Montserrat Volcano Observatory, Montserrat, West Indies

Scientific Report 35
21 September 1996


The reporting period began with activity at similar levels as obtained during the previous week. Dome growth continued with VT swarms and rockfalls being the most common seismic events. At about mid-week a period of near-continuous rockfall and pyroclastic flow activity began which, after a brief period of calm, went into an abrupt, small explosive eruption. This style of activity had previously not been observed at the volcano. A large elongated u-shaped scar feature was created by the explosive event on the eastern flank of the dome and extensive damage was caused to structures in Long Ground village. After the period of heightened seismic activity during the early parts of the week, the period ended in relative quiet with seismicity being at a low level and rockfall and pyroclastic flow activity much scaled down.

Visual observations

Visibility was generally poor for most of the reporting period although there were some periods when brief glimpses were obtained of the dome. Rockfall and pyroclastic flow activity during the early part of the week was concentrated on the eastern flank of the volcano. Some of the rockfalls produced small ash clouds which were blown to the west. Several small pyroclastic flows occurred on 15 September, the largest reaching beyond the Tar River Soufriere. Ash clouds from rockfalls and flows were generally blown to the northwest over Garibaldi Hill; a change in wind direction at 12 noon caused some ash to be blown towards the north of the volcano. A period of intense ash and steam venting occurred from 12:50 to 13:20 from the highest part of the dome to the west of the active area.

A period of near-continuous rockfall activity on 16 September produced an ash cloud which drifted in a northwesterly direction depositing ash in Plymouth, Lovers Lane and surrounding areas. Despite generally poor visibility on 17 September, excellent views were obtained of the rockfalls and pyroclastic flows which developed in the Tar River Valley during the afternoon period. The active face continued to be the eastern flank of the dome; flows were generated by gravity collapse from this area. Ash clouds associated with the rockfall and pyroclastic flow activity produced ashfalls in the area of Plymouth and its environs during most of the afternoon period.

The heightened period of activity which began at 23:42 on 17 September and eventually culminated in a small explosive eruption was poorly observed since it occurred during the night. Reports and visual observations by MVO staff throughout the night indicate that a laterally direct blast was projected towards Long Ground village and an eruption column was sustained for a short period of time. Several houses were set alight by large hot projectiles which were blasted eastwards from the summit dome. The blasted material covers the areas west of Long Ground into the Hermitage area, but is not present on any of the other flanks of the volcano. More than half of the houses in Long Ground were impacted by blocks falling through roofs, at least seven houses were completely burnt. Gravel-sized material of both pumiceous and dense nature was deposited in areas across the central corridor and in Cork Hill, Richmond Hill and Foxes Bay. All ash erupted during the night was blown westwards over Plymouth and Richmond Hill and both of these areas received heavy ashfall. Reports from the NOAA Satellite Analysis Branch indicate that the ash column attained a height of at least 37,000 ft and caused the closure of the airport in Guadeloupe. A major collapse scar was produced on the eastern flank of the dome, cutting deep into the new dome. Some material was eroded from Castle Peak and a large volume of material was deposited in the Tar River Valley. The delta at the mouth of the Tar River Valley was enlarged and vegetation was completely removed from the valley itself. Estimates are that about 25-30% of the new dome rock was removed.

Visibility was generally poor for the remainder of the week although some brief views were obtained of the u-shaped scar feature which was excavated by the 17-18 September event. Several small rockfalls and pyroclastic flows were observed from the inner steep sided walls of the scar, particularly on the northern and western parts. These generated small ash clouds and deposited new debris at the base of the valley. Unstable blocks were observed on the northern face of the lava dome from Windy Hill on 21 September and these were expected to produce further rockfalls in the next few days.

Field surveys of the distribution and dimensions of the air-borne particles carried out on 19 September indicate that pumice clasts of up to 95 grams were deposited at 3km with clast size falling to 3.5 grams at 6km. Maps of the lithic and ash distribution which resulted from the 17-18 September event has been prepared and will be published in a separate report at a later date, along with a fuller volcanological account of the events of 17-18 September


Rockfalls and short swarms of volcano-tectonic earthquakes have been the most dominant aspects of seismicity during most of this reporting period. Long period and hybrid events remain at background levels and tremor has been generally low.

Earthquake types: 15 to 21 September 1996

These earthquake counts are made manually from the helicorder record for the Gages seismic station.

