(i) Volcanic Eruption. A Volcanic Eruption involves the escape at the surface of molten rock (magma) which has risen from a zone of melting several tens of
kilo¬metres below the surface. The magma generally contains a much larger volume of gas than liquid and the gas, before it emerges at the surface, is under very
high pressure. The more gas present the more violently explosive will be the eruption.

(ii)  Area Affected. The area severely affected by a large volcanic eruption will normally be no more than 10 km is radius around the active crater or vent. With¬in
this area certain parts can be specified (especially the deeper valleys originating at or close to the crater) which are particularly liable to suffer damage as will be
explained in detail later.

iii)  Glowing Avalanche is a type of eruption involving the emission of dense clouds of solids (ash and cinder or lava blocks) suspended in very hot gas. Because of
their high density, glowing avalanches flow downward away from the active vent in much the same way as milk boils over the side of a saucepan. They follow the
easiest route downhill i.e. the major valleys down the flanks of the volcano.

(iv)  Ash fall consists of material blasted vertically upward from the vent in a high pressure gas column. This rises to many thousands of metres in height and the
fragments fall as showers. The larger blocks tend to be blown less high and therefore fall closer to the active vent, whilst the fine dust may be carried downwind for
hundreds of kilometres.

(v)  Mudflow is composed of solids (ranging in size from dust to blocks of several metres in diameter) suspended in water. A mudflow is caused either by the
ejection of a crater lake, or by temporary damming of a river, or by torrential rain (which often accompanies large eruptions) which washes newly fallen ash off the
upper slopes of the volcano. Mudflows, like glowing avalanches, follow the easiest route downhill. Their temperature will not exceed that of boiling water (100 c) i.
e. they will be considerable cooler than glowing avalanches.


(i) The size of an eruption depends upon the volume of magma within reach of the surface. This volume in large eruptions may amount to several cubic kilometres.

(ii) The explosiveness of an eruption depends upon the abundance of gas which may range from an insignificant amount (as in the very mild eruption in St Vincent in
1971) to enormously more than the volume of magma (as in the violent eruptions of 1902 in St Vincent and Martinique). From historic accounts and from
reconstruc¬tions of prehistoric activity it is clear that West Indian eruptions are among the world’s largest and most violent.

(iii) The frequency of West Indian eruptions fortunately is relatively low. At a single volcano they recur at intervals ranging from several tens to several thousands of
years. Those relatively long periods of dormancy are compensated by the extreme violence of the larger eruptions.

(iv) Physical characteristics of a typical large eruption

In a typical explosive eruption there are two main directions in which material is ejected from the vent. Firstly there is a vertical eruption column which during the
strongest explosions may rise 10   20 kilometres high carrying large (e.g. football sized) blocks or clots of semi consolidated lava one or two kilometres high and
finer ash to the full height of the eruption column; secondly, there are glowing avalanches which travel downward away from the active vent and which follow the
easiest routes downhill on the flanks of the volcano. These avalanches consist of expanding dense clouds of semi molten lava fragments and ash in hot gas. They
travel very quickly: speeds of over 100 kph (60 mph were recorded during the 1902 eruption of Mount Pelee in Martinique. In moderate sized eruptions these
glowing avalanches may sweep down only one or two of the larger valleys which begin at low points on the rim of the active crater. If, however sufficiently large
volumes of glowing avalanche material are erupted in a short space of time the avalanches will radiate in all directions down the flanks of the volcano.

It is the flowing avalanches which have caused the major loss of life and destruction of property by West Indian volcanoes in historic time. Human casualties have
been caused either by dismemberment (in the same way as from a bomb explosion) as was common in the city of St. Pierre in Martinique in 1902 where most
buildings disintegrated and were razed to foundation level by the blast of the avalanche.  Alternatively death may be caused by asphyxia¬tion as appears to have been
the predominant case in St Vincent in 1902 where casualties were attributed to the inhaling of very hot dust laden gas.

In summary, the main features of glowing avalanches which make them so lethal are that:

(a) They travel too rapidly to allow people to escape their paths between the time of their emergence from the summit and their arrival at the foot of the volcano;

(b) They are capable of striking with sufficient physical force to destroy ordinary wooden or masonry buildings or at least to set them on fire. Thus there is no
simple alternative but to evacuate threatened areas before the volcano emits glowing avalanches.

