Go to 'figures/overhead' section.
Soufriere stratovolcano, St. Vincent
Reference: Francis, Peter, Volcanoes, a Planetary Perspective, 1993, Clarendon Press.
Three primary environments: mid ocean ridges (basalt), plumes (basalt; Hawaii), arcs (andesite-rhyolite).
The most common kinds of volcanic rock are:
Basalt Basaltic Andesite Andesite Dacite Rhyolite SiO2 < 52 52-55 55 - 63 63-68 >68Correlative variation with increasing viscosity of the magma:
from shield to central (cone, stratovolcano) morphology of the volcano
increasing volatile content (90% H2O) of magma
increasing vesicularity of lavas
increasing explosivity, and decreasing grain size of pyroclastic ash material
increasing danger to local populace
increasing economic importance, natural fertilizer, base metals.
Types of Volcanic Activity
activity is the mildest form of volcanic eruption. Lavas are at high temperature
and have low viscosity. Gas and liquid phases separate easily. Lava
may be sprayed high into the air in the form of fire fountains but does
not fragment easily. If the eruption rate is low the basalt sprayed
falls back to the ground as solid cindery scoria
material. At higher rates the volcanic material falls back as aggregated
plastic material to form spatter cones. At
very high rates the lava fails to cool at all and on falling back
flows away as clastogenic lava.
Strombolian activity consists of intermittent, discrete explosive bursts, ejecting pyroclasts a few tens or hundreds of metres into the air. Each burst lasts only a few seconds, and pauses between bursts last twenty minutes of more. The lava is more viscous than Hawaiian lava and more scoria deposits are produced. Little fine grained material is produced but bombs may be prominent.
Vulcanian eruptions (Vulcano) have columns that rise much higher than strombolian eruptions, sometimes reaching 10- 20 kilometres. The erupted material is composed of fine fragments of older material. They are noisy and messy but their effects are not widespread. Deposits are finer grained than strombolian tephra. A component of ground-water interacting with the magma may be involved with some violent vulcanian eruptions. Large bombs are prominent and pyroclastic flows of nuées ardente type (Mt Pelée, Mt Lamington, Mt St Helens) are by-products of vulcanian (Peléan) explosions. Volcanism is commonly andesitic and associated with growing lava domes. Peléan volcanic activity involves the generation of nuées ardentes or glowing avalanches (ash flows) of the kind that destroyed St. Pierre, Martinique in 1902. Nuées are blasted sideways and are composed of large fragments mixed with fine volcanic dust which avalanche down slope under gravity and roll over the ground at high speed.
Vesuvian or sub-Plinian activity (Vesuvius, Italy; Sunset crater, Arizona) is a step up from vulcanian activity, and the eruption columns give rise to extensive sheets of tephra deposits. The tephra may include new magmatic material rather than shattered bits of old rock.
Plinian eruptions (Vesuvius AD 79 destroyed Pompeii) are driven by powerful thermally convecting eruption columns that rise up into the stratosphere as high as 45 km. Generally plinian eruptions involve silicic magma, although the Tarawera eruption of New Zealand was basaltic. The silicic Taupo event of New Zealand, which covered the whole of the North Island with tephra such that the tephra 100 kilometres away from the vent was 25 cm thick, would be classified as ultraplinian. Plinian eruptions build volcanic sheets rather than steep sided cones, and the vent may be a negative topographic feature (caldera). The eruptive material is usually cold by the time it returns to the ground, and plinian deposits are therefore rarely welded.
Where water is involved in the volcanic event, the eruptions are said to be phreatomagmatic. In such eruptions, steam fragments the magma as it expands explosively, producing highly fragmented fine grained ash (.24 mm versus 1.6 mm). Accretionary lapilli (fragments mantled by fine ash) are also typical. Phreatomagmatic reactions can be a prolonged self-sustaining phenomena. Surtseyan eruptions are wet equivalents of basaltic strombolian events, and may evolve into strombolian type activity, whereas phreatoplinian eruptions are the wet equivalent of silicic sub-plinian and plinian events. Typically large blocks are present in the fine ash near the vent. No phreatoplinian eruptions have been observed in progress, but the Oruanui ash of the North Island of New Zealand is thought to represent such an event on the basis of its extreme fine grain size and wide dispersion.
