Montserrat Volcano Observatory, Montserrat, West Indies

Scientific Report 65
24 May 1997

This report covers the 14-day period to 24:00 on 24 May.


Activity has been at a higher level during this period and is characterised by a switch in the area of dome growth from the Galway's area to the north side of the dome, with an increase in the rockfall activity on the eastern and northern slopes of the dome from 11 May. The number of pyroclastic flows in the Galway's area decreased and instead small to moderate sized pyroclastic flows ran down the northern slope of the dome, with debris piling up against Farrell's wall. By the 19 May, the dome was flush with the low notch in the crater wall at the top of Tuitt's Ghaut, and a few small boulders had fallen into the top of the ghaut.

Visual observations

Visual observations have been hampered by low cloud for most of the reporting period. However brief views of the dome were possible on the 15, 18 and 23 May, and the dome was clear for a few hours on the 19 May, allowing a dome survey to be made.

At the start of this period, the main area of active dome growth was still focused on the southern side of the dome, above the Galway's area, with rockfalls and few small pyroclastic flows shed both into White River and down the south side of Tar River valley as far as the base of Perches Mountain. On the White River side, the flows didn't go beyond the base of the talus slope. No detailed dome survey had been possible during the first half of May, and the rate of dome growth was estimated to be low (around 2 m3/s).

This pattern of activity established during the past few weeks was suddenly interrupted at 07:55 on May 13 by the occurrence of a moderate size pyroclastic flow discharged down the Tar River valley, producing an ash cloud of 10,000 ft estimated height. This pyroclastic flow, the first in this area since the end of March, originated from the summit region, eroding a narrow channel on the eastern flank of the dome, where Castle Peak remains have been buried. This flow split into two branches which travelled either side of the upper break-in-slope, and eventually reached the delta, where it fanned out. Vigorous steaming was observed from near where this flow originated. Samples collected on the delta on the 30 May from the deposits of this flow show dense dome rocks and moderately vesicular pumice along with white, hydothermally altered litholigies.

Loud, roaring sounds from the dome were first noticed on 12 May. This roaring could be heard periodically from different areas (Whites, Harris, Farrell's) around the volcano in the following days, and are thought to be the result of vigorous steam venting of ash and gas, possible indicating increased gas pressures within the dome.

On 14 and 15 May, small, nearly continuous rockfalls were seen on both the south-east and north-east flanks of the dome. Some of them developed in small pyroclastic flows, which travelled half way down the Tar River valley, to the lower second break-in-slope. Very small pyroclastic flows were also seen on the Galway's side. These small pyroclastic flows were followed by an increase in gas and steam emissions, and an increase in audible venting.

On 15 May, more observations of the dome were made from Whites. An increase in rockfall activity was noted in the Tar River valley, accompanied by degradation of the north-eastern flank of the dome. A period of moderate pyroclastic flow activity occurred at 20:40 for 70 minutes, on the eastern face of the dome down Tar River valley. These flows created a small scar about 40 m to the north of the scar formed by the 13 May flow, and travelled down the north side of the valley to the delta. During this small collapse, pulsing, roaring sounds from the dome were heard at Harris Lookout.

Pyroclastic flow activity continued on 16 May, with small to moderate flows in the Tar River valley. These flows originated from the summit, and travelled down the north and north-west sides of the dome towards Farrell's wall, where they were deflected down the inside crater wall into the northern Tar River and stopped 300 m from the delta. Further erosion occurred on the northern face, with material piling up against the crater rim between Hermitage and Farrell's. Brief views of the whole dome from the MVO and the east revealed that a large, new spine had grown at the summit of the dome where a small spine was seen the morning before.

Despite persistent low cloud, observations on 17 and 18 May showed that activity was concentrated on the northern and eastern flanks of the dome, with three major rockfall chutes developed on the east, north-east and north sides. Rockfall material piled up against the Farrell's wall at the base of one of these chutes near the top of White's ghaut. Most of the rockfalls and pyroclastic flows originate from the summit area. Several small rockfalls were also heard on Galway's side, where new, relatively fine-grained rockfall deposits were noticed on the entire talus slopes. The largest pyroclastic flow which occurred on the 18 May at 08:20 was initiated by a large LP earthquake. Observers at Whites saw signs that the entire dome was being shaken just before the flow started. This flow travelled from the summit down both the 15 May chute and the north-eastern chute, and passed down the northern side of the Tar River valley before stopping after the second break-in-slope, a few hundred meters from the delta. As observed during the previous days, steam and gas emissions from the dome increased after some of the pyroclastic flows. Views at night from Harris Lookout and Whites, showed intense glowing of the entire north-west face, above Farrell's and also incandescent rockfalls and small pyroclastic flows on the north-east and the eastern faces, down the 13 and 15 May chutes.

