Las car (Chile) 4 May 2005 erup tion sends ash over 1,000 km SE, ? of the way to Bue nos Ai res . . . . . . . . 2
Fernandina (Gal?pagos Is lands) Lava flows down S flank from cir cumfer ential vents near cal dera rim . . . . 3
Anatahan (Mariana Is lands) Ex plo sive erup tion on 6 April 2005 is sues high est ash plume re corded here . . . 4
Vailulu?u (Sa moa) ALIA cruise dis closes new cone in the sum mit crater . . . . . . . . . . . . . . . . . . . . 7
Awu (In do ne sia) Sta ble dur ing mid- to late Au gust 2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Karthala (Como ros) 16 April 2005 seis mic ity lead ing to erup tion; near-source tephra 1.5 m thick. . . . . . . 9
Ol Doinyo Lengai (Tan za nia) Tall hor ni to al most reaches sum mit elevation; more lava spills over rim. . . . 11
Edi tors: Rick Wunderman, Ed ward Venzke, and Gari May berry
Vol un teer Staff: Rob ert An drews, Catherine Gal ley, Wil liam Henoch, Clem ent Pryor, Ste ven Bentley, and Jerome Hudis
Global Vol can ism Program ? Na tional Mu seum of Natu ral His tory, Room E-421, PO Box 37012 ? Wash ing ton, DC 20013-7012 ? USA
Tele phone: (202) 633-1800 ?  Fax: (202) 357- 2476  ?  Email:  gvn@vol cano.si.edu  ?  URL: http://www.vol cano.si.edu/
Sub scrip tions are pro vided by the Ameri can Geo physi cal Un ion (see the box on the last page for details).
Data are pre limi nary and sub ject to change; con tact the origi nal source or the Global Vol can ism Program bef ore us ing.
Bul le tin of the Global Vol can ism Net work
Vol ume 30, Num ber 4, April 2005
Las car
north ern Chile
23.37?S, 67.73?W; sum mit elev. 5,592 m
All times are lo cal (= UTC - 3 hours)
Las car, the most ac tive vol cano in north ern Chile,
erupted on 4 May 2005. Al though the erup tion was sub -
stan tial, thus far there is an ab sence of re ports from any one
who saw the erup tion at close range. Pre lim i nary as sess -
ments came mainly from sat el lite sen sors and dis tant af fects 
wit nessed in Ar gen tina. This re port is based on one sent to
us by Chil ean Observatorio Volcanol?gico de los An des del 
Sur (OVDAS) sci en tists Jos? An to nio Naranjo and Hugo
Moreno, dis cuss ing events around 4 May, with brief com -
ments on some of Las car?s be hav ior in the past sev eral
years, and sug ges tions for fu ture mon i tor ing.
Las car sits ~ 70 km SW of the in ter sec tion be tween
Chile, Ar gen tina, and Bolivia, ~ 300 km in land from the
Chil ean port city of Antofagasta. This part of the coast lies
along the Atacama desert, and on flat ter rain tens of ki lo me -
ters W of Las car re sides a large salt pan, the Salar de
Atacama (about 50 x 150 km). The set tle ment of Toconao
is ~ 33 km NW of Las car. Pre vi ous re ports dis cussed field
observations dur ing 13 Oc to ber 2002 to 15 Jan u ary 2003,
and fine ash dis charged from fumaroles on 9 De cem ber
2003 (Bul le tin v. 28, no. 3, and v. 29, no. 1).
Naranjo and Moreno con cluded that at roughly 0400 on
4 May an ex plo sive erup tion ejected an ash cloud to a ten ta -
tive al ti tude on the or der of 10 km that dis persed to the SE.
About 2 hours later the cloud be gan drop ping ash on Salta,
Ar gen tina. Sat el lite im ages por trayed the ash cloud?s dis -
persal. An avi a tion ?red alert? was is sued by the Bue nos Ai -
res Vol ca nic Ash Cen ter; they saw the plume over Ar gen -
tina at al ti tudes of 3-5 km.
Shortly af ter at mo spheric im pacts of the 4 May erup tion 
be came ap par ent, the Bue nos Ai res VAAC no ti fied
OVDAS that NW Ar gen tine cit ies had re ported fall ing ash.
These cit ies, all SE of Las car, in cluded Jujuy, Salta, San ti -
ago del Estero, and Santa Fe?lo ca tions with re spec tive ap -
prox i mate dis tances from Las car of 260, 275, 580, and
1,130 km. The Ar gen tine prov ince of Chaco, along the
coun try?s NE mar gin, was also noted as re ceiv ing ash. Bue -
nos Ai res (~ 1,530 km SE of Las car) re mained ~ 400 km
be yond the point of the far thest de tected ashfall.
Pa tri cia Lobera, a pro fes sor in Talabre, Ar gen tina, 17
km E of Las car, said that erup tion noises were not heard
there on the morn ing of 4 May. When ob serv ers saw the
plume from Talabre that morn ing they re port edly thought
the plume looked sim i lar to those on pre vi ous days.
Re motely sensed hot spots were de tected on a GOES
sat el lite im age for 0339 (0639 UTC) on 4 May, show ing the 
first ev i dence of an erup tion. In a later im age, at 0409, the
ther mal anom aly had in creased, and the im age sug gested a
grow ing, ash-bear ing cloud then trending ~ 23 km to the
SE. The ther mal anom aly di min ished in in ten sity by 0439,
re main ing di min ished there af ter, but by that time the
plume?s lead ing mar gin ex tended over ~ 100 km SE and its
tail had de tached from the vol cano. At 0509 the plume
reached 170 km SE. Ac cord ing to a press re port, at around
0600 ash fell in Salta (~ 275 km SE of Las car).
Rosa Marquilla, a ge ol o gist at the Uni ver sity of Salta,
re ported that res i dents there no ticed a mist at trib uted to the
erup tion, which hung over the city un til at least to 1600, af -
ter  which,  the sky grad u al ly cleared.  Pre l im i  nary
de scrip tion of the pe trog ra phy of the ash that fell in Salta
came from Ricardo Pereyra (Uni ver sity of Salta) who saw
crys tal frag ments (py rox enes, feld spars, and mag ne tite) and 
frag ments of vol ca nic glass con tain ing plagioclase
mircrolites. Lithic frag ments were not ob served.
The OVDAS au thors con cluded that, ap par ently since
the year 2000, Las car un der went con stant de gas sing from
an open vent within the ~ 780-m-di am e ter ac tive cen tral
crater. Spo radic ex plo sions as in July 2000 and Oc to ber
2002, and in this case, 4 May 2005, could be due to di verse
causes. For ex am ple, there may have been tem po rarily ob -
structed con duits at depth, lo cal col lapses block ing the vent
at the crater floor, or fresh magma in jec tion con tact ing
ground wa ter. Extrusion of a vis cous dome lava also might
ex plain the sud den ex plo sions. That cir cum stance would
pre sum ably lead to vis i bly in creased fumarolic out put.
Naranjo and Moreno had sev eral sug ges tions for on go -
ing mon i tor ing. First, they sug gested de vel op ing closer
long-term con tacts, including peo ple able to vi su ally mon i -
tor the vol cano di rectly, as well as con tin ued sys tem atic
con tact with the Bue nos Ai res VAAC and their sat el lite an -
a lysts. They rec om mended on go ing re la tions with the Uni -
ver sity of Ha waii (MODVOLC) pro gram to re motely sense
hot-spots. They went on to sug gest a cam paign of ste reo ae -
rial pho tog ra phy to de tect changes in the ac tive crater. They 
ad vo cated no ti fy ing lo cal in hab it ants of the pos si bil ity of
ash falls be fore an other ex plo sive ep i sode. They pointed
out that moun tain eers should be made aware of el e vated
risks within 8 km of the ac tive crater.
Back ground. Las car is the most ac tive vol cano of the
north ern Chil ean An des.  The andesi tic-to-daci tic
stratovolcano con tains six over lap ping sum mit crat ers.
Prom i nent lava flows de scend its NW flanks. An older,
higher stratovolcano 5 km to the E, Volc?n Aguas
Calientes, dis plays a well-de vel oped sum mit crater and a
prob a ble Ho lo cene lava flow near its sum mit (de Silva and
Fran cis, 1991). Las car con sists of two ma jor ed i fices; ac tiv -
ity be gan at the east ern vol cano and then shifted to the
west ern cone. The larg est erup tion of Las car took place
about 26,500 years ago, and fol low ing the erup tion of the
Tumbres sco ria flow about 9,000 years ago, ac tiv ity shifted
back to the east ern ed i fice, where three over lap ping crat ers
were formed. Fre quent small-to-mod er ate ex plo sive erup -
tions have been re corded from Las car in his tor i cal time
since the mid-19th cen tury, along with pe ri odic larger erup -
tions that pro duced ashfall hun dreds of ki lo me ters away
from the vol cano. The larg est his tor i cal erup tion of Las car
took place in 1993, producing pyroclastic flows to 8.5 km
NW of the sum mit and ashfall in Bue nos Aires.
Ref er ences: Gardeweg, M., 1989, Informe preliminar
sobre la evoluci?n de la erupci?n del volc?n L?scar (II
Regi?n): noviembre 1989: Servicio Nacional de Geolog?a y 
Miner?a, Informe In?dito (un pub lished re port), 27 p.
Gardeweg, M., and Lindsay, J., 2004, Las car Vol cano,
La Pacana Cal dera, and El Tatio Geo ther mal Field:
IAVCEI Gen eral As sem bly Puc?n 2004, Field Trip
Guide-A2, 32 p.
Gardeweg, M., Me dina, E., Murillo, M., and Espinoza,
A., 1993, La erupci?n del 19-20 de abril de 1993: VI
informe sobre el comportamiento del volc?n L?scar (II
Regi?n): Servicio Nacional de Geolog?a y Miner?a, Informe 
In?dito (un pub lished re port), 20 p.
2    Las car Smithsonian Institution ? Bulletin of the Global
In for ma tion Con tacts: Jos? An to nio Naranjo and Hugo 
Moreno, Programa Riesgo Volcanico, Servicio Nacional de 
Geologia y Mineria, Avda. Santa Maria 0104, Casilla 1347, 
San ti ago, Chile; Gustavo Alberto Flow ers, Bue nos Ai res
Vol ca nic Ash Ad vi sory Cen ter (Bue nos Ai res VAAC),
Servicio Meteorol?gico Nacional-Fuerza A?rea Ar gen tina,
25 de mayo 658, Bue nos Ai res, Ar gen tina (URL: http://
www.meteofa.mil.ar/vaac/vaac.htm).
