Smithsonian Institution
Global Volcanism Network Bulletin v. 20, no. 9, September 1995

Soufriere Hills (Montserrat)  Phreatic eruptions continue; new lava
dome in summit crater

Soufriere Hills
Montserrat, West Indies
16.72N, 62.18W; summit elev. 915 m
All times are local (= GMT - 4 hours)
Following the formation of Vent 3 and significant ashfall on 22
August (Bulletin v. 20, no. 8), more than 6,000 residents of
southern Montserrat were evacuated to safe areas in the N part of
the island. Press sources estimated that by late August ~3,000
people had left for neighboring islands. Vent 4 opened on 27 August
and produced mainly steam emissions with some minor ash through 30
August. Although seismicity was high from 30 August through 1
September, steam and ash emissions remained low (Bulletin v. 20,
no. 8).
From 0500 on 1 September through 0500 on the 3rd, only 19 shallow
earthquakes occurred beneath the volcano. During that same period,
17 episodes of gas venting were recorded; at least six of those
episodes produced some ash, and the two events on 2 September each
decayed into a long-period signal of ~10 minutes duration. Venting
continued to enlarge vents 2 and 3, but emissions from Vent 4
remained low. A helicopter observation flight on the afternoon of
2 September was in progress when an emission episode began at 1606
with increased steaming that developed rapidly into a small steam-
and-ash plume. The emission occurred from a narrow part of the main
group of vents that extend SE from Vent 1. Mud on the floor of the
vent was expelled during the episode, forming a small mudflow that
moved down the S side of the moat and over the area of Vent 2. A
gas-and-ash emission at 1912 on 2 September, similar in size and
duration to emissions in recent days, was widely observed because
of clear conditions. Lightning associated with this activity lasted
~1.5 hours, and an SO2 odor was detected. Installation of a
hardened EDM (electronic distance meter) station in the Tar River
area was completed on 2 September.
During 3-4 September there were four gas-venting episodes, twelve
volcano-tectonic (VT) earthquakes, and four long-period
earthquakes. Aerial observations on the morning of 3 September
revealed that the area around the S end of the main group of vents
had been enlarged. The moat pond in the NW corner was still
present, and fragmental material had collapsed into Vent 1.
Afternoon observations showed no new mudflows, and the S moat
appeared dry.
On the afternoon of 3 September, scientists at the volcano
observatory completed an assessment of the current volcanism since
21 August and prospects for future activity. The rate of eruption
signals increased slightly after 21 August, but the size of the
eruptions did not. No change in the style of eruptions was
anticipated, but areas downwind could be subject to ashfall and
temporary darkness. Eruptions were thought likely to be
concentrated along the linear vent chain on the W side of Castle
Peak dome. The amount of shallow seismicity decreased below that
prior to 21 August. SO2 flux remained near detection limits since
21 August. The rate of long-period seismic events showed no clear
pattern, although a slight decrease may have occurred. Initial EDM
results indicated no movement of the SE flank of Castle Peak dome
or at a site in Upper Gages. Electronic tiltmeters have detected no
large-scale deformation since they stabilized on 5 August. Ash
samples analyzed through 27 August revealed no juvenile material.
The scientists concluded the following: ". . . eruptions to date
have been entirely phreatic, with no direct evidence of magmatic
involvement. So long as this behavior pattern persists, it only
constitutes a significant hazard to areas within 1.5 km of Castle
Peak dome and the areas S of White's Bottom ghaut. All ghauts
[ephemeral watercourses] that  originate on the flanks of the
Soufriere Hills volcano are subject to flooding and should be
avoided."  Based on this advice, the government approved re-
occupation of the areas immediately S of the Belham Valley River
from which residents were evacuated on 23 August. All other
residents from areas closer to the crater, evacuated since 21
August, were required to stay in the northern third of the island.
Controlled entry restrictions were relaxed in most areas to allow
residents to prepare for an approaching hurricane. Following
passage of the hurricane, on 6 September the remaining evacuation
orders were lifted.
Activity during 4-8 September was consistent at a low and generally
declining level. At about 1530 on 8 September there was a
significant steam explosion. Two hours later, at about 1730, two
large ash eruptions produced a vertical plume that formed a
mushroom cloud, which drifted to St. Peters (~30 km NNW) and to the
N. Soufriere Hills continued to have intermittent swarms of
earthquakes from the summit and nearby areas, including three
events felt in Woodlands on 11 September. Occasional steam
eruptions produced falls of fine ash in communities around the
volcano, and morphological changes were continuing in the summit
area. These developments suggested to volcanologists that magma was
close to the surface under the volcano and that a magmatic eruption
was still a possibility.
Two weeks later, on 25 September, a lava dome began growing in the
W part of the moat near the linear chain of vents. An explosion
between 1100 and 1200 on 27 September caused ashfall on the S part
of the island, with minor ashfall also reported in the St. Georges
area. Minor explosive activity continued through the end of
Information Contacts: Soufriere Hills Volcano Observatory, c/o
Chief Minister's Office, P.O. Box 292, Plymouth, Montserrat;
Seismic Research Unit, University of the West Indies, St.
Augustine, Trinidad; United Nations Department of Humanitarian
Affairs, Palais des Nations, 1211 Geneva 10, Switzerland;
Associated Press; Caribbean News Agency (CANA), Barbados.