Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Soufriere Hills (Montserrat) Dome growth continues Soufriere Hills Montserrat, West Indies 16.72 N, 62.18 W; summit elev. 915 m The dome continued to extrude in the breached summit crater. During January, subtle to dramatic variations occurred in the location, style, and rate of growth (with some areas undergoing up to 1-m vertical rise per day). Numerous spines grew, fell, and shattered. Besides obtaining the first samples of the new dome, fieldworkers established that the emplacement of the old dome (Castle Peak) was accompanied by one or more pyroclastic flows and lahars. The total seismic energy release for the last week of January was the highest since early December. Cumulative deformation measurements suggested inflation of the edifice. Dome growth and visible observations. On 28-30 December 1995, the dome=FEs E side grew 3 m upwards. Local avalanches accompanied this growth, but by 1 January both the growth and avalanches in this area temporarily slowed to a stop. Adjacent Castle Peak, on the new dome=FEs S side where a small spine had formed on 26 December, volume increased without vertical growth in the week ending on 3 January. During that same week, significant steam escaped near the N part of the dome; the nearly continuous steam plume was sometimes charged with small amounts of ash. Although dome growth appeared slow during part of the week ending on 10 January, it did not cease. Observers noted local avalanches (off both the dome=FEs N and E sides) coupled with suggested swelling and new lava extrusions in the dome=FEs central region. During this same week, the September spine appeared to move and tilt, and in the subsequent week the spine was pushed S by new dome growth. A new spine was identified on 10 January in the dome=FEs center; this spine grew relatively rapidly until it fell down on 13 or 14 January. Other spines on the new dome also appeared to undergo a growth spurt. Another spine appeared just after 14 January in the N dome area. Large parts of this spine had fallen by 18 January, possibly contributing to airborne ash seen on two occasions that day; the next day large blocks of the broken spine lay on the talus slope. Yet another spine appeared around 18 January along the S edge of the dome; it grew for two days prior to fall and breakup. Spalling material created a substantial talus pile in the S moat. On 20-21 January another spine grew in roughly the same spot. During the next two days this spine reached 25 m in height and 15 m in basal diameter prior to its partial collapse (an event correlated with significant ash emission on 23 January). Late in the week ending on 24 January, growth of the dome=FEs S edge included growth of spines, spalling debris, slow swelling, and vertical growth. Steam emissions were generally high during mid-January, and observers first saw new dome material piling up against the crater=FEs W wall at the base of Chances Peak. The other side of the new dome extended a formidable distance up Castle Peak. The most pronounced dome growth in the final week of January took the form of swelling on the dome=FEs steeply sloping N margin. Although early in the week mass wasting repeatedly sent debris into the adjacent moat, later in the week this took place less frequently. On 26 January a spine was again noted in the dome=FEs S area, but growth there on 28 January was manifested as swelling. Beginning on about 29 January on the N and S ends of the new dome observers saw two elongate ridges trending NE-SW. These appeared as rough mirror images of each other, their forms resembling whale backs. Lofted ash and mass wasting. Airborne ash seen during January was mainly attributed to mass wasting. For example, a small amount of ash fell in Plymouth early on 4 January; the source was thought to have been a crater-confined rock avalanche off the dome. Minor ash fell in Plymouth four times on 12 January, and one time on both 15 and 16 January. Four ash emission events on 24 January were all associated with major rock-fall events on the lava dome. In some cases very minor ash emissions also issued directly from the dome and some of the material involved in mass wasting may have been dislodged by small explosions. Visual observations into the crater have enabled good correlation between seismic signals and rock-fall events. During the week ending on 24 January, heat production from the dome appeared somewhat higher than in the past. During January dome