Fernandina
Gal?pagos Is lands, Ec ua dor
0.37?S, 91.55?W; sum mit elev. 1,476 m
All times are lo cal (= UTC - 6 hours)
On the morn ing of 13 May 2005, a new erup tion started
on un in hab ited Fernandina vol cano (fig ure 1). Fernandina
last erupted in 1995 (fig ure 2), and had been quiet and
seem ingly un changed when a team from the Ec ua dor ian In -
sti tute of Geo phys ics (IG) flew over it in late March 2005.
On 11 May an M 5.0 earth quake oc curred with an epi cen ter 
~ 30 km E of Fernandina?s cen ter. Only two other earth -
quakes have been lo cated by the U.S. Geo log i cal Sur vey
(USGS) within 100 km of Fernandina in last 4.5 years (M
4.0 on 23 Feb ru ary 2005 and M 4.6 on 16 April 2005), both 
hav ing epi cen ters ~ 70-80 km SE of Fernandina?s cen ter. A
seis mic sta tion, in stalled by the IG in 1996 on the NE coast
of the is land, was out of service at the time of eruption.
Gal?pagos Na tional Park work ers in west ern Gal?pagos 
were ap par ently the first to wit ness the erup tion, and IG
tech ni cians rec og nized it on sat el lite im ag ery. The Uni ver -
sity of Ha waii pres ents hotspot im ages on their website.
Their GOES data lacked hotspots at 0930, but a clear and
strong one had de vel oped on the S flank by 0945. Fran cisco 
Dousdebes (of Met ro pol i tan Tour ing) placed the erup tion?s 
start time at 0935. S-flank hotspots were com par a tively ex -
ten sive by 1015. The Wash ing ton VAAC is sued their first
full ad vi sory at 1315. Their no tices re ported that the W-di -
rected plume rose to ~ 5 km al ti tude, and the S-di rected
plume went to 9 km; both were vis i ble as late as 1745 on 13 
May, de pict ing the lead ing por tions of Fernandina? s ash
plume more than 200 km from the vol cano
An over flight of the erup tion on the 13th by the Na -
tional Park re sulted in a re port by Patricio Ram?n and Hugo 
Yepes, and the erup tion was con firmed by Wash ing ton
Tapia, di rec tor of the Gal?pagos Na tional Park. That eve -
ning, Gal?pagos res i dent Greg Estes tele phoned Den nis
Geist to re port that the erup tive source was a ?cir cumfer -
ential vent near [the] sum mit, S side . . . 6 km long with an
erup tive zone 50 m across.? It was un cer tain how this fis -
sure was re lated to the 1981 erup tion site (fig ure 2 and Bul -
le tin v. 9, no. 3). IG also noted that tephra had fallen on
neigh bor ing Isabela Is land, in the ar eas of the vol ca noes
Wolf and Ecuador (~ 40 km from the vent, fig ure 1).
At 0537 on the sec ond morn ing, 14 May, the Wash ing -
ton VAAC re ported low level ash/steam not vis i ble in in fra -
red im ag ery, but at 0746, 1? hours af ter sun rise, a plume of 
ash ex tended ~ 130 km to the W and was mov ing at 18 km/
hour at 1,800 m el e va tion. The GOES ther mal anom aly was 
greatly di min ished by 0930, and re mained low to non-ex is -
tent un til re sump tion around 1415. That af ter noon, an over -
flight by Godfrey Merlen, Wacho Tapia, and Alan Tye
(Charles Dar win Re search Sta tion) re sulted in the fullest
report to date.
They said that al though the vent area was ob scured by
clouds, to pog ra phy sug gested a 4.5 km long fis sure vent
near the S rim, with ac tiv ity hav ing pro gressed from SW
(near the first and up per most flows of the 1995 ra dial fis -
sure erup tion) to the E (fig ure 1). The lava flows ?had be -
gun to pond on the gen tler outer skirt of the is land,? and
were then 5.5 km from the coast (~ 5 km from the vents).
They thought it un likely that the flows would reach the sea.
A fol low-up news re port in El Comercio (Quito) quoted
Tapia as iden ti fy ing five flows down the S flank. Only one
re mained in can des cent. At 1745 on 14 May, Wash ing ton
VAAC re ported a plume re main ing to the NW, but?lack ing
de tect able ash?they dis con tin ued ad vi so ries. Ther mal
anom a lies on the GOES sat el lite re mained strong, however, 
until the next morning.
The re port also noted that, ?As on pre vi ous erup tions,
such as that on Cerro Azul in 1998, lava pass ing through
veg e tated ar eas has caused small fires, but these have not
spread far from the lava tongues them selves be fore go ing
out. Most of the new flows have passed over unvegetated
older lava, and dam age to Fernandina?s veg e ta tion is
limited.?
The team also flew over Alcedo vol cano on Isabela,
where Pro ject Isabela staff had re ported in creased fumarole 
ac tiv ity. Steam was ris ing from the ?new? fumarole sites
(ac tive since the 1990s) and from the area of sul fur de pos its 
and fumaroles in the south west ern area of the rim, but this
ac tiv ity did not ap pear unusual.
 Volcanism Network, Volume 30, Number 4, April 2005 Fernandina    3
Fig ure 1. Sketch map of Fernandina, show ing the con spic u ous sum mit
cal dera, and in di cat ing the flow fields and cir cumfer ential vent area from
the 13 May 2005 erup tion (as mapped on 14 May by air borne
re con nais sance and  re ported by the Charles Dar win Re search Sta tion). 
Key fea tures in clude the ac tive cir cumfer ential fis sure vent and two main
ar eas im pacted by lava flows. The east ern area contained lava flows still
mo bile on 14 May; flows to the W had al ready cooled by 14 May.
On the in dex map of the Gal?pagos Is lands, the larg est is land, Isabela, is
~ 130 km long and lies to the E of Fernandina is land. ?La
Cumbre??Span ish for the sum mit, peak, or top?has been mis tak enly
ap plied to the vol cano, ap par ently be cause the sum mit was so la beled on
an old map. The is land has also been called Narborough. The in dex map is
in com plete in its por trayal of both vol ca noes and is lands of the
ar chi pel ago. Re vised from Bul le tin v. 20, no. 1.
On 15 May, the GOES ther mal anom aly was gone be -
fore noon, but re turned near mid night (about 2330), over a
smaller area, and it re mained through sun rise (0615) on 16
May. Small anom a lies were vis i ble the next sev eral nights
(when con trast with ad ja cent cold flows was stron gest), but
there was no ob vi ous ev i dence of con tin ued feed ing of the
new flows.
The com plex ther mal anom a lies de tected in MODIS
sat el lite im ag ery (pro vided by the Uni ver sity of Ha waii),
were abun dant around the time of erup tion. They spread
over Fernandina?s rim, in some cases in the caldera, and
broadly over the S flank. They con tin ued through at least
the rest of May.
The Wash ing ton VAAC re ported that a weak hotspot
started 29 May 2005 at 1945 (30 May at 0145 UTC) and a
very short nar row plume of ash and gases ap peared in
multi-spec tral im ag ery at 2145 (30 May at 0345 UTC). No
ground con fir ma tion of an erup tion was avail able, and there 
was a layer of low-level weather cloud over the is land. At
that time, the plume ap peared to dis si pate as it moved away
at ~ 18 km/hour.
Back ground. Fernandina, the most ac tive of Gal?pagos
vol ca noes and the one clos est to the Gal?pagos man tle
plume, is a ba saltic shield vol cano with a deep 5 x 6.5 km
sum mit cal dera. The vol cano dis plays the clas sic ?over -
turned soup bowl? pro file of Gal?pagos shield vol ca noes.
Its cal dera is elon gated in a NW-SE di rec tion and formed
dur ing sev eral ep i sodes of col lapse. Cir cumfer ential fis -
sures sur round the cal dera and were in stru men tal in growth
of the vol cano. Re port ing has been poor in this un in hab ited
west ern end of the ar chi pel ago and even a 1981 erup tion
was not wit nessed at the time. In 1968 the cal dera floor
dropped 350 m fol low ing a ma jor ex plo sive erup tion. Sub -
se quent erup tions, mostly from vents lo cated on or near the
cal dera bound ary faults, have pro duced lava flows in side
the cal dera as well as those in 1995 that reached the coast
from a SW-flank vent. Col lapse of a nearly 1 km? sec tion of 
the E cal dera wall dur ing an erup tion in 1988 pro duced a
de bris-av a lanche de posit that cov ered much of the cal dera
floor and absorbed the caldera lake.
In for ma tion Con tacts: Patricio Ram?n and Hugo
Yepes, Geo phys i cal In sti tute (IG), Escuela Polit?cnica
Nacional, Apartado 17-01-2759, Quito, Ec ua dor (URL:
http://www.igepn.edu.ec/); Alan Tye, Charles Dar win Re -
search Sta tion, Puerto Ayora, Santa Cruz, Galapagos Is -
lands, Ec ua dor (URL: http://www.darwinfoundation.org/);
Wash ing ton Vol ca nic Ash Ad vi sory Cen ter (VAAC), Sat el -
lite Anal y sis Branch, NOAA/NESDIS E/SP23, NOAA Sci -
ence Cen ter Room 401, 5200 Auth Road, Camp Springs,
MD 20746 USA (URL: http://www.ssd.noaa.gov/); Tom
Simkin, Dept. of Min eral Sci ences, Na tional Mu seum of
Nat u ral His tory, Smith so nian In sti tu tion, Wash ing ton, DC
20013-7012, USA (Email: simkin@nmnh.si.edu); Na tional 
Earth quake In for ma tion Cen ter, U.S. Geo log i cal Sur vey,
Box 25046, DFC, MS 966, Den ver, CO 80225-0046, USA
(http://neic.usgs.gov/); MODIS Ther mal Alert Sys tem; Uni -
ver sity of Ha waii and Manoa, 168 East-West Road, Post
602, Ho no lulu, HI 96822, USA (URL: http://modis.higp.
ha waii.edu).
