Sakurajima rises from Kagoshima Bay, which fills the Aira Caldera near the southern tip of Japan's Kyushu Island. Frequent explosive and occasional effusive activity has been ongoing for centuries. The Minamidake summit cone has been the location of persistent activity since 1955; the Showa crater on its E flank has also been intermittently active since 2006. Numerous explosions and ash-bearing emissions have been occurring each month at either Minamidake or Showa crater since the latest eruptive episode began in late March 2017. This report covers ongoing activity from January through June 2019; the Japan Meteorological Agency (JMA) provides regular reports on activity, and the Tokyo VAAC (Volcanic Ash Advisory Center) issues tens of reports each month about the frequent ash plumes.

From January to June 2019, ash plumes and explosions were usually reported multiple times each week. The quietest month was June with only five eruptive events; the most active was March with 29 (table 21). Ash plumes rose from a few hundred meters to 3,500 m above the summit during the period. Large blocks of incandescent ejecta traveled as far as 1,700 m from the Minamidake crater during explosions in February and April. All the activity originated in the Minamidake crater; the adjacent Showa crater only had a mild thermal anomaly and fumarole throughout the period. Satellite imagery identified thermal anomalies inside the Minamidake crater several times each month.

Table 21. Monthly summary of eruptive events recorded at Sakurajima's Minamidake crater in Aira caldera, January-June 2019. The number of events that were explosive in nature are in parentheses. No events were recorded at the Showa crater during this time. Data courtesy of JMA (January to June 2019 monthly reports).

There were eight eruptive events reported by JMA during January 2019 at the Minamidake summit crater of Sakurajima. They occurred on 3, 6, 7, 9, 17, and 19 January (figure 76). Ash plume heights ranged from 600 to 2,100 m above the summit. The largest explosion, on 9 January, generated an ash plume that rose 2,100 m above the summit crater and drifted E. In addition, incandescent ejecta was sent 800-1,100 m from the summit. Incandescence was visible at the summit on most clear nights. During an overflight on 18 January no significant changes were noted at the crater (figure 77). Infrared thermal imaging done on 29 January indicated a weak thermal anomaly in the vicinity of the Showa crater on the SE side of Minamidake crater. The Kagoshima Regional Meteorological Observatory (KRMO) (11 km WSW) recorded ashfall there during four days of the month. Satellite imagery indicated thermal anomalies inside Minamidake on 7 and 27 January (figure 77).

Figure 76. Incandescent ejecta and ash emissions characterized activity from Sakurajima volcano at Aira during January 2019. Left: A webcam image showed incandescent ejecta on the flanks on 9 January 2019, courtesy of JMA (Explanation of volcanic activity in Sakurajima, January 2019). Right: An ash plume rose hundreds of meters above the summit, likely also on 9 January, posted on 10 January 2019, courtesy of Mike Day.

Figure 77. The summit of Sakurajima consists of the larger Minamidake crater and the smaller Showa crater on the E flank. Left: The Minamidake crater at the summit of Sakurajima volcano at Aira on 18 January 2019 seen in an overflight courtesy of JMA (Explanation of volcanic activity in Sakurajima, March 2019). Right: Two areas of thermal anomaly were visible in Sentinel-2 satellite imagery on 27 January 2019. "Geology" rendering (bands 12, 4, and 2) courtesy of Sentinel Hub Playground.

Activity increased during February 2019, with 15 eruptive events reported on days 1, 3, 7, 8, 10, 13, 14, 17, 22, 24, and 27. Ash plume heights ranged from 600-2,300 m above the summit, and ejecta was reported 300 to 1,700 m from the crater in various events (figure 78). KRMO reported two days of ashfall during February. Satellite imagery identified thermal anomalies at the crater on 6 and 26 February, and ash plumes on 21 and 26 February (figure 79).

Figure 78. An explosion from Sakurajima at Aira on 7 February 2019 sent ejecta up to 1,700 m from the Minamidake summit crater. Courtesy of JMA (Explanation of volcanic activity in Sakurajima, February 2019).

Figure 79. Thermal anomalies and ash emissions were captured in Sentinel-2 satellite imagery on 6, 21, and 26 February 2019 originating from Sakurajima volcano at Aira. Top: Thermal anomalies within the summit crater were visible underneath steam and ash plumes on 6 and 26 February (closeup of bottom right photo). Bottom: Ash emissions on 21 and 26 February drifted SE from the volcano. "Geology" rendering (bands 12, 4, and 2) courtesy of Sentinel Hub Playground.

The number of eruptive events continued to increase during March 2019; there were 29 events reported on numerous days (figures 80 and 81). An explosion on 14 March produced an ash plume that rose 3,500 m above the summit and drifted E. It also produced ejecta that landed 800-1,100 m from the crater. During an overflight on 26 March a fumarole was the only activity in Showa crater. KRMO reported 14 days of ashfall during the month. Satellite imagery identified an ash plume on 13 March and a thermal anomaly on 18 March (figure 82).

Figure 80. A large ash emission from Sakurajima volcano at Aira was photographed by a tourist on the W flank and posted on 1 March 2019. Courtesy of Kratü.

Figure 81. An ash plume from Sakurajima volcano at Aira on 18 March 2019 produced enough ashfall to disrupt the trains in the nearby city of Kagoshima according to the photographer. Image taken from about 20 km away. Courtesy of Tim Board.

Figure 82. An ash plume drifted SE from the summit of Sakurajima volcano at Aira on 13 March (left) and a thermal anomaly was visible inside the Minamidake crater on 18 March 2019 (right). "Geology" rendering (bands 12, 4, and 2) courtesy of Sentinel Hub Playground.

A decline in activity to only ten eruptive events on days 7, 13, 17, 22, and 25 was reported by JMA for April 2019. An explosion on 7 April sent ejecta up to 1,700 m from the crater. Another explosion on 13 April produced an ash plume that rose 2,200 m above the summit. Most of the eruptive events at Sakurajima last for less than 30 minutes; on 22 April two events lasted for almost an hour each producing ash plumes that rose 1,400 m above the summit. Ashfall at KRMO was reported during seven days in April. Two distinct thermal anomalies were visible inside the Minamidake crater on both 12 and 27 April (figure 83).

Figure 83. Two thermal anomalies were present inside Minamidake crater at the summit of Sakurajima volcano at Aira on 12 (left) and 27 (right) April 2019. "Geology" rendering (bands 12, 4, and 2) courtesy of Sentinel Hub Playground.

There were 15 eruptive events during May 2019. An event that lasted for two hours on 12 May produced an ash plume that rose 2,900 m from the summit and drifted NE (figure 84). The Meteorological Observatory reported 14 days with ashfall during the month. Two thermal anomalies were present in satellite imagery in the Minamidake crater on both 17 and 22 May.

Figure 84. An ash plume rose 2,900 m above the summit of Sakurajima at Aira on 12 May 2019 (left); incandescent ejecta went 1,300 m from the summit crater on 13 May. Courtesy of JMA (Explanation of volcanic activity in Sakurajima, May 2019).

During June 2019 five eruptive events were reported, on 11, 13, and 24 June; the event on 11 June lasted for almost two hours, sent ash 2,200 m above the summit, and produced ejecta that landed up to 1,100 m from the crater (figure 85). Five days of ashfall were reported by KRMO.

Figure 85. A large ash plume on 11 June 2019 rose 2,200 m above the summit of Sakurajima volcano at Aira. Courtesy of Aone Wakatsuki.

Geologic Background. The Aira caldera in the northern half of Kagoshima Bay contains the post-caldera Sakurajima volcano, one of Japan's most active. Eruption of the voluminous Ito pyroclastic flow accompanied formation of the 17 x 23 km caldera about 22,000 years ago. The smaller Wakamiko caldera was formed during the early Holocene in the NE corner of the Aira caldera, along with several post-caldera cones. The construction of Sakurajima began about 13,000 years ago on the southern rim of Aira caldera and built an island that was finally joined to the Osumi Peninsula during the major explosive and effusive eruption of 1914. Activity at the Kitadake summit cone ended about 4850 years ago, after which eruptions took place at Minamidake. Frequent historical eruptions, recorded since the 8th century, have deposited ash on Kagoshima, one of Kyushu's largest cities, located across Kagoshima Bay only 8 km from the summit. The largest historical eruption took place during 1471-76.

Information Contacts: Japan Meteorological Agency (JMA), Otemachi, 1-3-4, Chiyoda-ku Tokyo 100-8122, Japan (URL: http://www.jma.go.jp/jma/indexe.html); Tokyo Volcanic Ash Advisory Center (VAAC), 1-3-4 Otemachi, Chiyoda-ku, Tokyo, Japan (URL: http://ds.data.jma.go.jp/svd/vaac/data/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground); Mike Day, Minnesota, Twitter (URL: https://twitter.com/MikeDaySMM, photo at https://twitter.com/MikeDaySMM/status/1083489400451989505/photo/1); Kratü, Twitter (URL: https://twitter.com/TalesOfKratue, photo at https://twitter.com/TalesOfKratue/status/1101469595414589441/photo/1); Tim Board, Japan, Twitter (URL: https://twitter.com/Hawkworld_, photo at https://twitter.com/Hawkworld_/status/1107789108754038789); Aone Wakatsuke, Twitter (URL: https://twitter.com/AoneWakatsuki, photo at https://twitter.com/AoneWakatsuki/status/1138420031258210305/photo/3).

After a large, deadly explosive and effusive eruption during 1963-64, Indonesia's Mount Agung on Bali remained quiet until a new eruption began in November 2017 (BGVN 43:01). Lava emerged into the summit crater at the end of November and intermittent ash plumes rose as high as 3 km above the summit through the end of the year. Activity continued throughout 2018 with explosions that produced ash plumes rising multiple kilometers above the summit, and the slow effusion of the lava within the summit crater (BGVN 43:08, 44:02). Information about the ongoing eruptive episode comes from Pusat Vulkanologi dan Mitigasi Bencana Geologi (PVMBG), also known as the Indonesian Center for Volcanology and Geological Hazard Mitigation (CVGHM), the Darwin Volcanic Ash Advisory Center (VAAC), and multiple sources of satellite data. This report covers the ongoing eruption from February through May 2019.

Intermittent but increasingly frequent and intense explosions with ash emissions and incandescent ejecta characterized activity at Agung during February through May 2019. During February, explosions were reported three times; events on seven days in March were documented with ash plumes and ashfall in surrounding villages. Five significant events occurred during April; two involved incandescent ejecta that traveled several kilometers from the summit, and ashfall tens of kilometers from the volcano. Most of the five significant events reported in May involved incandescent ejecta and ashfall in adjacent villages; air traffic was disrupted during the 24 May event. Ash plumes in May reached altitudes over 7 km multiple times. Thermal activity increased steadily during the period, according to both the MIROVA project (figure 44) and MODVOLC thermal alert data. MAGMA Indonesia reported at the end of May 2019 that the volume of lava within the summit crater remained at about 25 million m³; satellite information indicated continued thermal activity within the crater. Alert Level III (of four levels) remained in effect throughout the period with a 4 km exclusion radius around the volcano.

Figure 44. Thermal activity at Agung from 4 September 2018 through May 2019 was variable. The increasing frequency and intensity of thermal events was apparent from February-May. Courtesy of MIROVA.

Steam plumes rose 30-300 m high daily during February 2019. The Agung Volcano Observatory (AVO) and PVMBG issued a VONA on 7 February (UTC) reporting an ash plume, although it was not visible due to meteoric cloud cover. Incandescence, however, was observed at the summit from webcams in both Rendang and Karangasem City (16 km SE). The seismic event associated with the explosion lasted for 97 seconds. A similar event on 13 February was also obscured by clouds but produced a seismic event that lasted for 3 minutes and 40 seconds, and ashfall was reported in the village of Bugbug, about 20 km SE. On 22 February a gray ash plume rose 700 m from the summit during a seismic event that lasted for 6 minutes and 20 seconds (figure 45). The Darwin VAAC reported the plume visible in satellite imagery moving W at 4.3 km altitude. It dissipated after a few hours, but a hotspot remained visible about 10 hours later.

Figure 45. An ash plume rose from the summit of Agung on 22 February 2019, viewed from the Besakih temple, 7 km SW of the summit. Courtesy of PunapiBali.