An explosive volcanic eruption involves three stages:
1) fragmentation of the lava by bubble growth (vesiculation by decompression or first stage boiling, or crystallization or second boiling);
2) blasting of the fragmented mass through the vent to the surface when the pressure within the magma exceeds the strength of the surrounding rock causing expansion of the growing bubbles to 4/5ths of the volume fraction of the magma;
3) ascent of the eruption column.
Pyroclastic flows and surges
It is generally recognized
that there are two major types of flows composed of a mixture of hot pyroclasts
1) Pyroclastic flows are high particle concentration solid-gas mixtures. Flow is laminar, with clast support enhanced by fluidization of the mixture.
2) Pyroclastic surges are low particle concentration mixtures which flow in a turbulent regime where particle support is accomplished by fluid turbulence. They have low flow density (more gas) and moderate clast density.
flows are merely volcanic avalanches propelled by gravity. The flows
are not composed of liquids, but of gas fluidized
Flows composed of vesiculated low density pumice fragments, dust and gas are called ignimbrites. Low density pumice clasts tend to float up within such flows whereas lithic clasts tend to sink. They have high flow density (more fragments, less gas) compared to pyroclastic surges, and a low clast density. Pyroclastic flows that contain unvesiculated, dense lava clasts are called nuées ardentes (Peléan volcanic activity) and form block and ash deposits. They are coarser and have a high clast density as well as a high flow density.
can be loose sandy ash or solid glassy rock similar to a lava flow. They
tend to be restricted to topographic lows but can sweep over irregular
topography. The base of the flow, layer 1, is a pyroclastic surge deposit,
rich in crystals, fine grained, and sometimes exhibiting cross bedding.
Layer 2a is rarely more than a metre thick, shows reverse grading, and
grades into layer 2b, a poorly sorted, reverse graded mixture of
pumice clasts and dusty ash. Layer 3 is fine ash winnowed out of layer
2 by the escaping, fluidizing gas. It is often termed co-ignimbrite
ash. Lag breccias may be present near the
source area. Ignimbrites formed from a high eruption column will
have longer time to cool and will be less likely to weld. The most densely
welded ignimbrites consist of nothing but glass and crystals. More
commonly only the pumice clasts are glassy (fiamme).
On a microscopic scale tiny glass shards are flattened and molded
over one another (eutaxitic texture). Welded
ignimbrites may undergo rheomorphic flow and resemble lavas.
Pyroclastic flows can form by collapse of a growing dome or lava flow, or collapse of an eruption column. Eruption columns are turbulent jets in which convection of hot magmatic gases is aided by heat transfer between small, hot pyroclasts and entrained air. The mass eruption rate is controlled by the vent radius and the volatile content of the magma. If the eruption column is stable it forms a convecting plinian eruption with the formation of a tephra fall deposit; however, if as a consequence of increasing vent radius it becomes unstable it will collapse to form a pyroclastic flow. An analogy would be a pan of rice boiling over. Where there are no associated air fall deposits, the flows are likely formed directly from the crater without formation of a column.
The eruption of the 1980 Mt. St. Helens pyroclastic flow ended a long period of relatively minor volcanic activity all around the world. Although there had been a variety of locally important eruptions before 1980, the last large eruptions were in 1932 (Cerro Azul, Chile) and especially in 1912 (Valley of Ten Thousand Smokes, Alaska) and 1902 (Mt. Pelée, Martinique, Soufriere, St. Vincent and Santa Maria, Guatemala). Since 1980 there have large and tragic eruptions in 1982 (El Chichon, Mexico and Galunggung, Indonesia), 1985 (Ruiz, Colombia) and two more in 1991 (Cerro Hudson, Chile and Pinatubo, Philippines). In the case of Mount St. Helens volcanologists had never before observed such a giant doming on the side of a volcano, and the resulting collapse and sideways explosion was completely unanticipated. The experience monitoring MSH allowed all of the later eruptions there to be forecast accurately, and permitted scientists to successfully warn residents near Pinatubo in the Phillipines of the impending major eruption. The Pinatubo eruption was one of the largest this century, but the loss of life was minimized due to the management of the crisis by volcanologists; it was their finest hour!
