) ......... 8



INTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTIONSan Salvador volcano is one of many volca-

noes along the volcanic arc in El Salvador(figure 1). This volcano, having a volume of about110 cubic kilometers, towers above San Salvador,the country’s capital and largest city. The city has apopulation of approximately 2 million, and apopulation density of about 2100 people per squarekilometer. The city of San Salvador and othercommunities have gradually encroached onto thelower flanks of the volcano, increasing the risk thateven small events may have serious societalconsequences. San Salvador volcano has noterupted for more than 80 years, but it has a longhistory of repeated, and sometimes violent, erup-tions. The volcano is composed of remnants ofmultiple eruptive centers, and these remnants arecommonly referred to by several names. Thecentral part of the volcano, which contains a largecircular crater, is known as El Boquerón, and itrises to an altitude of about 1890 meters. ElPicacho, the prominent peak of highest elevation(1960 meters altitude) to the northeast of the crater,and El Jabali, the peak to the northwest of thecrater, represent remnants of an older, largeredifice. The volcano has erupted several timesduring the past 70,000 years from vents central tothe volcano as well as from smaller vents and

fissures on its flanks [1] (numerals in brackets referto end notes in the report). In addition, severalsmall cinder cones and explosion craters arelocated within 10 kilometers of the volcano. Sinceabout 1200 A.D., eruptions have occurred almostexclusively along, or a few kilometers beyond, thenorthwest flank of the volcano, and have consistedprimarily of small explosions and emplacement oflava flows. However, San Salvador volcano haserupted violently and explosively in the past, evenas recently as 800 years ago. When such eruptionsoccur again, substantial population and infrastruc-ture will be at risk.

Volcanic eruptions are not the only events thatpresent a risk to local communities. Anotherconcern is a landslide and an associated debris flow(a watery flow of mud, rock, and debris--alsoknown as a lahar) that could occur during periodsof no volcanic activity. An event of this typeoccurred in 1998 at Casita volcano in Nicaraguawhen extremely heavy rainfall from HurricaneMitch triggered a landslide that moved down slopeand transformed into a rapidly moving debris flowthat destroyed two villages and killed more than2000 people. Historical landslides up to a fewhundred thousand cubic meters in volume havebeen triggered on San Salvador volcano by torren-tial rainstorms and earthquakes, and some have

Volcano Hazards in the San SalvadorVolcano Hazards in the San SalvadorVolcano Hazards in the San SalvadorVolcano Hazards in the San SalvadorVolcano Hazards in the San SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El SalvadorRegion, El SalvadorByByByByBy J.J. Major, S.P. Schilling, D.J. Sofield J.J. Major, S.P. Schilling, D.J. Sofield J.J. Major, S.P. Schilling, D.J. Sofield J.J. Major, S.P. Schilling, D.J. Sofield J.J. Major, S.P. Schilling, D.J. Sofield11111, C.D. Escobar, C.D. Escobar, C.D. Escobar, C.D. Escobar, C.D. Escobar22222, and C.R. Pullinger, and C.R. Pullinger, and C.R. Pullinger, and C.R. Pullinger, and C.R. Pullinger22222

1 1 1 1 1 GeoEngineers, Inc., Tacoma, WA, 98402, U.S.A.2 2 2 2 2 Servicio Nacionale de Estudios Terretoriales, Ave. Roosevelt y 55 Ave. Norte, Torre El Salvador (IPSFA)


transformed into debris flows that have inundatedpopulated areas down stream. Destructive rainfall-and earthquake-triggered landslides and debrisflows on or near San Salvador volcano in Septem-ber 1982 and January 2001 demonstrate that suchmass movements in El Salvador have also beenlethal.

This report describes the kinds of hazardousevents that occur at volcanoes in general and thekinds of hazardous geologic events that haveoccurred at San Salvador volcano in the past. Theaccompanying volcano-hazards-zonation mapsshow areas that are likely to be at risk whenhazardous events occur again.

VOLCANIC PHENOMENAVOLCANIC PHENOMENAVOLCANIC PHENOMENAVOLCANIC PHENOMENAVOLCANIC PHENOMENAVolcanoes pose a variety of geologic hazards--

both during eruptions and in the absence oferuptive activity (figure 2). Many of the hazardousevents depicted in figure 2 have occurred at San

Salvador volcano in the past and will likely occurin the future. Most of these events are caused bythe eruption of molten rock, or magma, but some,like debris avalanches (landslides) and lahars, canoccur without eruptive activity. The nature andscale of eruptive activity depends in part on thesize and type of volcano, the composition of themagma, and on interactions between magma andwater.

Two types of volcanoes are present in the SanSalvador region: composite volcanoes and monoge-netic cones. Composite volcanoes eruptepisodically over time intervals of thousands tohundreds of thousands of years and can have awide range of eruption styles. San Salvadorvolcano is a composite volcano that has been activeepisodically for more than 70,000 years.

Monogenetic volcanoes typically erupt foronly brief intervals of time--weeks to perhapscenturies--and they generally have a narrowerrange in eruptive behavior. The magma of many

Figure 1. Location of major cities and significant Quaternary volcanoes in El Salvador. Circles indicate major cities,triangles indicate major volcanoes. Lake Coatepeque and Lake Ilopango are large silicic calderas.

Lake Coatepeque

Lake Ilopango

Santa AnaV. Santa Ana

V. Izalco

V. San Vicente

V. San Miguel

V. Cochague

San Vicente

San Miguel

San Salvador

V. San Salvador

PacificOcean

GUATAMALA EL

SALVADOR

HONDURAS

90ºW. 88ºW.

14ºN.

0 40 km


Magma

Pyroclastic Flow

Pyroclastic Flow

Eruption Column

Lahar(DebrisFlow)

Lava Flow

Prevailing Wind

MagmaReservoir

Dome Collapse

Fumaroles

Eruption Cloud

Ash (Tephra)FallAcidRain

Bombs

LavaDome

Conduit

Crack

GroundWater

Landslide(DebrisAvalanche)

Magma Silica (SiO2)Types Content

Rhyolite >68%Dacite 63-68%Andesite 53-63%Basalt <53%

Figure 2. Simplified sketch showing hazardous events associated with volcanoes like San Salvador. Events such aslahars and landslides (debris avalanches) can occur even when the volcano is not erupting. Inset box showsclassification of magma types on the basis of silica content. Illustration by Bobbie Meyers, modified from USGS FactSheet 002-97.


monogenetic volcanoes is basaltic in composition,but some magmas may have higher silica contentsthat range from andesite to dacite (see insetfigure 2). The more silica in the magma, the moreexplosive an eruption. In the San Salvador region,eruptions of monogenetic volcanoes have producedexplosion craters, cinder cones, and lava flows.The rocks from most of these monogenetic centershave andesitic compositions (silica contentsranging from about 54% to 61%), but a few havebasaltic compositions (<53% silica content).Prominent cinder cones in the area include Cerro ElPlayón, Montána Quezaltepeque, Plan del Hoyo,Cerro Las Viboras and Laguna Caldera (Plates 1and 2). Prominent explosion craters includeLaguna de Chanmico, Resumidero crater, craterLas Granadillas, and other unnamed craters mainlyon the northwest flank of the volcano (Plates 1and 2). Lava flows erupted from monogeneticcenters or from fissures on the volcano’s flanks arefound mainly to the north and extend a fewkilometers northwest of Quezaltepeque.