Persistent steam plumes rose 50-500 m from the summit during March 2019. An explosion on 4 March was recorded for just under three minutes and produced ashfall in Besakih (7 km SW); no ash plume was observed due to fog. A short-lived ash plume rose to 3.7 km altitude and drifted SE on 8 March (UTC) 2019. The seismic event lasted for just under 4 minutes. Ash emissions were reported on 15 and 17 March to 4.3 and 3.7 km altitude, respectively, drifting W (figure 46). Ashfall from the 15 March event spread NNW and was reported in the villages of Kubu (6 km N), Tianyar (14 km NNW), Ban, Kadundung, and Sukadana. MAGMA Indonesia noted that two explosions on the morning of 17 March (local time) produced gray plumes; the first sent a plume to 500 m above the summit drifting E and lasted for about 40 seconds, while the second plume a few hours later rose 600 m above the crater and lasted for 1 minute and 16 seconds. On 18 March an ash plume rose 1 km and drifted W and NW. An event on 20 March was measured only seismically by PVMBG because fog prevented observations. An eruption on 28 March produced an ash plume 2 km high that drifted W and NW. The seismic signal for this event lasted for about two and a half minutes. The Darwin VAAC reported the ash plume at 5.5 km altitude, dissipating quickly to the NW. No ash was visible four hours later, but a thermal anomaly remained at the summit (figure 47). Ashfall was reported in nearby villages.

Figure 46. Ash plumes from Agung on 15 (left) and 17 (right) March 2019 resulted in ashfall in communities 10-20 km from the volcano. Courtesy of PVMBG and MAGMA Indonesia (Information on G. Agung Eruption, 15 March 2019 and Gunung Agung Eruption Press Release March 17, 2019).

Figure 47. A thermal anomaly was visible through thick cloud cover at the summit of Agung on 29 March 2019 less than 24 hours after a gray ash plume was reported 2,000 m above the summit. "Atmospheric Penetration" rendering (bands 12, 11, and 8A) courtesy of Sentinel Hub Playground.

The first explosion of April 2019 occurred on the 3rd (UTC); PVMBG reported the dense gray ash plume 2 km above the summit drifting W. A few hours later the Darwin VAAC raised the altitude to 6.1 km based on infrared temperatures in satellite imagery. The seismic signal lasted for three and a half minutes and the explosion was heard at the PGA Post in Rendang (12 km SW). Incandescent material fell within a radius of 2-3 km, mainly on the S flank (figure 48). Ashfall was reported in the villages of Telungbuana, Badeg, Besakih, Pempatan, Teges, and Puregai on the W and S flanks (figure 49). An explosion on 11 April also produced a dense gray ash plume that rose 2 km above the summit and drifted W. A hotspot remained about six hours later after the ash dissipated.

Figure 48. Incandescent ejecta appeared on the flanks of Agung after an eruption on 4 April 2019 (local time) as viewed from the observation post in Rendang (8 km SW). Courtesy of Jamie Sincioco.

Figure 49. Ashfall in a nearby town dusted mustard plants on 4 April 2019 from an explosion at Agung the previous day. Courtesy of Pantau.com (Photo: Antara / Nyoman Hendra).

PVMBG reported an eruption visible in the webcam early on 21 April (local time) that rose to 5.5 km altitude and drifted SW. The ash spread W and S and ash fell around Besakih (7 km SW), Rendang (8 km SW), Klungkung (25 km S), Gianyar (20 km WSW), Bangli (17 km WNW), Tabanan (50 km WSW), and at the Ngurah Rai-Denpasar Airport (60 km SW). About 15 hours later a new explosion produced a dense gray ash plume that rose to 3 km above the summit and produced incandescent ejecta in all directions as far as 3 km away (figure 50). The ash spread to the S and ashfall was reported in Besakih, Rendang, Sebudi (6 km SW), and Selat (12 km SSW). Both of the explosions were heard in Rendang and Batulompeh. The incandescent ejecta from the explosions remained within the 4-km exclusion zone. A satellite image on 23 April showed multiple thermal anomalies within the summit crater (figure 51). A dense gray plume drifted E from Agung on 29 April (30 April local time) at 4.6 km altitude. It was initially reported by ground observers, but was also visible in multispectral satellite imagery for about six hours before dissipating.

Figure 50. An explosion at Agung on 21 April 2019 sent incandescent eject 3,000 m from the summit. Courtesy of MAGMA Indonesia (Gunung Agung Eruption Press Release April 21, 2019).

Figure 51. Multiple thermal anomalies were still present within the summit crater of Agung on 23 April 2019 after two substantial explosions produced ash and incandescent ejecta around the summit two days earlier. "Atmospheric Penetration" rendering (bands 12, 11, and 8A) courtesy of Sentinel Hub Playground.

PVMBG reported an eruption on 3 May 2019 that was recorded on a seismogram with a signal that lasted for about a minute. Satellite imagery reported by the Darwin VAAC showed a growing hotspot and possible ash near the summit at 4.3 km altitude moving NE. A few days later, on 6 May, a gray ash plume rose to 5.2 km altitude and drifted slowly W before dissipating; it was accompanied by a seismic signal that lasted for about two minutes. Explosions on 12 and 18 May produced significant amounts of incandescent ejecta (figure 52). The seismic signal for the 12 May event lasted for about two minutes; no plume was observed due to fog, but incandescent ejecta was visible on the flanks and the explosion was heard at Rendang. The Darwin VAAC reported an ash plume from the explosion on 17 May (18 May local time) at 6.1 km altitude in satellite imagery moving E. They revised the altitude a short while later to 7.6 km based on IR temperature and movement; the plume drifted N, NE, and E in light and variable winds. A few hours after that it was moving NE at 7.6 km altitude and SE at 5.5 km altitude; this lasted for about 12 hours until it dissipated. Ashfall was reported in villages downwind including Cutcut, Tongtongan, Bonyoh (20 km WNW), and Temakung.

Figure 52. Explosions on 12 (left) and 18 (right) May (local time) 2019 produced substantial ejecta on the flanks of Agung visible from a distance of 10 km or more in PVMBG webcams. The ash plume from the 18 May event resulted in ashfall in numerous communities downwind. Courtesy of PVMBG (Information Eruption G. Agung, May 13, 2019, Information Eruption G. Agung, May 18, 2019).

The initial explosion on 18 May was captured by a webcam at a nearby resort and sent incandescent ejecta hundreds of meters down the NE flank within 20 seconds (figure 53). Satellite imagery on 3, 8, 13, and 18 May indicated multiple thermal anomalies growing stronger at the summit. All of the images were captured within 24 hours of an explosive event reported by PVMBG (figure 54).

Figure 53. The 18 May 2019 explosion at Agung produced an ash plume that rose to over 7 km altitude and large bombs of incandescent material that traveled hundreds of meters down the NE flank within the first 20 seconds of the explosion. Images taken from a private webcam located 12 km NE. Courtesy of Volcanoverse, used with permission.

Figure 54. Satellite images from 3, 8, 13, and 18 May 2019 at Agung showed persistent and increasing thermal anomalies within the summit crater. All images were captured within 24 hours of explosions reported by PVMBG. "Atmospheric Penetration" rendering (bands 12, 11, and 8A) courtesy of Sentinel Hub Playground.

PVMBG issued a VONA on 24 May 2019 reporting a new ash emission. They indicated that incandescent fragments were ejected 2.5-3 km in all directions from the summit, and the seismic signal lasted for four and a half minutes (figure 55). A dense gray ash plume was observed from Tulamben on the NE flank rising 2 km above the summit. Satellite imagery indicated that the plume drifted SW and ashfall was reported in the villages of Besakih, Pempatan, Menanga, Sebudi, Muncan, Amerta Bhuana, Nongan, Rendang, and at the Ngurah Rai Airport in Denpassar. Additionally, ashfall was reported in the districts of Tembuku, Bangli, and Susut (20 km SW). The Darwin VAAC reported an ash plume visible in satellite imagery at 4.6 km altitude along with a thermal anomaly and incandescent lava visible in webcam imagery. The remains of the ash plume were about 170 km S of the airport in Denpasar (60 km SW) and had nearly dissipated 18 hours after the event. According to a news article several flights to and from Australia were cancelled or diverted, though the International Gusti Ngurah Rai (IGNR) airport was not closed. On 31 May another large explosion produced the largest ash plume of the report period, rising more than 2 km above the summit (figure 56). The Darwin VAAC reported its altitude as 8.2 km drifting ESE visible in satellite data. It split into two plumes, one drifted E at 8.2 km and the other ESE at 6.1 km altitude, dissipating after about 20 hours.

Figure 55. A large explosion at Agung on 24 May 2019 produced incandescent ejecta that covered all the flanks and dispersed ash to many communities to the SW. Courtesy of PVMBG (Gunung Agung Eruption Press Release 24 May 2019 20:38 WIB, Kasbani, Ir., M.Sc.).

Figure 56. An explosion at Agung on 31 May 2019 sent an ash plume to 8.2 km altitude, the highest for the report period. Courtesy of Sutopo Purwo Nugroho, BNPB.

Geologic Background. Symmetrical Agung stratovolcano, Bali's highest and most sacred mountain, towers over the eastern end of the island. The volcano, whose name means "Paramount," rises above the SE caldera rim of neighboring Batur volcano, and the northern and southern flanks extend to the coast. The summit area extends 1.5 km E-W, with the high point on the W and a steep-walled 800-m-wide crater on the E. The Pawon cone is located low on the SE flank. Only a few eruptions dating back to the early 19th century have been recorded in historical time. The 1963-64 eruption, one of the largest in the 20th century, produced voluminous ashfall along with devastating pyroclastic flows and lahars that caused extensive damage and many fatalities.

Information Contacts: Pusat Vulkanologi dan Mitigasi Bencana Geologi (PVMBG, also known as Indonesian Center for Volcanology and Geological Hazard Mitigation, CVGHM), Jalan Diponegoro 57, Bandung 40122, Indonesia (URL: http://www.vsi.esdm.go.id/); MAGMA Indonesia, Kementerian Energi dan Sumber Daya Mineral (URL: https://magma.vsi.esdm.go.id/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground); The Jakarta Post, Mount Agung eruption disrupts Australian flights, (URL: https://www.thejakartapost.com/news/2019/05/25/mount-agung-eruption-disrupts-australian-flights.html); PunapiBali (URL: http://punapibali.com/, Twitter: https://twitter.com/punapibali, image at https://twitter.com/punapibali/status/1098869352588288000/photo/1); Jamie S. Sincioco, Phillipines (URL: Twitter: https://twitter.com/jaimessincioco. Image at https://twitter.com/jaimessincioco/status/1113765842557104130/photo/1); Pantau.com (URL: https://www.pantau.com/berita/erupsi-gunung-agung-sebagian-wilayah-bali-terpapar-hujan-abu?utm_source=dlvr.it&utm_medium=twitter); Volcanoverse (URL: https://www.youtube.com/channel/UCi3T_esus8Sr9I-3W5teVQQ); Sutopo Purwo Nugroho, BNPB (Twitter: @Sutopo_PN, URL: https://twitter.com/Sutopo_PN ).

Frequently active, Indonesia's Mount Kerinci on Sumatra has been the source of numerous moderate explosive eruptions since its first recorded eruption in 1838. Intermittent explosions with ash plumes, usually multiple times per month, have characterized activity since April 2018. Similar activity continued during February-May 2019, the period covered in this report with information provided primarily by the Indonesian volcano monitoring agency, Pusat Vulkanologi dan Mitigasi Bencana Geologi (PVMBG, also known as Indonesian Center for Volcanology and Geological Hazard Mitigation, CVGHM), MAGMA Indonesia, notices from the Darwin Volcano Ash Advisory Center (Darwin VAAC), and satellite data. PVMBG has maintained an Alert Level II (of 4) at Kerinci for several years.

On 13 February 2019 the Kerinci Volcano Observatory (KVO), part of PVMBG, noted a brownish-white ash emission that was drifting NE about 400 m above the summit. The seismicity during the event was dominated by continuous volcanic tremor. A brown ash emission was reported on 7 March 2019 that rose to 3.9 km altitude and drifted NE. Ash also drifted 1,300 m down the SE flank. Another ash plume the next morning drifted W at 4.5 km altitude, according to KVO. On 10, 11, and 13 March KVO reported brown ash plumes drifting NE from the summit at about 4.0-4.3 km altitude. The Darwin VAAC observed continuous ash emissions in satellite imagery on 15 March drifting W at 4.3 m altitude that dissipated after about 3 hours (figure 10). A gray ash emission was reported on 19 March about 600 m above the summit drifting NE; local news media noted that residents of Kayo Aro reported emissions on both 18 and 19 March (figure 11). An ash emission appeared in satellite imagery on 25 March (figure 10). On 30 March the observatory reported two ash plumes; a brown emission at 0351 UTC and a gray emission at 0746 UTC that both drifted NE at about 4.4 km altitude and dissipated within a few hours. PVMBG reported another gray ash plume the following day at a similar altitude.