Block and Ash Pyroclastic Flows (Nuées ardentes)
There are three types of
these clast-rich pyroclastic flows:
1) Merapi, gravitational collapse of lava flows and domes;
2) Peléan, explosive events on growing lava domes;
and 3) Soufriere, eruption column collapse.
Merapi-type nuees are essentially hot avalanches. Silicic lavas are too viscous to flow and therefore grow lava domes which from time to time become unstable and crash downslope under the influence of gravity.
Peléan lava domes are represented by the 1902 and 1929 Mont Pelée eruptions. Some Peléan nuées were initiated by explosive events, vertically or laterally directed. In a Peléen or Mount Lamington nuée the core is composed of a dense avalanche of fast-moving incandescent debris, a ground hugging pyroclastic torrent of everything from fine dust to lava boulders more than a meter in diameter. A lower density component, the lateral equivalent of the avalanche, is a pyroclastic surge (see below). The cloud itself is similar to the fine ash component winnowed from a pumice flow and gives rise to thin deposits of fine ash equivalent to those of ignimbrite layer 3.
In the Mayon Soufriere-type eruption of 1968 incandescent blocks were hurled to 600 meters and the eruption column rose about 10 km. Nuées involved avalanching of material which had initially had been ejected vertically upwards from the crater but which had collapsed back on itself. In block and ash flows reverse grading is typical, and is the result of mechanical segregation. Blocks exhibit prismatic jointing that formed after the deposit had been emplaced.
Surges are low density
pyroclastic flows. They are deflated and have less momentum than pyroclastic
flows. Base surges
are associated with hydrovolcanic explosions and develop from the collapse
of overloaded eruption columns. Best known in small basaltic eruptions
(although Santorini is rhyodacite). They are turbulent not laminar like
pyroclastic flows, and are wet and sticky.
Ground surges form directly from the crater, or by collapse of the outer part of an eruptive column, or as a flow front in the turbulent head of a pyroclastic flow. In this case they would be found at the base of an ignimbrite flow. They need not however form at the same time as a pyroclastic flow.
Ash cloud surges may form when the condition of the eruption column may be close to the boundary between the plinian convection regime and the pyroclastic flow collapse regime. They are similar to ground surges, finely laminated, sometimes cross bedded and rich in crystals and lithics, but are found within or on top of ignimbrites or nuees ardentes, and form by elutriation of material into the turbulent overriding ash cloud. They occur above the pyroclastic flow or as its lateral equivalent. Ash clouds can become detached and even resegregate in the bottoms of adjacent valleys. Ash Clouds extend over a wider area than their associated pyroclastic flows. (Smith, A.L., et al. , 1981. Pyroclastic flows and surges: examples from the lesser Antilles. in Self, S. and Sparks, R.S.J., eds., Tephra Studies, D. Reidel Pub. Co. , v., p. 421-425.)
Intrusive Intrusion breccias
Pyroclastic Subterranean Explosion breccias
deposit Strombolian |Surtseyan
>64 mm bombs Sub-Plinian |Surt. or Phreato.
>2 mm lapilli Plinian |Phreatoplinian
<2 mm ash Ultra-Plinian |Phreatoplinian
Pyroclastic Ignimbrite: pumice; ash
flow deposit Scoria and ash
Vesicular andesite and ash
Block and ash
Pyroclastic Base surge
surge deposit Ground surge
Ash cloud surge
Submarine Pillow breccia and hyalotuffs
Subaqueous pyroclastic flow
Epiclastic Subaerial and subaqueous volcanic sediments and lahars
Sigurdsson, H. and Carey, S. 1991. Caribbean volcanoes: a field guide. Field Trip B1: Guidebook. GAC Ann. Meet., 101p.
The St. Pierre event of the island
of Martinique was a low concentration turbulent surge
similar to the overpressured blast surge at Mount St. Helens. Turbulent
flows show normal grading and presence of cross stratification indicative
of traction related deposition from low concentration flows. The
turbulent low density flow was stratified in terms of both particle
concentration and size. Transport of particles was by turbulent flow suspension
and traction processes such as saltation and rolling. As the flow
moves down slope, gravity segregation leads to an increase in particle
size and concentration of material at the base. Transport of large
particles by suspension to distances as far as St. Pierre is unlikely in
a turbulent flow. The largest particles are more likely carried by
load. If flow densities are adjusted for the presence of fine ash,
then the settling velocities of large particles in the flow is substantially
reduced. This allows for turbulent suspension of particles to take place
at lower flow velocities (Lajoie et al, 1989, J. Volc Geotherm Res.