Clear visibility in the morning on 19 May revealed a new prominent dark extrusion growing at the top of the dome towards the north-west. Intense, vigorous ash and steam emissions were generated at the top of the dome in the area of the extrusion, with fist-to-head sized fragments intermittently thrown up to a height of about 60 m, suggesting that there was high gas pressure within the dome. The most active area was localised on a large sector spreading from the 15 May chute, on the eastern side, to the margin of the September 96 scar on the north-west. The remnant wall of the September 96 scar prevented material from reaching the rim of Gages wall. The activity was characterised by periods of continuous low-level rockfalls with small pyroclastic flows into the Tar River valley, following hybrid swarms. The northern side of the crater filled up, with very small amount of dome material falling into the top of Tuitt's Ghaut. The pyroclastic flows on the northern side travelled to the edge of the crater and were diverted down the northern side of the Tar River valley. The 13 May chute was still very active, and some small to large rockfalls were also seen travelling down the south-eastern side of the dome into the upper part of Tar River valley.

On 21 and 22 May, rockfalls were observed on the eastern, northern and north-western flanks of the dome, with blocks reaching about 100 metres down Tuitt's Ghaut around 13:00 on the 22 May, after a rockfall on the northern flank. A few small spines were observed on the upper northern flank, with most of the rockfalls originating from this area. At 14:30, regular pyroclastic flows were generated from the upper parts of the dome, flowing mainly to the north and producing almost continuously ash, with a vertical plume observed on several occasions emanating from the summit of the dome. 1 mm-sized ash fragments were reported at Farrells, and due to a change in the wind direction, the ash fell in the north from Salem up to St Johns. Lapilli up to 4 mm diameter were collected at Dyers.

On 23 May the top of the dome was visible from Chances Peak. There were several small spines and large blocks in the summit area, with a cleft in the middle separating the southern lobe from the new extrusion in the north. Vigorous venting gas emissions were noted in the area of current growth.


The seismicity during this period showed changes associated with the switch in the area of dome growth and an increase in the level of volcanic activity. There was a general decline in the level of long-period earthquake activity, which peaked on 5 May, and a corresponding decline in the number of rockfalls which were triggered by LP earthquakes.

Hybrid earthquakes appeared again during this period, usually occurring in swarms (Table 2). Each swarm lasted for a few hours, with typically about 10 earthquakes per hour, although three swarms between 19 and 21 May were significantly more intense, with up to 35 earthquakes per hour. Some of the hybrid earthquakes triggered rockfalls from the dome, the first time that this has been observed since December 1996.

Two accelerometers were installed temporarily at Chances Peak on May 18, to enable measurement of strong ground motion in the vicinity of the dome. The accelerometers captured 9 hybrid earthquakes between 19 May to 23 May, including some moderate amplitude events. The range of peak accelerations in individual events ranged from .02g, just above the trigger level, to over 0.2g. Values in excess of about 0.05g could be significant to local stability of the dome and parts of the crater wall, explaining the correlation between some hybrid events and rockfalls noted above.

The level of rockfall activity remained high and most of the rockfall signals occurred in concentrated periods of minor dome collapse. There was only one moderate pyroclastic flow, on 15 May in the Tar River valley. The change in focus of dome growth was clearly displayed in the rockfall amplitudes recorded by the broadband network, with the ratio of maximum rockfall amplitudes at Galway's Estate and Long Ground used to differentiate between Tar River and White River pyroclastic flows. Activity on the south side of the dome decreased during 14 and 15 May, and pyroclastic flows large enough to trigger the event recording system were almost non-existent on the south side for the rest of the reporting period.

Table 1: Earthquake types

These earthquake counts are of events that triggered the broadband network's event recording system between 0000 and 2400 each day (local time).

Date		VT	Hybrid	LP	Dome RF	LPRF*	HYRF*

11 May 97	0	5	25	86	9	0
12 May 97	0	4	37	106	27	0
13 May 97	2	25	25	108	21	0
14 May 97	0	82	15	81	10	0
15 May 97	1	42	9	40	3	0
16 May 97	0	27	35	68	11	0
17 May 97	0	5	16	82	15	0
18 May 97	0	73	27	53	19	3
19 May 97	2	151	19	83	14	2
20 May 97	0	159	7	42	2	4
21 May 97	0	122	12	53	8	1
22 May 97	3	23	17	87	15	1
23 May 97	1	138	8	88	5	6
24 May 97	0	64	9	138	7	7

*LPRF: LP earthquake followed by rockfall signal. HYRF: Hybrid earthquake followed by rockfall signal. The LPs, hybrids and rockfalls in these signals are also counted in their respective columns.