Figure 10. Sentinel-2 satellite imagery of Kerinci from 15 (left) and 25 (right) March 2019 showed evidence of ash plumes rising from the summit. Kerinci's summit crater is about 500 m wide. "Geology" rendering (bands 12, 4, 2), courtesy of Sentinel Hub Playground.

Figure 11. Dense ash plumes from Kerinci were reported by local news media on 18 and 19 March 2019. Courtesy of Nusana Jambi.

Activity continued during April with a brown ash emission reported on 3 April by several different agencies; the Darwin VAAC and PVMBG daily reports noted that the plume was about 500 m above the summit (4.3 km altitude) drifting NE. KVO observed two brown ash emissions on 13 April (UTC) that rose to 4.2 km altitude and drifted NE. Satellite imagery showed minor ash emissions from the summit on 14 April; steam plumes 100-500 m above the summit characterized activity for the remainder of April (figure 12).

Figure 12. A dilute ash emission rose from the summit of Kerinci on 14 April 2019 (left); only steam emissions were present on a clear 29 April in Sentinel-2 imagery (right). "Geology" rendering (bands 12, 4, 2), courtesy of Sentinel Hub Playground.

Ashfall on the NE and S flanks within 7 km of the volcano was reported on 2 May 2019. According to a news article, at least five villages were affected late on 2 May, including Tanjung Bungo, Sangir, Sangir Tengah, Sungai Rumpun, and Bendung Air (figures 13 and 14). The smell of sulfur was apparent in the villages. Brown ash emissions were observed on 3 and 4 May that rose to 4.6 and 4.1 km altitude and drifted SE. The Darwin VAAC reported an emission on 5 May, based on a pilot report, that rose to 6.7 km altitude and drifted NE for about an hour before dissipating. A brown ash emission on 10 May rose 700 m above the summit and drifted SE. Satellite imagery captured ash emissions from the summit on 14 and 24 May (figure 15). For the remainder of the month, 300-700-m-high dense steam plumes were noted daily until PVMBG reported white and brown plumes on 26 and 27 May rising 500-1,000 m above the summit. Although thermal anomalies were not reported during the period, persistent weak SO₂ emissions were identified in TROPOMI instrument satellite data multiple times per month (figure 16).

Figure 13. Ashfall was reported from five villages on the flanks of Kerinci on 2 May 2019. Courtesy of Uzone.

Figure 14. An ash plume at Kerinci rose hundreds of meters on 2 May 2019; ashfall was reported in several nearby villages. Courtesy of Kerinci Time.

Figure 15. Ash emissions from Kerinci were captured in Sentinel-2 satellite imagery on 14 (left) and 24 (right) May 2019. The summit crater is about 500 m wide. "Geology" rendering (bands 12, 4, 2), courtesy of Sentinel Hub Playground.

Figure 16. Weak SO2 anomalies from Kerinci emissions were captured by the TROPOMI instrument on the Sentinel-5P satellite multiple times each month from February to May 2019. Courtesy of NASA Goddard Space Flight Center.

Geologic Background. Gunung Kerinci in central Sumatra forms Indonesia's highest volcano and is one of the most active in Sumatra. It is capped by an unvegetated young summit cone that was constructed NE of an older crater remnant. There is a deep 600-m-wide summit crater often partially filled by a small crater lake that lies on the NE crater floor, opposite the SW-rim summit. The massive 13 x 25 km wide volcano towers 2400-3300 m above surrounding plains and is elongated in a N-S direction. Frequently active, Kerinci has been the source of numerous moderate explosive eruptions since its first recorded eruption in 1838.

Information Contacts: Pusat Vulkanologi dan Mitigasi Bencana Geologi (PVMBG, also known as Indonesian Center for Volcanology and Geological Hazard Mitigation, CVGHM), Jalan Diponegoro 57, Bandung 40122, Indonesia (URL: http://www.vsi.esdm.go.id/); MAGMA Indonesia, Kementerian Energi dan Sumber Daya Mineral (URL: https://magma.vsi.esdm.go.id/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground); Global Sulfur Dioxide Monitoring Page, Atmospheric Chemistry and Dynamics Laboratory, NASA Goddard Space Flight Center (NASA/GSFC), 8800 Greenbelt Road, Goddard, Maryland, USA (URL: https://so2.gsfc.nasa.gov/); Nuansa Jambi, Informasi Utama Jambi: (URL: https://nuansajambi.com/2019/03/20/gunung-kerinci-semburkan-asap-tebal/); Kerinci Time (URL: https://kerincitime.co.id/gunung-kerinci-semburkan-abu-vulkanik.html); Uzone.id (URL: https://news.uzone.id/gunung-kerinci-erupsi-5-desa-tertutup-abu-tebal).

Suwanosejima is an active volcanic island south of Japan in the Ryuku islands with recent activity centered at Otake crater. The current eruption began in October 2004 and activity has mostly consisted of small ash plumes, ballistic ejecta, and visible incandescence at night. This report summarizes activity during January through June 2019 and is based on reports by the Japan Meteorological Agency (JMA), and various satellite data.

Thermal activity recorded by the MIROVA system was low through January and February after a decline in November (figure 36), shown in Sentined-2 thermal infrared imagery as originating at a vent in the Otake crater (figure 37). During January an explosive event was observed at 1727 on the 3rd, producing a gray plume that rose 600 m above the crater. A white gas-and-steam plume rose to 1.5 km above the crater and nighttime incandescence was observed throughout the month. Reduced activity continued through February with no reported explosive eruptions and light gray plumes up to 900 m above the crater. Incandescence continued to be recorded at night using a sensitive surveillance camera.

Figure 36. MIROVA log radiative power plot of MODIS thermal infrared data at Suwanosejima during September 2018 through June 2019. There was reduced activity in 2019 with periods of more frequent anomalies during March and June. Courtesy of MIROVA.

Figure 37. A Sentinel-2 thermal satellite image shows Suwanosejima with the active Otake crater in the center with elevated temperatures shown as bright orange/yellow. There is a light area next to the vent that may be a gas plume. False color (urban) satellite image (bands 12, 11, 4) courtesy of Sentinel Hub Playground.

There was an increase in thermal energy detected by the MIROVA system in mid-March and there was a MODVOLC thermal alert on the 15th. Occasional small explosions occurred but no larger explosive events were recorded. A white plume was noted on the 27th rising to 900 m above the crater and an event at 1048 on the 30th produced a light-gray plume that rose to 800 m. Incandescence was only observed using a sensitive camera at night (figure 38).

Figure 38. Incandescence from the Suwanosejima Otake crater reflecting in clouds above the volcano. Courtesy of JMA (Volcanic activity of Suwanosejima March 2019).

No explosive events were observed through April. A white gas-and-steam plume rose to 1,200 m above the crater on the 19th and incandescence continued intermittently. Minor explosions were recorded on 5, 30, and 31 May, but no larger explosive events were observed during the month. The event on the 30th produced ash plume that reached 1.1 km above the crater. Similar activity continued through June with one explosive event occurring on the 2nd. Overall, there was a reduction in the number of ash plumes erupted during this period compared to previous months (figure 39).

Figure 39. Observed activity at Suwanosejima for the year ending in July 2019. The black vertical bars represent steam, gas, or ash plume heights (scale in meters on the left axis), yellow diamonds represent incandescence observed in webcams, gray volcano symbols along the top are explosions accompanied by ash plumes, red volcano symbols represent large explosions with ash plumes. Courtesy of JMA (Volcanic activity of Suwanosejima June 2019).

Geologic Background. The 8-km-long, spindle-shaped island of Suwanosejima in the northern Ryukyu Islands consists of an andesitic stratovolcano with two historically active summit craters. The summit of the volcano is truncated by a large breached crater extending to the sea on the east flank that was formed by edifice collapse. Suwanosejima, one of Japan's most frequently active volcanoes, was in a state of intermittent strombolian activity from Otake, the NE summit crater, that began in 1949 and lasted until 1996, after which periods of inactivity lengthened. The largest historical eruption took place in 1813-14, when thick scoria deposits blanketed residential areas, and the SW crater produced two lava flows that reached the western coast. At the end of the eruption the summit of Otake collapsed forming a large debris avalanche and creating the horseshoe-shaped Sakuchi caldera, which extends to the eastern coast. The island remained uninhabited for about 70 years after the 1813-1814 eruption. Lava flows reached the eastern coast of the island in 1884. Only about 50 people live on the island.

Information Contacts: Japan Meteorological Agency (JMA), Otemachi, 1-3-4, Chiyoda-ku Tokyo 100-8122, Japan (URL: http://www.jma.go.jp/jma/indexe.html); Hawai'i Institute of Geophysics and Planetology (HIGP) - MODVOLC Thermal Alerts System, School of Ocean and Earth Science and Technology (SOEST), Univ. of Hawai'i, 2525 Correa Road, Honolulu, HI 96822, USA (URL: http://modis.higp.hawaii.edu/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground).

The Great Sitkin volcano is located about 40 km NE of Adak Island in the Aleutian Islands and has had a few short-lived eruptions over the past 100 years. Prior to the latest activity in early June 2019 described below, small phreatic explosions occurred in June and August 2018 (BGVN 43:09). An eruption in 1974 produced a lava dome in the center of the crater. The Alaska Volcano Observatory (AVO) is the primary source of information for this September 2018-June 2019 reporting period.

Low-level unrest occurred from September 2018 through February 2019 with slightly elevated seismic activity (figure 6). Small explosions were seismically detected by AVO on 30 October, 5 and 16 November, and 11 December 2018, but they were not seen in regional infrasound data and satellite data did not show an ash cloud.

On 1, 7, and 9 June 2019, AVO reported small steam explosions as well as slightly elevated seismic activity. Steam plumes and surficial evidence of an explosion were not observed during these events. On 18 June 2019 weakly elevated surface temperatures were recorded, field crews working on Adak observed some steam emissions, and a gas flight was conducted. Elevated concentrations of carbon dioxide detected above the lava dome were likely associated with the steam explosions earlier in the month (figures 7 and 8). From 23 June through the end of the month seismicity began to decline back to background levels.

Figure 6. A steam plume was seen at the summit of Great Sitkin on 7 December 2018. Photo by Andy Lewis and Bob Boyd; courtesy of AVO/USGS.

Figure 7. Some degassing was observed on the southern flank of the Great Sitkin during an overflight on 18 June 2019. Photo by Laura Clor; image courtesy of AVO/USGS.

Figure 8. View of Great Sitkin with white plumes rising from the summit on 20 June 2019. Photo by Laura Clor, courtesy of AVO/USGS.

Geologic Background. The Great Sitkin volcano forms much of the northern side of Great Sitkin Island. A younger parasitic volcano capped by a small, 0.8 x 1.2 km ice-filled summit caldera was constructed within a large late-Pleistocene or early Holocene scarp formed by massive edifice failure that truncated an ancestral volcano and produced a submarine debris avalanche. Deposits from this and an older debris avalanche from a source to the south cover a broad area of the ocean floor north of the volcano. The summit lies along the eastern rim of the younger collapse scarp. Deposits from an earlier caldera-forming eruption of unknown age cover the flanks of the island to a depth up to 6 m. The small younger caldera was partially filled by lava domes emplaced in 1945 and 1974, and five small older flank lava domes, two of which lie on the coastline, were constructed along northwest- and NNW-trending lines. Hot springs, mud pots, and fumaroles occur near the head of Big Fox Creek, south of the volcano. Historical eruptions have been recorded since the late-19th century.

Information Contacts: Alaska Volcano Observatory (AVO), a cooperative program of a) U.S. Geological Survey, 4200 University Drive, Anchorage, AK 99508-4667 USA (URL: https://avo.alaska.edu/), b) Geophysical Institute, University of Alaska, PO Box 757320, Fairbanks, AK 99775-7320, USA, and c) Alaska Division of Geological & Geophysical Surveys, 794 University Ave., Suite 200, Fairbanks, AK 99709, USA (URL: http://dggs.alaska.gov/).

Ibu volcano on Halmahera island in Indonesia began the current eruption episode on 5 April 2008. Since then, activity has largely consisted of small ash plumes with less frequent lava flows, lava dome growth, avalanches, and larger ash plumes up to 5.5 km above the crater. This report summarizes activity during December 2018 through June 2019 and is based on Volcano Observatory Notice for Aviation (VONA) reports by MAGMA Indonesia, reports by Pusat Vulkanologi dan Mitigasi Bencana Geologi (PVMBG) and Badan Nasional Penanggulangan Bencana (BNPB), and various satellite data.