Hazardous phenomena at compositeHazardous phenomena at compositeHazardous phenomena at compositeHazardous phenomena at compositeHazardous phenomena at compositevolcanoesvolcanoesvolcanoesvolcanoesvolcanoes

TTTTTephraephraephraephraephra

As magma nears the surface of a volcano,gases dissolved in the magma are released. If thegas is released rapidly then the magma can bebroken explosively into small fragments and bedispersed into the atmosphere. Fragments fromsuch eruptions, which range in size from micro-scopic ash to meter-sized blocks, are collectivelycalled tephra. Tephras form deposits that blanketbroad areas downwind from a volcano. A deposit’sthickness and particle size generally decrease awayfrom the vent, but a deposit can cover large areastens to hundreds of kilometers from the source.The largest tephra fragments, called ballisticprojectiles, fall to the ground within a few kilome-ters of the vent.

Tephra falls seldom threaten life directly,except within a few kilometers of a vent. Largeballistic fragments are capable of causing death orinjury by impact. Large projectiles may also stillbe hot when they land and can start fires if they fallonto combustible material. Most injuries andfatalities from tephra falls occur when the tephra

accumulations are thick, or are saturated by rain-fall, and thus are heavy enough to collapse buildingroofs. Fine tephra suspended in the air can irritateeyes and respiratory systems and exacerbatepulmonary problems, especially in the elderly andinfants.

Indirect effects of tephra falls can be perhapsmore disruptive than the direct effects of tephrafalls. Even thin accumulations of tephra fall cansignificantly disrupt social and economic activitiesover broad regions. Tephra plumes can create tensof minutes or more of darkness, even on sunnydays, and tephra falls can reduce visibility andnavigability on highways. Tephra ingested byvehicle engines can clog filters and increase wear.Deposits of tephra can short-circuit or breakelectric transformers and power lines, especially ifthe tephra is wet, sticky, and heavy. Tephra cancontaminate surface-water drinking supplies, plugstorm- and sanitary-sewer systems, and clogirrigation canals. Even thin tephra accumulationsmay ruin sensitive crops. A serious potentialdanger of tephra stems from the threat of evensmall, dilute tephra clouds to jet aircraft that flyinto them. Ingestion of even small amounts oftephra into jet engines can cause them to malfunc-tion and lose power.

Lessons learned from the 1980 eruption ofMount St. Helens in the United States can helpgovernments, businesses, and citizens to preparefor future tephra falls. Communities downwind ofMount St. Helens experienced significant disrup-tions in transportation, business activity, andservices from fallout of as little as 5 millimeters oftephra. The greater the amount of tephra fall, thelonger it took for a community to recover. Asperceived by residents, tephra falls of less than 5millimeters were a major inconvenience, whereasfalls of more than 150 mm constituted a disaster.Nonetheless, all of the downwind communitiesaffected by Mount St. Helens resumed normalactivity within about two weeks of the event.

Eruptions of San Salvador volcano andassociated cinder cones have deposited severaltephra layers over the past 70,000 years [1].Although many of the layers are relatively thin(less than several centimeters thick) beyond thevolcano, an eruption that occurred about 30,000 to40,000 years ago deposited a tephra layer that is as


much as 1 meter thick within 10 kilometers of thevolcano. This tephra layer is composed mainly ofpumice, a light frothy fragment of explodedmagma, which indicates that gas-rich magmaintruded the volcano and erupted violently. Thecompositions, textures, and distributions of othertephra layers, especially those erupted from coneson the flanks of the volcano, indicate that someexplosive eruptions were phreatomagmatic andinvolved interactions of magma and water.

Pyroclastic flow and pyroclastic surgePyroclastic flow and pyroclastic surgePyroclastic flow and pyroclastic surgePyroclastic flow and pyroclastic surgePyroclastic flow and pyroclastic surge

Sometimes the mixture of hot gases andvolcanic rock particles produced by an explosiveeruption is denser than air, and instead of risingabove the vent to produce tephra, this densemixture behaves like a fluid, stays close to theground, and flows downslope. If the mixture ismade up mostly of rock particles, then it has a highdensity and its path will be confined to topographi-cally low areas, much as topography will controlthe flow of water. This type of dense flow is calleda pyroclastic flow. However, if the mixture ismade up mostly of gas with a small proportion ofrock fragments, then it will have a lower densityand its path will be less controlled by topography.This type of gas-rich mixture is called a pyroclas-tic surge. Pyroclastic flows and surges alsoproduce ash clouds that can rise thousands ofmeters into the air, drift downwind, and transporttephra for tens of kilometers or more from avolcano.

Pyroclastic flows and surges often occurtogether, and both are exceedingly hazardous.They move at such rapid speeds, 50 to 150 kilome-ters per hour, that escape from their paths isdifficult or impossible. Temperatures in pyroclasticflows and surges commonly are several hundreddegrees Celsius or more. Owing to their highdensity, high velocity, and high temperature,pyroclastic flows can destroy all structures and killall living things in their paths by impact, burial, orincineration. Although pyroclastic surges are moredilute and less dense than pyroclastic flows, surgescan affect larger areas and still be very destructiveand lethal. People and animals caught in pyroclas-tic surges can be killed directly by trauma, severeburns, or suffocation.

There have been at least two eruptive episodesat San Salvador volcano that have producedpyroclastic flows and surges. One episode isrelated to an eruption that produced the thickpumice tephra-fall deposit about 30,000 to 40,000years ago. A younger eruptive episode in 1200A.D. also produced pyroclastic flows and surges.

Lava flows and lava domesLava flows and lava domesLava flows and lava domesLava flows and lava domesLava flows and lava domes

Lava is liquid magma that reaches Earth’ssurface nonexplosively. Depending on its viscosityand rate of discharge, lava will form a bulbous lavadome over a vent or a lava flow that can travelseveral kilometers or more down slope from a vent.Lava flows commonly move down slope as streamsof molten rock a few to tens of meters thick. Therate at which lavas flow depends chiefly upon theirchemical composition. Basaltic lavas of the kindobserved in Hawaii can move rapidly, at tens ofmeters per minute, whereas andesitic lavas of thekind commonly erupted by the volcanoes in ElSalvador are more sluggish and move at most at afew tens of meters per hour. Although lava flowscan be extremely destructive, they typically are notlife threatening. People and animals can usuallywalk out of the path of an advancing flow. How-ever, fronts on sluggish lava flows moving acrosssteep slopes can sometimes collapse and generateblocks of hot debris that cascade downslope, breakapart, and form hazardous, fast-moving pyroclasticflows or surges.

Lava domes can pose a greater hazard thanlava flows. Lava domes form from lava that ismore viscous than that in lava flows, and as a resultthey can grow on steep slopes or construct steep-sided landforms. As lava domes grow, they canbecome unstable and collapse repeatedly, alsogenerating blocks of hot debris that cascadedownslope, break apart, and form hazardous, fast-moving pyroclastic flows or surges.

Lava flows extend down all flanks of SanSalvador volcano, but many of the flows areconcentrated on the north and northwest flanks ofthe volcano. Although some individual lava flowsare difficult to follow completely, fields of lavaflows extend from about 5 to more than 10kilometers from the summit crater. No lava domeshave been recognized at the volcano.


VVVVVolcanic gasesolcanic gasesolcanic gasesolcanic gasesolcanic gases

All magmas contain gases that are releasedboth during and between eruptions. Volcanic gasesconsist mainly of steam but also include carbondioxide and compounds of sulfur and chlorine, aswell as minor amounts of several other gases.

Generally, volcanic gases are diluted rapidlydownwind from the vent, but within a few kilome-ters of a vent they can endanger life and health ifconcentrations are high and exposure is prolonged.Eyes and lungs of people and animals can beinjured by acids, ammonia, and other compounds.People and animals can suffocate in denser-than-airgases like carbon dioxide, which pond and accumu-late in closed depressions.

