Montserrat Volcano Observatory, Montserrat, West Indies

Scientific Report 97
This report covers the month of October 1998


The activity this month has been at a slightly increased level. Five small dome collapse events occurred, each producing significant deposits up to 3 km from the dome. Pyroclastic flows occurred down most of the main drainages of the volcano and, although ash fell predominantly to the west and north-west of the volcano, light ashfall was also experienced in the north of the island. Dome collapses were commonly followed by periods of volcanic tremor and associated ash venting; in some cases swarms of volcano-tectonic earthquakes occurred shortly after collapses. The activity is leading to gradual erosion of the dome. Some large fractures were noted in the carapace and these could lead to larger dome collapses in the future.

Visual Observations

Observations this month have been dominated by intermittent small volume pyroclastic flows originating from all flanks of the dome. The first significant event at 08:01 on 13 October produced pyroclastic flows in Tuitt's Ghaut and Tyer's Ghaut. The flow in Tyer's Ghaut extended to within 200m of the Molyneux road while the flow in Tuitt's Ghaut reached Paradise Estate. Volcanic tremor was recorded after the collapse and lasted for the remainder of the reporting period. This tremor was correlated with vigorous ash venting from high up on the northern flank of the dome. Observations from the ground and helicopter indicated that the ash cloud associated with the collapse event rapidly reached 25,000 feet. The cloud drifted towards the north-west and deposited ash in areas extending from Woodlands to Weekes with most of the ash falling over Cork Hill, Old Towne, Olveston, Salem and the Belham Valley. Ashfall was also reported on Nevis.

At 09:16 on 18 October, there was a further collapse of material from the dome. The ash cloud from this event rose to around 7,000 feet and moved to the west across Plymouth. The exact direction of the collapse was uncertain because of low cloud that hampered visual observations. The collapse was followed by a period of volcanic tremor that lasted for several hours.

Another small dome collapse occurred at 22:41 on 20 October. The ash cloud from this event was estimated as around 8,000 feet and again drifted slowly to the west and north-west. Observations from St. George's Hill on the following morning revealed that the pyroclastic flows from this event had travelled down towards Plymouth as far as Upper Parsons. Several lobes in the deposit were observed on the Gages fan. The fall-out from this event included some coarse lithic fragments to 4 mm diameter in Old Towne, and up to 5 mm on the road between Cork Hill and Plymouth.

At 00:51 on 26 October, another small collapse of the dome occurred. The seismic signal lasted for about 12 minutes, and was followed by an extended period of volcanic tremor. Reports were received of thunder from the resultant ash cloud, and there was subsequent wet ash fall as far north as St. Peters. Information received from NOAA satellite images suggested that the ash cloud reached to between 20,000 and 25,000 feet high. Observations during the early hours of the morning from Salem suggested that there were two ash cloud lobes: one to the south of Belham Valley and one over the Salem-Old Towne area. Measurements of ash thicknesses later in the day showed that deepest measured ash fall was at Richmond Hill where 25 mm had accumulated. About 4 mm had fallen in Old Towne. The ash was almost entirely fine grained, with common accretionary lapilli. During an observation flight the following morning, new deposits were seen in the Tar River area. Steaming could be seen at the edge of the delta, indicating that the pyroclastic flows had travelled into the sea. The flows also reached as far north as the Tar River Estate House. On the south-west side of the volcano down the White River, a thin deposit of ash from the pyroclastic flows could be seen as far as about 700 m from the old coastline at O'Garras. The date on which these latter deposits were emplaced is unknown.

At 04:18 on 31 October a further small collapse of the dome occurred. This produced an ash plume which first drifted to the west, and then towards the north and north-east, depositing ash in occupied areas in the north of the island. An observation flight later in the day revealed that new deposits had formed in the White River and Gages valleys. The flow in the White River reached as far as the old location of the Galways Soufriere while the Gages valley flow did not extend beyond the top of the Gages fan. The White River deposit had numerous large angular blocks resting on the surface of the flow. A large fissure was observed within the dome near to its contact on the south-east flank of Chances Peak. This crack extended up from the base of the dome where it rests against Chances Peak to the top of the dome in the Galways area. At the foot of this crack a triangular shaped opening had developed and appeared to have been the source of the White River pyroclastic flow deposits. Collapse of the material in the Gages valley had produced a deep gully in the dome that extended to the back of the dome behind the Galway's area.

Unusual wind directions in the latter part of the month meant that the volcanic plume was directed to the north; strong sulphurous odours were thus reported in St. John's and in other parts of the north of the island.

On 27 October, temperature measurements were made in the pyroclastic deposits in the area of the Farm River in Trant's. The maximum temperature recorded was 239 *C at 2.25 m beneath the surface; at 1.00 m depth the temperature was 86 *C; at a second location, the temperature reached 146 *C at 1.44 m depth.

Also on 27 October, unusually clear conditions in the early evening enabled observers in Old Towne and Salem to see three small glowing areas on the dome behind Gages Mountain. It is thought that this was the incandescent interior of the dome that had been exposed during the dome collapses of the preceding weeks.

Small mudflow signals were recorded on 5 and 8 October, but no significant changes in the deposits were observed.


