During mid-September 2002 through February 2003 at Soufrière Hills, the dome continued to grow, producing numerous rockfalls and small-to-moderate pyroclastic flows. Most of the activity was concentrated on the NE and N flanks, producing numerous pyroclastic flows in White's Ghaut, the Tar River Valley, and Tuitt's Ghaut. Pyroclastic flows and rockfalls also traveled down the W and NW flanks. Ashfall affected surrounding areas, accumulating in thicknesses up to 9 mm. The Washington VAAC issued notices to the aviation community almost daily. Seismicity was dominated by rockfalls (table 42).

Table 42. Seismicity at Soufrière Hills during 13 September 2002-28 February 2003. *During some weeks, the number of seismic events was under-represented because of problems with the seismic stations. Courtesy MVO.

Activity during September 2002. Lava-dome growth was directed to the NE during 13-20 September, with frequent rockfalls and small pyroclastic flows sending material to a sector extending from the central Tar River Valley on the E flank to the NE flanks above Tuitt's Ghaut. Some material tumbled through a notch onto the N flank. A major change in direction of extrusion followed a hybrid earthquake swarm between 0703 and 1515 on 19 September. Growth of the previously active NE lobe stagnated during 21-22 September. A near-vertical spine was extruded in the central area around the 21st, possibly indicating a switch in growth direction. On 26 September a swarm of 36 hybrid events occurred between 0330 and 1112. The same day observations revealed a large new dome lobe that had extruded towards the W in the area previously known as Gages Wall. Material spalling off of this lobe produced rockfalls and small pyroclastic flows down Gages Valley that reached up to 1 km.

Notable pyroclastic flows occurred on the evening of 25 September and the morning of the 27th. Growth and rockfall activity then changed towards the N flanks, suggesting a possible stagnation of the recently extruded western lobe. Spectacular incandescence and semi-continuous rockfall activity were observed on the NE and N flanks of the dome on the night of 26-27 September.

On 27 September a 4-hour-period of heightened activity occurred in the afternoon and evening, with small semi-continuous pyroclastic flows traveling down the N flanks and eastwards into the upper portions of Tuitt's Ghaut and then into White's Bottom Ghaut. A newly extruded lobe was visible on 28 September almost directly to the NW with a broad headwall over the N, NW, and W flanks. On the evening of 29 September there was another period of heightened activity on the N flanks that lasted 1.5 hours, with pyroclastic flows just reaching the sea along White's Bottom Ghaut. It was estimated that during this event only 2-3 x 10⁶ m³ of the N edge of the active NW lobe was shed.

The Washington VAAC reported that a low-level ash cloud from an emission at 1510 on 29 September was visible over eastern Puerto Rico on satellite imagery through the following day. On 30 September a light dusting of white ash fell in eastern Puerto Rico at Roosevelt Roads Naval Air Station.

Activity during October 2002. Observations on 1 October revealed that re-growth of the collapsed area had occurred. A brief period of heavy rain on 2 October triggered a moderate-sized mudflow down the Belham Valley. Analysis of seismic data suggested that pyroclastic-flow activity on 2 October began at 1310, and sustained dome collapse continued for 6 hours. Low-energy pyroclastic flows were observed reaching the sea on the Tar River's flanks throughout the collapse, and ash clouds were produced that drifted to the NW. Heavy ashfall occurred in the residential areas of Salem, Old Towne, and Olveston, with deposits up to 9 mm thick. Subsequent observations revealed that this collapse was confined to the E flanks, and that this was again a relatively small event (less than 5 x 10⁶m³ of material was shed off of the E side of the dome complex).

According to the Washington VAAC, after daybreak on 3 October there were several reports of ashfall in Puerto Rico, and visible satellite imagery at 1115 confirmed that an ash cloud around 2.4 km altitude covered most of the island. At 1615 the area of very thin ash was not visible on satellite imagery. By the next day, ash from the previous day's emissions had drifted W, and around 0902 it was located over southern Puerto Rico. A thin plume of ash also extended SSW of St. Croix island.