Anatahan
Mariana Is lands, USA
16.35?N, 145.67?E; sum mit elev. 788 m
All times are lo cal (= UTC + 10 hours)
Anatahan?s third his tor i cal erup tion be gan on 5 Jan u ary
2005, and is de scribed in Bul le tin v. 29, no. 12. Fur ther de -
tails and sat el lite im ages were pre sented in Bul le tin v. 30,
no. 2, which cov ered events un til mid-Feb ru ary 2005. A
5-6 April 2005 erup tion cloud rose to at least 15 km al ti -
tude, which was the high est yet seen at the volcano.
Anatahan erupted al most con -
tin u ously after 5 Jan u ary 2005,
when it started its third erup tion
in re corded his tory. An im age
col lected by the Ozone Mon i tor -
ing In stru ment on NASA?s Aura
sat el lite shows at mo spheric sul fur 
di ox ide (SO2) con cen tra tions be -
tween 31 Jan u ary and 4 Feb ru ary
2005 ( f ig  ure  3) .  A long SO 2
plume ex tends NE and SW of
Anatahan, and the edge of the
plume cov ers Guam (the south -
ern most is land) and the other
Mariana Is lands immediately to
Anatahan?s N and S.
Volcanogenic SO2 com bines
with wa ter to cre ate a sul fu ric
acid haze. Called ?vog,? this haze
can cause i l l  ness  and  make
breath ing dif fi cult. Vol ca nic haze
grew so thick dur ing the first
week of Feb ru ary that the Na -
tional Weather Ser vice is sued a
vol ca nic haze ad vi sory for Guam, 
where sev eral i l l  nesses were
reported.
4    Anatahan Smithsonian Institution ? Bulletin of the Global
Fig ure 2. A 2002 In ter na tional Space Sta tion pho to graph of Fernandina, look ing obliquely to wards the E (N is
to wards the left). La bels show key fea tures de vel oped in 1995, 1981, and 1968 erup tions. Note the is land?s
coast line in the lower-right cor ner and along much of the left mar gin. De spite the steep walls bound ing the 850 m
deep, 5 x 6.5 km cen tral cal dera, it sup ports both an i mal and plant pop u la tions. Im age ISS05E06997 (Vis i ble
Earth v1 ID 18002) with con trast en hanced and la bels added by Bul le tin ed i tors.
Af ter mid-Feb ru ary 2005, erup tive ac tiv ity at Anatahan
steadily de clined to less than 5% of the peak level at tained
since the erup tion started on 5 Jan u ary. Ash erup tions con -
tin ued, and the 2003 crater floor was al most en tirely cov -
ered by fresh lava out to a di am e ter of ~ 1 km. A MODIS
im age taken at 0115 on 18 Feb ru ary showed a plume of
steam and vog ex tend ing about ~ 170 km SW of Anatahan.
Seis mic and acous tic re cords dur ing the last week of Feb ru -
ary 2005 showed very low lev els of ac tiv ity. Seis mic am pli -
tudes dur ing 23-28 Feb ru ary were sim i lar to those re corded
prior to the 5 Jan u ary erup tion. NASA MODIS (Mod er ate
Res o lu tion Im ag ing Spectroradiometer) im ag ery taken on
28 Feb ru ary showed a faint plume of vog and steam
trending W of Anatahan.
Dur ing the first two weeks of March 2005 vol ca nic and
seis mic ac tiv ity in creased rel a tive to the pre vi ous weeks.
Dur ing 14-17 March, seis mic ity in creased and steam rose a
few hun dred me ters above the vol cano. The in ner E crater
had been nearly filled with lava flows and lapilli since early 
January.
A small erup tion be gan on 18 March at 1544 ac cord ing
to seis mic data. On 19 March the Wash ing ton VAAC is -
sued an ad vi sory that an ash plume was vis i ble on sat el lite
im ag ery be low 4 km al ti tude. Small ex plo sions that be gan
late on 20 March lasted for 14 hours. No emis sions were
vis i ble on sat el lite im ag ery, but oth ers were, later in March
and April.
A strong out burst ap par ently be gan on 21 March, a day
when seis mic ity in creased sig nif i cantly. Seis mic am pli -
tudes peaked on the 25th and faded out on the 26th. Near
the peak on the 25th, the U.S. Air Force Weather Agency
(AFWA) de tected a hot spot on the is land on sat el lite im ag -
ery and re ported an ash plume briefly reach ing ~ 5.8 km al -
ti tude. The plume height soon dropped to be low 3 km al ti -
tude, and by near mid day on the 27th the plume had
changed from ash and steam to steam and vog. On the 27th
the plume extended ~ 240 km SW.
On 5 April at about 2200 seis mic sig nals be gan to in -
crease slowly, and the Wash ing ton VAAC be gan to see in -
creased ash on sat el lite im ag ery. On 6 April 2005 around
0300 an ex plo sive erup tion be gan and pro duced an ash
plume to an ini tial height of ~ 15.2 km al ti tude, the high est
in re corded his tory from the vol cano. Seis mic ity peaked at
the same time.
The AFWA re ported an up per level ash plume at ~ 15.2
km al ti tude blow ing E to SE and a lower level ash plume at
~ 4.6 km al ti tude blow ing SW; the up per plume ex tended
more than 465 km. Earth Probe TOMS data on 6 April at
1046 showed a com pact sul fur-di ox ide cloud drift ing E of
Anatahan fol low ing the eruption.
Chuck Sayon, the su per in ten dent of Amer i can Me mo -
rial Park noted, ?On Saipan at around 10 AM the skies
dark ened and light ash started fall ing . . . park op er a tion[s]
have been re stricted to in door ac tiv i ties due to ir ri ta tion to
eyes and breath ing as ash starts to lightly coat the area.
Schools are closed as well as the air port un til further notice
. . ..?
On 6 April dur ing 0400 to 0900 the seis mic ity at
Anatahan de creased to near back ground. The seis mic ity
surged for about 1 hour, with am pli tudes about one-half
those reached dur ing the ear lier erup tion, and sub se quently
dropped again to near back ground. Prior to the 6 April
erup tion, dur ing 31 March to 4 April the am pli tudes of har -
monic tremor var ied, reach ing a 2-month high on the 3rd.
Small ex plo sions oc curred ev ery one min ute to sev eral min -
utes, prob a bly as so ci ated with cin der-cone for ma tion.
Steam-and-ash plumes drifted ~ 200 km, and vog drifted
~ 400 km at altitudes below ~ 2.4- 4.6 km.
The U.S. Geo log i cal Sur vey (USGS) (in con junc tion
with the Com mon wealth of the North ern Mariana Is lands)
stated that the ?erup tion of 6 April 2005 was the larg est his -
tor i cal erup tion of Anatahan and ex pelled roughly 50 mil -
lion cu bic me ters of ash. The erup tion col umn and the am -
pli tude of har monic tremor both grew slowly over about 5
hours and both peaked about 0300 on 6 April lo cal time . . .. 
The peak of the erup tion lasted about one hour and then the
ac tiv ity de clined rap idly over the following hour.?
The 6 April 2005 erup tion?s plume was cap tured on sat -
el lite im ages. The im age showed a plume that was tan or
brown in color and clearly ash laden (fig ure 4).
Fig ure 5 shows SO2 con cen tra tions in the at mo sphere
on 7 April 2005, over 30 hours af ter the large 6 April erup -
tion. SO2 emis sions from the erup tion were mea sured by
the Ozone Mon i tor ing In stru ment (OMI) on NASA?s EOS/
Aura sat el lite. OMI de tects the to tal col umn amount of SO2
be tween the sen sor and the Earth?s sur face and maps this
quan tity as it or bits the planet. A new per spec tive on the
ver ti cal dis tri bu tion of the SO2 is re vealed by com bin ing the 
OMI data with co in ci dent mea sure ments made by the Mi -
cro wave Limb Sounder (MLS), also part of the Aura
mission.
 Volcanism Network, Volume 30, Number 4, April 2005 Anatahan    5
Fig ure 3. Anatahan?s at mo spheric SO2 as im aged by Aura?s OMI
in stru ment on 31 Jan u ary 2005. The OMI (Ozone Mon i tor ing In stru ment) 
mea sures SO2 in Dobson Units. Dobson Units, which de rive from
spec tro scopic mea sure ment tech niques, can be thought of as the mass of
mol e cules per unit area of Earth?s at mo spheric col umn. One Dobson Unit
equals 0.0285 grams of SO2 per square me ter. The Ozone Mon i tor ing
In stru ment (OMI) that cre ated this im age tracks global ozone change and
mon i tors aero sols like sul fates in the at mo sphere. It was added to the Aura
sat el lite as part of a col lab o ra tion be tween the Neth er lands? Agency for
Aero space Pro grams and the Finn ish Me te o ro log i cal In sti tute. NASA
de scribes the Eos sys tem Aura as ?A mis sion ded i cated to the health of the
Earth?s at mo sphere.? NASA im age cour tesy Si mon Carn (Joint Cen ter for
Earth Sys tems Tech nol ogy (JCET), Uni ver sity of Mary land Bal ti more
County (UMBC)).
The MLS data criss cross the OMI im age and clearly
show that some, but not all, of the SO2 mea sured by OMI to 
the vol cano?s E was in the up per tro po sphere or above. At
these al ti tudes, SO2?and the sul fate aero sols that form
from it?can stay in the at mo sphere and af fect the cli mate
for a lon ger pe riod of time. A weaker SO2 sig nal was also
mea sured in the same re gion dur ing the night time MLS
over pass, which crosses the im age from up per right to
lower left. The day time data, run ning from up per left to
lower right, co in cide with the OMI mea sure ments. The
MLS data west of Anatahan show no sig nif i cant SO2 sig nal, 
in di cat ing that the SO2 mea sured by OMI in this region was 
in the lower troposphere.
MLS mea sures ther mal emis sions from the Earth?s
limb, so un like the OMI sen sor it also col lects data at night.