During December PVMBG reported ash plumes ranging from 200 to 800 m above the crater. There were 11 MODVOLC thermal alerts that registered during 1-12 December. An explosion on 12 January 2019 produced an ash plume that reached 800 m above the crater and dispersed to the S (figure 15). A report released for this event by Sutopo at BNPB said that Ibu had erupted almost every day over the past three months; an example given was of activity on 10 January consisting of 80 explosions. There were four MODVOLC thermal alerts through the month.

Figure 15. An eruption at Ibu at 1712 on 21 January 2019 produced an ash plume that rose to 800 m above the crater. Courtesy of BNPB (color adjusted).

Throughout February explosions frequently produced ash plumes as high as 800 m above the crater, and nine MODVOLC thermal alerts were issued. Daily reports showed variable plume heights of 200-800 m most days throughout the month. Wind directions varied and dispersed the plumes in all directions. A VONA released at 1850 on 6 February reported an ash plume that rose to 1,925 m altitude (around 600 m above the summit) and dispersed S. Activity continued through March with the Darwin VAAC and PVMBG reporting explosions producing ash plumes to heights of 200-800 m above the crater and dispersing in various directions. There were ten MODVOLC alerts through the month.

Similar activity continued through April, May, and June, with ash plumes reaching 200-800 m above the crater. There were 12, 6, and 15 MODVOLC Alerts in April, May, and June, respectively.

Planet Scope satellite images show activity at a two vents near the center of the crater that were producing small lava flows from February through June (figure 16). Thermal anomalies were frequent during December 2018 through June 2019 across MODVOLC, MIROVA, and Sentinel-2 infrared data (figures 17 and 18). Sentinel-2 data showed minor variation in the location of thermal anomalies within the crater, possibly indicating lava flow activity, and MIROVA data showed relatively constant activity with a few reductions in thermal activity during January and February.

Figure 16. Planet Scope natural color satellite images showing activity in the Ibu crater during January through June 2019, with white arrows indicating sites of activity. One vent is visible in the 21 February image, and a 330-m-long (from the far side of the vent) lava flow with flow ridges had developed by 24 March. A second vent was active by 12 May with a new lava flow reaching a maximum length of 520 m. Activity was centered back at the previous vent by 23-27 June. Natural color Planet Scope Imagery, copyright 2019 Planet Labs, Inc.

Figure 17. Examples of thermal activity in the Ibu crater during January through May 2019. These Sentinel-2 satellite images show variations in hot areas in the crater due to a vent producing a small lava flow. Sentinel-2 false color (urban) images (bands 12, 11, 4) courtesy of Sentinel Hub Playground.

Figure 18. MIROVA log radiative power plot of MODIS thermal infrared at Ibu from September 2018 through June 2019. The registered energy was relatively stable through December, with breaks in January and February. Regular thermal anomalies continued with slight variation through to the end of June. Courtesy of MIROVA.

Geologic Background. The truncated summit of Gunung Ibu stratovolcano along the NW coast of Halmahera Island has large nested summit craters. The inner crater, 1 km wide and 400 m deep, contained several small crater lakes through much of historical time. The outer crater, 1.2 km wide, is breached on the north side, creating a steep-walled valley. A large parasitic cone is located ENE of the summit. A smaller one to the WSW has fed a lava flow down the W flank. A group of maars is located below the N and W flanks. Only a few eruptions have been recorded in historical time, the first a small explosive eruption from the summit crater in 1911. An eruption producing a lava dome that eventually covered much of the floor of the inner summit crater began in December 1998.

Information Contacts: Pusat Vulkanologi dan Mitigasi Bencana Geologi (PVMBG, also known as Indonesian Center for Volcanology and Geological Hazard Mitigation, CVGHM), Jalan Diponegoro 57, Bandung 40122, Indonesia (URL: http://www.vsi.esdm.go.id/); Badan Nasional Penanggulangan Bencana (BNPB), National Disaster Management Agency, Graha BNPB - Jl. Scout Kav.38, East Jakarta 13120, Indonesia (URL: http://www.bnpb.go.id/); MAGMA Indonesia, Kementerian Energi dan Sumber Daya Mineral (URL: https://magma.vsi.esdm.go.id/); Darwin Volcanic Ash Advisory Centre (VAAC), Bureau of Meteorology, Northern Territory Regional Office, PO Box 40050, Casuarina, NT 0811, Australia (URL: http://www.bom.gov.au/info/vaac/); Hawai'i Institute of Geophysics and Planetology (HIGP) - MODVOLC Thermal Alerts System, School of Ocean and Earth Science and Technology (SOEST), Univ. of Hawai'i, 2525 Correa Road, Honolulu, HI 96822, USA (URL: http://modis.higp.hawaii.edu/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground); Planet Labs, Inc. (URL: https://www.planet.com/).

The Ebeko volcano, located on the northern end of the Paramushir Island in the Kuril Islands, consists of many craters, lakes, and thermal features and has been frequently erupting since late February 2017. Typical activity includes ash plumes, explosive eruptions, and gas-and-steam activity. The previous report through November 2018 (BGVN 43:12) described frequent ash explosions that sometimes caused ashfall in Severo-Kurilsk (7 km E). The primary source of information is the Kamchatka Volcanic Eruptions Response Team (KVERT). This report updates the volcanic activity at Ebeko for December 2018 through May 2019.

Frequent moderate explosive activity continued after November 2018. Volcanologists in Severo-Kurilsk observed explosions sending up ash, which drifted N, NE, and E, resulting in ash falls on Severo-Kurilsk on 28 different days between December 2018 and March 2019. On 25 December 2018 an explosion sent ash up to a maximum altitude of 4.5 km and then drifted N for about 5 km. Explosions occurring on 8-10 March 2019 sent ash up to an altitude of 4 km, resulting in ashfall on Severo-Kurilsk on 9-10 March 2019. An ash plume from these explosions rose to a height of 2.5 km and drifted to a maximum distance of 30 km ENE.

Satellite data analyzed by KVERT registered 12 thermal anomalies from December 2018 through May 2019. According to satellite data analyzed by MIROVA (Middle InfraRed Observation of Volcanic Activity), only one thermal anomaly was recorded from December 2018-May 2019, and no hotspot pixels were recognized using satellite thermal data from the MODVOLC algorithm.

Geologic Background. The flat-topped summit of the central cone of Ebeko volcano, one of the most active in the Kuril Islands, occupies the northern end of Paramushir Island. Three summit craters located along a SSW-NNE line form Ebeko volcano proper, at the northern end of a complex of five volcanic cones. Blocky lava flows extend west from Ebeko and SE from the neighboring Nezametnyi cone. The eastern part of the southern crater contains strong solfataras and a large boiling spring. The central crater is filled by a lake about 20 m deep whose shores are lined with steaming solfataras; the northern crater lies across a narrow, low barrier from the central crater and contains a small, cold crescentic lake. Historical activity, recorded since the late-18th century, has been restricted to small-to-moderate explosive eruptions from the summit craters. Intense fumarolic activity occurs in the summit craters, on the outer flanks of the cone, and in lateral explosion craters.

Information Contacts: Kamchatka Volcanic Eruptions Response Team (KVERT), Far Eastern Branch, Russian Academy of Sciences, 9 Piip Blvd., Petropavlovsk-Kamchatsky, 683006, Russia (URL: http://www.kscnet.ru/ivs/kvert/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Hawai'i Institute of Geophysics and Planetology (HIGP) - MODVOLC Thermal Alerts System, School of Ocean and Earth Science and Technology (SOEST), Univ. of Hawai'i, 2525 Correa Road, Honolulu, HI 96822, USA (URL: http://modis.higp.hawaii.edu/).

Klyuchevskoy has had alternating eruptive and less active periods since August 2015. Activity has included lava flows, a growing cinder cone, thermal anomalies, gas-and-steam plumes, and ash explosions. Though some eruptions occur near the summit crater, major explosive and effusive eruptions have also occurred from flank craters (BGVN 42:04 and 43:05). Intermittent moderate gas-and-steam and ash emissions were previously reported from mid-February to mid-August 2018. The Kamchatka Volcanic Eruptions Response Team (KVERT) is the primary source of information for this September 2018-June 2019 reporting period.

KVERT reported that moderate gas-and-steam activity, some of which contained a small amount of ash, and weak thermal anomalies occurred intermittently from the beginning of September 2018 through mid-April 2019. On 21-22 April 2019 webcam data showed a gas-and-steam plume extending about 160 km SE (figure 31). Moderate Strombolian-type volcanism began late April 2019 and continued intermittently through June 2019. On 11-12 June webcam data showed explosions that sent ash up to a maximum altitude of 6 km, with the resulting ash plume extending about 200 km WNW.

Figure 31. Gas-and-steam plume containing some amount of ash rising from the summit of Klyuchevskoy on 22 April 2019. Photo by A. Klimova, courtesy of Institute of Volcanology and Seismology (IVS FEB RAS).

Thermal anomalies were noted by KVERT during two days in September 2018, six days in April 2019, eleven days in May 2019, and six days in June 2019. MIROVA (Middle InfraRed Observation of Volcanic Activity) analysis of MODIS satellite data showed infrequent weak thermal anomalies December 2018 through early May 2019.

Geologic Background. Klyuchevskoy (also spelled Kliuchevskoi) is Kamchatka's highest and most active volcano. Since its origin about 6000 years ago, the beautifully symmetrical, 4835-m-high basaltic stratovolcano has produced frequent moderate-volume explosive and effusive eruptions without major periods of inactivity. It rises above a saddle NE of sharp-peaked Kamen volcano and lies SE of the broad Ushkovsky massif. More than 100 flank eruptions have occurred during the past roughly 3000 years, with most lateral craters and cones occurring along radial fissures between the unconfined NE-to-SE flanks of the conical volcano between 500 m and 3600 m elevation. The morphology of the 700-m-wide summit crater has been frequently modified by historical eruptions, which have been recorded since the late-17th century. Historical eruptions have originated primarily from the summit crater, but have also included numerous major explosive and effusive eruptions from flank craters.

Information Contacts: Kamchatka Volcanic Eruptions Response Team (KVERT), Far Eastern Branch, Russian Academy of Sciences, 9 Piip Blvd., Petropavlovsk-Kamchatsky, 683006, Russia (URL: http://www.kscnet.ru/ivs/kvert/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/).

Yasur volcano on Tanna Island has been characterized by Strombolian activity with large incandescent bombs, frequent explosions, lava fountaining, and ash emissions for much of its known eruptive history. Melanesians from nearby islands are believed to have settled Tanna in about 400 BCE; it is now part of the nation of Vanuatu, independent since 1980. The Kwamera language (or Tannese) spoken on the SE coast of the island is thought to be the source of the name of the island. No known oral history describes volcanic activity; the first written English-language documentation of activity dates to 5 August 1774, when Captain James Cook saw "a great fire" on Tanna Island. Cook realized that it "was a Volcano which threw up vast quantities of fire and smoak and made a rumbling noise which was heard at a good distance" (The Captain Cook Society) (figure 51).

Figure 51. Incandescence, steam, and dark ash from Yasur fill the sky in this sketch representing Captain James Cook's landing in the 'Resolution' at Tanna Island on 5 August 1774. The form of the volcano is behind the ship, the incandescence is in the upper right next to the ship's masts. "Landing at Tanna" by William Hodges, 1775-1776, National Maritime Museum, Greenwich, London. The Maritime Museum noted that this is one of a group of panel paintings produced by Hodges of encounters with islanders during the voyage, in which the European perception of each society at the time is portrayed. Image taken from Wikimedia Commons.

Based on numerous accounts from ships logs and other sources, volcanic activity has been continuous since that time. During periods of higher activity, multiple vents within the summit crater send ejecta 100 m or more above the crater rim, with large bombs occasionally landing hundreds of meters away. Continued activity during February-May 2019 is covered in this report with information provided by the Geo-Hazards Division, Vanuatu Meteorology and Geo-Hazards Department (VMGD) which monitors the volcano and satellite data; photographs from tourists also provide valuable information about this remote location.

VMGD has maintained Alert Level 2 at Yasur since October 2016, indicating that it is in a major state of unrest. There is a permanent exclusion zone within 395 m of the eruptive vents where access is prohibited due to multiple hazards, primarily from large incandescent bombs up to 4 m in diameter which have been ejected from the vents onto the crater rim in the past, resulting in fatalities (BGVN 20:08).

