THE GEOLOGY OF MONTSERRAT
Montserrat island (with an area of 102 sq.km and a pre-1995 eruption population of 10,500, is composed of three major volcanic centers, ranging in age from Pleistocene to Recent The northern end of the island is formed by the Silver Hill complex described by Rea (1970, 1974) as composed of two-pyroxene andesite lavas and agglomerates. 39Ar/40Ar dating of these lavas have yielded ages of 2,580±60 ka and 1,160±46 ka (Harford et al., 2002). (The notation ka means thousands of years, so these ages could also be written as 2,640,000 to 2,520,000 years and 1,206,000 to 1,114,000 years). The central part the island is dominated by the Centre Hills complex characterized by steep-sided peaks (remnants of Pelean domes) surrounded by aprons of block and ash flow and surge deposits, a distinctive pumice lapilli fall deposit, reworked equivalents of the pyroclastic deposits as well as mudflow/lahar deposits. Petrographically the complex is composed of two-pyroxene and hornblende-hypersthene andesites (Rea, 1970, 1974), and has yielded 39Ar/40Ar ages of 954±12 ka, 871±10 ka, 826±12 ka, 663±49 ka, and 550±23 ka (Harford et al., 2002).
The southern part of the island consists of the South Soufriere Hills – Soufriere Hills complex (SSH-SH complex) of two morphologically distinct centers (see map). The SSH center measures 4 km east-west by 2.5 km north-south and is bounded by the White River in the west and tributaries of Dry Ghaut in the north. It has a summit crater with a north-south dimension of about 1 km and is open to the east. The crater is occupied by the Raspberry Hill dome complex composed of two nested domes, the oldest in the west and the youngest in the east (Rea, 1970). The deposits from the SSH have been 39Ar/40Ar dated at 131±7 ka ( Landing Bay ), 130±5 ka (Roche’s Bluff), 129±14 ka (Shoe Rock), 128±17 ka (Roche’s Bluff), and 128±27 ka (White River Pyroclastic Fall series)(Harford et al., 2002).
The active SH center, measuring 9 km east-west by 6 km north-south is larger and more complex than the SSH, and before the present eruption rose to an altitude of just over 1000m at Chance’s Peak. The superstructure of this center comprises four domes, Gage’s, Chance’s, Galway ‘s and Perche’s. On its eastern flanks is the steep-walled elliptical depression of English’s Crater that contained the Castle Peak dome. English’s Crater is now believed to be a sector or flank collapse scar with a corresponding submarine avalanche deposit offshore in the Atlantic . The Castle Peak dome was destroyed during the present eruption and has now been replaced by the 1995-present dome. Since European settlement in 1635 AD (Fergus, 1994), southern Montserrat has undergone four episodes of volcanic activity. The first three were volcano-seismic crises, in 1897-98 (Robson and Tomblin, 1966; Wadge and Isaacs, 1988), 1933-37 (Perret, 1939) and 1966-67 (Shepherd et al., 1971). The fourth episode which started in 1992 also involved an initial increase in seismic activity (Ambeh et al., 1995), but on August 18, 1995 explosive pyroclastic activity centered on English’s Crater commenced (Young et al., 1998), and is still on. 39Ar/40Ar dating has provided weighted mean ages of 151±4 ka for Gage’s dome, 112±9 ka for Galway ‘s dome, and 24±2 ka for Perche’s dome (Harford et al., 2002). No ages were obtained for Chance’s dome but it is believed by Harford, based on petrography, to be younger than 110 ka (Harford et al., 2002). An age of 331±35 ka (Harford et al., 2002) from the xenocrystal component of Gage’s dome suggests that the current superstructure of the SH is underlain by deposits from an earlier phase of activity.
The four small hills, Garibaldi Hill, Richmond Hill and St. George’s Hill on the southwest side of the island, and Roche’s Bluff on the southeastern coast are now regarded as uplifted volcanic sequences (Harford et al., 2002) rather than parasitic centers (Rea, 1970,1974). Garibaldi Hill has been 39Ar/40Ar dated at 282±8 ka, and is thought to represent part of the older superstructure of the SH. The deposits below Roche’s Bluff lava flow (dated at 130 ka) have yielded an 39Ar/40Ar age of 1021±20 ka and are now interpreted as forming part of an intensely disrupted submarine fan from the Centre Hills complex (Harford et al., 2002), rather than part of a parasitic center of the SSH (Rea, 1970, 1974).