Early in October the NW extrusion lobe of the lava dome grew to the NW, but later growth remained more centralized and there was noticeable bulking up of the lobe's summit area. Talus continued to accumulate behind the NW buttress and in the head of Tyre's Ghaut. Minor mudflow activity occurred on 9 October. The growth of the lava dome towards the NW prompted the evacuation of populated areas along the fringes of the lower part of the Belham Valley (~300 people) on 8 and 9 October, and the area was declared part of the Exclusion Zone. A relatively small pyroclastic flow traveled NNE down the flanks on 13 October.

On the afternoon of 22 October intense rainfall at midday produced large mudflows NW in the Belham Valley. At the peak of flow, the entire width of the valley floor at Belham Bridge was flooded and standing waves up to 2.5 m high were observed. By 1430, pyroclastic-flow activity began. For several hours, pyroclastic flows from the N flank of the dome were channeled NE into the upper parts of Tuitt's Ghaut, from where they crossed over into White's Bottom Ghaut. Flows also occurred on the dome's E flank in the Tar River Valley.

The volcano was observed using a remote camera and during a flight on 31 October. The active extruded lobe in the NW continued to steadily grow, bulking out on the N and W sides. Rockfalls and pyroclastic flows traveled down the E and N flanks, particularly within Tuitt's Ghaut and the Tar River Valley. A considerable amount of debris also spalled off the W flank of the active extruded lobe and accumulated in the upper parts of Fort Ghaut.

Activity during November 2002. During early November lava-dome growth on the N part of the dome was less directed, with rockfalls dispersed over the summit and flanks. The lobe shed rockfall debris predominately down Tuitt's Ghaut and Tar River Valley, although also onto the NW flank and into the top of Gage's Valley. According to the Washington VAAC, on 8 November strong pyroclastic flows produced ash-and-gas clouds to a height of ~1.5 km.

On 8 and 9 November pyroclastic flows traveled 900-1,000 m NW into Tyer's Ghaut at the headwaters of the Belham Valley. During 12-15 November, the size and energy of the pyroclastic flows increased slightly. During 15-19 November, small pyroclastic flows traveled 1-1.5 km from the dome every few hours in Tuitt's Ghaut to the NE and in the Tar River Valley to the E. On 29 November the active lobe had a broad whaleback-shaped upper surface, which was oriented towards the NNE.

During 29 November-6 December a number of small, short-lived spines formed at the base of the active lobe in the N part of the dome complex, shedding material E into White's Ghaut and the Tar River Valley. Lava blocks continued to spall off the front of the lobe, shedding material NE into Tuitt's Ghaut and onto the northern talus slope. An average of one moderate-sized pyroclastic flow occurred per day and traveled no farther than 1-1.5 km from the lava dome into Tuitt's and White's ghauts and into the Tar River Valley. During 5-6 December, rockfalls and small pyroclastic flows occurred more frequently on the northern talus slope and on the NW, at the top of Tyer's Ghaut.

Activity during December 2002. A sustained dome collapse began on 8 December at 2045, producing energetic pyroclastic flows down White's Ghaut to the sea at Spanish Point. Ash clouds rose to ~3 km altitude and drifted WNW. In Plymouth and Richmond Hill 4 mm of ash was deposited. Seismicity returned to background levels on 9 December by 0045, and several days of weak tremor occurred.

The collapse scar on the dome's NNE flank, estimated to have had a volume of 4-5 x 10⁶ m³, was being filled rapidly with freshly extruded lava. Observations on 13 December revealed a large amount of fragmental lava extruded in a northerly direction on the summit. A large spine was also extruded on the NW side of the summit.

During late December spectacular incandescence of the dome was observed on most nights. Activity increased during 18-20 December, and on 19 December mudflows occurred in White River, Tar River Valley, and Fort Ghaut. During 20-27 December extrusion occurred on the N, and occassionally NW, sides of the summit. A large spine was pushed up at the back of the active extruded lobe during the night of 26-27 December, but was not visible by 2 January. The Washington VAAC reported that on 28 December around 1130 a 3-km-high ash cloud generated from pyroclastic flows drifted over the islands of St. Kitts and Nevis.