The greatest hazards arising directly fromgases emitted by San Salvador volcano are likely tobe confined to the summit crater, and thus ofconcern to those who work or recreate within thecrater. Outside the summit crater, direct hazardsfrom volcanic gases are likely to be minor.

A widerspread, but indirect, hazard arisingfrom volcanic gases emitted by San Salvadorvolcano involves formation of acid rain.Compounds of sulfur are one of the main gasesemitted by volcanoes, and excessive acidificationof rainfall can occur when sulfur compoundscombine with water vapor and droplets and formsulfuric acid that is deposited during storms. Ifsuch acid is sufficiently concentrated it can damagecrops, reduce land productivity, and pollute surfacewater. In El Salvador, acid rain resulting fromemission of volcanic gases has damaged coffeecrops locally, particularly around Santa Anavolcano [2].

Debris avalanche, landslide, and laharDebris avalanche, landslide, and laharDebris avalanche, landslide, and laharDebris avalanche, landslide, and laharDebris avalanche, landslide, and lahar

The slopes of a volcano can become unstableand fail catastrophically, generating a rapidlymoving landslide called a debris avalanche.Slope instability at volcanoes can be caused bymany factors. Magma rising upward through avolcano can push aside older volcanic rock anddeform and steepen the flanks of a volcano, orwarm acidic ground water can circulate throughcracks and porous zones inside a volcano, alterstrong rock to weak slippery clay, and graduallyweaken the volcano so that it is susceptible to

debris avalanches. A volcano’s slopes can also failwithout direct involvement of magma. Unexpectedearthquakes, torrential rains, or steam explosionscan trigger slope failures, but these failures arecommonly smaller in volume than those triggeredby magmatic intrusion. A debris avalanche canattain speeds in excess of 150 kilometers per hour;generally, the larger the avalanche, the faster andfarther it can travel. Small-volume debris ava-lanches typically travel only a few kilometers fromtheir source, but large-volume debris avalanchescan travel tens of kilometers from a volcano.Debris avalanches destroy everything in their pathsand can leave deposits of 10 meters to more than100 meters thick on valley floors.

Deposits of debris avalanches have beenrecognized at numerous volcanoes around theworld, including volcanoes in El Salvador. How-ever, no deposits of large debris avalanches haveyet been recognized at San Salvador volcano,suggesting that large parts of the volcano have notcollapsed catastrophically. Nevertheless, SanSalvador volcano is a large volcano, and thepossibility of a future flank collapse cannot bedismissed.

Lahars, also called mudflows and debrisflows, are flowing masses of mud, rock, and waterthat look much like flowing concrete. They areproduced when water mobilizes large volumes ofloose mud, rock, and volcanic debris. Commonly,landslides and debris avalanches transform intolahars as they move down valley. Lahars, likefloods, inundate floodplains and submerge struc-tures in low-lying areas. They can travel many tensof kilometers at speeds of tens of kilometers perhour. Lahars can destroy or damage everything intheir paths through burial or impact. They followriver valleys and leave deposits of muddy sand andgravel that can range from a few to tens of metersthick. They are particularly hazardous becausethey travel farther from a volcano than any otherhazardous phenomenon except tephra, and theyaffect stream valleys where human settlement isusually greatest. In some instances, landslides andlahars can clog a channel or block a tributarychannel and impound a lake behind the blockage.Commonly, the impounded water will spill over theblockage, quickly cut a channel, catastrophicallydrain the lake, and generate a flood that moves


down valley endangering people and property.Breaching of the blockage may occur within hoursto months after impoundment.

Like floods, lahars range greatly in size. Thesmallest lahars recur most frequently (perhapsevery few years), whereas the largest recur on theorder of centuries to millennia. The size of laharsis controlled by both the amount of water and theamount of loose sediment or volcanic debrisavailable. Large debris avalanches or eruptions candump tens to hundreds of millions of cubic metersof sediment into channels and produce large lahars.Small debris avalanches or eruptions producesmaller lahars. Lahars have occurred on SanSalvador volcano, and historical landslides andlahars of more than 100,000 cubic meters involume have failed from the volcano’s steep upperslopes and traveled more than 4 kilometers fromtheir sources [1].

Landslides and lahars can cause problems longafter the event that formed them ends. Oncelandslides and lahars fill stream channels withsediment, the streams begin to erode new paths,and the new stream channels can be highly unstableand shift rapidly as sediment is eroded and movedfarther down valley. Rapid stream shifting cancause rapid and dramatic bank erosion.Furthermore, because stream channels are cloggedwith sediment, they have less ability to conveywater. As a result, relatively small floods, whichmay have previously passed unnoticed, can posepotentially significant threats to people living inlow-lying areas. In general, people living in low-lying areas along river valleys are most susceptibleto these secondary affects from landslides andlahars, but areas on higher ground adjacent to riverchannels apparently safe from flooding may bethreatened by bank erosion. Examples from manyvolcanoes around the world show that the effects ofsediment deposition by landslides and lahars instream channels can persist for years todecades [3].

Hazardous phenomena at monogeneticHazardous phenomena at monogeneticHazardous phenomena at monogeneticHazardous phenomena at monogeneticHazardous phenomena at monogeneticvolcanoesvolcanoesvolcanoesvolcanoesvolcanoes

Monogenetic volcanoes in the San Salvadorarea have dominantly andesitic composition.Although their compositions are similar to those of

many eruptive products from San Salvadorvolcano, the effects of their eruptions generally donot extend far from the source. Future eruptions ofmonogenetic craters and cinder cones may involvemodest explosions and emplacement of lava flows.Perhaps the most dangerous events associated withmonogenetic volcanoes occur when rising magmainteracts with surface water or shallow groundwater and produces steam explosions andpyroclastic flows or pyroclastic surges, which cantravel rapidly outward for several kilometers fromthe vent. Deposits and explosion craters producedby interactions of magma and water are found onthe northwest side of San Salvador volcano.

PAST EVENTS APAST EVENTS APAST EVENTS APAST EVENTS APAST EVENTS AT SAN SALT SAN SALT SAN SALT SAN SALT SAN SALVVVVVADORADORADORADORADORVOLCANOVOLCANOVOLCANOVOLCANOVOLCANO

San Salvador volcano has erupted intermit-tently for more than 70,000 years, and historicalobservations of eruptions date back nearly 500years [1]. However, only three eruptions haveoccurred since the early 1500’s, and those eruptionsconsisted of a series of small explosions of lowviscosity basaltic magma and emplacement ofbasaltic lava flows on the flanks of the volcano.Most of the information about San Salvador’s pastbehavior comes from studies of deposits producedby prehistoric events. Many details of past erup-tions as well as the precise age of the volcano areunknown, but it is clear that the volcano hasexhibited a wide range of eruptive behaviors--fromhighly explosive events to relatively quiet lavaflows (figure 3).

The bulk of the volcano was constructed morethan 70,000 years ago. The oldest rocks exposed atSan Salvador underlie deposits from a 72,000-year-old eruption of Coatepeque, a caldera that islocated about 50 kilometers west of San Salva-dor [4]. These old rocks of San Salvador volcanoconsist of blocky basaltic to andesitic lavas andtephras, and they are exposed at the bases of ElPicacho and El Jabali and in scattered outcropsaround the volcano. El Picacho and El Jabaliconsist entirely of layered volcanic rocks that dipaway from the center of the volcano, suggestingthat these two peaks are remnants of an ancestralcone, which is called the San Salvador edifice.Presently, El Picacho and El Jabali mark the


Eruptions of explosion craters, cinder cones, lava flows. Precise timing is unknown; activity is broadly bracketed by Ilopango caldera tephra deposits TBJ and TB2

Eruptions of flank cinder cones and lava flows; eruption of summit cratercinder cone; landslides

Eruptions of flank explosion craters; cinder cones, and lava flows

Eruptions of tephras, cinder cones, lava flows, explosion craters

Violent explosive eruptions of tephra, pyroclastic flows,pyroclastic surges

Activity

EXPLANATION

Years ago

Eruptions of tephra and lava flows; construction of ancestral San Salvadorvolcano. Precise timing unkown.