It is de signed to mea sure ver ti cal pro files of at mo spheric
gases that are im por tant for study ing the Earth?s ozone
layer, cli mate, and air qual ity, such as SO2. These im ages,
de rived from pre lim i nary, unvalidated OMI and MLS data,
show MLS SO2 col umns (filled cir cles) mea sured ev ery
165 km along the Aura or bit, plot ted over the OMI SO2
map. The MLS SO2 col umns shown here are de rived from
pro file mea sure ments made from the up per tro po sphere
into the strato sphere (~ 215-0 hPa (hectoPascal, 10
2
 Pa) or
~ 12 km al ti tude and above), and the cir cles do not rep re -
sent the ac tual size of the MLS footprint, which is roughly
165 x 6 km.
Anatahan?s mor pho log i cal changes were high lighted in
be fore (pre-erup tion) ver sus af ter (post-erup tion) im ages.
Seis mic ity de creased at Anatahan af ter 6 April and dur ing
7-11 April was at very low lev els, near back ground. On 11
April, a steam-and-ash plume rose ~ 2.7 km al ti tude and
drifted ~ 280 km WSW.
Oc ca sional data from Anatahan re vealed that seis mic ity
ap peared to in crease dur ing 24-25 April. Dur ing 20-25
April, a con tin u ous thin plume of ash-and-steam rose to less 
than ~ 3 km al ti tude and drifted more than 185 km from the
vol  cano.  Har  monic  t remor
dropped dra mat i cally on 1 May
af ter be ing at high lev els for sev -
eral days. Dur ing 27 April to 1
May, the main ash-and-steam
plume rose to ~ 3 km al ti tude Ac -
cord ing to a news ar ti cle, the vol -
ca nic  plume from Anatahan
reached Philippine airspace on 4
May.
On 5 May an ex ten s ive
ash-and-steam plume to 4.5 km
al ti tude was vis i ble in all di rec -
tions. Ash ex tended 770 km N,
130 km S (to north ern Saipan),
and 110 km W. Vog ex tended in a 
broad swath from 3,000 km W,
over the Phil ip pines, to 1,000 km
N of Anatahan. By 9 May har -
monic tremor am pli tude had de -
creased to near-back ground lev -
els, with a cor re spond ing drop in
erup tive ac tiv ity. As of 10 May
AFWA was re  por t  ing ash to
about 3 km al ti tude ex tend ing 400 
km W and an area of vog less than 
half that noted on 5 May.
On 11 May AFWA re ported
thick ash ris ing to 4.2 km al ti tude
and mov ing WNW. The thick ash
ex tended in a tri an gu lar shape
6    Anatahan Smithsonian Institution ? Bulletin of the Global
Fig ure 4. A ma jor erup tion from Anatahan on 6 April 2005 sent an ash plume to ~ 15 km. The erup tion was
con sid ered the larg est since Anatahan?s first re corded erup tion on 10 May 2003. This Mod er ate Res o lu tion
Im ag ing Spectroradiometer (MODIS) im age was ac quired by NASA?s Terra sat el lite at 0035 UTC, about 8 hours
af ter the erup tion be gan. By this time, the ash plume had spread S to en tirely cover Saipan and Tinian, is lands
im me di ately to the S. Cour tesy of the MODIS Rapid Re sponse Im age Gal lery, spon sored in part by NASA.
Fig ure 5. Anatahan?s 5-6 April 2005 erup tion in jected sig nif i cant SO2
high into the at mo sphere. This OMI im age de picts the con cen tra tions
found over 30 hours af ter the erup tion, a time when the SO2 formed two
sep a rate zones at dis tance from the source. Anal y sis sug gests that the
west erly zone of SO2 was prob a bly in the lower tro po sphere and the
east ern zone was prob a bly in the up per tro po sphere or above. Cour tesy of
Si mon Carn.
from the sum mit 444 km to the WSW through 510 km to
the NW. A layer of thin ash at 3 km al ti tude ex tended an -
other 1,000 km be yond the thick ash. A broad swath of vog
ex tended over 2,200 km W nearly to the Phil ip pines and
over 1,400 km NNW of Anatahan. Al though the ash plume
di min ished over the next few days and was not as thick, it
re mained sig nif i cant, ris ing to 2.4 km and ex tend ing 370
km WNW on the 13th. Sci en tific per son nel from the Emer -
gency Man age ment Of fice and the USGS work ing the next
day at a spot 2-3 km W of the ac tive vent heard a con tin u -
ous roar ing sound. They also saw ash and steam ris ing by
pure con vec tion, not explosively, to 3 km altitude.
Back ground. The elon gated, 9-km-long is land of
Anatahan in the cen tral Mariana Is lands con sists of two co -
alesc ing vol ca noes with a 2.3 x 5 km, E-trending sum mit
de pres sion formed by over lap ping sum mit cal de ras. The
larger west ern cal dera is 2.3 x 3 km wide and ex tends E
from the sum mit of the west ern vol cano, the is land?s
788-m-high point .  Ponded lava f lows over lain by
pyroclastic de pos its fill the cal dera floor, whose SW side is
cut by a fresh-look ing smaller crater. The sum mit of the
lower east ern cone is cut by a 2-km-wide cal dera with a
steep-walled in ner crater whose floor is only 68 m above
sea level. Sparse ness of veg e ta tion on the most re cent lava
flows on Anatahan in di cated that they were of Ho lo cene
age, but the first his tor i cal erup tion of Anatahan did not oc -
cur un til May 2003, when a large ex plo sive erup tion took
place forming a new crater inside the eastern caldera.
Ref er ence: Chadwick, W.W., Embley, R.W., John son,
P.D., Merlea, S.G., Ristaub, S., and Bobbitta, A., 2005, The 
sub ma rine flanks of Anatahan vol cano, Com mon wealth of
the North ern Mariana Is lands: Jour. of Vol ca nol ogy and
Geo ther mal Res. (In press, June 2005).
In for ma tion Con tacts: Juan Takai Camacho and
Ramon Chong, Emer gency Man age ment Of fice of the
Com mon wealth of the North ern Mariana Is lands (CNMI/
EMO), PO Box 100007, Saipan, MP 96950, USA (URL:
http://www.cnmiemo.org/; Email: juantcamacho@hotmail.
com and rcchongemo@hotmail.com); Si mon Carn, Joint
Cen ter for Earth Sys tems Tech nol ogy (JCET), Uni ver sity
of Mary land Bal ti more County (UMBC), 1000 Hill top Cir -
cle, Bal ti more, MD 21250, USA; Ha wai ian Vol cano Ob -
ser va tory (HVO), U.S. Geo log i cal Sur vey, PO Box 51, Ha -
waii Na tional Park, HI 96718, USA (URL: http://hvo.wr.
usgs.gov/; Email: hvo-info@hvomail.wr.usgs.gov);
Charles Holliday ,  U.S. Air Force Weather Agency
(AFWA), Offutt Air Force Base, Ne braska 68113, USA
(Email: Charles.Holliday@afwa.af.mil); Randy White and
Frank Trusdell, U.S. Geo log i cal Sur vey, 345 Middlefield
Road, Menlo Park, CA 94025-3591 USA (URL http://hvo.
wr.usgs.gov/cnmi/up date.html; Email: rwhite@usgs.gov;
trusdell@usgs.gov,); Saipan Tri bune, PMB 34, Box 10001, 
Saipan, MP 96950, USA (URL: http://www.saipantribune.
com/); Op er a tional Sig nif i cant Event Im ag ery (OSEI) team, 
World Weather Bldg., 5200 Auth Rd Rm 510 (E/SP 22),
NOAA/NESDIS, Camp Springs, MD 20748, USA (URL:
http://www.osei.noaa.gov/, Email: osei@noaa.gov); Wash -
ing ton Vol ca nic Ash Ad vi sory Cen ter (VAAC), Sat el lite
Anal y sis Branch, NOAA/NESDIS E/SP23, NOAA Sci ence 
Cen ter Room 401, 5200 Auth Road, Camp Springs, MD
20746 USA (URL: http://www.ssd.noaa.gov/); Chuck
Sayon, Amer i can Me mo rial Park, Saipan, MP 96950, USA; 
NASA's Earth Ob ser va tory (URL: http://earthobservatory.
nasa.gov/).
Vailulu?u
Sa moan Is lands, USA
14.21?S, 169.06?W; sum mit elev. -590 m
Ac cord ing to Hubert Staudigel (Scripps In sti tu tion of
Ocean og ra phy) and Stan ley Hart (Woods Hole Ocean o -
graphic In sti tute), Vailulu?u seamount, the most ac tive Sa -
moan sub ma rine vol cano, erupted be tween April 2001 and
April 2005. It formed a 293-m-tall lava cone, which was
named Nafanua af ter the Sa moan God dess of War. This
new cone had been grow ing in side the 2-km-wide cal dera
of Vailulu?u at a min i mum rate of about 20 cm/day since
April 2001. Nafanua was dis cov ered dur ing a 2005 div ing
ex pe di tion with the Na tional Oce anic and At mo spheric
Agency (NOAA) re search sub mers ible Pis ces V, launched
from the Uni ver sity of Ha waii re search ves sel Kaimikai O
Kanaloa  (KOK).  I t  is  lo  cated  in the or ig i  nal ly
1,000-m-deep W crater of Vailulu?u (figures 6 to 9).
Seis mic mon i tor ing dur ing April-June 2000 showed
sub stan tial seis mic ity, ~ 4 earth quakes per day with
hypocenters be neath Nafanua (Konter and oth ers, 2004;
Bul le tin v. 26, no. 6), which can now be in ter preted as
pre-erup tion seis mic ac tiv ity. These ob ser va tions are con -
sis tent with pre vi ous re ports high light ing the vol ca nic and
hy dro ther mal ac tiv ity of Vailulu?u (Hart and oth ers, 2000;
Staudigel and oth ers, 2004). The sci en tists sug gested that
 Volcanism Network, Volume 30, Number 4, April 2005 Vailulu?u    7
Fig ure 6. The route of the 2005 cruise of the re search ves sel Kilo Moana.
Vailulu?u, to wards the E end of the Sa moan hotspot trail was vis ited on
cruise days 1-4, 4-8 April 2005. Cour tesy of H. Staudigel and S. Hart.