incandescence was reported and some material within the rock falls was hot. Rockfalls and avalanches remained confined within the crater area, although the moat continued to gradually fill with debris. Field studies. Good conditions on 8 January enabled the collection of a sample from the part of the new dome located within the 18 July vent, an area thought to have been extruded in late November. The crystal-rich sample contained dominant plagioclase, subordinate pyroxene and hornblende; parts of the sample were sent to four labs for further analysis. Around the same time, other fieldwork in flanking drainages (Hot River and Fort Ghaut) found new exposures and established that several charcoal-bearing pyroclastic units (including at least one pyroclastic flow unit) were erupted during the growth of Castle Peak dome. These were also found adjacent to deposits having the character of lahars. Deformation. EDM lines composed a network consisting of four surveyed triangles around the volcano. The lines continued to be measured routinely. Dry-tilt sites at Amersham and Brodericks on the volcano=FEs W side were re-occupied during the early part of the week ending 10 January; neither showed any change since their last occupation in October. During January, the Spring Hill tiltmeter failed and was moved to a new site, but the Long Ground tiltmeter continued to indicate angular stability. Looked at in the short term, EDM measurements taken during the first half of January did not show any changes in slant distance (above the error of the method); however, during the week ending 17 January it was reported that a slow shortening had occurred on many of the lines towards the volcano. The shortening indicated swelling of the edifice. In the week ending on 24 January, it was reported that slant line distance in the NE sector (Tar River to Castle Peak area) underwent a 2.5-cm shortening over the period of a month. During the shorter interval of the final week of January, no changes above the error of the method were detected in slant distance measurements on two deformation triangles in the volcano=FEs S to SW and NE sectors (the Galways-Chance=FEs Peak-O=FEGarra=FEs and Long Ground-White=FEs Yard-Castle Peak triangles). Seismicity. During January, broadband tremor commonly registered on the Gages station. Tremor was generally absent at the other stations, although the Long Ground station also registered some tremor in mid-January. The number of daily earthquakes typically measured in the thousands (eg. 5,000 to 6,000 events at the closest station on 27 January), too numerous to count on a real- time basis. Instead, MVO often quantified seismicity for rapid dissemination by using located events. These are events for which a hypocenter (the earthquake focus, the point at which the first motion originates) was calculated based on one S-wave and the records from four stations. An MVO report on 3 January stated that long-period events recorded at most seismic stations had been occurring at a rate of 10-15/day. The hypocenters for these events could not be found but they were thought to be at very shallow depths in the crater area. Later reports in January did not quantify the rate of occurrence for long-period events. Late on 5 January, broadband tremor picked up slightly in amplitude at Gages station. Then, small long-period events occurred for about the next 12 hours. This was followed by an 8- hour swarm of >300 hybrid events with virtually identical waveforms <3 km beneath the volcano. Lower amplitude, regular hybrid events occurred every 1-2 minutes until 8 January. A smaller series of similar hybrid events took place on 12-15 January. Some initially small hybrid events that first appeared on 23 January grew in amplitude and rate of occurrence (to 6- 7/minute) and continued until at least 31 January. During the last week of January these hybrid events formed the dominant seismic activity. Repeated shallow hybrid events in early January within the crater preceded new dome growth by a few days. This had happened on at least two previous occasions. During the first week of January, shallow (0-7 km depth) volcano- tectonic earthquakes with epicenters scattered around the volcano continued at a rate of 2-3/day. An exception was 1 January, when a cluster of 17 volcano-tectonic earthquakes took place just N of the active crater at 1-3 km depths. They occurred in an eight- hour period following an M 5.0 earthquake that struck at a depth of 25 km, centered ~55 km N