Ilopango Tierra Blanca 4 (TB4)

San Andrés Tuff 1200 A.D.

Coatepeque Arce tephra fall 72ka



Ilopango Tierra Blanca Joven (TBJ) 260 A.D.

0

1,000

2,000

4,000

6,000

8,000

10,000

20,000

40,000

60,000

80,000

Explosive eruption of Boquerón; emplacement of pyroclastic flows, tephra fall

Violent explosive eruptions of ancestral San Salvador volcano; emplacement of pumicefall. Eruptions of tephras and lava flows during construction of Boquerón edifice

Figure 3. Summary of eruptive history of San Salvador volcano, based largely on Sofield (1998). Timescale isapproximately calendar years. Regional fall deposits having known dates are shown as dashed lines; regional falldeposits having approximate dates are shown as dotted lines.


location of an older volcanic edifice that has beendeeply incised. Beyond the volcano, outcropsexpose sections of a series of andesitic and basaltictephra-fall deposits. Some of these fall deposits areonly a few centimeters thick and are separated byerosional surfaces, suggesting that they are depositsfrom many small explosive eruptions that occurredduring a long interval of time.

A distinctive gray, dacitic pumice-fall deposit,known as the G1 unit, marks a significant episodeof explosive activity at San Salvador volcano. TheG1 deposit is widespread and locally is more than 1meter thick within 10 kilometers of the volcano’scrater. Near El Picacho, this distinctive fall depositis interlayered with pyroclastic-flow and surgedeposits, and multiple flow and surge depositsextend about 6 kilometers from the summit. Rockfragments within these flow and surge depositsconsist of the older San Salvador lavas. The natureand volume (about 2 to 8 cubic kilometers) of theflow and surge deposits indicate a large explosiveeruption that may have largely destroyed the olderSan Salvador edifice and formed the crater nowdefined by the peaks of El Picacho and El Jabali.Although the exact timing of this event is un-known, stratigraphic relations with tephra-falldeposits from eruptions of Ilopango caldera, whichis located on the eastern outskirts of San Salvadorcity, help constrain the time of this eruption. TheG1 unit lies between Ilopango caldera tephra-falldeposits Tierra Blanca 3 (TB3) and Tierra Blanca 4(TB4). On the basis of paleosols, buried soilhorizons, formed on these tephra deposits, TB4 hasbeen estimated to be about 40,000 to 50,000 yearsold.

Eruptions subsequent to the G1 fall depositproduced tephras and lava flows that now largelyfill the crater formed during the G1 eruption andform a cone called El Boquerón. El Boquerón iscomposed of multiple blocky lava flowsinterlayered with tephra-fall deposits, all of whichare chemically distinct from the lava flows of theancestral San Salvador edifice. Lava flows from ElBoquerón spilled over the north, east, and southrims of the San Salvador edifice. In addition toeruptions from the central crater, several smallerexplosion craters, cinder cones, and lava flowserupted on the volcano’s flanks.

El Boquerón volcano exploded violently about800 years ago in an eruption that was perhapssimilar to, but smaller than, the eruption thatproduced the G1 fall deposit. Pyroclastic-flow andtephra-fall deposits, known as the San Andrés tuff,are found at the rim of El Boquerón, and thesedeposits have been correlated with similar deposits,known as the Talpetate tephra fall, on the westernflank of San Salvador volcano. The Talpetatetephra has been dated archeologically at 1200 A.D.Together, the Talpetate fall and San Andrés tuffdeposits have a volume of about 0.5 cubic kilome-ters, indicating that this explosive eruption wasroughly one tenth as large as the eruption thatproduced the G1 fall deposit. Sedimentary struc-tures in the Talpetate fall deposit suggest that it wasproduced by an eruption in which the magmainteracted with water.

Some of the youngest volcanic features anddeposits at San Salvador volcano are along thenorthern and northwestern flanks of the volcano.These features and deposits consist of explosioncraters, cinder cones, and lava flows that aregenerally concentrated along two prominent faultzones. Several explosion craters formed between260 A.D and 1200 A.D. They are older than theSan Andrés tuff deposits, but overlie the distinctiveTierra Blanca Joven (TBJ) tephra deposit, theyoungest regional deposit associated with eruptionsof Ilopango caldera [4]. Many of these explosioncraters show evidence that they are products oferuptions in which the magma interacted withwater.

Many monogenetic cinder cones and associ-ated lava flows are on the northern andnorthwestern flanks of San Salvador volcano andbeyond (Plates 1 and 2). Many of these cindercones and lava flows are younger than the SanAndrés tuff deposit, and thus are less than 800years old. Three prominent lava flows formed onand beyond the northwestern flank of San Salvadorvolcano within the past 500 years. In addition tothese volcanic events, landslide deposits youngerthan the Talpetate tephra are present on the north-ern and northwestern flanks.

The most recent volcanic activity at SanSalvador began in June, 1917, when, following anearthquake, steam billowed from El Boquerón


crater and several fissures opened along the north-western flank of the volcano. Within a month ofthis activity, a lake in the crater of El Boquerón hadboiled off, and small explosions formed a cindercone called Boqueróncito. The eruption lastedseveral months, constructed Boqueróncito andextruded a lava flow on the northwestern flank ofthe volcano. After the 1917 eruption, fumarolicactivity continued until the late 1970’s.

Although no eruptive activity has occurred atSan Salvador volcano for more than 80 years, lethalvolcano-related events have occurred. In 1982,heavy rainfall triggered numerous landslides in ElSalvador, and at San Salvador volcano a rainfall-triggered landslide from the flank of El Picachoswept along a channel and into the northwest partof the city. The landslide, having a volume ofbetween 200,000 to 300,000 cubic meters, rapidlytransformed into a lahar that traveled more than 4kilometers from its source. Near the base of thevolcano the lahar destroyed or buried severalhomes and killed between 300 and 500 people.

More than 30 volcanic events have occurred atSan Salvador volcano in the past 40,000 years(figure 3). Hence, the average apparent frequencyof eruptions is roughly 1 per 1300 years. Thevolcano has undoubtedly erupted more frequently,because some eruptions do not leave conspicuousdeposits in the geologic record. If we separateeruptive events into discrete, datable time periods,then we establish the following apparent frequen-cies of eruptions: between 40,000 years ago and260 A.D. about 13 identifiable events indicate aneruption frequency of about 1 event per 3000 years.Between 260 A.D. and 1200 A.D. about 9 identifi-able eruptive events indicate an eruption frequencyof about 1 event per 100 years. Between 1200 A.D.and 1917 about 9 identifiable eruptive eventsindicate an eruption frequency of about 1 event per80 years. These eruption frequencies are onlystatistically averaged values. Many of theseeruptive events are likely associated with a singleeruptive episode, such as eruptions of cinder conesand lava flows, rather than distinct events widelyseparated in time. The apparent eruption frequen-cies should not be interpreted to mean that the timebetween eruptions is necessarily decreasing. Theapparent decrease with time is related primarily tobetter preservation of younger deposits. Neverthe-

less, even accounting for the imperfections of thegeologic record, evidence clearly indicates that SanSalvador volcano erupts sufficiently frequently,with an annual probability of perhaps 1 in 1000,that potential hazards from future eruptions need tobe taken seriously, as densely populated areas willbe at risk.