Fig ure 7. Bathymetry of Vailulu'u and nearby Ta'u Is land, based on a
SeaBeam bathymetric sur vey per formed dur ing R/V Mel ville's AVON 2
and 3 cruises, aug mented with sat el lite-de rived bathymetry from Smith
and Sandwell (1996). The in set shows the gen eral lo ca tion of Vailulu'u
with re spect to the Sa moan Ar chi pel ago; two other newly mapped and
dredged seamounts (Malumalu and Muli, AVON 3 cruise) are shown as
well. Scale: 10' = 18 km. From Hart and oth ers (2000).
con tin ued vol ca nic ac tiv ity could bring the sum mit re gion
of Vailulu?u to a wa ter depth of ~ 200 m. At that point,
Nafanua would over top the crater rim and fur ther growth
would re quire a build-up of the lower flanks, ar eas that rise
from the 5,000-m-deep floor of the ocean.
Staudigel and Hart teamed up in April 2005 on the Ha -
wai ian Re search Ves sel Kilo Moana to study the Sa moan
hotspot thought to un der lie Vailulu?u. They named their ex -
pe di tion ALIA af ter the an cient twin-hulled ca noe that Sa -
moan war riors used to ex plore the SW Pa cific. The Kilo
Moana left Pago Pago on 4 April 2005 to study ac tive and
ex tinct un der wa ter vol ca noes along the chain of Sa moan is -
lands. The ex pe di tion in ves ti gated pre vi ously un charted
seamounts and the sub ma rine por tions of some is lands,
scat tered over al most 600 nau ti cal miles, from its most re -
cent and quite ac tive Vailulu?u sub ma rine volcano in the E
to Combe Island in the W.
The Nafanua cone was first mapped by us ing the cen ter
beam of the re search ves sel KOK in sev eral cross ings of the 
W crater. An ac tive hy dro ther mal sys tem was ap par ent
from ev i dence such as the murky wa ter that lim ited vis i bil -
ity dur ing two sub mers ible dives, sev eral mi cro bial biomats 
cov er ing pil low lavas that were cen ti me ters thick, and a
large num ber of dif fuse vents. A dive on 30 March 2005 to
ex am ine Nafanua re ported ?that it must have grown in the
last 4 years be cause CTD (con duc tiv ity-tem per a ture-depth) 
cross ings in 2001 still were con sis tent with the old crater
mor phol ogy ... the basal por tion of the cone dis played rel a -
tively large pil lows, and higher up pil lows look al most like
very fluid pahoehoe that col lapsed and/or transitioned into
aa flows. Nafanua . . . grew very fast with abun dant brec cia
ma te rial from col laps ing and draining pillows, in particular
in the summit region.?
On 1 April, an other dive along the outer flanks of
Vailulu?u found that dur ing the up-slope tran sit, ob serv ers
saw a few ad di tional ar eas of ac tive vent ing and many sites
where there had been vent ing in the past. Large and per -
fectly formed pil low lavas were pres ent in most sites, with a 
few ar eas be ing dom i nated by bro ken ta lus frag ments and
some hav ing com pletely black glassy pil lows with no ox i -
da tion, ap par ent ev i dence for rel a tively re cent for ma tion.
The to pog ra phy was ex tremely rough, the slope be ing
punc tu ated with nu mer ous fis sure systems and edifices of
pillow lava.
A pri mary plan for the ALIA ex pe di tion was to study
the wa ter in and around the seamount for sev eral days us ing 
a CTD probe. To sam ple the in side of the vol cano for a full
tidal cy cle, the sci en tists var ied the depth of the CTD be -
tween 40 and 930 m (al most to the crater floor), col lect ing
var i ous data, in clud ing vis i bil ity. At Vailulu?u, the
particulates given off by hy dro ther mal vent ing are flushed
out of its cal dera dur ing each tidal cy cle into the sur round -
ing wa ter. In 2005, a dense layer of particulates was found
in the wa ter within the crater, but the wa ter was clear out -
side the crater rim. This con trasts with ob ser va tions seen
from the cruise in 2000, when there was a dense ring of
particulates around the whole vol cano. It ap pears that in
2005 the particulates were ris ing above the crater and then
8    Vailulu?u Smithsonian Institution ? Bulletin of the Global
Fig ure 8. Bathymetry of the Vailulu?u crater be tween the 1999 and 2005
sur veys, show ing the emer gence of Nafanua. Cour tesy of H. Staudigel
and S. Hart.
Fig ure 9. Bathymetric map of the Vailulu?u seamount from multibeam
data dur ing the April 2005 sur vey. Note the new in ner cone named
Nafanua. Con tour in ter val is 20 m. Cour tesy of H. Staudigel and S. Hart.
later sink ing, in stead of forming the widespread ring
observed in 2000.
In ad di tion, the ex pe di tion crew con ducted dredg ing of
the new sum mit of Nafanua. Staudigel and Hart noted that
the rocks from the first dredge haul were quite newly
formed, con tain ing pris tine ol iv ine-phyric vol ca nic rocks.
Abun dant large ves i cles in the rocks from Nafanua sug gest
a  vol  a  t i le-r ich magma ca pa ble  of  sub ma r ine lava
fountaining and ex plo sive outgassing in shal lower wa ter.
Dredg ing from a sec ond site, out side of Vailulu?u, re cov -
ered rocks that were both much older and far less fragile.
Back ground. A mas sive vol ca nic seamount, not dis cov -
ered un til 1975, rises 4,200 m from the sea floor to a depth
of 590 m about one-third of the way be tween Ta?u and
Rose is lands at the east ern end of the Amer i can Samoas.
The ba saltic seamount, named Vailulu?u, is con sid ered to
mark the cur rent lo ca tion of the Sa moan hotspot. The sum -
mit of Vailulu?u con tains a 2-km-wide, 400-m-deep
oval-shaped cal dera. Two prin ci pal rift zones ex tend east
and west from the sum mit, par al lel to the trend of the Sa -
moan hotspot, and a third less prom i nent rift ex tends SE of
the sum mit. The rift zones and es carp ments pro duced by
mass wast ing phe nom ena give the seamount a star-shaped
pat tern. On July 10, 1973, ex plo sions from Vailulu?u were
re corded by SOFAR (hydrophone re cords of un der wa ter
acous tic sig nals). An earth quake swarm in 1995 may have
been re lated to an erup tion from the seamount. Tur bid wa -
ter  above the  sum mit  shows evidence of  ongoing
hydrothermal plume activity.
Ref er ences: Hart, S.R., Staudigel, H., Koppers, A.A.P.,
Blusztajn, J., Baker, E.T., Work man, R., Jack son, M.,
Hauri, E., Kurz, M., Sims, K., Fornari, D., Saal, A., and Ly -
ons, S., 2000, Vailulu?u un der sea vol cano: The new Sa moa: 
Geo chem is try, Geo phys ics, Geosystems (G3), Amer i can
Geo phys i  ca l  Un ion,  v .  1 ,  no.  12,  doi :  10.1029/
2000GC000108.
Konter, J.G., Staudigel, H., Hart, S.R., and Shearer, P.
M., 2004, Seafloor seis mic mon i tor ing of an ac tive sub ma -
rine vol cano: Lo cal seis mic ity at Vailulu?u Seamount, Sa -
moa: Geo chem is try, Geo phys ics, Geosystems (G3), Amer i -
can Geo phys i cal Un ion, v. 5, no. 6, QO6007, doi: 10.1029/
2004GC000702.
Lippsett, L., 2002, Voy age to Vailulu?u: Search ing for
Un der wa ter Vol ca noes. Woods Hole Ocean o graphic In sti -
tu tion, Fathom on line mag a zine (URL: http://www.fathom.
com/fea ture/122477/).
Staudigel, H., Hart, S.R., Koppers, A., Con sta ble, C.,
Work man, R., Kurz, M., and Baker, E.T., 2004, Hy dro ther -
mal vent ing at Vailulu?u Seamount: The smok ing end of the 
Sa moan chain: Geo chem is try, Geo phys ics, Geosystems
(G3), Amer i can Geo phys i cal Un ion, v. 5, no. 2, QO2003,
doi: 10.1029/2003GC000626.
In for ma tion Con tacts: Hubert Staudigel, In sti tute of
Geo phys ics and Plan e tary Phys ics, Scripps In sti tu tion of
Ocean og ra phy, Univ. of Cal i for nia, San Diego, La Jolla,
CA 92093-0225, USA (Email: hstaudigel@ucsd.edu; URL: 
http://earthref.org/PACER/team/hstaudigel/hstaudigel.htm; 
http://igpp.ucsd.edu); Stan ley R. Hart, Woods Holes
Ocean o graphic In sti tute, Ge ol ogy and Geo phys ics Dept.,
Woods Hole, MA 02543, USA (Email: shart@whoi.edu);
ALIA Ex pe di tion, Sa moan Seamounts, R/V Kilo Moana
(KM0506), sup ported by the San Diego Super com puter
Cen ter and the Scripps In sti tu tion of Ocean og ra phy (URL:
http://earthref.org/ERESE/pro jects/ALIA/).
Awu
Great Sangihe Is land, In do ne sia
3.67?N, 125.50?E; sum mit elev. 1,320 m
All times are lo cal (= UTC + 8 hours)
Awu?s erup tion on 6 June 2004 and its el e vated seis mic -
ity in early Au gust 2004 was pre vi ously re ported (Bul le tin
v. 29, no. 10). This re port cov ers the last half of Au gust
2004, which had not been re ported on pre vi ously. Since the
6 June erup tion, ob ser va tion of the sum mit failed to re veal
any sig nif i cant changes (ta ble 1). The haz ard sta tus of Awu
dur ing this Au gust re port re mained at Level 2, hav ing been
el e vated to 4 (the high est on a scale of 1 to 4) at the time of
the 6 June erup tion and then lowered on 14 June.
Back ground. The mas sive Gunung Awu stratovolcano
oc cu pies the north ern end of Great Sangihe Is land, the larg -
est of the Sangihe arc. Deep val leys that form pas sage ways
for lahars dis sect the flanks of the 1,320-m-high vol cano,
which was con structed within a 4.5-km-wide cal dera. Awu
is one of In do ne sia?s dead li est vol ca noes; pow er ful ex plo -
sive erup tions in 1711, 1812, 1856, 1892, and 1966 pro -
duced dev as tat ing pyroclastic flows and lahars that caused
more than 8,000 fa tal i ties. Awu con tains a sum mit crater
lake that was 1 km wide and 172 m deep in 1922, but was
largely ejected during the 1966 eruption.