Table 2: Earthquake swarms (times are UTC)

Date		Strt	Stop	Durn 	#vt 	#hy 	#Gap to next (hrs)

18 May 97	06:18	11:25	5.12	0	53	14.45
18 May 97	20:45	00:34	3.85	0	17	12.72
19 May 97	09:28	11:04	1.60	0	19	11.70
19 May 97	21:10	02:36	5.43	0	127	16.20
20 May 97	13:22	22:28	9.1	0	158	23.67
21 May 97	12:02	21:11	9.15	0	122	25.52
22 May 97	13:33	19:04	5.52	3	23	17.30
23 May 97	06:51	13:52	7.02	1	94	17.62
24 May 97	00:28	04:11	3.72	0	44	14.77
24 May 97	15:14	21:21	6.12	0	61	20.32

Ground Deformation

A GPS occupation was made on Chances Peak on 23 May running base stations at Harris and Windy Hill. The result from this occupation suggests that Chances has moved a further 3.5cm since 28 April when it was last occupied, representing a significant increase in the rate of movement. The direction of movement has also changed. Up to the occupation of 28 April movement was towards 255 degrees; it appears now to have changed towards 205 degrees, possibly reflecting the shift in activity to the north of the dome complex (NNE of Chances summit). It is stressed however that this is a single measurement and more are required to confirm the change in motion of the site.

A GPS occupation of FT3 (Farrells crater wall) was made on 12 May running base stations at Harris and Windy Hill. FT3 shows a further 1.2cm of movement to the NW, consistent with the movement direction since January 1997. The total movement of the pin since that time is now nearly 20cm. Two separate, single-prism reflectors were installed on this visit close to the GPS pin. Unfortunately low cloud prevented EDM measurements being made to them before the switch in activity to the north. The reflectors are now covered in ash and the site is extremely dangerous due to the proximity of pyroclastic flow activity.

GPS occupations including Hermitage and FT1 (on the flank of Farrells midway between Farrells target and FT3) were made on 17 and 21 May. The 17 May occupation had suggested an abrupt movement of the Hermitage site, although this was not confirmed by the occupation of 21 May. FT1 had only been occupied once before in mid-January, and a conclusive movement direction cannot yet be given.

Crack 2 on Chances Peak was measured on 23 May. Four measurements of crack width are taken between nails in four trees in a straight line crossing the crack. The crack has developed into a broad fracture zone, so individual measurements may increase or decrease depending on the motion of the part of the fracture zone on which a tree is sited. All movements were small with 1-3cm of opening and closing being recorded across the crack. Shear rate along the crack has increased significantly with a further 6cm of dextral movement (eastern block moving south relative to western block) being recorded since 28 April. This is the highest shear rate on the crack (0.24 cm/day) since mid-February (0.44 cm/day) and the total shear movement is now nearly 30cm. The region where the crack intersects the crater wall has collapsed in an 8m wide slump, which is slowly sliding down the wall towards the crater floor

A telemetered extensometer was installed across part of Crack 2 on May 23; patterns to date suggest a diurnal change of about 5 mm.

Telemetered electronic tiltmeters were installed at locations on Chances Peak, with one (CP2) installed on 18 May at the Antilles Radio house west of the summit, and the other (CP3) on 22 May on a wall mount in the Cable and Wireless structure east of the summit, on the shoulder of the Galways Wall. Within about a day of installation the tiltmeters began to show a rhythmic pattern of repetitive cycles of inflation and deflation. Each cycle lasted about 12 to 18 hours, with a period of inflation followed by a shorter period of deflation. The amount of deflation is generally less than the amount of preceding inflation, so that overall, the trend is toward accumulative inflation. The near-real time deformation occurs in concert with seismicity, with the hybrid earthquake swarms correlating with peak inflations, and with peak rockfall/pyroclastic flow events correlating with deflations. The RSAM patterns show a strong correlation with tilt, with the higher spikes in the RSAM pattern reflecting rockfalls, and the lower intensity patterns the sum of hybrid events and lesser rockfall activity. After a few days of observation of the pattern, it became clear to MVO staff that the tilt data, in concert with seismic data (e.g., RSAM), provided a predictive capability: once the inflation episode began to plateau , we could expect rockfalls and pyroclastic flows to peak within the next few hours. On that basis, a field mission to recover accelerometer data was cancelled on May 21, and since then, other field missions to close-in areas have been choreographed to coincide (when possible) to the time of early inflation, when likelihood of pyroclastic flows is minimal. Thus the data are being used in a practical manner to reduce the risk to field parties.