FUTURE ACTIVITY AFUTURE ACTIVITY AFUTURE ACTIVITY AFUTURE ACTIVITY AFUTURE ACTIVITY AT SAN SALT SAN SALT SAN SALT SAN SALT SAN SALVVVVVADORADORADORADORADORVOLCANOVOLCANOVOLCANOVOLCANOVOLCANO

On the basis of eruptive activity during thepast 2000 years, future eruptive activity at SanSalvador volcano may involve violent eruptionsfrom the summit crater, and likely will involverelatively small explosions and tephra fall from thesummit crater and monogenetic centers, andemplacement of lava flows on or beyond the flanksof the volcano. The volcano has erupted violentlyat least twice from the central summit crater, onceas recently as 800 years ago, and could do so againin the future. Explosive eruptions are more danger-ous than those that generate lava flows or cindercones. Such explosive eruptions can producepyroclastic flows and pyroclastic surges thatsimultaneously affect multiple sectors of thevolcano, as well as produce thick tephra falls andlahars that could affect areas more than 10 kilome-ters from the volcano. If magma rising into thevolcano or a nearby monogenetic center interactswith ground water or shallow surface water, itcould produce energetic steam explosions anddestructive pyroclastic surges. Landslides andlahars, triggered by any of several mechanisms, canoccur on any flank of the volcano.

The primary effects of future eruptions orlandslides will likely be confined to within about10 kilometers of the summit of the volcano. How-ever, large lahars could travel more than 10kilometers away from the summit, monogeneticcenters could erupt beyond a 10-kilometer radius,and heavy tephra fall could be carried more than 10kilometers downwind.

Upper-level wind patterns in Guatemalabetween 3000 and 15,000 meters altitude arestrongly seasonal [5]. Similar wind patterns arelikely in El Salvador. From January to March,westerly winds dominate. April and May are


transitional months in which westerly winds giveway to more northerly and easterly winds. Junethrough October are characterized by easterlywinds, and November and December are transi-tional months during which westerly windsgradually become dominant. The strong seasonalityof these winds will influence areas affected bytephra falls. Erupted tephras will likely fall east-ward of the volcano from January through March,potentially cover broad regions to the east, south,and west in April and May, affect areas west of thevolcano from June through October, and possiblyareas west, north, and east of the volcano inNovember and December. Surface winds may alsoaffect tephra distributions, and their patterns arediurnal as well as seasonal [5]. Therefore, allsectors around San Salvador volcano can beaffected by tephra fall, but some areas are morelikely to be affected than others depending uponthe season in which an eruption occurs.

The primary effects of eruptions and landslidesare serious, but secondary effects can be equally assevere, can affect areas beyond the zone of primaryimpact, and can linger for several years. Suchsecondary effects, which are associated chieflywith sediment deposited in river channels bylandslides and lahars, involve reworking andredistribution of sediment, bank erosion, loss ofchannel capacity, and enhanced hazards of floodsin low-lying areas. Secondary effects that occur inthe aftermath of an eruption of San Salvadorvolcano or a large landslide can affect areas manytens of kilometers down stream from the volcano.

EVENTS AEVENTS AEVENTS AEVENTS AEVENTS AT OTHER VOLCANOES CANT OTHER VOLCANOES CANT OTHER VOLCANOES CANT OTHER VOLCANOES CANT OTHER VOLCANOES CANAFFECT THE SAN SALAFFECT THE SAN SALAFFECT THE SAN SALAFFECT THE SAN SALAFFECT THE SAN SALVVVVVADOR REGIONADOR REGIONADOR REGIONADOR REGIONADOR REGION

San Salvador volcano and nearby vents are notthe only sources of volcanic hazards in the region.The most devastating volcanic events that haveaffected the San Salvador region are related to largeexplosive eruptions from Ilopango caldera, whichis located on the eastern outskirts of San Salvadorcity (figure 1). Four explosive eruptions from thiscaldera within approximately the past 40,000 to50,000 years left tephra-fall and pyroclastic-flowdeposits that are as much as several meters thick inthe San Salvador region. Deposits of the youngestexplosive eruption from Ilopango, the regional

Tierra Blanca Joven (TBJ) unit, are dated at 260A.D. [4], and these deposits extend from severalkilometers east of San Vicente volcano (figure 1) toseveral kilometers west and northwest of SanSalvador volcano. Such large catastrophiceruptions of Ilopango occurred about once every10,000 to 15,000 years during the past 40,000 to50,000 years, so the annual probability of anothereruption of this magnitude at Ilopango is very low.Depending upon the season and prevailing winddirections, tephra from eruptions of othervolcanoes, such as Santa Ana, San Vicente, or SanMiguel (figure 1), for example, could affect the SanSalvador region.

VOLCANO-HAZARD-ZONAVOLCANO-HAZARD-ZONAVOLCANO-HAZARD-ZONAVOLCANO-HAZARD-ZONAVOLCANO-HAZARD-ZONATION MAPSTION MAPSTION MAPSTION MAPSTION MAPSThe accompanying volcano-hazard-zonation

maps (plates 1 and 2) show areas that could beaffected by future hazardous geologic events at ornear San Salvador volcano. Individual eventstypically affect only part of a hazard zone. Thelocation and size of an affected area will depend onthe location of an erupting vent or landslide, thevolume of material involved, and the character ofan eruption, especially its explosivity.

Potentially hazardous areas around SanSalvador volcano are divided into proximal-volcanic, lahar, and regional-volcanic hazard zonesdepending on distance from the volcano and thetype of hazard. The lahar hazard zones are subdi-vided further on the basis of the relative degree ofhazard from lahars of various volumes. Hazard-zone boundaries are drawn on the basis of (1) themagnitude of past events at the volcano, as inferredfrom deposits; (2) mathematical models that usecalibrations from other volcanoes to predict theprobable extent of lahars; and (3) our experienceand judgement derived from observations andunderstanding of events at similar volcanoes.

Although we show sharp boundaries forhazard zones, the limit of the hazard does not endabruptly at these boundaries. Rather, the hazarddecreases gradually as distance from the volcanoincreases, and for lahars decreases rapidly withincreasing elevation above channel floors. Areasimmediately beyond outer hazard zones should notbe regarded as hazard-free, because the limits ofthe hazard can only be located approximately,


especially in areas of low relief. Many uncertain-ties about the source, size, and mobility of futureevents preclude locating the boundaries of zero-hazard zones precisely. Furthermore, we showonly those hazards associated with events thatoriginate at San Salvador and distal monogeneticvolcanoes. Events not directly associated withactivity at San Salvador volcano or surroundingmonogenetic volcanoes may increase the hazardswithin apparently low-hazard zones shown onplates 1 and 2. A good example of this is the lethalJanuary 2001 landslide from Balsamo Ridge,located south of San Salvador volcano, that devas-tated Las Colinas neighborhood near Nueva SanSalvador. Plate 1 shows this area as having arelatively low degree of hazard from landslides andlahars that originate on San Salvador volcano. LasColinas and other communities at the base of thesteep ridge to the south of the volcano clearly arelocated in a geologically hazardous area, buthazards from regional landslides are not defined inour analysis of volcano-related events.

Users of the hazard maps in this report shouldbe aware that the maps do not show all hazardousareas subject to landslides and lahars from SanSalvador volcano. The volcano is extensivelyincised, and landslides could occur in any drainage.For this report, we defined zones of inundationfrom lahars of various volumes for prominentchannels directed toward populous areas. Otherchannels for which we have not modeled laharinundation should not be considered as areasdevoid of lahar hazard. Landslides and lahars fromother unmapped channels could just as wellthreaten life and property.