In for ma tion Con tacts: Dali Ahmad, Di rec tor ate of
Vol  ca  nol  ogy and Geo log i  cal  Haz ard  Mit  i  ga  t ion
(DVGHM), Jalan Diponegoro 57, Bandung 40122, In do ne -
sia (Email: dali@vsi.dpe.go.id; URL: www.vsi.dpe.go.id).
Karthala
Grand Comore Is land, Como ros
11.75?S, 43.38?W; sum mit elev. 2,361 m
All times are lo cal (= UTC + 3 hours)
Af ter a long pe riod of qui es cence fol low ing the 1991
phreatic erup tion, Karthala?s seis mic ity re bounded start ing
in July 2000 (Bul le tin v. 25, no.10). In Oc to ber 2000, more
than 20 seis mic events per day were recorded.
The lo cal ob ser va tory and a key source for this re port is
the Karthala Vol cano Ob ser va tory (KVO; Netter and
Chemin?e, 1997). They main tain close ties with the Cen tre
Na tional de Doc u men ta tion et de Re cher che Scientifique
des Comores (CNDRS), Re union Is land Uni ver sity, the
Institut de Phy sique du Globe de Paris, Pi ton de la
Fournaise Volcanological Ob ser va tory, and var i ous
universities in France.
Ac tiv ity dur ing Oc to ber 2000-March 2004. Be tween
Oc to ber 2000 and Jan u ary 2003, rel a tively low seis mic ity
 Volcanism Network, Volume 30, Number 4, April 2005 Karthala    9
Date Vol ca nic A Vol ca nic B Tec tonic
08 Aug-15 Aug 2004 ? ? 75
16 Aug-22 Aug 2004 2 1 81
23 Aug-29 Aug 2004 2 ? 102
Ta ble 1. Seis mic ity at Awu dur ing Au gust 2004 as re ported by DVGHM.
was detected be neath Karthala?s sum mit. The seis mic ity
slowly in creased. Dur ing Jan u ary in stru ments re corded 51
earth quakes on the 5th, 58 on the 10th, and 50 on the 11th.
Dur ing the month of April 2003 in stru ments reg is tered 732
(i.e. av er ag ing ~ 24 each day).
Seis mic in stru ments de tected sev eral short earth quake
swarms, each com prised of ~ 150 earth quakes. These
swarms took place on 25 March and in April 2003, and
each lasted sev eral hours. More over, seis mol o gists wit -
nessed an other swarm con sist ing of 183 events on 15 May.
Ex cept for that lat ter swarm, Karthala?s seis mic ity was rel a -
tively quiet for 35 days af ter the 25 April swarm. A photo
of the Chahal? crater from the year 2003, well be fore the
April 2005 erup tion, ap pears in fig ure 10. (For a map of
Karthala?s sum mit, see Bul le tin v. 16, no. 8.)
Dur ing the time in ter val from early June 2003 to Jan u -
ary 2004 in stru ments reg is tered three pe ri ods with el e vated
seis mic ity. The first in ter val spanned 121 days from June
un til the end of Sep tem ber 2003 and in cluded 6,315 earth -
quakes. Within that in ter val there was a ma jor cri sis on 6th
Sep tem ber, com prised of 345 events, some be ing felt by lo -
cal res i dents (Bul le tin v. 28, no. 8).
The sec ond in ter val be gan on 11 Oc to ber 2003, reach -
ing its peak on 4 Jan u ary 2004 (253 events) and stopped on
31 Jan u ary 2004. Dur ing this in ter val of 113 days, in stru -
ments reg is tered 4,431 earth quakes. The third in ter val, dur -
ing the time pe riod of 3 Feb ru ary to 5 March 2004, con -
tained fewer earth quakes. In stru ments re corded 832 events
in 31 days with a max i mum of 143 events per day. Af ter the 
third in ter val, KVO re corded only low seis mic ity un til early 
2005, when daily events rose to 50-60.
Erup tion dur ing April 2005. A seis mic cri sis be gan at
0812 on 16 April. Al though in stru ments ini tially re ceived
only short-pe riod events, start ing at 0914 they also reg is -
tered many long-pe riod ones. From 1055 on 16 April a con -
tin u ous sig nal was re corded, which was in ter preted as
tremor mark ing the be gin ning of the erup tion. At around
1400 that day in hab it ants heard a rum bling com ing from the 
vol cano. A few min utes later they ob served an ash col umn
above the sum mit. The first ash-fall de pos its be gan to form
around 1600, de vel op ing on the is land?s east ern side. Ac -
cord ing to the firsts re ports, ash de po si tion in creased and
con tin ued through the night accompanied by a strong smell
of sulfur.
On the morn ing of 17 April ash falls con tin ued on the
east ern part of the is land and were heavy enough to re quire
in hab it ants to use um brel las to get about. At mid day,
Jean-Marc Heintz, a pi lot for Comores Avi a tion, flew over
the west flank and ob served a large plume in the di rec tion
of the Chahal? crater. He also clearly ob served air borne
molten ejecta.
Around 1300, ob serv ers saw a very dark plume, spread -
ing into a mush room shape and ac com pa nied by light ning
flashes. Some in hab it ants pan icked and fled the is land?s
east ern vil lages. In the af ter noon, res i dents heard rum bling.
Dur ing the eve ning, sig nif i cant rain fall gen er ated small
mudflows, and the rumbling became stronger.
At that time, au thor i ties evac u ated some east ern vil lages 
(ac cord ing to Agence France Presse (AFP) this af fected
~ 10,000 peo ple). Ash there started to fall on the is land?s
west ern and north ern parts, no ta bly, on the coun try?s cap i -
tal city of Moroni (~ 10 km NW of the sum mit) and on the
Hahaya air port (~ 20 km N of Moroni, ~ 25 km NW of the
sum mit). Fig ure 11 shows a photo with the base of a vig or -
ous plume over the E flanks on the afternoon of 17 April.
KVO au thor i ties some times wit nessed a red color at the
plume?s base, in ter preted as a sign of an on go ing mag matic
erup tion. At 2105 the KVO seis mic net work re corded a
dras tic de crease in the am pli tude of the tremor. Dur ing the
night of 17-18 April, wide vari a tions of the tremor am pli -
tude were re corded with a max i mum at 0140 on 18 April
and a min i mum at 0430 on 18 April. There af ter, the tremor
am pli tude did not in crease. Dur ing the night of 17-18 April
the plume and falling ash disappeared.
On an over flight of the Chahal? crater at 0830 on 18
April, KVO per son nel ob served ma jor mod i fi ca tions at the
sum mit (fig ures 12-14). A lava lake (fig ure 12) had re -
placed the wa ter-bear ing lake (fig ure 10) that had oc cu pied
the crater since 1991.
On 19 April a new over flight re vealed the crater floor
con tain ing the lava lake, with its chilled sur face emit ting
steam (fig ure 13). Lava re mained con fined to Chahal?
crater. Around the cal dera area, and par tic u larly on its N,
10    Karthala Smithsonian Institution ? Bulletin of the Global
Fig ure 10. Karthala?s ~ 300-m-di am e ter Chahal? crater as seen on 15
Au gust 2003, more than a year prior to the April 2005 erup tion. The photo
was shot by the au to matic cam era lo cated at the sum mit, look ing from the
NNE to wards the SSW. The 2005 erup tion dra mat i cally changed this
scene, re plac ing the green lake seen here with a lava lake, and blan ket ing
con sid er able ar eas with tephra. Cour tesy of Nicolas Villenueve.
Fig ure 11. A phreatic erup tion as seen from Karthala?s east ern slopes on
the af ter noon of 17 April 2005. The vent lies be low the white-col ored
zone in the cen ter-right por tion of the photo. With en large ment, many
parts of the im age re cord the de scent of large pieces of ejecta. Photo credit
to school teacher Dan iel Hoffschir.
ob serv ers saw con spic u ous tephra de pos its; most of the
veg e ta tion had been destroyed (figure 14).
On 20 April a field ex cur sion found that ash de pos its
var ied in thick ness from a few mil li me ters on the coast to
~ 1.5 m at the sum mit. Near the sum mit the ob serv ers rec -
og nized some post-erup tive evap o ra tion and geo ther mal
pro cesses. Spe cif i cally, al though the lava lake?s sur face had 
frozen, there re mained suf fi cient heat un der the sur face that 
ground wa ter mi grat ing to wards to the crater?s floor evap o -
rated into steam. Dur ing an other field sur vey on 8 May, ob -
serv ers noted the re newed presence of lake water inside the
crater.
Back ground. The south ern most and larg est of the two
shield vol ca noes form ing Grand Comore Is land (also
known as Ngazidja Is land), Karthala con tains a 3 x 4 km
sum mit cal dera gen er ated by re peated col lapse. Elon gated
rift zones ex tend to the NNW and SE from the sum mit of
the Ha wai ian-style ba saltic shield, which has an asym met ri -
cal pro file that is steeper to the south. The lower SE rift
zone forms the Mas sif du Badjini, a pen in sula at the SE tip
of the is land. His tor i cal erup tions have mod i fied the mor -
phol ogy of the com pound, ir reg u lar sum mit cal dera. More
than twenty erup tions have been re corded since the 19th
cen tury from both sum mit and flank vents. Many lava flows 
have reached the sea on both sides of the is land, in clud ing
dur ing many 19th-cen tury erup tions from the sum mit cal -
dera and vents on the north ern and south ern flanks. An
1860 lava flow from the sum mit cal dera trav eled ~ 13 km to 
the NW, reach ing the western coast north of the capital city
of Moroni.
Ref er ence: Netter, C., and Chemin?e, J. (eds.), 1997,
Di rec tory of Vol cano Ob ser va to ries, 1996-1997: World
Or ga ni za tion of Vol cano Ob ser va to ries (WOVO), WOVO/
IAVCEI/UNESCO, Paris, 268 p.
In for ma tion Con tacts: Nicolas Villeneuve (CREGUR,
Cen tre de Recherches et d?Etudes en G?ographie de
l?Universit? de la R?un ion), Hamidou Nassor, and Pat rick
Bach?lery (LSTUR, Laboratoire des Sci ences de la Terre),
Universit? de La R?un ion BP 7151, 15 Av e nue, Ren?