of Port of Spain, Trinidad. This larger earthquake may have been the trigger for the 1 January seismicity. Three small, 1-3 km deep, volcano-tectonic earthquakes struck SW of the island during mid-January. Very occasional, small, long- period earthquakes started to appear on the evening of 28 January and a solitary volcano-tectonic earthquake took place early on the 29th. This M 2-2.5 earthquake was located beneath the crater area at a depth of 2.8 km; it was felt by Long Ground area residents. Crisis management. The eruption driving the current crisis began on 18 July 1995 (Bulletin v. 20, no. 6). According to the Montserrat Government Information Unit (on 13 February 1996), during the crisis there has been no official off-island evacuation. However, a phased relocation of 6,000 residents from the southern half of the island to the northern half immediately followed a large phreatic eruption on the morning of 21 August 1995. Ash from that eruption=FEs cloud, and from a density current that flowed down the flanks of the volcano, caused darkness in the capital (Plymouth) and surrounding areas, and ultimately deposited several millimeters of ash there. The relocation order was partially lifted on 3 September, a day before the passage of Hurricane Luis. A change in eruptive style in mid-November ultimately lead to the extrusion of lava at the surface. On 1 December 1995 a second relocation of 4,000 residents took place. The relocation lasted a month for residents on the island=FEs SW side and about a month- and-a-half for those on the island=FEs SE side. Some preparatory steps for future emergencies included the continued relocation of Glendon Hospital and newly acquired school buses to move residents. Information Contacts: Montserrat Volcano Observatory (MVO), c/o Chief Minister=FEs Office, P.O. Box 292, Plymouth, Montserrat. Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Virunga Volcanoes (Zaire) High levels of seismicity during September 1995 Virunga Volcanoes Zaire 1.41 S, 29.20 E; summit elev. 3,058 m (Nyamuragira) Volcanic seismicity in the Nyamuragira/Nyiragongo area during 15- 30 September was very high (figure 4). This activity was characterized by A-type (high-frequency) and C-type (low- frequency) events. Hypocenters were principally concentrated ESE of the summit of Nyamuragira, NE of Nyiragongo, at depths of 0-20 km. Depths for all of the earthquakes decreased from W to E and from S to N, suggesting a volcanic conduit rising in a generally NE direction towards the surface on the ESE flank of Nyamuragira. The level of volcanic tremor was very low. The low tremor level does not signify an overall reduction in the Nyiragongo lava lake activity, but no fresh lava was apparent in the crater. Seismic recordings in the first half of October were severely impaired by frequent power disruptions to the observatory and the regular discharge of the batteries. The number of earthquakes decreased significantly compared to September; events were mainly centered at greater depths (5-30 km) SE of Nyiragongo. Problems with the transmitter at the Kunene seismic station prevented acquisition of better data; scientists were unable to visit the station regularly due to a lack of tires for their vehicle. Seismicity during 15 November-2 December remained high, mainly consisting of A- and C-type earthquakes, and volcanic tremor. The distribution of hypocenters was similar to that observed during September. Each series of earthquakes was followed by up to several hours of tremor. Five seismic stations were operating during this period, four of them (Kibati, Rusayo, Buhimba, and Kunene) telemetered to the Goma observatory, and the fifth (Mt. Goma) connected by cable. However, the transmitter from Kunene was intermittent. Tremor amplitude remained low. Monitoring of both active Virunga volcanoes is done from a small observatory building located in Goma, ~18 km S of the Nyiragongo crater. Goma is the city where the major encampment of Rwandan civil-war refugees is located. A previous lava lake in the deep summit crater of Nyiragongo, active since 1894, drained suddenly on 10 January 1977, killing about 70 people. Lava lake activity resumed in June 1982, but had ceased by early 1983. The lava lake was again activated after an eruption that began in June 1994 (Bulletin v. 19, nos. 6-8). Historical eruptions from Nyamuragira (14 km NW of Nyiragongo) have occurred within the summit caldera and from numerous flank fissures and cinder cones. Twentieth century flank lava flows extend >30 km from the summit. Nyamuragira also began erupting in July 1994, producing lava fountaining, lava flows, and ash emission. Information Contacts: Wafula Mifundi and Mahinda Kasereka, Goma Volcano Observatory, Departement de Geophysique, Centre de Recherche en Sciences Naturelles, Lwiro, D.S. Bukavu, Zaire. Figure 4. Seismicity in the Nyamuragira/Nyiragongo area, 1 September-15 October 1995. Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Ruapehu (New Zealand) Geochemical analyses of lake water; record glacial retreat continues Ruapehu North Island, New Zealand 39.28 S, 175.57 E; summit elev. 2,797 m No significant eruptions are known to have occurred at Ruapehu since November 1995. During a crater lake inspection on 18 January 1996, in conditions of limited visibility, the crater floor was occupied by a 70-m diameter lake whose surface lay around 80 m below the overflow level. The lake was chiefly green- gray in color; the E part of the lake, beneath Pyramid Peak, was mostly clear. Vigorous fumaroles were present above the lake=FEs W and N shores, and beneath Pyramid Peak. At the Peak=FEs foot, new crater wall exposures revealed 20 m of steep fans composed of ash and deposits tentatively identified as lake sediments. Both of these deposits were being dissected and eroded into mudflows. A zone of 1995 ejecta, 5 m thick, included a scoria-and-block unit (probably erupted on 11 October) and fine-medium ash layers. Glaciers in the crater basin and Whangaehu continued their rapid, and in the experience of the IGNS observers, unprecedented, retreat. A tilt-leveling survey was conducted using four Dome benchmarks falling along a 45-m-long line radial to the crater. The benchmarks had moved since last measured. Assuming site instability at these benchmarks had developed since the latest pre-eruption survey (April 1994), the Dome=FEs deformation averaged 30 +- 10 microrad deflation. Both earthquakes and tremor started to decline in mid-October. >From 31 October-8 November, long-duration events almost completely ceased, while short-duration events continued unchanged. During the month of November through 18 January, the previously seen 2-Hz tremor ceased, replaced mainly by higher- frequency tremor. This 7-Hz tremor had amplitudes 3-4x higher than seen prior to the commencement of activity in September 1995. Numerous, short-duration high-frequency earthquakes continued to be recorded at the Dome station through December, January, and into February 1996, defining Ruapehu=FEs most recent background seismicity. Maximum fumarole temperatures in one area (station A, beneath peg I) had fallen from 281 degrees C (6 December 1995) to 92 degrees C (18 January). Despite the lower temperatures, the discharges were still vigorous but notably water-rich. H2/Ar temperatures fell only slightly over this period (from 397 degrees C to 392 degrees C), suggesting that the cooling reflects near-surface quenching by shallow groundwater. Analytical results for lake water samples are given in Table 1. These include samples taken on 6 December (by sampler suspended from a helicopter), and on 20 December 1995, and 18 January 1996 (from the lakeshore). A dilution trend with time is evident in the data, but it is not clear whether this represents an artifact of sampling or decreased magmatic input into the lake. So far, ratios of SO4/Cl and Mg/Cl show inconsistent trends. Much of the volcano=FEs discharged heat and gases evaded the lake, to be released instead beyond the lake=FEs NE and SW margins. A COSPEC flight on 31 December measured an SO2 flux of 1,295 +- 160 tons/day; this value was presumed to indicate continuing, steady state, open vent degassing. Information Contacts: P.M. Otway, S. Sherburn, and I. A. Nairn, Institute of Geological & Nuclear Sciences (IGNS), Private Bag 2000, Wairakei, New Zealand. Table 1. Lake water temperature and chemical analyses from Ruapehu, 6 and 20 December 1995, and 18 January 1996. Courtesy of IGNS. Date Temp Mg SO4 Cl SO4/CL Mg/Cl (deg C) 06 Dec 57.7 903 11,400 12,536 0.909 0.072 20 Dec 60.0 615 7,070 8,127 0.870 0.076 18 Jan 49.6 367 5,780 5,664 1.021 0.065 Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Rabaul (Papua New Guinea) Explosive eruptions from Tavurvur Rabaul New Britain Island, Papua New Guinea 4.27 S, 152.20 E; summit elev. 688 m Low-level activity from Tavurvur throughout January continued with plumes containing low to moderate ash contents. The plumes were released at intervals of 3-10 