Activity during January-February 2003. Activity escalated to very high levels on the night of 27 December. During 27 December-10 January continuous rockfalls and numerous pyroclastic flows spalled off the active extruded lobe on the NNE side of the lava dome. Activity decreased on the night of 2 January to moderate levels on the 3rd.

During mid-January, activity generally declined to a moderate level. During 15-17 January almost all pyroclastic flows occurred in the Tar River Valley, with only minor rockfalls traveling down the dome's NE and N sides. Lava extrusion occurred NE of the lava-dome complex that was associated with rockfalls and small pyroclastic flows down Tar River Valley, White's Ghaut, Tuitt's Ghaut, and on the northern talus slopes. On 18, 20, and 24 January small pyroclastic flows traveled ~1 km down Tyer's Ghaut.

Activity increased during late January. Growth of the active extrusion lobe continued on the N side of the lava dome. The direction of growth was generally towards the NNE, although the focus of rockfall and pyroclastic-flow activity varied from day to day. A pulse of activity occurred at midday on 30 January, during which pyroclastic flows simultaneously descended several flanks of the lava dome traveling to the Tar River Valley, White's Ghaut, Tuitt's Ghaut, and W to Fort Ghaut.

During 31 January-14 February activity remained moderate. Growth of the lava dome was focused on a large, steep lobe directed to the NE. A small amount of rockfall material was directed W towards Fort Ghaut. Rockfalls and small pyroclastic flows also occurred off the N flank of the dome onto the area of Riley's Estate.

During 19-25 February pyroclastic flows and rockfalls were concentrated more on the E flank of the lava dome and in the Tar River Valley, although there were several periods of activity on the N flank, with pyroclastic flows in Tuitt's Ghaut and at the top of Farrell's Plain.

Activity increased slightly during 21-28 February. During an observation flight on 27 February lava-dome growth was concentrated towards the NE. Pyroclastic flows and rockfalls traveled down the lava dome's E and NE flanks via the Tar River Valley and Tuitt's Ghaut. There were also several periods of activity on the N flank, with pyroclastic flows at the top of Farrell's Plain.

SO₂ emission rates varied throughout the report period (table 43), and were especially high following the dome-collapse event on 9 December (2,350 tons per day average).

Table 43. SO₂ emission rates at Soufrière Hills during 13 September 2002 through 28 February 2003. Courtesy MVO.

Geological Summary. The complex, dominantly andesitic Soufrière Hills volcano occupies the southern half of the island of Montserrat. The summit area consists primarily of a series of lava domes emplaced along an ESE-trending zone. The volcano is flanked by Pleistocene complexes to the north and south. English's Crater, a 1-km-wide crater breached widely to the east by edifice collapse, was formed about 2000 years ago as a result of the youngest of several collapse events producing submarine debris-avalanche deposits. Block-and-ash flow and surge deposits associated with dome growth predominate in flank deposits, including those from an eruption that likely preceded the 1632 CE settlement of the island, allowing cultivation on recently devegetated land to near the summit. Non-eruptive seismic swarms occurred at 30-year intervals in the 20th century, but no historical eruptions were recorded until 1995. Long-term small-to-moderate ash eruptions beginning in that year were later accompanied by lava-dome growth and pyroclastic flows that forced evacuation of the southern half of the island and ultimately destroyed the capital city of Plymouth, causing major social and economic disruption.

Information Contacts: Montserrat Volcano Observatory (MVO), Mongo Hill, Montserrat, West Indies (URL: http://www.mvo.ms/); Washington Volcanic Ash Advisory Center (VAAC), Satellite Analysis Branch (SAB), NOAA/NESDIS E/SP23, NOAA Science Center Room 401, 5200 Auth Road, Camp Springs, MD 20746, USA (URL: http://www.ospo.noaa.gov/Products/atmosphere/vaac/); Associated Press.