Soufriere on the Caribbean island of St. Vincent is a stratovolcano. Pyroclastic flows here are basaltic and the lavas basaltic andesite and andesite. The explosive events are phreatomagmatic, and are controlled by crater morphology and the presence of crater lakes.
Soufriere stratovolcano, St Vincent
Effects of the 1902 Soufriere pyroclastic flow, St Vincent - 1
Effects of the 1902 Soufriere pyroclastic flow, St Vincent - 2
May 8,1902, Eruption of Mont Pelée
The destruction of St Pierre
The destruction of St. Pierre and the death of its 30,000 inhabitants on the morning of May 8, 1902, remains as one of the worst volcanic catastrophies in historic times. Only the eruption of Krakatau (1883) and Tambora (1815) in Indonesia have surpassed it in terms of the loss of human life (Tilling 1989). As a result, the deposits of the eruption have received considerable attention in an attempt to understand the nature of this destructive phenomenon. The pioneering observations of Lacroix (1904) and Perret (1937) on the eruptions of Mont Pelee mark the beginning of the modern study of pyrociastic flows and surges. Since that time, considerable progress has been made in understanding the processes by which pyroclastic flows are generated and the physical nature of the flows themselves. Despite the significant advances that have taken place, the interpretation of the May 8, 1902, eruption and its deposits remains controversial. There is no question that St. Pierre was destroyed by some type of flow that was a mixture of hot gases and volcanic particles, but the origin, flow regime and path which it took from the summit are still open to question. During the field excursion, we will examine the deposits of the May 8, 1902, eruption in detail. In the next section, a review of the eruption and the current models for the origin of the flowage phenomena are presented as a framework for examination of the deposits.
Precursory Activity and Governmental Response to the Disaster
Prior to the devastating eruption of 1902 there had
been some historic activity from Mont Pelée. In January of 1792,
some minor activity was reported and a small eruption occurred in
August 1851, which lasted until October that year. This eruption
caused minor ash fall on St. Pierre, and left a steaming crater lake
(I'Etang) of about 100 m diameter in the summit of the volcano. A four-man
scientific commission was set up after the 1851 activity. The tone
of their report was one of reassurance, as they emphasized that Mont Pelées
activity only formed a "picturesque decoration" to the city of St.
Pierre. The first signs of renewed activity was
increased fumarolic emission at Mont Pelée in 1889. In February
1902, sulfurous gases were being emitted in large volumes and were
noted especially in Le Precheur and in St. Pierre, where the fumes killed
birds and tarnished silver. Local earthquakes were felt in
Le Precheur on April 22, 1902, and steam was seen rising from the volcano
on April 23, 1902. Upper and lower Etangs were boiling and venting
much steam. Explosive activity was first noted on the morning of 25 April,
when a great noise was heard, together with rumblings, and an ash
cloud rose over the volcano, with fallout of fine ashes over the town of
Le Precheur. The first ballot in the elections for a representative for
the legislature was on April 27th. Sugarmill owner Fernand Clerc
gained a majority of 348 votes over his opponent Louis Percin, but
an absolute majority was not gained, and thus the elections were re-scheduled
for Sunday, May llth. The political preoccupation contributed
to the Government's decision to discourage the population from abandoning
St. Pierre before the elections. During the next few days,
rumblings were frequent from the volcano, ash fall continued on Le Precheur,
and the Rivière Blanche was in flood.