minutes, sometimes accompanied by weak roaring or explosion sounds. Incandescent lava fragments were ejected during some ash emissions. Large explosions occurred on 3 (2), 17 (2), 19 (1), and 24 (1) January. These explosions deposited lava blocks, as large as 60-80 cm in diameter, 1-1.5 km from the vent. The plumes rose 400-1,000 m above Tavurvur and were generally blown SE, but sometimes in a broad arc extending from the SW to the N. Ashfalls were recorded at Talwat village, in the Kokopo area, and in Rabaul Town. No vapor emissions were observed from Vulcan. Seismicity was higher in January compared to December, with three high-frequency earthquakes, 2,401 low-frequency earthquakes, and 1,404 explosion events. Discontinuous non-harmonic tremor also occurred during the month. High-frequency earthquakes that could be located occurred NE of the caldera. Except for three low- frequency earthquakes, which originated NW of the caldera, all the other seismicity was associated with eruptive activity at Tavurvur. The increase in the number of both types of events was also accompanied by an increase in their amplitudes. Information Contact: Ben Talai, Rabaul Volcano Observatory, P.O. Box 386, Rabaul, Papua New Guinea. Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Semeru (Indonesia) Explosions, pyroclastic flows, and lava avalanches continue Semeru E Java, Indonesia 8.11 N, 112.92 E; summit elev. 3,676 m Volcanic activity continued until the end of 1995 at a level of intensity comparable to August-October (Bulletin v. 20, no. 9). During November and December, small to moderate explosions and avalanches occurred from the Jonggring Seloko summit crater. The average plume height was 300-500 m. Pyroclastic avalanches descended along the Kembar river to a distance of 500-1,000 m from the summit. The number of lava avalanches increased during November, traveling down the Kembar river up to 300 m from the summit. On 27 December an incandescent lava flow traveled 500 m; during this event volcanic tremor was recorded with a maximum amplitude of 3 mm. The frequency of volcanic earthquakes (both A- and B- types) during the November-December period ranged from 1 to 10 events/day. Volcanic tremor was recorded on 1, 2, 17, and 18 November; tremor from 15 to 18 December increased to 3-8 events/day (figure 5). Explosion earthquakes were variable (31- 136 events/day), with two minima on 5 and 25 December (24 and 19 events/day, respectively) (figure 5). Typical activity for this volcano was observed by Steve O=FEMeara on 13 November from the rim of Tengger Caldera. Thick cauliflower-shaped columns of gray ash rose up to 200 m high every 15-20 minutes, and the sky to the SW was spotted with ash clouds from previous eruptions. One event lasted ~5 minutes, generating a dark gray ash cloud that caused ashfall on the S slopes before detaching from the summit. Another eruption cloud spilled over the SW rim and flowed downslope. Information Contact: Wimpy S. Tjetjep (Director), Volcanological Survey of Indonesia (VSI), Jalan Diponegoro 57, Bandung, Indonesia; Steve O=FEMeara, P.O. Box 218, Volcano, Hawaii 96785 USA (Email: 102022.111@compuserve.com). Figure 5. Eruptive activity at Semeru as detected by seismograph, November - December 1995: Daily number of tremor and volcanic earthquakes (top), explosions and avalanches (bottom). Courtesy of VSI. Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Merapi (Indonesia) Increased seismicity related to lava avalanches and rockfalls Merapi Java, Indonesia 7.54 S, 110.44 E; summit elev. 2,911 m During November-December 1995 glowing avalanches down the Boyong, Krasak, and Bebeng rivers reached up to 2.5 km from the source. Seismic activity was dominated by multiphase earthquakes, low- frequency earthquakes, and lava avalanches (rockfalls). The number of multiphase earthquakes decreased from 924 in November to 152 in December; low-frequency earthquakes also decreased from 74 in November to 42 in December. Seismicity associated with lava avalanches and rock falls increased from 816 events in November to 1,078 in December (figure 6). A deep volcanic earthquake (A- type) and two tremor events were recorded in November, three shallow volcanic earthquakes (B-type) occurred in December. Inflation increased since 17 November from 2.5 to 10.8 microrad/day. Tilt data collected at two stations in the summit area during November and December indicated inflation of 60 and 320 microrad, respectively. The geomagnetic intensity in early