Satellite and ground based information all support high levels of thermal activity during February -May 2019. MODVOLC thermal alerts were issued 11 times in February, 27 times in March, and 20 times each in April and May. The MIROVA graph also indicated the ongoing consistently high levels of thermal energy throughout the period (figure 52). Plumes of SO₂ emissions are common from Vanuatu's volcanoes; newer higher resolution data available beginning in 2019 reveal a persistent stream of SO₂ from Yasur on a near-daily basis (figure 53).

Figure 52. The MIROVA graph of thermal energy at Yasur from 3 September 2018 through May 2019 indicates the ongoing activity at the volcano. Courtesy of MIROVA.

Figure 53. The SO2 plumes from Yasur were persistent during January-May 2019 when they were visible many days of each week throughout the period. Top left: On 12 January plumes were visible drifting E from both Ambrym (top) and Yasur (bottom). Top right: Plumes drifted W from three Vanuatu volcanoes on 7 February, Gaua (top), Ambrym (middle) and Yasur (bottom). Bottom left: On 12 March N drifting plumes could be seen from Ambae (top) and Yasur (bottom). On 27 April, only Yasur had an SO2 plume drifting W. Courtesy of Goddard Space Flight Center.

Satellite imagery confirmed that the heat sources from Yasur were vents within the summit crater of the pyroclastic cone. Both northern and southern vent areas were active. On 7 March 2019 the N vent area had a strong thermal signal. Ten days later, on 17 March, similar intensity thermal anomalies were present in both the N and S vent areas (figure 54). On 6 April the S vent area had a stronger signal, and gas emissions from both vents were drifting N (figure 55). Satellite imagery from 21 May 2019 indicated a strong thermal signal inside the crater in the area of the vents, and included a weaker signal clearly visible on the inside E crater rim. Strong Strombolian activity or spatter sending large incandescent bombs as far as the crater rim are a likely explanation for the signal (figure 56), underscoring the hazardous nature of approaching the crater rim.

Figure 54. Strong thermal anomalies from the crater of Yasur's pyroclastic cone seen in satellite images confirmed the ongoing high level of activity. Left: 7 March 2019, a strong thermal anomaly from the N vent area, shown with "Geology" rendering (bands 12, 4, 2). Right: 17 March 2019, thermal anomalies at both the N and S vent areas, shown with "Atmospheric Penetration" rendering (bands 12, 11, 8A). The crater is about 500 m in diameter. Sentinel-2 satellite imagery courtesy of Sentinel Hub Playground.

Figure 55. Strong thermal anomalies (left) and gas emissions (right) at Yasur were captured with different bands in the same Sentinel-2 satellite image on 6 April 2019. Left: The thermal anomaly in the S vent area was stronger than in the N vent area, "Atmospheric Penetration" rendering (bands 12, 11, 8A). Right: Gas plumes drifted N from both vent areas, "Natural color" rendering (bands 4, 3, 2). The crater is about 500 m in diameter. Sentinel-2 satellite imagery courtesy of Sentinel Hub Playground.

Figure 56. Thermal activity from the crater of Yasur on 21 May 2019 produced a strong thermal signal from the center of the crater and a weaker signal on the inside E crater rim, likely the result of hazardous incandescent bombs and ejecta, frequent products of the activity at Yasur. Left: "Atmospheric Penetration" rendering (bands 12, 11, 8A). Right: "Geology" rendering (bands 12, 4, 2). The crater is about 0.5 km in diameter. Sentinel-2 satellite imagery courtesy of Sentinel Hub Playground.

Tourists visit Yasur on a regular basis. A former lake on the N side of Yasur has left ripples in the sand deposits over older volcanic rocks on the N side of the volcano (figure 57) since it drained in 2000 (BGVN 28:01). Visitors are allowed to approach the S rim of the crater where incandescence from both the N and S vents is usually visible (figure 58). Incandescent spatter from the convecting lava in the vents is highly dangerous and unpredictable and often covers the inner slopes of the rim as well as sending bombs outside the crater (figure 59).

Figure 57. The pyroclastic cone of Yasur viewed from the north on 6 May 2019. Ripples in volcaniclastic sand in the foreground are remnants of a lake that was present on the N side of the volcano until a natural dam breached in 2000. Copyrighted photo by Nick Page, used with permission.

Figure 58. Two glowing vents were visible from the south rim of Yasur on 6 May 2019. The S vent area is in the foreground, the N vent area is in the upper left. Copyrighted by Nick Page, used with permission.

Figure 59. Incandescent spatter at Yasur on 6 May 2019 sent fragments of lava against the inside crater wall and onto the rim. The convecting lava in the vent can be seen in the lower foreground. Copyrighted photo by Nick Page, used with permission.

Geologic Background. Yasur, the best-known and most frequently visited of the Vanuatu volcanoes, has been in more-or-less continuous Strombolian and Vulcanian activity since Captain Cook observed ash eruptions in 1774. This style of activity may have continued for the past 800 years. Located at the SE tip of Tanna Island, this mostly unvegetated pyroclastic cone has a nearly circular, 400-m-wide summit crater. The active cone is largely contained within the small Yenkahe caldera, and is the youngest of a group of Holocene volcanic centers constructed over the down-dropped NE flank of the Pleistocene Tukosmeru volcano. The Yenkahe horst is located within the Siwi ring fracture, a 4-km-wide, horseshoe-shaped caldera associated with eruption of the andesitic Siwi pyroclastic sequence. Active tectonism along the Yenkahe horst accompanying eruptions has raised Port Resolution harbor more than 20 m during the past century.

Information Contacts: Geo-Hazards Division, Vanuatu Meteorology and Geo-Hazards Department (VMGD), Ministry of Climate Change Adaptation, Meteorology, Geo-Hazards, Energy, Environment and Disaster Management, Private Mail Bag 9054, Lini Highway, Port Vila, Vanuatu (URL: http://www.vmgd.gov.vu/, https://www.facebook.com/VanuatuGeohazardsObservatory/); Hawai'i Institute of Geophysics and Planetology (HIGP) MODVOLC Thermal Alerts System, School of Ocean and Earth Science and Technology (SOEST), Univ. of Hawai'i, 2525 Correa Road, Honolulu, HI 96822, USA (URL: http://modis.higp.hawaii.edu/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground); Global Sulfur Dioxide Monitoring Page, Atmospheric Chemistry and Dynamics Laboratory, NASA Goddard Space Flight Center (NASA/GSFC), 8800 Greenbelt Road, Goddard, Maryland, USA (URL: https://so2.gsfc.nasa.gov/); The Captain Cook Society (URL: https://www.captaincooksociety.com/home/detail/225-years-ago-july-september-1774); Royal Museums Greenwich (URL: https://collections.rmg.co.uk/collections/objects/13383.html); Wikimedia Commons, (URL: https://commons.wikimedia.org/wiki/File:The_Landing_at_Tana_one_of_the_New_Hebrides,_by_William_Hodges.jpg); Nick Page, Australia,Flickr: (URL: https://www.flickr.com/photos/152585166@N08/).

With historical eruptions reported back to 1842, Papua New Guinea's Bagana volcano on the island of Bougainville has been characterized by viscous andesitic lava flows down the steep flanks of its cone, along with intermittent ash plumes and pyroclastic flows. Ongoing thermal anomalies and frequent ash plumes have been typical of activity during the current eruption since it began in early 2000. Activity declined significantly in December 2018 and remained low through May 2019, the period covered in this report (figure 33). Information for this report comes primarily from satellite images and thermal data.

Figure 33. The MIROVA plot of radiative power at Bagana from 1 September 2018 through May 2019 shows a marked decline in thermal activity during December 2018 after ash explosions and satellite observations of flows during the previous months. Courtesy of MIROVA.

The last ash emission at Bagana was reported on 1 December 2018 by the Darwin Volcanic Ash Advisory Center (VAAC). A Sentinel-2 satellite image showed a linear thermal anomaly trending NW from the summit on 14 December (BGVN 50:01). On 8 January 2019, an image contained a dense steam plume drifting E and a very faint thermal anomaly on the N flank a few hundred meters from the summit. A more distinct thermal anomaly at the summit appeared on 22 February 2019 (figure 34). A visitor to the region photographed incandescence on the flank, likely from the volcano, at dawn around 19 February 2019 (figure 35).

Figure 34. Sentinel-2 satellite imagery revealed thermal anomalies at Bagana in January and February 2019. Left: a very faint thermal anomaly was N of the summit at the edge of the E-drifting steam plume on 8 January 2019. Right: A thermal anomaly was located at the summit, at the base of the NE-drifting steam plume on 22 February 2019. Sentinel-2 satellite images with "Atmospheric Penetration" rendering (bands 12, 11, 8A). Courtesy of Sentinel Hub Playground.

Figure 35. A visitor near Bagana spotted incandescence on the flank at dawn, possibly from a lava flow. Posted online 19 February 2019. Courtesy of Emily Stanford.

Two faint thermal anomalies were visible at the summit in satellite imagery on 19 March; a single one appeared on 29 March 2019 (figure 36). No thermal anomalies were recorded in Sentinel-2 images during April or May, but steam plumes and gas emissions were visible through cloud cover on multiple occasions (figure 37).

Figure 36. Faint thermal anomalies at Bagana were recorded in satellite imagery twice during March 2019. Left: 19 March, two anomalies appear right of the date label. Right: 29 March, a small anomaly appears right of the date label. Sentinel-2 image rendered with "Atmospheric Penetration" (bands 12, 11, 8A). Courtesy of Sentinel Hub Playground.

Figure 37. Steam and gas emissions at Bagana were recorded in satellite imagery during April and May 2019. Left: A steam plume drifted NW from the summit on 23 April, visible through dense cloud cover. Right: A gas plume drifted SW from the summit on 18 May. Sentinel-2 image with "Geology" rendering (bands 12, 4, 2). Courtesy of Sentinel Hub Playground.

Geologic Background. Bagana volcano, occupying a remote portion of central Bougainville Island, is one of Melanesia's youngest and most active volcanoes. This massive symmetrical cone was largely constructed by an accumulation of viscous andesitic lava flows. The entire edifice could have been constructed in about 300 years at its present rate of lava production. Eruptive activity is frequent and characterized by non-explosive effusion of viscous lava that maintains a small lava dome in the summit crater, although explosive activity occasionally producing pyroclastic flows also occurs. Lava flows form dramatic, freshly preserved tongue-shaped lobes up to 50 m thick with prominent levees that descend the flanks on all sides.

Information Contacts: MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Darwin Volcanic Ash Advisory Centre (VAAC), Bureau of Meteorology, Northern Territory Regional Office, PO Box 40050, Casuarina, NT 0811, Australia (URL: http://www.bom.gov.au/info/vaac/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground); Emily Stanford (Twitter: https://twitter.com/NerdyBatLady, image posted at https://twitter.com/NerdyBatLady/status/1098052063009792001/photo/1).

Ambae (Aoba) is a large basaltic shield volcano in the New Hebrides arc, part of the multi-island country of Vanuatu. Its periodic phreatic and pyroclastic explosions originating in the summit crater lakes have been recorded since the 16th century. A pyroclastic cone appeared in Lake Voui during November 2005-February 2006 (BGVN 31:12, figure 30); an explosive eruption from a new pyroclastic cone in the lake began in mid-September 2017 (BGVN 43:02). Activity included high-altitude ash emissions (9.1 km), lava flows, and Strombolian activity. Intermittent pulses of ash emissions during the following months resulted in extensive ashfall and evacuations; multiple communities were affected by lahars. The most recent episode of the eruption from July to September 2018 (BGVN 44:02) resulted in 11-km-altitude ash plumes and the evacuation of the entire island due to heavy ashfall and lahars. This report covers activity from February to May 2019, with information provided by the Vanuatu Geohazards Observatory of the Vanuatu Meteorology and Geo-Hazards Department (VMGD) and satellite data from multiple sources.

Activity diminished after the extensive eruptive phase of July-September 2018 when substantial ash plumes and ashfall resulted in evacuations. An explosion with an ash plume on 30 October 2018 was the last activity reported for 2018. Thermal alerts were reported by the Hawai'i Institute of Geophysics and Planetology (HIGP) MODVOLC thermal alerts system through January 2019, and the Log Radiative Power graph prepared by the MIROVA project showed decreasing thermal anomalies into June 2019 (figure 92). Satellite images recorded in April and May 2019 (figure 93) showed the configuration of the summit lakes to be little changed from the previous November except for the color (BGVN 44:02, figure 89). No ash emissions or SO₂ plumes were reported during the period. VMGD noted that the volcano remained at Alert Level 2 through May 2019 with a 2-km-radius exclusion zone around the summit.