Proximal volcanic hazard zoneProximal volcanic hazard zoneProximal volcanic hazard zoneProximal volcanic hazard zoneProximal volcanic hazard zoneThe proximal volcanic hazard zone includes

areas immediately surrounding San Salvadorvolcano, and extends about 5 to 7 kilometers fromthe summit depending upon local topography [6].This zone delineates areas subject to devastatingvolcanic phenomena including pyroclastic flowsand surges, debris avalanches, lava flows, andballistics. Owing to the high speed anddestructiveness of many of these phenomena,escape or survival is unlikely in the proximalhazard zone. Therefore, evacuating this hazard

zone during periods of volcano unrest isrealistically the only way to protect lives. Debrisavalanches and lahars will originate in the proximalarea, and deposits from small slides and flows maybe restricted to this zone. Large debris avalanchesand lahars, however, will travel away from thevolcano and flow onto adjacent lowlands. Theextent of inundation from lahars of variousvolumes is the basis for defining lahar hazardzones.

Lahar hazard zonesLahar hazard zonesLahar hazard zonesLahar hazard zonesLahar hazard zonesLahar hazard zones lie along the primary

channels that drain San Salvador volcano. Depend-ing on the distance from the volcano, these areaswill be affected a few minutes to about one hourafter a the onset of a lahar. Beyond 10 kilometersfrom the volcano’s summit escape may be possibleif people are given sufficient warning. Within 10kilometers of the volcano lahars may happen tooquickly to provide effective warning.

We used a mathematical technique calibratedwith data from other volcanoes [7] to estimatepotential areas of inundation from lahars of variousvolumes. For each channel analyzed, we definefour to five nested hazard zones that depict antici-pated inundation by hypothetical “design” laharshaving different volumes. The largest design lahar,2 million cubic meters, reflects our estimate of thelargest probable lahar generated by a debris ava-lanche that might descend suddenly from SanSalvador volcano [7]. However, a debris avalancheof sufficient volume to generate a lahar of this sizerequires failure of a large part of the volcano’supper flank, and such an event would likely berestricted to the eastern flank of the volcano [7].We therefore use 1 million cubic meters as ourlargest design lahar from other source areas.Debris avalanches of 1 million cubic meters ormore require catastrophic failures of relativelylarge pieces of the volcano, but avalanches of thissize might occur on any flank of the volcano.Avalanches might occur in conjunction withvolcanic activity, such as intrusion of magma intothe edifice, that would be detected by monitoring.However, the possibility that large flank failurescould be triggered by mechanisms other thanmagma intrusion, such as strong earthquakes or


torrential rains, cannot be dismissed. In general,landslides and lahars triggered by mechanismsother than volcanic activity are most likely to besmaller than 1 million cubic meters in volume.

The intermediate (300,000 to 500,000 cubicmeters) and smallest (100,000 cubic meters) designlahars are more typical lahar volumes for a small tomoderate eruption or for a landslide that occurswithout warning. Lahars of these sizes haveoccurred historically at San Salvador and othervolcanoes in El Salvador, and lahars of these sizesand smaller are the most likely sizes to occur again.

Large lahars are less likely to occur than smalllahars. Thus, the nested lahar-hazard zones showthat the likelihood of lahar inundation decreases asdistance from the volcano and elevation above thevalley floors increases. To our knowledge, no laharas voluminous as 1 million cubic meters has everoccurred at San Salvador volcano. The annualprobability of a lahar of this size is difficult toestimate, but is probably less than 1 in 40,000 [8].Smaller landslides and lahars triggered by earth-quakes or torrential rains are much more likely tooccur but would probably inundate only parts ofthe design hazard zones adjacent to stream chan-nels. Lahars of about 300,000 cubic meters or lessmay have an annual probability of about 1 in 100 toperhaps as great as 1 in 10 [8].

In general, lahar hazard zones are within about10 kilometers of the summit crater. Even thelargest volume “design” lahars of 2 million cubicmeters extend no more than about 15 kilometersfrom the summit crater. Local topography plays alarge role in controlling lahar runout. Althoughlandslides and lahars originate in and flow alongsteeply incised drainages on the flanks of thevolcano, these channels abruptly shallow and thetopography abruptly flattens near the base of theedifice. As a result, lahars rapidly spill out ofchannels, spread, and stop. The most distanthazard zones are associated with the deepestincised channels in which lahars will remainconfined, such as in the southwestern andnortheastern sectors of the edifice. Despite theirrelatively short runout distances, even the smallestlahars can be devastating. The city of San Salvadorand surrounding communities have encroachedonto the lower flanks of the volcano, and the lahar

hazard zones extend well into areas that are nowdensely settled.

Regional volcanic hazard zoneRegional volcanic hazard zoneRegional volcanic hazard zoneRegional volcanic hazard zoneRegional volcanic hazard zoneEruptions from monogenetic cones located

beyond the flanks of San Salvador volcano haveaffected areas more than 10 kilometers from thesummit of the volcano. Eruptions from thesemonogenetic vents have produced lava flows,pyroclastic surges, ballistics projectiles, and tephrafalls. Much of this activity has occurred to thenorth and northwest of San Salvador volcano. Wedefine a regional volcanic hazard zone (plate 2) bydetermining the distribution of monogenetic ventsin the area and assuming that future monogeneticvents or fissures will erupt only within this area.Lava flows, pyroclastic flows and surges, andballistic projectiles produced by eruptions fromthese monogenetic centers are assumed to travel amaximum of 5 kilometers from their source vent.Therefore, we establish a hazard zone boundaryeither 5 kilometers downslope from where ventsmay open, or where significant topographic fea-tures would likely stop or divert lava flows. Theregional hazard zone encompasses land that islargely populated. An eruption from a monogeneticvent within this zone will cause significant societaldisruption, because lava flows, pyroclastic flows,and pyroclastic surges destroy everything in theirpaths.

At least 8 lava flows have occurred in the past1700 years, suggesting that the annual probabilityof a lava flow occurring in the local area north ornorthwest of the volcano is about 1 in 200.However, because only a relatively small area ofthe regional hazard zone will be affected duringany given eruptive episode, and because large areaswithin the hazard zone never have been covered bylava flows, the annual probability for any specificpoint in the zone being inundated by renewedvolcanism is less than 1 in 200.

HAZARD FORECASTS ANDHAZARD FORECASTS ANDHAZARD FORECASTS ANDHAZARD FORECASTS ANDHAZARD FORECASTS ANDWARNINGSWARNINGSWARNINGSWARNINGSWARNINGS

Scientists normally can recognize and monitorseveral indicators of impending volcanic eruptions.Magma rising into a volcano prior to an eruption


causes changes that can usually be detected byvarious geophysical instruments and visual obser-vations. Swarms of small earthquakes aregenerated as rock breaks to make room for risingmagma or as heating of fluids increases under-ground pressures. Heat from the magma canincrease the temperature of ground water and raisetemperatures of hot springs and steaming fromfumaroles; it can also generate small steam explo-sions. The composition and volume of gasesemitted by fumaroles can change as magma nearsthe surface, and injection of magma into a volcanocan cause swelling or other types of surface defor-mation.

El Salvador has a national seismic network, soa significant swarm of earthquakes at San Salvadorvolcano would be noticed quickly. At other volca-noes similar to San Salvador, notable increases inseismicity have occurred days to months beforeeruptions. An increase in seismicity near thevolcano should prompt deployment of additionalseismometers to better locate earthquakes, andstimulate other monitoring efforts that examinesigns of volcanic unrest.

Periods of unrest at volcanoes produce timesof great uncertainty. During the past few decadessubstantial advances have been made in volcanomonitoring and eruption forecasting, but stillscientists can often make only very general state-ments about the probability, type, and scale of animpending eruption. Precursory activity can gothrough accelerating and decelerating phases, andsometimes will die out without an eruption. Gov-ernment officials and the public must realize thelimitations in forecasting eruptions and must beprepared to cope with such uncertainty.