Cassin, 97715 Saint-Denis, Re union Is land (Email:
Nicolas .Vil leneuve@univ-re union.f r ;  Pat  r ick .
Bachelery@univ-re union.fr); Fran?ois Sauvestre and
Hamid Soul?, CNDRS, BP 169, Moroni, R?publique
F?d?rale  Is lamique des Comores (Email :
obs_karthala@ifrance.com; URL: http://vol cano.ipgp.
jussieu.fr:8080/karthala/stationkar.html).
Ol Doinyo Lengai
Tan za nia, east ern Af rica
2.751?S, 35.902?E; sum mit elev. 2,960 m
Al times are lo cal (= GMT + 3 hours)
Al though lava vent ing at Ol Doinyo Lengai con tin ued
in ter mit tently after Feb ru ary 2004 (Bul le tin v. 29, no. 2), no 
sig nif i cant changes were de tected un til July 2004, a time
when vig or ous vent ing emit ted sub stan tial amounts of the
low-vis cos ity carbonatitic lava typ i cal at this vol cano
(?flash floods? of lava). This sum mary re port cov ers the
time in ter val from Feb ru ary 2004 through early Feb ru ary
2005 based on ob ser va tions made by Fred er ick Belton,
Celia Nyamweru, Bernhard Donth, and Christoph Weber.
Websites de voted to Ol Doinyo Lengai, in clud ing pho to -
graphs, in for ma tion on the evo lu tion, re cent his tory, and
cur rent sta tus of the vol cano are main tained by Belton,
Nyamweru, and Weber.
 Volcanism Network, Volume 30, Number 4, April 2005 Ol Doinyo Lengai    11
Fig ure 12. On 18 April 2005 the Karthala erup tion gen er ated a  lava lake 
in the Chahal? crater. In this photo, taken the morn ing of 18 April,
con sid er able por tions of the lava lake?s sur face still re mained mol ten and
in can des cent.. The lake?s sur face only re mained mol ten for a few hours.
This ae rial photo was taken look ing from the N. Cour tesy of Hamid Soul?.
Fig ure 13. A 19 April 2005 ae rial pho to graph of Karthala taken from the
SE cen tered on Chahal? crater. The lava lake?s sur face had chilled and it
emit ted white va por. Much of the sum mit area dis plays the re cently
de pos ited smooth-sur faced tephra blan ket. Cour tesy of Nicolas
Villenueve.
Fig ure 14. Tephra de pos its left by Karthala?s mid-April 2005 erup tion
al tered the land scape and de stroyed veg e ta tion. This pic ture was taken at
the en trance to the first cal dera on the west ern trail, viewed look ing to the
S. Cour tesy of Nicolas Villenueve.
A map, ther mal data, and some new el e va tion es ti -
mates. In Feb ru ary 2005 Weber and oth ers col lected lo ca -
tion data with a global po si tion ing sys tem (GPS) re ceiver.
Weber used this to cre ate a sketch map of the ac tive crater
(figure 15).
In July 2004 Belton com pleted the third of a se ries of
dis tance mea sure ments across crater out flow ar eas at the
crater rim (ta ble 2). Due to the un usu ally strong erup tion on 
15 July 2004 (fig ure 16), de pos its com pris ing the E over -
flow wid ened by 3 or 4 m (grow ing from 44 to 47 m, fig ure
15).  Later, in Jan u ary 2005, ob serv ers no ticed a fourth area 
of over flows had be come es tab lished on the N crater rim,
with lavas pour ing over the rim at two ad ja cent points there
(figure 15).
Dur ing 3-7 Feb ru ary 2005 Weber and oth ers com pleted
a se ries of lava and fumarole tem per a ture mea sure ments
that ap pear as ta bles 3 and 4. The ta bles in di cate the hot test
lava and fumarole tem per a tures at cracks were 588?C (at
T49C, Feb ru ary 2004) and 150?C (at T49G, June 2004), re -
spec tively. The hor ni tos T49C and T49G both lie near
T49B, a hor ni to de lin eated on fig ure 15.
Weber?s team GPS mea sure ments sug gested a sum mit
el e va tion of 2,960 ? 5 m. This is con sis tent with GPS mea -
sure ments taken in Oc to ber 2000, by a sci en tific group led
by Joerg Keller, of 2,950-2,960 m (Bul le tin v. 25, no. 12).
In ad di tion, the tall est hor ni to in the N-cen tral crater rises to 
nearly this el e va tion (see dis cus sion of T49/T56B, be low).
Dur ing ob ser va tions in Feb ru ary 2004, Weber mea -
sured the tall est hor ni to at the T49 lo ca tion (part of T56B)
in the cen ter area of the ac tive crater. GPS read ings on top
of T56B yielded an el e va tion of 2,886 m. This is only 4 m
be low the el e va tion of the sum mit and within the stated, ?
5m un cer tainty of that mea sure ment. The top of T49 is also
~ 33 m above the ad ja cent crater floor to the N. In ad di tion,
when he mea sured on 3-7 Feb ru ary, Weber found hor ni to
T58C (a then re cent fea ture) had grown to reach an
elevation of ~ 2,870 m.
Ob ser va tions dur ing Feb ru ary 2004 to Feb ru ary 2005.
Dur ing Feb ru ary 2004 vis its, T56B did not erupt, but in -
stead a new vent erupted at the T49 lo ca tion (~ 10 m E of
T49B, see also Bul le tin, v. 29, no. 2). This new vent was
called T49G (figure 15).
A group from Vol cano Ex pe di tions In ter na tional (VEI)
spent 24-30 June 2004 on Lengai and found much of the
scene at the vents in the crater sim i lar to that noted in Feb -
ru ary 2004. They noted that half of the up per 10 m of hor ni -
to T56B had col lapsed on its E side, and an ac tive lava lake
had formed in side this hor ni to with lava es cap ing sev eral
times through the col lapsed open ing to its E and flow ing
out ~ 200 m. The lava was rich in gas with a tem per a ture of
560?C. The hor ni to T58B was also ac tive and spat tered
12    Ol Doinyo Lengai Smithsonian Institution ? Bulletin of the Global
Fig ure 15. Sketch map of crater fea tures at Ol Doinyo Lengai sur veyed
with a global po si tion ing sys tem (GPS) dur ing 3-7 Feb ru ary 2005. Dur ing 
the course of the re port in ter val, new vents de vel oped at T49G and T58C
(amid the N-cen tral group of hor ni tos; T49G sits ~ 10 m E of T49B).
These new vents produced com par a tively vig or ous erup tions. Cour tesy of 
Chris Weber (Vol cano Ex pe di tions In ter na tional, VEI).
Fig ure 16. A time ex po sure pho to graph of Ol Doinyo Lengai taken just
af ter sun set on 15 July 2004. At that time the newly formed vent T58C
("Charg ing Rhino") is sued co pi ous lava. This photo was taken look ing
ap prox i mately NW from the SE part of the crater rim. Cour tesy of
Fred er ick Belton.
Crater rim
over flow area
Date of
mea sure ment
Width (m)
NW over flow 02 Aug 2003 135
NW over flow 29 Jun 2004 135
NW over flow 29 Jul 2004 135
E over flow 2 Aug 2003 44
E over flow 29 Jun 2004 44
E over flow 29 Jul 2004 47
W over flow 2 Aug 2003 17
W over flow 29 Jun 2004 18
W over flow 29 Jul 2004 18
Ta ble 2. For Ol Doinyo Lengai, the width of the three ex tant lava out flows
at the points where they spilled from the ac tive crater (?over flows,? fig ure
15), as mea sured dur ing 2 Au gust 2003-29 July 2004. Two ad di tional
small over flows formed later, by Jan u ary 2005, on the N crater rim. The
3-m E-over flow in crease oc curred dur ing the erup tion of T58C on 15 July
2004. Cour tesy of Fred er ick Belton.
lava many times dur ing these days of ob ser va tion. Some
lava flows from T58B reached about 150 m to the S.
Dur ing 2-3 July 2004, Belton ob served T58B erupt re -
peat edly, emit ting lava and strombolian dis plays. The es -
cap ing lava flowed S, pass ing near the base of hor ni to T47.
On 4 July, Belton saw some of the most in tense ac tiv ity of
the month. A se quence of lavas erupted on that day and
over the next few days. How ever, events in mid-July and
later were also un usu ally vigorous.
The 4 July 2004 ac tiv ity in cluded strong strombolian
erup tions at T58B and sev eral col lapses of its vent area,
which re leased large cas cades of lava onto the crater floor.
Si mul ta neously, a tube-fed erup tion of pahoehoe lava from
the new vent T49G flowed across the NW crater rim to spill 
down that flank. Early on 5 July nu mer ous erup tions of
T58B sent lava flow ing to ward T47 at an es ti mated ve loc ity 
of 10 m/sec. On 6 July, lava flowed out of the lake in T56B
and onto the crater floor mov ing E and en ter ing a cave in
T45 for a short distance.
Af ter very low ac tiv ity dur ing 7-10 July 2004, re newed
flows and spat ter came out on 11 July from T58B, and fre -
quent but short (usu ally ~ 2 min ute) ep i sodes of loud de gas -
sing and spat ter ing is sued from the lava lake in T56B. At
night, this vent emit ted in can des cent gas. This pat tern con -
tin ued un til the morn ing of 14 July, when erup tions at T58B 
be came more ex plo sive and it ex pelled small ash clouds.
On the morn ing of 15 July a col lapse in the vent area of
T58B re leased large rapid lava flows to the E. The ep i sodes
of de gas sing and spat ter ing from T56B in creased in fre -
quency un til 1500 on 15 July, when a small hole formed in
the crater floor just E of T58B.
Called T58C, the hole be came a newly opened vent. It
be gan emit ting vis i ble gas puffs mixed with spat ter. At this
time the de gas sing ep i sodes from T56B ceased. T58C then
be gan strong de gas sing and squirted up in ter mit tent lava
foun tains. The foun tains soon fed a large lava stream mov -
ing to ward the S crater wall.
By 1600 on 15 July 2004 a par ox ysm at T58C was in
prog ress, with lava form ing 10- to 12-m-tall foun tains and
?flash floods? that com pletely in un dated the cen tral-east ern
crater floor (in the area be tween T56B, T58B, T37, T37B,
T45, and T57). T58C also ejected strong jets of ash and gas. 