Figure 92. The MIROVA log radiative power plot for Ambae showed ongoing intermittent thermal anomalies from early September 2018 through May 2019. Courtesy of MIROVA.

Figure 93. Satellite imagery in April and May 2019 showed little change in the configuration of lakes at the summit of Ambae since November 2018 (see BGVN 44:02, figure 89). Left: 24 April 2019. Right: 29 May 2019. Sentinel-2 satellite imagery with "Natural Color" rendering (bands 4, 3, 2); courtesy of Sentinel Hub Playground.

Geologic Background. The island of Ambae, also known as Aoba, is a massive 2500 km3 basaltic shield that is the most voluminous volcano of the New Hebrides archipelago. A pronounced NE-SW-trending rift zone dotted with scoria cones gives the 16 x 38 km island an elongated form. A broad pyroclastic cone containing three crater lakes (Manaro Ngoru, Voui, and Manaro Lakua) is located at the summit within the youngest of at least two nested calderas, the largest of which is 6 km in diameter. That large central edifice is also called Manaro Voui or Lombenben volcano. Post-caldera explosive eruptions formed the summit craters about 360 years ago. A tuff cone was constructed within Lake Voui (or Vui) about 60 years later. The latest known flank eruption, about 300 years ago, destroyed the population of the Nduindui area near the western coast.

Information Contacts: Geo-Hazards Division, Vanuatu Meteorology and Geo-Hazards Department (VMGD), Ministry of Climate Change Adaptation, Meteorology, Geo-Hazards, Energy, Environment and Disaster Management, Private Mail Bag 9054, Lini Highway, Port Vila, Vanuatu (URL: http://www.vmgd.gov.vu/, https://www.facebook.com/VanuatuGeohazardsObservatory/); Hawai'i Institute of Geophysics and Planetology (HIGP) - MODVOLC Thermal Alerts System, School of Ocean and Earth Science and Technology (SOEST), Univ. of Hawai'i, 2525 Correa Road, Honolulu, HI 96822, USA (URL: http://modis.higp.hawaii.edu/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground).

Sangay is the southernmost active volcano in Ecuador, with confirmed historical eruptions going back to 1628. The previous eruption occurred during August and December and was characterized by ash plumes reaching 2,500 m above the crater. Lava flows and pyroclastic flows descended the eastern and southern flanks. This report summarizes activity during January through July 2019 and is based on reports by Instituto Geofísico (IG-EPN), Washington Volcanic Ash Advisory Center (VAAC), and various satellite data.

After the December 2018 eruption there was a larger reduction in seismicity, down to one event per day. During January, February, and most of March there was no recorded activity and low seismicity until the Washington VAAC reported an ash plume at 0615 on 26 March. The ash plume rose to a height of around 1 km and dispersed to the SW as seen in GOES 16 satellite imagery as a dark plume within white meteorological clouds. There was no seismic data available due to technical problems with the station.

More persistent eruptive activity began on 10 May with thermal alerts (figure 30) and an ash plume at 0700 that dispersed to the W. An explosion was recorded at 1938 on 11 May, producing an ash plume and incandescent material down the flank (figure 31). Two M 2 earthquakes were detected between 3.5 and 9 km below the crater on 10 May, possibly corresponding to explosive activity. By 17 May there were two active eruptive centers, the central crater and the Ñuñurcu dome (figure 32).

Figure 30. MIROVA log radiative power plot of MODIS thermal infrared at Sangay for the year ending June 2019. The plot shows the August to December 2018 eruption, a break in activity, and resumed activity in May 2019. Courtesy of MIROVA.

Figure 31. An explosion at Sangay on 10 May 2019 sent ballistic projectiles up to 650 m above the crater at a velocity of over 400 km/hour, an ash plume that rose to over 600 m, and incandescent blocks that traveled over 1.5 km from the crater at velocities of around 150 km/hour. Screenshots are from video by IG-EPN.

Figure 32. A photograph of the southern flank of Sangay on 17 May 2019 with the corresponding thermal infrared image in the top right corner. The letters correspond to: a) a fissure to the W of the lava flow; b) an active lava flow from the Ñuñurcu dome; c) the central crater producing a volcanic gas plume; d) a pyroclastic flow deposit produced by collapsing material from the front of the lava flow. Prepared by M. Almeida; courtesy of IG-EPN (special report No. 3 – 2019).

Activity at the central crater by 21 May was characterized by sporadic explosive eruptions that ejected hot ballistic ejecta (blocks) with velocities over 400 km/hour; after landing on the flanks the blocks travelled out to 2.5 km from the crater. Ash plumes reached heights between 0.9-2.3 km above the crater and dispersed mainly to the W and NW; gas plumes also dispersed to the W. The Ñuñurcu dome is located around 190 m SSE of the central crater and by 21 May had produced a lava flow over 470 m long with a maximum width of 175 m and an estimated minimum volume of 300,000 to 600,000 m³. Small pyroclastic flows and rockfalls resulted from collapse of the lava flow front, depositing material over a broad area on the E-SE flanks (figure 33). One pyroclastic flow reached 340 m and covered an area of 14,300 m2. During the 17 May observation flight the lava flow surface reached 277°C.

Figure 33. A view of the ESE flanks of Sangay on 17 May 2019. The area within the black dotted line is the main area of pyroclastic flow deposition from the Ñuñurco Dome. Photo by M. Almeida; courtesy of IG-EPN (special report No. 4 – 2019).

At the end of June activity was continuing at the central crater and Ñuñurco Dome. At least three lava flows had been generated from the dome down the SE flank and pyroclastic flows continued to form from the flow fronts (figure 34). Pyroclastic material had been washed into the Upano river and steam was observed in the Volcán River possibly due to the presence of hot rocks. Ash plumes continued through June reaching heights of 800 m above the crater (figure 35), but no ashfall had been reported in nearby communities.

Figure 34. Sentinel-2 natural color (left) and thermal (center) images (bands 12, 11, 4), and 1:50 000 scale maps (right) of Sangay with interpretation on the background of a 30 m numerical terrain model (WGS84; Zone 17S) (Prepared by B. Bernard). The dates from top to bottom are 17 May, 22 May, 27 May, 16 June, and 26 June 2019. Prepared by B. Bernard; courtesy IG-EPN (special report No. 4 – 2019).

Figure 35. Plots giving the heights and dispersal direction of ash plumes at Sangay during May and June 2019. Top: Ash plume heights measures in meters above the crater. Bottom: A plot showing that the dominant dispersal direction of ash plumes is to the W during this time. Courtesy of IG-EPN (special report No. 4 – 2019).

Geologic Background. The isolated Sangay volcano, located east of the Andean crest, is the southernmost of Ecuador's volcanoes and its most active. The steep-sided, glacier-covered, dominantly andesitic volcano grew within horseshoe-shaped calderas of two previous edifices, which were destroyed by collapse to the east, producing large debris avalanches that reached the Amazonian lowlands. The modern edifice dates back to at least 14,000 years ago. It towers above the tropical jungle on the east side; on the other sides flat plains of ash have been sculpted by heavy rains into steep-walled canyons up to 600 m deep. The earliest report of a historical eruption was in 1628. More or less continuous eruptions were reported from 1728 until 1916, and again from 1934 to the present. The almost constant activity has caused frequent changes to the morphology of the summit crater complex.

Information Contacts: Instituto Geofísico (IG-EPN), Escuela Politécnica Nacional, Casilla 17-01-2759, Quito, Ecuador (URL: http://www.igepn.edu.ec); Washington Volcanic Ash Advisory Center (VAAC), Satellite Analysis Branch (SAB), NOAA/NESDIS OSPO, NOAA Science Center Room 401, 5200 Auth Rd, Camp Springs, MD 20746, USA (URL: www.ospo.noaa.gov/Products/atmosphere/vaac, archive at: http://www.ssd.noaa.gov/VAAC/archive.html); Hawai'i Institute of Geophysics and Planetology (HIGP) - MODVOLC Thermal Alerts System, School of Ocean and Earth Science and Technology (SOEST), Univ. of Hawai'i, 2525 Correa Road, Honolulu, HI 96822, USA (URL: http://modis.higp.hawaii.edu/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Sentinel Hub Playground (URL: https://www.sentinel-hub.com/explore/sentinel-playground).

Observations of Ambrym were made by John Seach during a climb to the caldera during 11-15 December 2002. Lava lakes were visible in both Mbwelesu and Benbow craters that had been absent during a visit in February 2000 (BGVN 25:02) . Reports from local guides indicated that two lava lakes appeared in Mbwelesu crater during February 2001 and joined to form a single lava lake in August 2001. A lava lake reappeared in Benbow crater during June 2002. During November 2002 acid rain, for the third consecutive year, destroyed the mango crops between Sanesup and Lalinda on the W coast of Ambrym.

Activity at Mbwelesu Crater, 12 December 2002. Perfect visibility into the crater enabled detailed observations of the lava lake over 5 hours from the S side of the crater at an elevation of 950 m and over 300 m above the lava lake. The lava lake, located at the bottom of Mbwelesu Crater inside a circular pit (figures 6 and 7), had a diameter of 40-50 m, was in constant motion, and made continuous loud crashing sounds like waves at the beach. The lava lake was much more active than during previous visits in 1998 and 1999. Pele's hair littered the observation area, and white lithic blocks up to 30 cm in diameter were scattered on the rim.

Figure 6. Photo of the lava lake inside a circular pit within Mbwelesu Crater at Ambrym, 12 December 2002. The diameter of the lava lake is 40-50 m. Courtesy of John Seach.

Figure 7. Photo showing the violent degassing from the lava lake in Mbwelesu Crater at Ambrym, 12 December 2002. Courtesy of John Seach.

The surface of the lava lake was continuously disrupted by degassing. Bubbles caused the lake surface to blister and finally burst, splashing lava into the air. Up to eight large bubbles formed at any one time and covered over 80% of the lake surface. The cycle of bubble formation and rupture took about 3 seconds. Waves up to 10 m high formed due to the degassing and crashed onto the side of the pit. After lava waves hit the side of the pit there was a drain-back of lava into the main lake much like ocean waves receding off a beach. Jets of lava were regularly expelled from the lake surface and directed both vertically and at an angle towards the pit side. Fountains reached up to 40 m high. Blobs of molten lava spattered onto the side of the pit up to 20 m from the lava lake edge. This spatter was more erratic than lava fountains and sprayed over a greater area. When large amounts of lava were thrown onto the pit wall, some would cascade back into the lake via a lava stream, lava fall, or a wide curtain of orange flowing lava.

Crusting of the surface was observed when parts of the lake had a lower level of activity, most often in the NE part of the pit opposite the area of most vigorous degassing. Sometimes a lava fountain would burst through the crust, throwing darker pieces of lava high into the air. At times the orange lava lake surface was covered with black pieces of broken crust. Crusting lasted for only a few minutes at a time before it was disrupted by fountains or waves. Lava disappeared into the lava lake surface by subducting under layers of other lava. Some lava disappeared into overhangs on the side of the pit. Lava lake activity continued out of view for an unknown distance past these overhangs.

The lava lake level rose and fell over a period of less than an hour in response to changes in the surface degassing rate. When the rate of degassing was high the lake level was raised by 10 m. The changes appeared to be caused by inflation of the lake due to gas rather than any change in lava eruption rate. During a period of low lava lake activity, the whole lake surface tilted 5 m towards the N and then back to the S over a two-second period. Violent intra-crater winds were observed around the lava lake as reflected in their effects on gas emissions. These were also felt beside the lava lake in Benbow crater. Vapors emitted from the lake surface were white tinged with blue.

Two 15-m-diameter vents 100 m N of the lava lake and 60 m higher were separated by a thin wall. The W vent did not show any activity. The E vent made almost continuous loud degassing noises, and larger explosions ejected black ash 50 m into the air. Mbwelesu was approached again on 15 December, but rain the previous day and low clouds had filled the crater with white vapor, allowing only brief views of the still constantly active lava lake.

Activity at Mbogon Niri Mbwelesu, 12 December 2002. This small collapse pit has been re-named (formerly Niri Mbwelesu Taten) after a request by local residents. The new name comes from the local Port Vato language of W Ambrym, as did the previous name, but is more culturally appropriate. The translation of the new name is " mouth of the wild young pig" (Mbogon = mouth, Niri = son, Mbwelesu = wild pig).