Despite advances in volcano monitoring anderuption forecasting, it is still difficult, if notimpossible, to predict the precise occurrence oflandslides triggered by earthquakes or torrentialrains. Therefore, government officials and thepublic need to identify the locations of lahar hazardzones and realize that potentially lethal events inthese hazard zones can occur with little or nowarning.

PROTECTING COMMUNITIES ANDPROTECTING COMMUNITIES ANDPROTECTING COMMUNITIES ANDPROTECTING COMMUNITIES ANDPROTECTING COMMUNITIES ANDCITIZENS FROM VOLCANO-RELACITIZENS FROM VOLCANO-RELACITIZENS FROM VOLCANO-RELACITIZENS FROM VOLCANO-RELACITIZENS FROM VOLCANO-RELATEDTEDTEDTEDTEDHAZARDSHAZARDSHAZARDSHAZARDSHAZARDS

Communities, businesses, and citizens mustplan ahead to mitigate the effects of future volcaniceruptions, landslides, and lahars from San Salvadorvolcano. Long-term mitigation efforts must includeusing information about volcano hazards whenmaking decisions about land use and siting ofcritical facilities. Future development should avoidareas judged to have an unacceptably high risk orbe planned and designed to reduce the level of risk.

When volcanoes erupt or threaten to erupt, arapid, well-coordinated emergency response isneeded. Such a response will be most effective ifcitizens and public officials have a basic under-standing of volcano hazards and have planned theactions needed to protect communities.

Because a volcanic eruption can occur withindays to months after the first precursory activityand because some hazardous events, such aslandslides and lahars, can occur without warning,suitable emergency plans should be made inadvance. Although it has been more than 80 yearssince San Salvador volcano erupted and it isunknown when it will erupt again, public officialsneed to consider issues such as public education,land-use planning, communication and warningstrategies, and evacuations as part of a responseplan. Emergency plans already developed forfloods may apply to some extent, but may needmodifications for hazards from lahars. For habitatsin low-lying areas, a map showing the shortestroute to high ground will also be helpful forevacuations.

Knowledge and advance planning are the mostimportant items for dealing with volcano hazards.Especially important is a plan of action based onthe knowledge of relatively safe areas aroundhomes, schools, and workplaces. All of thevolcano hazards described in this report are serious,and many different hazardous phenomena mayaffect an area that extends about 7 kilometers fromthe summit of San Salvador volcano. Lahars posethe biggest threat to people living, working, orrecreating along channels that drain San Salvadorvolcano, even at distances of as much as


15 kilometers from the volcano. The best strategyfor avoiding a lahar is to move to the highestpossible ground. A safe height above riverchannels depends on many factors including thesize of the lahar, distance from the volcano, andshape of the valley. For areas beyond about 10kilometers from the summit of the volcano, all butthe largest lahars will rise less than about 20 metersabove the channel bottom. San Salvador volcanowill erupt again, and the best way to cope withfuture eruptions is through advance planning inorder to mitigate their effects.

REFERENCESREFERENCESREFERENCESREFERENCESREFERENCESBäcklin, C. and Finnson, H., 1994, Landslide hazard

at the San Salvador volcano: M.S. thesis, Depart-ment of Civil and Environmental Engineering,Royal Institute of Technology, Stockholm,Sweden, 136 p.

Baum, R.L., Crone, A.J., Escobar, D., Harp, E.L.,Major, J.J., Martinez, M., Pullinger, C.R., andSmith, M.E., 2001, Assessment of landslidehazards resulting from the February 13, 2001, ElSalvador earthquake: U.S. Geological SurveyOpen-File Report 01-119, 22 p.

Hart, W.J.E., and Steen-McIntyre, V., 1983, TierraBlanca Joven tephra from the A.D. 260 eruptionof Ilopango caldera, in Sheets, P.D., ed.,Archeaology and Volcanism in Central America:University of Texas Press, Austin, p. 14-34.

Hayashi, J.N., and Self, S., 1992, A comparison ofpyroclastic flow and debris avalanche mobility:Journal of Geophysical Research, v. 97,p. 9063-9071.

Iverson, R.M., Schilling, S.P., and Vallance, J.W.,1998, Objective delineation of lahar-hazard zonesdownstream from volcanoes: Geological Societyof America Bulletin, v. 110, p. 972-984.

Major, J.J., Pierson, T.C., Dinehart, R.L., and Costa,J.E., 2000, Sediment yield following severevolcanic disturbance—A two decade perspectivefrom Mount St. Helens: Geology, v. 28,p. 819-822.

Malin, M.C., and Sheridan, M.F., 1982, Computer-assisted mapping of pyroclastic surges: Science,v. 217, p. 637-640.

Mercado, R., Rose, W.I., Najera, L., Matías, O., andGirón, J., 1988, Volcanic ashfall hazards andupper wind patterns in Guatemala, preliminary

report: Publication of Department of GeologicalEngineering and Sciences, Michigan Technologi-cal University: Houghton, MI, 34 p.

Portig, W.H., 1976, The climate of Central America,in Schwerdtfeger, W., ed., World Survey ofClimatology, Climates of Central and SouthAmerica, v. 12: Elsevier, New York, p. 405-478.

Rose, W.I., Conway, F.M., Pullinger, C.R., Deino, A.,and McIntosh, W.C., 1999, An improved ageframework for late Quaternary silicic eruptions innorthern Central America: Bulletin of Volcanol-ogy, v. 61, p. 106-120.

Rymer, M.J., and White, R.A., 1989, Hazards in ElSalvador from earthquake-induced landslides, inBrabb, E.E., and Harrod, B.L., eds., Landslides:Extent and Economic Significance. Balkema,Rotterdam, p. 105-109

Sofield, D.J., 1998, History and hazards of VolcanSan Salvador, El Salvador: M.S. thesis, MichiganTechnological University, 116 p.

ADDITIONAL SUGGESTED READINGADDITIONAL SUGGESTED READINGADDITIONAL SUGGESTED READINGADDITIONAL SUGGESTED READINGADDITIONAL SUGGESTED READINGBlong, R.J., 1984, Volcanic hazards: Academic Press,

Orlando, FL., 424 p.Sigurdsson, H., Houghton, B., McNutt, S.R., Rymer,

H., and Stix, J., eds., 2000, Encyclopedia ofVolcanoes: Academic Press, San Diego, CA.,1417 p.

Tilling, R.I., ed., 1989, Volcanic hazards: Shortcourse in geology, v. 1, American GeophysicalUnion, Washington, D.C., 123 p.

END NOTESEND NOTESEND NOTESEND NOTESEND NOTES[1] The geologic data upon which this report is

based come largely from Sofield (1998);Bäcklin and Finnson (1994); communicationswith personnel at Centro de InvestigacionesGeotécnicas, San Salvador; and our ownreconnaissance investigations.

[2] Diario del Hoy reported on gas emissions,acid rain, and crop damage at coffee planta-tions around Santa Ana volcano in a storypublished on January 19, 2001.

[3] Analyses of limited data from volcanoesaround the world indicate that sediment yieldsfrom river channels filled with volcanic debrisby an eruption can remain higher than typical


background levels for years to decades afteran eruption. In some cases sediment yieldscan remain 10 to 100 times greater thantypical background levels for more than twodecades (Major et al., 2000). River channelsheavily clogged with sediment typically areunstable. Heavy sediment deposition causes ariver to wander across the valley floor, whichcan trigger significant bank erosion thatfurther adds to a river’s sediment load.

[4] Ages of eruptions from large silicic calderasin Central America are given in Rose et al.(1999). Detailed discussion of the TierraBlanca Joven (TBJ) tephra from Ilopangocaldera is given in Hart and Steen-McIntyre(1983).