In the morning of 2 May, rumblings increased, a glare was seen over the volcano, and some explosions continued, with further ash fall over Le Precheur village, which was now covered with an ash layer several centimetres thick. Very fine and light grey ash fell also on St. Pierre. That day, the newspaper Les Colonies announced a sight-seeing excursion to Mont Pelée to take place the following Sunday, 4May. However the exodus had begun from Le Precheur and its inhabitants fled into St. Pierre. Noises and ash fall were nearly continuous from then on. Ash fall caused crop failure, starvation of livestock, and people from country began to flock into town. Near midnight on 3 May, a very loud explosion occurred, with incandescence seen above the volcano, and the explosion was accompanied by heavier ash fall, which even extended to Fort-de-France, 30 km to the south. At Le Precheur, the remaining panic-stricken inhabitants rushed to the church and received holy communion from the priest. St. Pierre was covered with very fine grey ash, and all schools and shops were closed. Vegetables and other food was getting scarce in the city. The proper management of a volcanic crisis is a delicate balance between caution and acceptable risk. Caution must be taken to ensure the safety of the population. An acceptable risk is the occupation of a region until the hazard is imminent. History shows that it is not acceptable to evacuate volcanic regions for months or years, even when some risk is present, but hazard is not imminent. In St. Pierre, the response of the Government to the 1902 volcanic crisis was to discourage evacuation of the city, even though danger was clear, the population was in a state of panic and fatalities had already occurred nearby. It is generally believed that the Government's intransigence was caused by their determination to keep the population in St. Pierre until the critical elections sceduled for 11 May.
"A leading authority" was quoted in the newspaper Les Colonies, stating that there is no danger to St. Pierre from an eruption. The editor Andreus Hurard may have published this out of political expediency, to prevent unrest before the elections scheduled for 11 May. The newspaper supported the colonial and reactionary Progressive Party, which wished to maintain white supremacy over the island. The party had two elected deputies and one senator from Martinique to the Assembly in Paris. In 1899, however, the negro Amedee Knight won a surprise victory on behalf of the Radical Party, and he had hopes of winning all the political seats in Martinique for his party in the upcoming 1902 elections. Governor Louis M. Mouttet was an open supporter of the Progressive Party, and it has been speculated that the Government exerted pressure on the newspaper to dismiss the possible threat from the volcano. On the morning of May 3, the Governor arrived in St. Pierre to study the situation, and to confer with Mayor Fouche.
A Governor's Commission of Inquiry was set up to study the activity, and they were to publish their findings on 7 May. The Commission included Gaston J.M.T. Landes, professor of natural sciences at the Lycee of St. Pierre. At this stage, the threatening volcanic activity was considered as "a grandiose spectacle" that presented no threat.
On 4 May the ash fall had ceased, but the Rivière Blanche had completely dried up. That evening the activity was renewed, which caused evacuation of Fonds Corré near St. Pierre. During the night the wind changed direction, resulting in ash fall upon Macouba, east of the volcano. About noon on May 5, the Rivière Blanche flooded suddenly, and a hot mudfiow overwhelmed the Usine Guerin, a sugarmill on the banks of the river, which was buried under about 3 m of mud. Twenty-three workers were killed, including the owner, and these were the first victims of the volcano. The source of water may have been Etang Sec, the crater lake of Mont Pelée. The flood of mud and water caused a tidal wave in the ocean, and the sea drew back 20 to 30 m away from the St. Pierre waterfront, and then suddenly broke upon the shore and flooded the low- lying part of the city as a small tsunami. Activity continued on the 6th, with ash fall and flooding in the Rivière Blanche. In the afternoon of 6 May, the telegraph cable between St. Pierre and St. Lucia broke, probably because of submarine extension of the mudfiows. On the morning of May 7, a witness reported seeing a great cloud leave the summit and descend part way down the flanks of the volcano toward Fond Corré. Another one followed shortly thereafter and travelled in the same direction. These were most likely weak nuee ardentes or hot rock avalanches of older debris (Chretien and Bousse 1989). Nevertheless, an official communique stated that "the intensity of the eruption is decidedly declining". The height of the ash column above the volcano had decreased also, and steaming mudfiows in the Rivière Blanche no longer reached the ocean.
Mont Pelée remained relatively quiet and many tourists headed for the crater. The five-man Commission of Inquiry reported that the city was not threatened, and in a telegram, M. Landes stated that "in my opinion, our Montagne Pelée does not endanger the city of St. Pierre more than VesuviusendangersNapies". Yet, that afternoon,a panicstarted to spread among the population as explosions resumed. Now the Roxelane River, running through the center of the city, became flooded with muddy water. Shops were closed, and the unruly mob was calling out for food. Mayor Fouche reported to the Governor that he feared a riot, and asked for police and military reinforcement to deal with the crowds.