On 12 December excellent visibility enabled detailed observations into Mbogon Niri Mbwelesu. Observations were made from the N side of the pit. Loud crashing, degassing sounds were heard inside the pit, and a 10-m-diameter vent was observed on the floor about 180 m below. The pit glowed bright orange, but lava was not directly observed. This was the first time in 2002 that guides had observed the presence of lava in this pit. Loud degassing occurred every few seconds, and the larger explosions were accompanied by light brown emissions and ground shaking. Pungent sulfurous fumes were emitted from the pit, forcing the observer to use a respirator at times. Strong degassing of brown vapors was coming from the E side of the pit, 50 m below the rim. The W inside wall of the pit was coated with red and yellow deposits.

Activity at Niri Mbwelesu Crater, 12 December 2002. On 12 December excellent views were obtained into Niri Mbwelesu. A recent large landslide on the W wall of the crater had covered the previously lava-filled vent. Rockfalls were heard regularly inside the crater and degassing occurred about every 30 seconds. About every 20 minutes larger explosions were heard at the crater; some were audible over 3 km away.

Activity at Benbow Crater, 13 December 2002. Benbow was climbed from the S on 13 December. The observer free-climbed 165 m down to the floor of the first level, and then another 45 m further down to the edge of the lava lake pit in the N of the crater. Inside Benbow there were two active pits. The larger pit, in the middle of the crater, contained a crusted lava lake and two active vents. The SW vent was 25 m in diameter and was full of vapor but emitted no sounds. The NW vent was 10 m in diameter, glowed red, and loudly degassed. The N crater in Benbow contained an active lava lake. The observer climbed to the rim and was able to view the lake surface, ~50 m below, for a few seconds before retreating. The lava lake was in constant motion and lava was ejected in to the air. Violent winds (over 80 km/hour) were generated inside the pit and made observations on the edge dangerous. At times the pit was filled with white and blue-tinged vapors which made breathing difficult. The lava lake made continuous rumbling and sloshing noises. On a wall next to the lava lake pit there was dripping water with a pH of 3.5 and 700 ppm total dissolved solids.

Visit to Ambrym, 15-20 August 2001. Jeff and Raine Williams, sailing aboard the S/Y Gryphon, visited Ambrym Island during 15-20 August 2001. One day was spent hiking to the Mbwelesu crater with a guide from the village of Ranvetlam. Their report has been reduced here to basic observations; a more poetic and complete description of their hike can be found on their website. After leaving Ranvetlam, they began a steep climb through jungle and gardens, continuing through coconut groves and thick woods of breadfruit trees and wild nut trees. After an hour they were still passing through the garden plots of villagers. At higher altitudes the vegetation changed to bananas, kava, and lap-lap plants; wild tree ferns and palm trees were abundant.

After about 90 minutes they emerged from the jungle onto a lava flow at the lower limit of the high central 'ash plain' plateau. They climbed along this "50-yard wide, black gravel road," also described as a "wild orchid-lined highway," through the jungle to the ash plain itself, where the tops of Marum and Benbow could be seen shrouded in clouds and mist. The hike continued across ~1.5 km of the ash plain before passing along a lava gully onto the final ridge, a 1-m-wide path of loose cinders and stone. They climbed to the rim and looked down the sheer, nearly vertical cliffs into the crater, where they heard rumbling and splashing sounds of the active lava lake. Although the weather was cold and windy, the fog cleared enough for the visitors to briefly observe bright red lava in the crater three times within 30 minutes. The 11-km-long hike to the crater took four hours, and another 3 hours to return.

Geologic Background. Ambrym, a large basaltic volcano with a 12-km-wide caldera, is one of the most active volcanoes of the New Hebrides arc. A thick, almost exclusively pyroclastic sequence, initially dacitic, then basaltic, overlies lava flows of a pre-caldera shield volcano. The caldera was formed during a major plinian eruption with dacitic pyroclastic flows about 1900 years ago. Post-caldera eruptions, primarily from Marum and Benbow cones, have partially filled the caldera floor and produced lava flows that ponded on the caldera floor or overflowed through gaps in the caldera rim. Post-caldera eruptions have also formed a series of scoria cones and maars along a fissure system oriented ENE-WSW. Eruptions have apparently occurred almost yearly during historical time from cones within the caldera or from flank vents. However, from 1850 to 1950, reporting was mostly limited to extra-caldera eruptions that would have affected local populations.

Information Contacts: John Seach, PO Box 16, Chatsworth Island, NSW, 2469, Australia (URL: http://www.volcanolive.com/); Jeff and Raine Williams, P.O. Box 729, Funkstown, MD 21734, USA.

The last Cotopaxi report (SEAN 01:03) described a decline in activity during December 1975. Beginning in October 2001, anomalous seismic activity was registered. Seismicity increased further during November 2001-January 2002, and at times was up to seven times the normal level (tables 1 and 2). During this period, other seismic signals were registered that were distinct from those during the 13 previous years of monitoring, including: tornillos, explosion events, bands of harmonic tremor sometimes lasting a few minutes, and deep, high-energy long-period (LP) events registered away from the volcano (at the Antisana and Guagua Pichincha stations). Seismic observations and statistics were compiled using station "VCl," located ~4 km NE of the volcano. Earthquake locations were determined using records from the seven seismic stations on different flanks of Cotopaxi, and for higher-energy events with stations of the National network.

Table 1. Monthly seismicity at Cotopaxi during 2001-2002. Data includes Total and Daily averages for long-period (LP) events, hybrid events, volcano-tectonic (VT) events, tornillo events, and all earthquakes. Courtesy IG.

Table 2. Comparison of average seismicity at Cotopaxi during 2001 and 2002. Courtesy IG.

On 5 and 29 January 2002, two seismic clusters lasted an average of 2 hours and were composed mainly of LP and VT earthquakes. Most of the earthquakes were located at depths of 1-10 km beneath the summit. On 5 and 13 January small fumaroles were reported in the crater, and visible defrosting occurred on the upper E flank. A visit to the summit on 13 January revealed increased fumarolic activity compared to previous months. On 19 and 20 January observers reported gray plumes rising as high as 1,000 m.

During February and March activity diminished, and no seismic clusters were registered. Most of the earthquakes were located 1-10 km beneath the volcano. On 5 February roaring noises were heard from Mulaló and the refuges located on the flanks of the volcano. Strong fumarolic activity was also reported. On 6 February steam plumes rose ~300 m above the summit. On 27 February a small steam plume was reported exiting from the NW side of the crater. On 7 and 10 March small steam plumes originated from the W side of the crater. On 28 March harmonic tremor lasted for ~10 minutes.

Activity remained low during April-June. On 17 April a band of harmonic tremor lasted ~6 minutes with a maximum frequency of 4.3 Hz. During the first days of April small steam plumes were reported. During May LP earthquakes lasted up to a minute and saturated the seismometer for several seconds. On 20 May a seismic cluster of LP earthquakes lasted ~2 hours. On 8 and 14 May a white steam plume from the NE side of the volcano reached up to 200 m high. During June VT events mostly occurred ~10 km N of the crater. On 30 June a band of harmonic tremor lasted ~7 minutes with a maximum frequency of 1.7-5.2 Hz. Visits to the summit on 1 and 2 June revealed that fumarolic activity had diminished ~40% since January.

During July seismicity was at a moderate level with respect to the rest of 2002. During the first days of the month a series of LP events were registered that were large enough to be detected at distant stations, such as Antisana and Guagua Pichincha. The earthquakes had maximum frequencies of ~2.1 Hz and were generally 1-2 km beneath the summit. However, some events were located at depths of ~10 km. On 18 July at 2000 a band of low-frequency tremor lasted ~4 minutes. About 5 hours later a seismic cluster began that lasted for ~8 hours. The cluster consisted of ~110 total events, mostly hybrid (HB) and volcano-tectonic (VT). The earthquakes were located 1-4 km beneath the summit, and 2 LP events were located ~10 km deep.

Visitors to the summit on 6 July reported fumarolic activity in the zone of Yanasacha, a slight sulfur smell on the NE side, and noise generated by an avalanche on the E side. At the end of July reports indicated defrosting in the W zone. During August moderate seismicity was dominated by LP events at a depth of ~10 km.

Seismicity was again high in September 2002. A small cluster of VT earthquakes on 15 September lasted ~7 hours. During the first days of the month a visit to the crater revealed new fumaroles in the E and S zones. Defrosting continued in the W zone and left 40% of the W wall open.

During October seismic activity was low but the number of hybrid events increased compared to the previous months. Tectonic events were registered in the S and N zones up to ~7 km from the summit. Deep LP events decreased by ~50% compared to previous months.

Seismicity remained low during November and December. Less than 10% of VT events were registered in the N sector. No fumarolic or other surface activity was observed. During December seismic events were located 1-7 km beneath the summit. On 7 December people in Yanahurco reported dark brown plumes rising from the crater.

Seismicity since 1989 clearly shows an increase in recent months (figure 1). The 2001 seismic events were registered at 1-10 km beneath the volcano, but ~90% occurred at 2-4 km and showed little migration. The 2002 activity was variable, from a high of 966 events in January to a low of 420 events in April. Mostly LP events occurred with some VT events during the first half of the year, and later mostly LP events with hybrids during the second half of the year. On the basis of 2002 seismic activity, a new injection of magma did not occur, and the anomalies in July and September were the result of the movement of gas from magma intrusion that occurred during the last months of 2001.

Figure 1. Graph of the total registered monthly events at Cotopaxi during 1989-2002. The activity increased beginning in November 2001 and has since remained above background levels. Courtesy of IG.

Geologic Background. Symmetrical, glacier-clad Cotopaxi stratovolcano is Ecuador's most well-known volcano and one of its most active. The steep-sided cone is capped by nested summit craters, the largest of which is about 550 x 800 m in diameter. Deep valleys scoured by lahars radiate from the summit of the andesitic volcano, and large andesitic lava flows extend to its base. The modern conical edifice has been constructed since a major collapse sometime prior to about 5000 years ago. Pyroclastic flows (often confused in historical accounts with lava flows) have accompanied many explosive eruptions, and lahars have frequently devastated adjacent valleys. The most violent historical eruptions took place in 1744, 1768, and 1877. Pyroclastic flows descended all sides of the volcano in 1877, and lahars traveled more than 100 km into the Pacific Ocean and western Amazon basin. The last significant eruption took place in 1904.

Information Contacts: Geophysical Institute (IG), Escuela Politécnica Nacional, Apartado 17-01-2759, Quito, Ecuador.

On 26 October 2002 at 2225 a swarm of earthquakes was recorded by the seismic network of the Catania Section of the National Institute of Geophysics and Volcanology (INGV-CT). This signaled the start of a new flank eruption that has formed fissures on the N and S sides of the volcano.

The lava supply from the main vents were cut off by 3 November. At that time both the N and S fissues stopped producing lava flows, although the S fissure continued to discharge fire fountains. After that, 20 m of downslope movement was observed at the most advanced flow front near Piano Provenzana on 5 November. This late movement was caused by channel emptying, and occurred when lava emerging at the main vent, ~5 km upstream, was completely crusted over. No further advancement of the lava flows was observed on the S or N flanks of the volcano after this date. However, while explosive and effusive activity stopped at the N fissure by 5 November, as of 11 November fire fountaining continued at the S vent located at 2,750 m elevation, near Torre del Filosofo. All data (gas emission, volcanic tremor, composition of the ash) suggested a steady state at this vent. Ash fallout caused intermittent disruption at the Catania airport and damage to buildings.

The eruption continued into December 2002. Lava flows and Strombolian activity continued on the S flank from vents at 2,750 m elevation. Ash emission from the 2,750 m cinder cone significantly declined on 17 December, allowing the local airport of Catania to reopen.

The two vents, which opened at the SE base of the 2,750 m cinder cone on 9-10 December, fed four major lava flows spreading S and SW. A lava flow spreading S on 13 December approached the Rifugio Sapienza and eventually crossed a road on 17 December. An overflow from the main lava channel covered a building and caused a strong explosion in the Rifugio Sapienza area during the night of 17 December, injuring 32 people. The explosion was not directly caused by the eruption, but by vaporization of oil or water inside the building while it was covered by the expanding lava flow. The effusion rate from the two vents gradually decreased, eventually causing the closure of the western vent and then the lack of supply to the lava flows spreading SW towards Monte Nero.

A new vent opened on 17 December at the S base of the 2,750 m cinder cone, a few meters W of the previous vents. A lava flow soon started from this vent, spreading SW towards Monte Nero. The new vent cut supply to the flows expanding S towards Rifugio Sapienza and formed a fan of thin lava flows spreading S, SSW and SW. The lower lava output produced shorter flows, which spread up to 2.5 km from the vent, without threatening the tourist facilities at Rifugio Sapienza. Lava flows spreading from the 17 December vent slowed down and crusted over on 22 December, when a new vent opened at the SW base of the 2,750 m cinder cone. A flow, again directed SW towards Monte Nero, originated from this vent and was expanding in this direction on 23 December.