Montserrat island is built on the south-central part of an elliptical submarine bank which at the 200m isobath measures 15 km east-west by 25 km north-south. The submarine bathymetry around the island, as well as the subaerial morphology and the location of epicenters associated with both the 1966-67, and 1992-95 seismic activity suggests that the island is traversed by a number of fault zones. One of these is morphologically represented by a linear topographic feature that crosses the island in an ENE-WSW direction from Farm River on the east coast to Bransby Point on the west. Earthquake epicenters, recorded between 1966 and 1967 (Tomblin, 1972), in the vicinity of Antigua , could define the easterly submarine extension of this fault. A possible westerly extension forms the northern boundary of the Montserrat Rise or Spur, a major submarine ridge extending 50 km WSW of the island (Fig. 3). Earthquake epicenters, particularly for the period July 27 to August 6, 1995 formed a belt parallel to this proposed fault trace (Aspinall et al., 1998).
The surface expression of a second possible fault zone which trends NW-SE can be seen as a scarp on the south side of Roche’s Bluff on the southeast coast and is particularly noticeable in the cliff line on the west coast at Old Road Bluff. The trace of this fault zone passes directly beneath SH. South of this fault line an 800m wide belt of topographic lows can be distinguished, represented by Dry Ghaut in the east, and the Belham River on the west. It has been proposed that prehistoric activity within this fault zone was responsible for the uplift of St. George’s , Richmond and Garibaldi Hills as well as Roche’s Bluff (Harford et al., 2002). More recent activity is suggested by the locations of earthquake epicenters during both the 1966-67 seismic crisis (Shepherd et al., 1971), and the initial stages of the current eruption, particularly during the period July 29 to August 15, 1995 (Aspinall et al., 1998). The intense hydrothermal alteration of the uplifted tuffs and sediments exposed on the east coast near Roche’s Bluff may also be associated with this fault. The trend of this fault is parallel to the trend of the “en echelon” fault zone formed by the Redonda, Bouillante-Montserrat, and Les Saintes fault systems described by Feuillet et al. (2002). GPS ground deformation studies of the Soufriere Hills volcano from October 1995 to July 1996 have suggested that up to 1m of expansion recorded during the current eruption was caused by the shallow (<3km) injection of a vertical dike into this zone (Mattioli et al., 1998).
Bathymetic studies around the island by Deplus and other (2001) have identified three separate debris avalanche deposits that radiate away from the island. The source areas of a fan to the southwest is just north of the now destroyed capital of Plymouth, a second to the south has a source in the White River area and a third to the west has a source in the tar River area which has to be the sector collapse scar of English’s Crater. The debris fans are smaller than those around south Dominica and are in shallower water at depths of less than 1000m. The fans extend up to 20km from the shoreline and have megablocks up to 400m across and 40m high.