That afternoon, the Governor returned to St. Pierre with his wife, and his arrival in the city somewhat reassured the population. During the following night, ceaseless explosions kept the population awake and in near-panic, with incandescence and columns of ash rising from the crater, until there was a lull about 4 am. Many people were getting ready to leave at daybreak.
Eruptive activity intensified on the morning of May 8 with the generation of several black clouds from the crater, beginning at 6 am. The clouds were vigorously ejected at an angle of 601-801 across the prevailing easterly winds toward the south. As the clouds spread out, St. Pierre was plunged into darkness at 6:30 am, but no ash fall was reported. The 8th of May was Ascension Day, and the Angelus bells were tolling, when suddenly a violent explosion occurred in the volcano, and a great black cloud descended from Mont Pelée. Explosions and a bright flash of light were associated with the opening phase and a shock wave propagated through the atmosphere. The cloud rapidly overwhelmed the city at 7:50 am, as shown by the broken clock on the Military Hospital in St. Pierre. The glowing cloud, or nuee ardente, spread from the crater and down the western and southwest slopes of the volcano, fanning out over the land and spread out to sea. The burning hot avalanche engulfed the city, broke down most walls and stripped roofs off buildings and set them on fire. The deadly cloud caused complete destruction in an area of about 58 kM2 west and southwest of the volcano. Within the city, there perished Governor Mouttet and his wife, most of the Commision of Inquiry, the Mayor and virtually all its residents plus the numerous refugees from Le Precheur and other regions on the slopes of the volcano. Amongst the victims were the passengers and crew of eighteen ships in the harbor of St. Pierre, as the vessels were either overturned by the black cloud or set on fire. The only vessel to escape was the English steamer Roddam, which managed to limp to St. Lucia where the captain and several crew members died from their burns.
When rescuers moved into the destroyed city, there was hardly a living soul to be found, but horribly burnt corpses were scattered in the rubble and on the streets. There were only two survivors in the city. One was the shoemaker Leon Compere, who lived near Morne Abel and was able to run from St. Pierre on the road toward Saint Denis. On Sunday, May 11, some visitors from Morne Rouge wandered through the ruins near the end of Rue Victor Hugo, where they heard faint moans from the direction of the prison ruins. Here, they found one prisoner alive in the dungeon. He was Auguste Cyparis, a native of Le Precheur, who had been sentenced to prison for assault and battery, and committed to eight days in the dungeon for running away to a festival at Le Precheur a few days before the eruption. Cyparis was badly burnt, but quickly recovered and became a celebrity and an international attraction in the Barnum and Bailey Circus as the "Prisoner of St. Pierre". On May 20, there was another violent eruption of Montagne Pelée, and a glowing avalanche invaded St. Pierre again, finishing off the destruction of those buildings left standing. This event was probably more intense than the 8 May eruption. Activity occurred again on 26 May, 6 June and 9 July. A spine also began to grow from the crater of Mont Pelde during the month of July. In mid August, activity was resumed, and an incandescent dome was seen in the summit region. There were numerous smaller explosions. On 30 August, about 1:50 pm the activity reached a climax, producing glowing avalanches thatwere much more extensive than in the 8 May event, asthe zone of destruction was in excess of 114 kM2. This time, the villages south and east of the volcano were badly affected. Thus, Morne Rouge was completely destroyed, as well as parts of Ajoupa Bouillon and Basse-Pointe, with a loss of over one thousand lives. Activity gradually subsided and ceased in late September 1902.
In November that year, the spine grew about 230 m in 20 days, averaging more than 10 m/day. The spine continued to rise but crumbling and disintegration kept pace with growth. Thus, in February 1903, the spine was reduced in height by about 151 m. The greatest height of the spine above the former summit of Mont Pelée was about 340 m on May 30,1903. It is estimated that a total of 850 m long column of rock was extruded to form the spine.
After a quarter century of dormancy, Mont Pelée resumed activity in 1929, which lasted until 1932. The eruption was mainly in the form of a dome extrusion, producing dome collapse and numerous minor glowing avalanches that descended the Rivière Blanche.
Soufriere volcano, St Vincent
Effects of the 1902 Soufriere pyroclastic flow, St Vincent - 1
Effects of the 1902 Soufriere pyroclastic flow, St Vincent - 2
Return to beginning.
Click here to return to course outline.
Return to course