December was -14.5 nTs; it then decreased to -1.5 nTs by the end of the month. The emission rate of SO2 during November fluctuated between 27 and 275 t/d, averaging 95 t/d, and the plume velocity was ~3.2-3.5 m/s. In December the emission rate decreased to 70 t/d, fluctuating between 18 and 156 t/d; plume speed was slightly higher at 3.3-3.6 m/s. Merapi is one of the most active volcanoes in Indonesia. The stratovolcano has an exposed, summit lava dome, the source of abundant glowing avalanches directed down its S to W slopes toward a region settled by at least 50,000 people. In historical time, instability of the growing dome has led to pyroclastic flows that have caused many fatalities. Collapse of the growing lava dome in November 1994 triggered pyroclastic flows that killed 64 people (Bulletin v. 19, nos. 10 and 12). Additional collapses were reported in January and August-October 1995 (Bulletin v. 20, nos. 2 and 10). Information Contact: VSI (see Semeru); Steve O=FEMeara (see Semeru). Figure 6. Monthly number of rockfall, multiphase, and low-frequency earthquakes at Merapi, June-December 1995. Courtesy of VSI. Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Krakatau (Indonesia) Steaming and fumarolic activity; cone description Krakatau Indonesia 6.10 S, 105.42 E; summit elev. 813 m During an approved visit on 6 November, the volcano was steaming but not erupting. A large sulfur-stained plug of lava, ~20% the diameter of the summit cone, was bulging out of a black cinder cone just below and W of the summit; it appeared to be inside the SW margin of a broad depression. A smaller sulfur-stained plug was farther S in another depression. The landing site on the SE shore was a black-sand beach with tiny dunes of white pumice. While climbing the SE slope of the older cone, the party crossed water-eroded fields of pyroclastic material dotted with volcanic bombs. The ascent to the summit went through deep cinder deposits covered with a blanket of loose breadloaf-sized stones. From the summit complicated internal crater structures could be seen. One symmetrical cone (~100 m across) had a sulfur-lined inner rim that was fuming. The largest and most active fumaroles were inside this cone=FEs S rim. A smaller cone within the larger one was almost horseshoe-shaped and steeper to the S. Bombs on the summit cone were as large as 1-2 m in diameter. The current activity began in March 1994 (Bulletin v. 19, no. 4) after almost continuous activity from November 1992-October 1993. Activity was intermittent during 1995 (Bulletin v. 20, nos. 3, 6, & 7). This renowned volcano lies in the Sunda Strait between Java and Sumatra. Caldera collapse, perhaps in 416 A.D., formed a 7- km-wide caldera. Three subsequent volcanoes coalesced into pre- 1883 Krakatau Island. Caldera collapse during the 1883 eruption destroyed all but a remnant of one volcano. The post-collapse cone of Anak Krakatau (Child of Krakatau), located within the 1883 caldera, has had frequent eruptions since 1927. Information Contact: Steve O=FEMeara (see Semeru). Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Soputan (Indonesia) Vapor emission and intense tremor; possible high ash cloud Soputan Sulawesi, Indonesia 1.11 N, 124.73 E; summit elev. 1,784 m Activity in late 1995 consisted of whitish vapor emission to 25- 100 m above the summit. During November occasional volcanic tremors were recorded with a maximum amplitude of 1.5 mm. Aviation reports on 7 November indicated increased eruptive activity with an ash cloud rising as high as 4.5 km altitude. Satellite imagery showed a possible ash cloud extending 90 km to the SW. On 5 December, an increase in tremor amplitude up to 5 mm followed a tectonic earthquake felt throughout the Mimahassa Peninsula on Sulawesi. The same day maximum tremor amplitude reached 200 mm and glow was observed from three points on the lava dome. About an hour later tremor reached a maximum amplitude of 40 mm. On 6 December, tremor was still being recorded, but maximum amplitude had decreased to 2 mm. Soputan, a small conical volcano on the rim of Tondaro Caldera, has had frequent historical eruptions. The eruption that began in October 1991 continued through at least mid-1993 with lava domes rising above the crater rim. Continuous endogenous growth and intense fumarolic activity were noted in July 1994 (Bulletin v. 19, no. 8). Information Contacts: VSI (see Semeru); Bureau of Meteorology, Northern Territory Regional Office, P.O. Box 735, Darwin, NT 0801 Australia. Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Lokon-Empung (Indonesia) Variable seismicity and minor vapor emissions Lokon-Empung N Sulawesi, Indonesia 1.36 N, 124.79 E; summit elev. 1,580 m Vapor emission was observed during November-December 1995 and whitish-gray plumes rose to 100 m above the active crater. Seismicity during September-October 1995 consisted of 1-8 A-type events/day. On 1 November there were 46 A-type events recorded, followed by very low seismicity over the next ten days. Activity then increased from 12 November through 31 December, but was highly variable with 4-21 events/day. B-type events remained at 0-8 events/day. The present activity is located at Tompaluan crater, in the saddle between the peaks of Lokon (1,579 m) and Empung (1,340 m). About 10,000 people evacuated following an explosion in October 1991 accompanied by a pyroclastic flow; the eruption ended in January 1992. Information Contact: VSI (see Semeru). Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Karangetang (Indonesia) Small ash plumes, incandescent ejecta, and increased seismicity Karangetang Siau Island, Indonesia 2.47 N, 125.29 E; summit elev. 1,784 m On 9 and 16 November "thunderclaps" were heard from the summit. A gray plume 500 m high was observed, and incandescent ejecta rose 10-50 m above the summit at night. On 17 December thunderclaps were heard again and ejecta rose 100 m above the summit. Seismicity increased from 26 October until the end of 1995. Daily counts of deep volcanic (A-type) earthquakes fluctuated up to 116 (figure 7). Karangetang volcano, the most active of Sulawesi province, is located at the N end of Siau Island and contains a line of summit craters. Activity during the summer of 1994 was characterized by the irregular occurrence of lava flows, explosions, Merapi-style pyroclastic flows and lahars (Bulletin v. 19, no. 8). Information Contact: VSI (see Semeru). Figure 7. A-type seismicity at Karangetang, September-December 1995. Courtesy of VSI. Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Suwanose-jima (Japan) Small eruptions in January; nine explosions throughout 1995 Suwanose-jima Ryukyu Islands, Japan 29.53 N, 129.72 E; summit elev. 799 m Monitoring from the Sakurajima Volcanological Observatory revealed nine explosions from Suwanose-jima in 1995. According to the Japan Meteorological Agency and the Kagoshima Prefectural Government, small eruptions during 10-13 January 1996 sent plumes 300-600 m above the volcano and caused ashfall to the S. Activity has been high since 1950, with 1-2 ash emissions every month, and some Strombolian explosions.=20 This 8 x 5 km volcano island consists of basaltic andesite and andesite. The largest historical eruption occurred in 1813-14; as a result, thick scoria beds covered the residential area and lava flows descended WSW from a new 300-m-diameter crater. Large eruptions also occurred in 1884-85 and a lava flow reached the E coast. The island was evacuated for about 70 years, but there are now around 50 residents. Information Contact: Volcano Research Center, Earthquake Research Institute, University of Tokyo, Yayoi 1-1-1, Bunkyo-ku, Tokyo 113, Japan (URL: http://hakone.eri.u-tokyo.ac.jp/vrc/vrc.html). Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Sakura-jima (Japan) Explosive eruptions and ashfalls continue Sakura-jima Kyushu, Japan 31.58 N, 130.67 E; summit elev. 1,117 m Minami-dake crater was slightly active throughout January. The monthly total number of eruptions was 60, including 42 explosive ones. At the seismic station 2.3 km NW of Minami-dake crater (Station B), 601 earthquakes and 684 tremors were recorded. The highest ash plume of the month rose 2,300 m above the summit crater on the 21st. Ashfall measured at the Kagoshima Local Meteorological Observatory, 10 km W form the crater, was 41 g/m^2. The VRC noted that there were more than 200 eruptions in 1995; total amount of erupted material was estimated at 3-4 million tons by the Sakurajima Volcanological Observatory, Kyoto University. The latter has been observing continuous uplift on the N side of the volcano, implying accumulation of magma beneath the volcano. Sakura-jima volcano, located within the Aira Caldera, lies in the N half of Kagoshima Bay. The volcano formed an island that was joined to the Osumi Peninsula during an eruption in 1914. Frequent historical eruptions have deposited ash on the city of Kagoshima, which is located across Kagoshima Bay 7 km W of the summit. Sakura-jima=FEs frequent eruptions present a significant hazard to air transportation using Kagoshima=FEs airport. Information Contacts: Volcanological Division, Seismological and Volcanological Department, Japan Meteorological Agency (JMA), 1- 3-4 Ote-machi, Chiyoda-ku, Tokyo 100 Japan; VRC (see Suwanose- jima). Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Kuju (Japan) Earthquake swarm and eruptions on 13-14 January; continuous plume Kuju Kyushu, Japan 33.08 N, 131.25 E; summit elev. 1,788 m Frequent earthquakes during the night of 13 January and through the next morning were centered 3-4 km NW of the Hosho dome near Sujiyu spa; eruptions caused minor ashfall around the volcano. Instruments recorded 526 earthquakes during the 13-14 January episode, some of which were felt by local residents. Some earthquakes on 27 January were centered SW of the active dome. Overall, there were 861 earthquakes detected in January, but no tremor. The plume height remained at 100-300 m throughout most of the month. Scientists at the University of Tokyo noted that vesiculated glass was again observed in the 13 January material, and deflation near the crater area was continuing. The Kuju volcano group consists of more than 10 stratovolcanoes and lava domes NE of Aso caldera. Hosho lava dome is situated near the center of the 20 x 15 km complex. The Geological Survey of Japan reports historical phreatic or hydrothermal eruptions at Kuju in 1662, 1675, and 1738. Molten sulfur effusion may have occurred in 1675 and 1738. The youngest known magmatic activity, ~1.7 ka, produced the Kuro-dake lava dome at the E end of the complex, accompanied by pyroclastic flows. Information Contacts: JMA (see Sakura-jima); VRC (see Suwanose- jima). Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Fukutoku-okanoba (Japan) Discolored seawater observed for the second time in three months Fukutoku-okanoba Volcano Islands, Japan 24.28 N, 141.52 E; summit elev. -14 m (submarine) On 12, 22, and 23 January, an aviator from the Japan Marine Safety Agency (JMSA) reported distinct discoloration of seawater to yellowish green. Similar discoloration was seen during 25-28 November 1995 (Bulletin v. 20, no. 11/12). Prior to that, discolored seawater was last seen at this location in September 1993. Water discoloration has often been observed at Fukutoku-okanaba (Shin-Iwo-jima), 5 km NE of Minami-Iwo-jima island. Eruptions produced ephemeral islands in 1904, 1914, 1974, and 1986. A dark plume rose above the water surface in November 1992, and floating pumice was observed. Information Contact: JMA (see Sakura-jima). Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Minami-Hiyoshi (Japan) Discolored seawater plume 6 km long Minami-Hiyoshi Volcano Islands, Japan 23.507 N, 141.905 E; summit elev. -30 m (submarine) On 12 January, an aviator from the JMSA observed seawater discolored to yellowish green in an area 500 m wide and 6 km long, flowing from Minami-Hiyoshi seamount to the S. Discolored water was last observed in February 1992 (Bulletin v. 17, no. 2). Periodic water discoloration and water-spouting has been reported over this submarine volcano since 1975, when detonations and an explosion were also reported. Located 90 km SE of Minami-Iwo- Jima, Minami-Hiyoshi lies near the SE end of a coalescing chain of youthful seamounts, and is the only historically active vent. The morphologically youthful seamounts Kita-Hiyoshi and Naka- Hiyoshi lie to the NW, and Ko-Hiyoshi to the SE. Information Contact: JMA (see Sakura-jima). Smithsonian Institution, Bulletin of the Global Volcanism Network Volume 21, Number 1, January 1996 Unknown Submarine Activity (W Pacific Ocean) Acoustic signals in late January; source not locatable Unknown Submarine Activity western Pacific Ocean The JMA Ocean Bottom Seismograph off the Boso Peninsula (E of Tokyo) detected T-phase-like signals after 22 January, and clear T-phase signals on 27 January (figure 8). According to tentative analyses of arrival times at the detectors, the signals were interpreted to have propagated from the S. As of mid-February, JMA had not determined a specific source for these signals. However, discolored seawater was observed above two submarine volcanoes in the Volcano Islands during January: Minami-Hiyoshi on 12 January, and Fukutoku-okanoba on 12, 22, and 23 January. Information Contact: JMA (see Sakura-jima). Figure 8. Example of T-phase signals (spikes) detected by the Ocean Bottom Seismograph off the Boso Peninsula, Japan, 27 January 1996. Courtesy of JMA.