MONTSERRAT ERUPTION SEQUENCE OF EVENTS 1992-2005
|
DATE |
ACTIVITY |
| 1992-1995 | Numerous precursor earthquake swarms |
| July 18, 1995 | Phreatic eruption – Start of eruption |
| July –Nov 1995 | Numerous phreatic eruptions |
| mid-Nov–Dec1995 | Initial dome growth |
| Dec 1-2, 1995 | Minor pyroclastic flows affect upper slopes of volcano |
| March 3, 1996 | Start of major pyroclastic flows down Tar River valley |
| May 1996 | Pyroclastic flows reach sea at mouth of Tar River |
| Sept 14, 1996 | Sub-plinian eruption, partial dome collapse |
| Nov-Dec 1996 | Growth of dome on western side of crater near crater wall connecting Chance’s and Galway ‘s domes |
| Feb 1997 | New dome overtops Galway ‘s wall |
| May 1997 | Dome growth switches to north and rate increases |
| June 1997 | Increase in dome growth and increase in production of pyroclastic flows down northern flank of volcano |
| June 25, 1997 | Pyroclastic flows move down northern valleys e.g Tuitts Ghaut, destroying various towns, e.g. Harris as well as airport airport, and surge-derived pyroclastic flow reaches vicinity of Cork Hill. Loss of 19 lives |
| July 4, 1997 | Pyroclastic flows begin to encroach on Plymouth |
| Aug 1997 | Sequences of Vulcanian explosions followed by fountain collapse and generation of semi-vesicular andesite pyroclastic flows |
| Sept 22-Oct 21 1997 | Second sequence of Vulcanian explosions |
| Dec 26, 1997 | Collapse of Galways wall and movememt of pyroclastic flows down western flanks of volcano |
| Feb 1998 | Vertical explosions resulted in heavy ash fall on central part of island |
| March 1998 | Cessation of magma dome growth |
| April 1998 | Dome substantially degraded by collapses |
| Nov 1999 | Renewed dome growth with minor ash venting and small collapse pyroclastic flows |
| Nov-July 2000 | Dome growth dominantly in east with growth and subsequent of numerous spines |
| Mar 20, 2000 | Major dome collapse with resulting pyroclastic flows |
| mid July 2000 | Dome growth switches to south and west side |
| Aug-Nov 2000 | Dome growth switches back to east and south east side, small pyroclastic flows affect northern flanks |
| Dec 8, 2000 | New dome overtops 95-98 dome |
| Jan- Mar 2001 | Continued dome growth mainly in east and south east with small collapses generating small pyroclastic flows in northern and eastern valleys (Tuiits Ghaut, White’s Ghaut , Tar River ) |
| Mar-May 2001 | Very little activity |
| May-July 2001 | Dome growth increased and switches to the south sector |
| July 29, 2001 | Major collapse, pyroclastic flows move down Tar River |
| Aug 2001 | Slowing of dome growth |
| Sept-Dec 2001 | Active dome growth which switches location from northeastern to eastern to westestern sectors accompanied by small to moderate collapses and generation of pyroclastic flows |
| Dec 28, 2001 | Collapse of NE flank of dome, significant pyroclastic flows moved down Tar river |
| Jan-June 2002 | Continued dome growth with generation and collapse of spines and continued switching of main locus of growth to different sectors; pyroclastic flows down Tar River and Tuitts |
| June-July 2002 | Slowing of dome growth |
| July-Oct 2002 | Renewed dome growth mainly in northern and northeastern sectors |
| July 23 and 26, 2002 | Minor collapses of dome and generation of pyroclatic flows in Tuits and White’s Ghauts |
| Aug 2002 | 95-98 dome almost completely covered by new dome |
| Aug –Dec 2002 | Continued growth of dome producing lobes in the northern, northeastern, northwestern and northern sectors respectively; numerous pyroclastic flows, with almost continuous activity on Sept 27, down northern flanks affecting Tuitts Ghaut, White’s Bottom Ghaut, Gages valley, Fort Ghaut , Tyer’s Ghaut and Tar River |
| Dec 2002-July 2003 | Continued growth of dome producing lobes and spines in northern- northeastern sector (Nov-April), eastern-southeastern (April-May), northeastern (May),eastern (June-July), northern (July) |
| Jan-July 2003 | Pyroclastic flows produced by dome collapse affecting White’s Ghaut, Tar River, Tuitts Ghaut, Fort Ghaut, Tyer’s Ghaut |
| March 7-14 2003 | General dome growth and pyroclastic flows down all flanks of volcano |
| July 11, 2003 | Dome collapse and significant pyroclastic flows affect Tar River , White’s Ghaut, Spanish Point; explosive vents formed in collapse scar |
| July 25-Aug 1 2003 | Ash venting and growth of small dome in collapse scar |
| Aug 2003-Feb 2003 | Very little activity |
| Mar 3, 2004 | Explosive activity and removal of new small dome that grew in collapse scar and portion of 95-98 dome; generation of pyroclastic flows |
| Mar 2004-July 2005 | No new dome growth, activity mainly steam and ash venting from collapse scar some of which is related to local weather conditions |
| June 20, 2005 | Phreatic eruption, pyroclastic flows down Tar River |
| July 9, 2005 | Phreatic eruption |