[5] Upper-level wind patterns in Guatemala aregiven in Mercado et al. (1988). Diurnal andseasonal surface-wind patterns in San Salva-dor are given in Portig (1976).

[6] The maximum extent of the proximal volcanichazard zone is estimated from the formulaH/L = 0.2, where H is the elevation differencebetween the summit rim of El Boquerón andthe hazard boundary line, and L is the horizon-tal distance from the center of the summitcrater to the hazard boundary line (see, forexample, Malin and Sheridan, 1982; Hayashiand Self, 1992; and Iverson et al., 1998). Thevalue 0.2 was selected because it approxi-mately encompasses the distal extent of lavaserupted from the central summit area, andbecause it is consistent with H/L ratios ofproximal hazardous phenomena at many othervolcanoes.

[7] Lahar hazard zones were constructed bymodeling lahar volumes of 100,000; 300,000;500,000; 1 million, and on the east flank ofthe volcano, 2 million cubic meters. Usingmathematical and digital cartographic tech-niques (Iverson et al., 1998), these volumeswere used to compute the estimated extent ofinundation down stream from a source area.Historical landslides from San Salvadorvolcano have had estimated volumes of asmuch as 300,000 cubic meters (Bäcklin andFinnson, 1994; Centro de InvestigacionesGeotécnicas, personal communication);regional earthquake- and rainfall-triggered

landslides have estimated volumes of morethan 10 million cubic meters, but most land-slides triggered by these mechanisms have hadvolumes of a few hundred to a few tens ofthousands of cubic meters (Rymer and White,1989; Baum et al., 2001; E.L. Harp and A.J.Crone, U.S. Geological Survey, personalcommunication). At Casita volcano in Nicara-gua, extremely heavy rainfall from HurricaneMitch triggered a landslide of about 1.5million cubic meters in volume, but as itmoved down slope it transformed into a laharthat scoured its channel and its volumeenlarged to more than 3 million cubic meters(K.M. Scott, U.S. Geological Survey, personalcommunication). On the basis of these data,we select a landslide and associated lahar of 1million cubic meters to be a probable maxi-mum size likely to be triggered in mostchannels at San Salvador volcano by earth-quakes or torrential rainfalls. Locally, we usea larger design lahar as noted below.

A volume of 10 million cubic meters isconsidered the largest likely debris avalanchefrom San Salvador volcano, on the basis of thefollowing analogy to the 1980 debris ava-lanche of Mount St. Helens and other geologicarguments. The 1980 Mount St. Helensavalanche removed about 2300 million cubicmeters from the north flank of the volcano,which had an average slope of about 30degrees. This avalanche removed about 25%of the cone’s total volume above the altitudeat which the failure plane intersected thelower north flank. In contrast to Mount St.Helens, San Salvador volcano has a broadertopographic profile and, with the exception ofEl Picacho, few slopes exceed about 20degrees. On El Picacho, slopes above 1400meters altitude equal or exceed 30 degrees,and the volume above 1400 meters is about950 million cubic meters. If the 25% valuefrom Mount St. Helens is applied to SanSalvador, then the maximum volume of a largedebris avalanche from El Picacho is slightlymore than 200 million cubic meters. Theanalogy with Mount St. Helens, however, isfor a debris avalanche triggered by localmagmatic intrusion. El Picacho is signifi-cantly separated from the central crater of San


Salvador, is composed of relatively unalteredlava flows, and is unlikely to produce such alarge debris avalanche unless the volcanoundergoes extreme deformation. Futuremagma intrusion is likely to occur in thecentral vent or along the weak fault zones onthe northwest flank of the volcano rather thanbeneath El Picacho. In our judgement, abetter estimate of the largest debris avalanchethat could come from El Picacho is perhaps onthe order of 10 million cubic meters, similar involume to a recent earthquake-triggeredlandslide along the Río Jiboa near San Vicentevolcano (Baum et al., 2001). San Salvadorvolcano does not have a geologic history ofproducing large landslides, and many of thelargest landslides generated throughout thecountry by strong earthquakes in 2001 wereless than 1 million cubic meters (E.L. Harpand A.J. Crone, U.S. Geological Survey,personal communication). Although we limitthe largest likely debris avalanche to 10million cubic meters, we think that a lahar ofthis volume is not likely to flow down a singlechannel. The east flank of San Salvadorvolcano is deeply dissected with closelyspaced channels. Thus a large debris ava-lanche would likely be dispersed amongseveral channels. Accordingly we selected 2million cubic meters as the largest laharvolume that might travel along any singlechannel beneath El Picacho.

[8] The annual probability of a lahar having ofvolume that equals or exceeds 1 million cubicmeters from San Salvador volcano is less than1 in 40,000. We estimate this probability onthe basis of the observation that no depositsfrom a lahar of this size are found in thegeologic record at least since the time of theexplosive eruption that emplaced the G1

tephra fall about 40,000 to 50,000 years ago.We estimate possible annual probabilities oflandslides and lahars having volumes of about300,000 cubic meters or less as follows.Historical earthquake-induced landslides haveoccurred throughout El Salvador at least adozen times from 1857 to 2001 (Rymer andWhite, 1989; Baum et al., 2001). Volumes ofthese landslides have ranged from a fewhundred to more than 10 million cubic meters,but most have had volumes of less than a fewto a few tens of thousands of cubic meters.Thus, earthquake-induced landslides of smallto moderate volume occur in El Salvadorabout once every 12 years. At San Salvadorvolcano, the rainfall that triggered the ap-proximately 300,000 cubic meter landslide in1982 was the greatest on record at somestations, and nearly equal to the greatest onrecord at others. The longest precipitationrecord in the area extends more than 50 years(Bäcklin and Finnson, 1994). Thus the 1982landslide and lahar perhaps had an annualprobability of less than 1 in 50. However, asimilarly sized landslide occurred in anadjacent gully sometime in the 1940’s (Centrode Investigaciones Geotécnicas, personalcommunication). At San Vicente volcano,landslides and lahars of comparable volumeshave occurred at least four times in the past225 years, and at least three times in the past65 years, suggesting that such events theremay have an annual probability of occurrenceof about 1 in 60 to 1 in 20. Although theseprobabilities are highly generalized across theregion rather than specific to San Salvadorvolcano, we conclude that the annual prob-ability of landslides and lahars ≤300,000 cubicmeters in size at San Salvador volcano isabout 1 in 100 to perhaps as great as 1 in 10.

2424242424 VVVVVolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El Salvadorolcano Hazards in the San Salvador Region, El SalvadorPrinted on recycled paper

Major, and others—

VOLCA

NO

HA

ZARD

S IN TH

E SAN

SAL

VOLCA

NO

HA

ZARD

S IN TH

E SAN

SAL

VOLCA

NO

HA

ZARD

S IN TH

E SAN

SAL

VOLCA

NO

HA

ZARD

S IN TH

E SAN

SAL

VOLCA

NO

HA

ZARD

S IN TH

E SAN

SALVV VVVA

DO

R REGIO

N, EL SA

LA

DO

R REGIO

N, EL SA

LA

DO

R REGIO

N, EL SA

LA

DO

R REGIO

N, EL SA

LA

DO

R REGIO

N, EL SA

LVV VVVAD

OR

AD

OR

AD

OR

AD

OR

AD

OR—

U.S. Geological Survey — Open-File Report 01-366

Documents

Volcano Hazards in the San Salvador Region, El Salvador · By J.J. Major, S.P. Schilling, D.J. Sofield, C.D. Escobar, and C.R. Pullinger U.S. GEOLOGICAL SURVEY Open-File Report 01-366