Date      VT        LP        Hybrid         RF        Tremor
15/09          72        14        15        84        Low
16/09          195       12        10        104       N/A
17/09          45        7         5         near-cont.     Low
18/09          4         1         1         77        Low
19/09          22        0         12        78        Low
20/09          0         0         0         46        N/A
21/09          46        0         25        49        Low

Rockfalls and pyroclastic flow signals were the most frequent seismic event recorded by the network occurring in close association with short volcano-tectonic swarms which were also quite common during the period. The periods of most intense rockfalls occurred immediately following short swarms of volcano-tectonic earthquakes and seemed to increase in frequency and duration immediately preceding the 17-18 September eruption. Periods of near-continuous rockfall and pyroclastic flow activity occurred between 12:00 and 14:40 on 16 September and between 11:30 and 20:30 on 17 September. Signals associated with such periods of high activity were of long duration with intermediate frequency and involving a typically cigar-shaped trace on the seismograms. The signals slowly decayed from a maximum amplitude but did not completely die out before another event started either with a slow build-up or a sharp onset. Signals with a sharp onset and which saturated the seismic records were generally correlated with larger rockfalls and with pyroclastic flow activity. Rockfall signals during the latter part of the period (18-21 September) were significantly reduced in frequency, amplitude and duration.

Volcano-tectonic earthquakes during this period continued to occur in short swarms the frequency of which increased from two to three each day during the period 15-17 September. The earthquakes were similar to those seen over the previous several weeks and were located at depths of less than 2 km (below sea level) beneath the crater. They are generally too small to be felt in occupied areas. The frequency of volcano-tectonic earthquakes also reduced after the explosive eruption of 17-18 September and by the end of the reporting period there were no events being recorded. None of the events during the latter part of the week occurred as swarms.

The levels of long period and hybrid events have been comparatively low throughout this period averaging less than 11 events per day. Hybrid activity increased during the latter part of the week with the marked reduction in other types of seismic signals. The level of tremor on the Gages seismometer has generally been low. A single mudflow signal which lasted for 40 minutes occurred on 20 September. Mudflow signals are very slow onset high frequency signals which maintain a maximum amplitude for a few minutes before a slow decay to background. The signals are often recorded on the Gages seismograph station during periods of heavy rainfall and/or following heavy ashfalls.

Ground Deformation

EDM measurements were severely hampered during this period by generally low cloud cover, ashing of various reflectors and generally poor weather. Attempts were made to occupy the eastern triangle to Castle Peak on 14 September but proved futile since ash cover on the reflector prevented a return signal. The northern triangle to Farrells was occupied on 16 September and showed a shortening of 2.8 cm on the St Georges Hill to Farrells line since the previous occupation in 22 August. The two other lines in this triangle; between Windy Hill and Farrells and St Georges Hill to Windy Hill did not show and change since the last occupation. The Castle Peak reflector was destroyed by the 17-18 September eruption and attempts may be made in the near future to replace this reflector. No EDM measurements were carried out on 17, 18 and 19 September due to heavy ashfalls during this period. EDM measurements between Lower and Upper Amersham and Amersham to Chances Peak showed no change since last occupation of these lines on 23 August.

The northern triangle was remeasured on 21 September; this was the only EDM line measured immediately before the eruption of 17-18 September. Both line lengths to the Farrells reflector site lengthened between 16 and 21 September. The changes observed were 4mm and 9mm respectively on line lengths between St Georges Hill-Farrells and Windy Hill-Farrells. The changes in line lengths are similar to those that have been observed on previous occupations of this triangle and are not considered to be related to the eruption. The triangle has always shown inconsistent changes in line lengths during the past ten months of its establishment.

A GPS survey of the lower flanks of the volcano was conducted on 15 September and the same network was re-occupied on 18 and 19 September. Initial processing of these data indicated some shortening of line lengths across the volcano between 25 August and 15 September and lengthening between 15 and 18 September. However, continued problems with processing of data from the 15 September survey means that these results are only preliminary. A GPS survey was initiated on a new network of stations on 21 September. The new configuration is a modification of the old one and now provides the MVO with coverage of the entire island and involves a rationalisation of the pre-existing network.

Gas Measurements

The COSPEC instrument is still out of operation and SO2 tube analysis for this period has still not been made.

Staff changes
Dr Gill Norton, BGS
Dr Peter Baxter

Dr Chris Kilburn, BGS
Dr Anne-Marie Lejeune, BGS

Visitors/Scientific Collaborators
Dr Glen Mattioli, University of Puerto Rico
Dr Alan Smith, University of Puerto Rico
Dr Jean-Jacques Jeremie, University of Guadeloupe

Montserrat Volcano Observatory