Tur bu lent rivers of lava flow ing at more than 10 m/sec
swept to ward the crater?s S wall and its E over flow and
com pletely sur rounded T37B and T45. Flow rate from the
vent was es ti mated to peak at 10 m
3
/s.
The mo men tum of the rap idly outflowing lava car ried it
nearly 3 m up the W (up stream) side of T45 and oblit er ated
the large cave within that cone. The as so ci ated surge of lava 
poured over the E crater rim and down the flank. It flooded
over a 3 m wide swath of veg e ta tion. This trig gered a huge
cloud of s team and smoke that  re sem bled a smal l
pyroclastic flow. The smoke cloud was ac com pa nied by a
loud siz zling sound. A brush fire burned along the crater
rim over flow as ad di tional floods of lava ar rived. These
larger-than-nor mal flows lasted for lit tle more than 30 sec
and were sep a rated by pe ri ods of re pose of 5 to 6 min. Af ter 
sun set, in can des cent gas flared from the vent dur ing the re -
pose pe ri ods. Weak strombolian activity was seen in T56B.
Early on 16 July 2004 the newly formed T58C was a
cir cu lar pit ~ 2 m in di am e ter with lava slosh ing vi o lently at 
a depth of ~ 2 m. Two small sub-vents on the N and S edges 
of the pit in ter con nected with the main vent. Ac tiv ity con -
tin ued spo rad i cally at T58B and T56B with strombolian ac -
tiv ity and lava flows. On 21 July there was an ex cep tion ally 
strong erup tion of T58B with loud ex plo sions, jet ting of
ash-poor clouds, and spat ter thrown to above-av er age
heights. Ex plo sions blasted a new vent in the up per E side
of T58B. At least four oval bombs 9-12 cm in length flew
through the air, along with a great deal of lapilli and ash.
Later ex am i na tion of the bomb?s in te ri ors re vealed that they 
 Volcanism Network, Volume 30, Number 4, April 2005 Ol Doinyo Lengai    13
Date Lo ca tion Tem per a ture (?C)
28 Aug 1999 T40 lava lake 529
01 Sep 1999 Pahoehoe flow in a tube
near T40
519
01 Sep 1999 Aa flow still in mo tion at
flat ter rains (60 cm thick)
516
03 Oct 2000 Pahoehoe flow in a tube
near T49B
507
03 Oct 2000 Aa flow still in slow
mo tion at flat ter rain (25
cm thick)
496
11 Feb 2004 Pahoehoe flow in a tube
near T49G
588
12 Feb 2004 Pahoehoe flow in a tube
near T49B
579
13 Feb 2004 Aa flow not in mo tion
any more at flat ter rain (15
cm thick)
490
26 Jun 2004 Pahoehoe flow in slow
mo tion (10 cm thick) flat
ter rain
560
03 Feb 2005 Pahoehoe flow (15 cm
thick) in mo tion trav el ing
within a levee.
561
03 Feb 2005 Aa flow not in mo tion
any more at flat ter rain (15
cm thick)
520
Ta ble 3. Re peated max i mum lava tem per a tures mea sured at Ol Doinyo
Lengai dur ing 28 Au gust 1999 to 3 Feb ru ary 2005. The mea sure ments
were made em ploy ing a dig i tal ther mom e ter (TM 914C with a stab feeler
of stan dard K type). The in stru ment was used in the 0-1200?C mode, and
at least four rep li cate mea sure ments were made at any one spot.
Cal i bra tion was by the delta-T method; un cer tain ties were ? 6?C in the
0-750?C range. Cour tesy of C. Weber.
Date Lo ca tion Tem per a ture (?C)
28 Aug 1999 F1 70
28 Aug 1999 Near T49 82
03 Oct 2000 Near T49C 75
03 Oct 2000 F1 69
20 Oct 2002 The hot test cracks in the
crater floor
124
20 Oct 2002 F1 78
30 Jun 2003 F1 86
30 Jun 2003 Near T49C 76
12 Feb 2004 F1 88
26 Jun 2004 F1 78
26 Jun 2004 Near T49C 150
04 Feb 2005 F1 84
Ta ble 4. Max i mum fumarole tem per a tures mea sured at cracks in Ol
Doinyo Lengai?s crater floor over a se ries of vis its dur ing 28 Au gust 1999
to 4 Feb ru ary 2005. Col lected us ing the dig i tal ther mom e ter with
pro ce dures and pa ram e ters noted with the pre vi ous ta ble. For lo ca tions,
see map (fig ure 15). Cour tesy of C. Weber.
all had an outer zone ~ 1.5- to 2-cm thick and a distinctive
inner core.
On 23 July 2004, a slop ing ~ 4 m
2
 oval sec tion of the
crater floor im me di ately SW of the new spat ter cone T58C
be gan to steam and vi brate. Tremor in creased and ground
move ment was vis i ble, man i fested as a small sec tion of
crater floor rap idly pushed out ward and then in ward sev eral 
cen ti me ters, like a mem brane vi brat ing in time to the de gas -
sing sounds of lava in T58C just be hind it. Abruptly this
por tion of the crater floor broke out ward, and a flood of
lava en sued. T58C was ob served to grow in height through
the time when Belton left the crater on 29 July 2004.
Ob ser va tions dur ing Jan u ary and Feb ru ary 2005.
Donth re ported that dur ing his visit on 10 Jan u ary 2005,
hor ni to T49B erupted to form many ef fu sive lava flows.
For the first time, lava es caped over the north ern edge of
the crater (see figure 15).
Dur ing Weber?s crater visit, 3-7 Feb ru ary 2005, the hor -
ni to T49B ac tively emit ted lava flows that trav eled to the N. 
Pahoehoe lava flows in mo tion within small lev ees on flat
ter rain were mea sured from 520?C up to a max i mum of
561?C (ta ble 3). The fumaroles at F1 had a max i mum tem -
per a ture of 84?C, and at hor ni to T46, a max i mum of 91?C
(ta ble 4). No change in dis tance was mea sured across the
CR1, CR2, and CR3 cracks cut ting the up per crater walls.
Add ing to vis i tor safety con cerns, which in clude al ti tude
sick ness, burns, falls, and im pact from ejecta, Weber?s team 
saw a spit ting co bra close to the sum mit. An over flight by
plane on 14 Feb ru ary showed no sub se quent change, but
did give an ex cel lent view of the crater and its central
hornitos (figure 16).
A flight on 14 Feb ru ary failed to re veal sub se quent
changes. But the ef fort pro vided an ex cel lent view of the
crater and its cen tral hor ni tos (fig ure 17).
Back ground. The sym met ri cal Ol Doinyo Lengai
stratovolcano is the only vol cano known to have erupted
carbonatite tephras and lavas in his tor i cal time. The prom i -
nent vol cano, known to the Maasai as ?The Moun tain of
God,? rises abruptly above the broad plain S. of Lake Nat -
ron in the Greg ory Rift Val ley. The cone-build ing stage of
the vol cano ended about 15,000 years ago and was fol -
lowed by pe ri odic ejec tion of natrocarbonatitic and nephe -
lin ite tephra dur ing the Ho lo cene. His tor i cal erup tions have
con sisted of smaller tephra erup tions and emis sion of nu -
mer ous natrocarbonatitic lava flows on the floor of the
sum mit crater and oc ca sion ally down the up per flanks. The
depth and mor phol ogy of the north ern crater have changed
dra mat i cally dur ing the course of his tor i cal erup tions, rang -
ing from steep crater walls about 200 m deep in the
mid-20th cen tury to shal low plat forms mostly fill ing the
crater. Long-term lava ef fu sion in the sum mit crater be gin -
ning in 1983 had by the turn of the cen tury mostly filled the
north ern crater; by late 1998 lava had begun overflowing
the crater rim.
In for ma tion Con tacts: Christoph Weber, Vol cano Ex -
pe  di  t ions  In  ter  na  t ional,  Muehlweg 11,  74199
Untergruppenbach, Ger many (URL: http://www.v-e-i.de,
Email: mail@v-e-i.de); Celia Nyamweru, De part ment of
An thro pol ogy, St. Law rence Uni ver sity, Can ton, NY
13617, USA (URL: http://it.stlawu.edu/~cnya/; Email:
cnyamweru@stlawu.edu); Fred er ick Belton, De vel op men -
tal Stud ies Department, PO Box 16, Mid dle Ten nes see
State Uni ver sity, Murfreesboro, TN 37132, USA (URL:
http://www.oldoinyolengai.org, Email: oldoinyolengai@
hotmail .com);  Ber nard Donth  (Emai l:  b .donth@
saarschmiede.com).
14    Ol Doinyo Lengai Smithsonian Institution ? Bulletin of the Global
Fig ure 17. An ae rial pho to graph taken look ing to wards the WSW at the
sum mit crater of Ol Doinyo Lengai on 14 Feb ru ary 2005. The sum mit,
which lies in the up per left cor ner has a re vised el e va tion based on GPS
(see text). In ad di tion, GPS el e va tions and un cer tain ties sug gest that in
2005 the sum mit was only mar gin ally higher than the top of the tall est
hor ni to (T56B). Copy righted photo pro vided cour tesy of T. Schulmeister
and C. Weber.
Sub scrip tions: The Bul le tin of the Global Vol can ism Net work (ISSN: 1050- 4818) is avail able by sub scrip tion from the
Ameri can Geo physi cal Un ion (2000 Flor ida Ave nue NW, Wash ing ton, DC 20009, phone 202:462- 6900 or 800:966- 2481, fax
202:328- 0566, Email: service@agu.org). An nual sub scrip tion price (2005) is $26 to US ad dresses and $43 to all other
coun tries. Back is sues can be or dered through AGU Sepa rates; con tact AGU for cur rent pric ing in for ma tion. Or ders must be
pre paid; make checks pay able to AGU; VISA/MC are ac cepted. The Smith sonian does not han dle any Bul le tin or ders.
Email Ac cess: The text of the Bul le tin is dis trib uted through the VOL CANO List serv (vol cano@asu.edu) maintained by
Jona than Fink at Ari zona State Uni ver sity (subscribe via listserv@asu.edu).