Seismicity was generally low over the reporting period. However, there were a number of small collapses of material from the dome that triggered both volcanic tremor lasting many hours and swarms of volcano-tectonic earthquakes. As in the previous month the tremor was observed to correlate with intensified ash and steam venting from the northern flanks of the dome.

Dome collapses occurred on 13, 18, 20, 26 and 31 October. These were marked by pyroclastic flow signals that lasted several minutes. The collapse on 13 October was preceded by a swarm of small volcano-tectonic earthquakes. During the collapse itself there were several much larger volcano-tectonic earthquakes, the first of which occurred approximately 30 seconds after the start of the collapse. Hypocentres for these events were tightly clustered directly under the lava dome.

The collapse on 18 October was accompanied by a much more intense swarm of earthquakes (Table 2). The first earthquake in the swarm occurred about 40 seconds after the beginning of the collapse. This event was one of the largest earthquakes recorded since the installation of the broadband network and was reported felt in the Woodlands area. The signal saturated on the broadband station at Roaches Yard. This earthquake was much richer in low frequencies than typical volcano-tectonic earthquakes on Montserrat, possibly suggesting a larger source dimension than normal. Hypocentres for the largest earthquakes in the swarm were located south of the volcano. At the start of the swarm, hypocentres were directly under Roaches Mountain then, as the swarm progressed, migrated to the area south of Chances Peak.

Examination of first motions shows that in many cases, the largest events in the swarm had compressional arrivals at all stations. Preliminary calculation of fault-plane solutions, using the first motions only and assuming a double-couple mechanism, shows that the largest events have remarkably stable mechanisms, consistent with oblique normal faulting in a northeast-southwest direction.

Table 1. Earthquake types These earthquake counts are of events that triggered the broadband network's event recording system between 00:00 and 00:00 each day (local time).

Date		Hybrid	LP	Rockfall	VT
01 October 98	0	0	2		5	
02 October 98	0	0	2		6
03 October 98	0	0	2		2
04 October 98	1	0	3		2
05 October 98	0	0	2		2
06 October 98	0	0	5		2
07 October 98	0	0	7		4
08 October 98	0	0	7		3
09 October 98	0	0	9		2
10 October 98	0	0	14		8
11 October 98	0	1	24		27
12 October 98	0	0	12		6
13 October 98	2	1	14		15	
14 October 98	0	0	9		3
15 October 98	0	0	5		5
16 October 98	0	0	9		2
17 October 98	0	0	7		9
18 October 98	0	0	9		54
19 October 98	0	0	13		13
20 October 98	1	0	14		5
21 October 98	1	0	11		5
22 October 98	1	0	13		9
23 October 98	1	0	22		11
24 October 98	0	0	13		16
25 October 98	0	6	7		30
26 October 98	1	6	37		5	
27 October 98	0	0	25		9
28 October 98	0	0	28		9
29 October 98	0	0	23		10
30 October 98	0	0	11		22
31 October 98	0	0	22		16

Table 2. Swarms

Start			Duration/hrs	Hybrid	LP	VT
13 October 98 02:49	5.10		0	0	11	
18 October 98 09:16	6.73		0	0	51	
25 October 98 06:14	11.32		0	0	24	

Ground Deformation

GPS sites on the volcano and in the north of the island were occupied this month. All the sites appear stable and there were no significant changes since last month's results. The EDM reflector on the northern flank was shot from Windy Hill. The line continues to shorten slowly. The site was destroyed by a pyroclastic flow later in the month.

Volume Measurements

A series of photographs and theodolite measurements were taken from sites around the volcano to help build up an accurate surface map of the dome.

Environmental Monitoring

Sulphur dioxide fluxes were measured using the miniCOSPEC machine this month. Results are shown in Table 3. These levels are similar to those measured over the past few months, although there was an apparent decrease later in the month.

A team from Cambridge University has been visiting this month to continue with FTIR measurements of the volcanic plume.

Table 3. Average daily sulphur dioxide fluxes measured by miniCOSPEC, October 1998

Date	Flux (tonnes/day)	
9 October	1300	
21 October	340	
30 October	280	

Sulphur dioxide was also measured at ground level by using several sets of diffusion tubes located around the island. The results are shown in Table 4. The tubes at Police HQ in Plymouth were inaccessible on 30 September as a result of mudflow deposits following Hurricane Georges. Hence the value for this location is over a longer time period. The levels in Plymouth continue to be high; elsewhere the average sulphur dioxide levels are very low.

Table 4. Sulphur dioxide diffusion tube results, October 1998. Levels in ppb.

		30-Sep to 16-Oct-98	9-Sep to 16-Oct-98	
Police HQ, Plymouth	-			134.5	
St. George's Hill	5.5			-	
Weekes			2.0			-	
Vue Pointe Hotel	1.6			-	

MVO Staff Changes

Richie Robertson (Seismic Research Unit)
Gill Norton (British Geological Survey)
Desmond Seupersad (Seismic Research Unit)
Keith Rowley (Independent)

Simon Young (British Geological Survey)
Sue Loughlin (British Geological Survey)
Godfrey Almorales (Seismic Research Unit)
Brian Baptie (British Geological Survey)

Marie Edmonds (Cambridge University)
Mike Burton (Cambridge University)

Montserrat Volcano Observatory