A third type of eruptive phenomenon which is smaller in numerous respects to glowing avalanche is the mudflow. This can originate by the ejection of a crater lake
(as in St Vincent in 1902) or it can be generated during or after the eruption of ash fall or glowing avalanche if heavy rain washes large quantities of watery mud
down the valleys on the flanks of the volcano. This mud may be very hot, i.e. at temperatures near boiling. Like the smaller glowing avalanches mudflows travel
down the main valleys on the volcano. However the extent of the areas they affect and also their temperature will appreciably less than those of glowing avalanches.

The time sequences in which the above types of activity occur varies considerably from one eruption to another even at the same volcano. The duration of a single
period of eruption may be as long as several years and may include periods of several weeks or months in which no significant emission occurs.


There is no way of reliably predicting the day or the hour of individual glowing avalanches. Once a West Indian volcano has entered a phase of strongly explosive
activity, however it is probable that glowing avalanches will be erupted and that there may be several such avalanches large enough to pose a threat to life and
property on the lower flanks of the volcano, at irregular intervals over the ensuing months. The sequence of events of events which lead up to the emission of
glowing avalanches in Martinique and St Vincent in 1902 consisted of several months of local earthquakes (i.e. tremors originating at shallow depth beneath the
volcano) and several days of increasingly large vertical explosions from which fine ash and pumice lumps showered the lower flanks of the volcano.

Thus for all practical purposes, the hazard to the local population becomes serious as soon as strongly explosive eruptions of any kind begin. In one or two historical
cases at volcanoes of similar type of these in the West Indies e.g. the 1929 eruption of Komagatake Volcano in Japan, glowing avalanche emission has begun within
a few hours of the first ex plosive activity. It is therefore desirable that the onset of explosive activity itself should be predicted in order to allow adequate time for an
orderly evacuation from the most threatened areas.


The damage caused in the direct path of a glowing avalanche will be virtually total: buildings will be razed and set on fire or they may be buried under hot volcanic
ash many metres will be equally decimated. On the flanks of a glowing avalanche e.g. on the ridges adjacent to the main valley down which the avalanche travels,
buildings will remain intact but their human occupants may be asphyxiated by the hot dust laden gas unless all doors and windows are well sealed.

Any human being exposed (i.e. outdoors) even to the fringes of the hot gas cloud will, if not asphyxiated probably suffer, severe burns both externally and to the
respiratory system. Thus the only effective protection from glowing avalanches consists in avoiding the areas where they are likely to occur.

The damage from ash fall and mudflow will in most cases be relatively small compared with that from glowing avalanches. With ash fall the occasional casualty can
be expected from larger fragments falling at high velocity. Damage to houses will consist mainly of the caving in of roofs under excessive loads of ash whilst
occasional fires may be started by red-hot fragments.

Mudflows represent a hazard to anyone in the valleys which they descend. Their temperature and/or depth may be sufficient to make it impossible to wade through
them and thus they may block escape routes around the flanks of a volcano, as was the case in the opening stage of the 1902 eruption in St Vincent.


The earliest warning of major volcanic eruptions will hope¬fully come weeks or months before the destructive climax e.g. in the form of small but frequent
earthquakes at shallow depth beneath the dormant volcano. Most of the earthquakes will be too small to be felt, but there will be many large enough to be detected
by sensitive seismographs. For this reason, at least one seismograph is kept in continu¬ous operation in each volcanic island served by the UWI Seismic Research
Unit. There may also be an increase in steam flow and a temperature at natural streams around the volcano before an eruption begins.

As soon as any abnormal activity of this kind is detected, additional monitoring equipment will be installed in the island of the Seismic Research Unit. The rate of
progress of events towards a possible volcanic eruption can then be followed, although this will probably not approach the relatively steady rate at which, for
example, a hurricane approaches. The period over which activity builds up to climax may vary from a few days to many months end may involve intervals of
accelerating and decelerating activity. However once events are being monitored as fully as possible there is a good chance that at least a few day's warnings can be
given of any destructive climax.


(i) Earthquake occurring around the volcano in nearby districts before an eruption.
(ii)  Steam issuing from the mouth of the volcano.
(iii) Floods of water and mud appearing in the rivers making them impassable.
(iv) First steam explosion after twelve hours of the first three signs.
(v) A few hours after the floods "pyroclastic flows" will cover much of the mountain.


Reports of unusual activities will generally come from:

(a) Residents of the danger areas;
(b) The Seismic Research Unit in Trinidad;
(c) The Nevis Disaster Coordinator.

These reports should be carefully examines and treated with great urgency.
Return to Home Page
Return to Previous Page