Dome Volume Measurements

A GPS - binocular survey of the dome was made from the helicopter on 19 May. The western side of the dome was obscured by cloud at the time of the survey, but 105 points were determined on the other faces of the dome. Short views of the western side of the dome from the helicopter and Chances Peak show no recent changes in this area, so data points from a survey of 7 April were used to define the surface here. Data from a theodolite survey on 19 and 20 May were included as were data from a comprehensive survey of the upper reaches of the Tar River Valley on 12 May. This latter survey was made from the Tar River GPS point using a pair of vector binoculars which have a working range of just over 2km and a field precision of below 1m in distance measurement and 0.25 degrees in dip and azimuth. The binoculars are currently on loan to MVO from Lancaster University, UK.

A summit height of 991m was obtained during the helicopter survey. The major changes since the last dome survey (15 April) are an inflation of the highest part of the dome immediately above Galway's, growth of the new extrusion in the northern summit area and talus accumulation in a zone 300m wide against the back of Farrells wall, due to the activity on the N and NE faces. Since the survey the area of degradation and rockfall activity has extended further round onto the eastern flanks of the dome, almost to the Castle Peak fumarole.

The volume of the dome from this survey was estimated at 60.1 million cubic metres. This represents an average extrusion rate since 15 April of 3.1 m3/s (270,000 m3/day), considerably above the long-term mean extrusion rate of 2.1 m3/s since around 100 days into the eruption.

Environmental Monitoring

The miniCOSPEC has now been returned to the MVO after being calibrated in Chicago. Several runs were completed on 24 May and an SO2 flux of 950 tonnes per day was calculated. This is substantially higher than the usual background flux of 200-300 tonnes per day reflecting the elevated level of volcanic activity.

Sulphur diffusion tubes were collected from five sites around the volcano on 17 May and have been sent to the UK for analysis. The results from sulphur diffusion tubes collected on 20 April and on 4 May show similar values to the results of previous weeks (Table 3).

Table 3:

SO2 concentrations (ppb) from diffusion tubes at sites around the volcano.
Recommended action level is 100ppb.

Location		6th April to 20th April		20th April to 4th May
Plymouth Police HQ		7.35				7.8
Upper Amersham			45.0				53.2
Lower Amersham			12.1				16.95
Weekes				0.0				0.0
Whites landfill			0.8				1.2

Ash was also collected on 17 May following several days of increased volcanic activity. The ash was at least 6mm thick at Upper Amersham, it was greater than 4mm thick at Lower Amersham, the Plymouth Police Headquarters and Dagenham. This compares with a maximum of 2mm of ash per week at Amersham during the previous two weeks. The level of volcanic activity remained high throughout the reporting period and this coupled with winds blowing from the south and south-east caused light ashfall in the safe zone on several occasions. In the late afternoon on 22 May, ash was reported falling in St John's (northern Montserrat).

Rainwater was collected from four sites west and northwest of the volcano on 17 May. The rainwater is more acidic than the previous week and chlorides and sulphates are present at substantially higher levels (Table 4)

Table 4 Rainwater geochemistry 11 to 17 May

Location		pH	Cond.	TDS	Flrd	Chlrd	Slpht
				mS/cm	g/l	mg/l	mg/l	mg/l
Weekes			3.82	0.272	0.136	0.65	50	37
Plymouth Police HQ	2.73	3.51	1.75	nd	710	560
Upper Amersham		2.42	2.45	1.22	nd	315	107
Lower Amersham		2.78	4.26	2.13	nd	760	97

nd: not detected

After a period of heavy rainfall and continued winds from the south and south-east, a rainwater sample was collected from Lawyers, 2km north of Salem, which had a pH of only 3.3. As a result the water sampling network has now been extended so that rainfall to the north of the volcano can be monitored regularly. New rainfall monitoring sites are Molyneux, Lawyers, MVO and Riley's Yard.

Staff Changes
Barry Voight
Rob Watts

Paul Jackson, SRU
Sayyaddul Arafin, SRU
Simon Young, BGS

Mr Lloyd Lynch is Chief Scientist in post until 6 June. The next victim will be Dr Willy Aspinall.

Montserrat Volcano Observatory