SO₂ emission measured daily during the eruption had significantly decreased as of 1 December, when the previous values of about 20,000 tons per day decreased to about 7,000 tons per day (figure 101). The lower gas output, the decrease in effusion rate, and the lower emission of ash from the summit, suggested a declining stage of the eruption.

Figure 101. A plot of SO2 flux at Etna during September-December 2002. Courtesy of INGV-CT.

Updated maps of the lava flows, and reports of the eruptive activity, gas emission and ash composition (in Italian), can be found on the INGV-CT website.

Geologic Background. Mount Etna, towering above Catania, Sicily's second largest city, has one of the world's longest documented records of historical volcanism, dating back to 1500 BCE. Historical lava flows of basaltic composition cover much of the surface of this massive volcano, whose edifice is the highest and most voluminous in Italy. The Mongibello stratovolcano, truncated by several small calderas, was constructed during the late Pleistocene and Holocene over an older shield volcano. The most prominent morphological feature of Etna is the Valle del Bove, a 5 x 10 km horseshoe-shaped caldera open to the east. Two styles of eruptive activity typically occur, sometimes simultaneously. Persistent explosive eruptions, sometimes with minor lava emissions, take place from one or more summit craters. Flank vents, typically with higher effusion rates, are less frequently active and originate from fissures that open progressively downward from near the summit (usually accompanied by Strombolian eruptions at the upper end). Cinder cones are commonly constructed over the vents of lower-flank lava flows. Lava flows extend to the foot of the volcano on all sides and have reached the sea over a broad area on the SE flank.

Information Contacts: Sonia Calvari, Istituto Nazionale di Geofisica e Vulcanologia (INGV), Sezione di Catania, Piazza Roma 2, 95123 Catania (URL: http://www.ct.ingv.it/).

During September-29 December 2002, seismicity at Karangetang was dominated by emission, multiphase and tectonic earthquakes (table 6). The S crater nearly always issued "white, thin ash plumes" that reached up to 500 m above the rim. At night, a light plume was visible rising 25-100 m. Loud noises were heard frequently, and the N crater emitted a "thin white ash plume" to 50 m. No ashfall was reported.

Table 6. Earthquakes recorded at Karangetang during 9 September-29 December 2002. No reports were issued for Karangetang during 25 November-22 December. Courtesy VSI.

During 9 September-13 October glowing avalanches flowed 25-250 m toward Nanitu river (West Siau), and toward Beha river as far as 400 m from the crater rim. By the week of 14-20 October, the lava avalanches extended ~1.5 km toward the Nanitu river, 1.0 km toward the Beha river (West Siau), and 750 m toward the Kahetang river.

On 12 September loud noises were accompanied by a 50-m-high gray ash plume. During 5-6 October, there were 2 volcanic tremor events. On 19 October at 1759 an explosion ejected glowing material to a height of 500 m; it landed inside the crater. A gray-black ash plume reached up to 750 m, drifted to the N, and fell on the sea.

Activity decreased during November, and loud sounds were rarely heard. On 15 November at 0248 an ash explosion produced glowing material up to ~200 m that fell around the crater. Some of the material entered the Batang, Beha, and Keting rivers, located 300-350 m away. Ash fell around Salili, Beong, Hiu, Ondong, Pehe, and Paniki villages to the SW. The Alert Level remained at level 3 through at least 29 December (on a scale of 1 to 4).

Geologic Background. Karangetang (Api Siau) volcano lies at the northern end of the island of Siau, about 125 km NNE of the NE-most point of Sulawesi island. The stratovolcano contains five summit craters along a N-S line. It is one of Indonesia's most active volcanoes, with more than 40 eruptions recorded since 1675 and many additional small eruptions that were not documented in the historical record (Catalog of Active Volcanoes of the World: Neumann van Padang, 1951). Twentieth-century eruptions have included frequent explosive activity sometimes accompanied by pyroclastic flows and lahars. Lava dome growth has occurred in the summit craters; collapse of lava flow fronts have produced pyroclastic flows.

Information Contacts: Volcanological Survey of Indonesia (VSI), Jalan Diponegoro No. 57, Bandung 40122, Indonesia (URL: http://www.vsi.esdm.go.id/).

Emissions were continuous through at least late October 2002 (table 4). During most of the period 9 September-27 October a "white-thin ash plume" rose 50-400 m and drifted toward the W or SW. No ashfall was reported. Kerinci remained at Alert Level 2 (on a scale of 1-4). No further reports were issued during 2002.

Table 4. Earthquakes registered at Kerinci during 9 September-27 October 2002. Courtesy VSI.

Geologic Background. Gunung Kerinci in central Sumatra forms Indonesia's highest volcano and is one of the most active in Sumatra. It is capped by an unvegetated young summit cone that was constructed NE of an older crater remnant. There is a deep 600-m-wide summit crater often partially filled by a small crater lake that lies on the NE crater floor, opposite the SW-rim summit. The massive 13 x 25 km wide volcano towers 2400-3300 m above surrounding plains and is elongated in a N-S direction. Frequently active, Kerinci has been the source of numerous moderate explosive eruptions since its first recorded eruption in 1838.

Information Contacts: Volcanological Survey of Indonesia (VSI), Jalan Diponegoro No. 57, Bandung 40122, Indonesia (URL: http://www.vsi.esdm.go.id/).

During 9 September through at least late December 2002, seismicity at Krakatau was dominated by A-and B-type volcanic earthquakes (table 2). Throughout the report period, clouds obscured the view of the summit. Krakatau remained at Alert Level 2.

Table 2. Earthquakes registered at Krakatau during 9 September-29 December 2002. No data were available during 16-29 September. Courtesy VSI.

Geologic Background. The renowned volcano Krakatau (frequently misstated as Krakatoa) lies in the Sunda Strait between Java and Sumatra. Collapse of the ancestral Krakatau edifice, perhaps in 416 or 535 CE, formed a 7-km-wide caldera. Remnants of this ancestral volcano are preserved in Verlaten and Lang Islands; subsequently Rakata, Danan, and Perbuwatan volcanoes were formed, coalescing to create the pre-1883 Krakatau Island. Caldera collapse during the catastrophic 1883 eruption destroyed Danan and Perbuwatan, and left only a remnant of Rakata. This eruption, the 2nd largest in Indonesia during historical time, caused more than 36,000 fatalities, most as a result of devastating tsunamis that swept the adjacent coastlines of Sumatra and Java. Pyroclastic surges traveled 40 km across the Sunda Strait and reached the Sumatra coast. After a quiescence of less than a half century, the post-collapse cone of Anak Krakatau (Child of Krakatau) was constructed within the 1883 caldera at a point between the former cones of Danan and Perbuwatan. Anak Krakatau has been the site of frequent eruptions since 1927.

Information Contacts: Volcanological Survey of Indonesia (VSI), Jalan Diponegoro No. 57, Bandung 40122, Indonesia (URL: http://www.vsi.esdm.go.id/).

Higher-than-normal activity continued at Lokon-Empung during August-December 2002. Throughout the report period a "white-thin ash plume" rose 25-75 m above the crater rim. No ashfall was reported. Seismicity was dominated by shallow volcanic and tectonic earthquakes (table 4).

Table 4. Earthquakes recorded at Lokon during 5 August-29 December 2002. No reports were issued during 11 November-22 December. Courtesy VSI.

During the week of 4-10 November, the hazard status was reduced from Alert Level 2 to 1 (on a scale of 1-4). On 23 December a "white-thick ash plume" rose 100-250 m over Tompaluan crater. No ashfall was reported. [A later report did note ashfall.] The same day, volcanic tremor with an amplitude of 0.5-2 mm occurred. A total of 42 emissions were reported during 23-29 December. The Alert Level returned to 2 by the end of the report period.

Geologic Background. The twin volcanoes Lokon and Empung, rising about 800 m above the plain of Tondano, are among the most active volcanoes of Sulawesi. Lokon, the higher of the two peaks (whose summits are only 2 km apart), has a flat, craterless top. The morphologically younger Empung volcano to the NE has a 400-m-wide, 150-m-deep crater that erupted last in the 18th century, but all subsequent eruptions have originated from Tompaluan, a 150 x 250 m wide double crater situated in the saddle between the two peaks. Historical eruptions have primarily produced small-to-moderate ash plumes that have occasionally damaged croplands and houses, but lava-dome growth and pyroclastic flows have also occurred. A ridge extending WNW from Lokon includes Tatawiran and Tetempangan peak, 3 km away.

Information Contacts: Volcanological Survey of Indonesia (VSI), Jalan Diponegoro No. 57, Bandung 40122, Indonesia (URL: http://www.vsi.esdm.go.id/).

Satellite data interpreted by Simon Carn indicate that anomalous degassing may have begun from a volcano in Vanuatu in mid-December 2002. SO₂ signals were noted in data from both the Global Ozone Monitoring Experiment (GOME) on the ERS-2 satellite and the Earth Probe Total Ozone Mapping Spectrometer (TOMS). Although GOME is more sensitive to SO₂ than TOMS, its spatial resolution is very poor, so distinguishing the source of emissions between Ambrym and Lopevi is impossible using the available imagery.

However, on 14 December John Seach noted a strong sulfurous smell on the W side of Ambrym caldera. The wind was blowing from the direction of Lopevi at the time, and white emissions were noticed on Lopevi's active crater on the NW flank of the volcano. Seach did not note unusual emissions from Ambrym during his 11-15 December 2002 visit, so the editors are attributing this activity to Lopevi unless other data are found that identify Ambrym as the source.

GOME data indicate SO₂ emissions over Vanuatu on 13, 19, 22, and 25 December 2002, then again during 4, 7, 11, 14, 17, and 20 January 2003. Data are only collected every third day, so degassing could be continuous, with a possible lull in late December. After 11 January GOME signals became very weak. TOMS data also indicated SO₂ originating from the region on 19, 21, and 25 December, and again during 4, 5, 6, 8, 9, 10, 11, and 12 January, with nothing really evident since then. On a couple of days, particularly 4 January, the anomaly seen in TOMS imagery seemed to be originating from Ambrym.

The SO₂ mass detected by TOMS immediately E of Lopevi and Ambrym on 8 January was estimated at less than 5,000 tons, a low value. Combining the two datasets indicates that the most significant SO₂ emissions occurred around 25 December 2002 and 4-11 January 2003. After mid-January the activity seemed to be tapering off.

Geologic Background. The small 7-km-wide conical island of Lopevi, known locally as Vanei Vollohulu, is one of Vanuatu's most active volcanoes. A small summit crater containing a cinder cone is breached to the NW and tops an older cone that is rimmed by the remnant of a larger crater. The basaltic-to-andesitic volcano has been active during historical time at both summit and flank vents, primarily along a NW-SE-trending fissure that cuts across the island, producing moderate explosive eruptions and lava flows that reached the coast. Historical eruptions at the 1413-m-high volcano date back to the mid-19th century. The island was evacuated following major eruptions in 1939 and 1960. The latter eruption, from a NW-flank fissure vent, produced a pyroclastic flow that swept to the sea and a lava flow that formed a new peninsula on the western coast.

Information Contacts: Simon A. Carn, TOMS Volcanic Emissions Group, Joint Center for Earth Systems Technology (NASA/UMBC), University of Maryland Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, USA (URL: https://jcet.umbc.edu/); John Seach, PO Box 16, Chatsworth Island, NSW 2469, Australia (URL: http://www.volcanolive.com/).

Accounts from ship-based observers and satellite imagery have revealed significant morphological changes to McDonald Island due to volcanic activity prior to 6 November 2001. A comparison of November 2001 satellite imagery with 1980 aerial photographs was described in AUSGEO News 68 (December 2002). Tourist reports were published in the Australian Antarctic Division's Antarctic Non-government Activity News (ANAN), no. 89 (January 2003). Geoscience Australia's National Mapping reports the elevation of McDonald Island as 230 m, but the activity described below has most likely increased this value.

A photograph taken on 9 November 2000 (BGVN 26:02) was similar to previous photos and descriptions. In addition, thermal alerts for nearby Heard Island occurred frequently in November and December 2000, an indication not only of eruptive activity there, but clear weather during which any significant activity at McDonald would likely have been detected in infrared satellite imagery. Combined, these observations place the eruptive activity after 9 November 2000, and probably after 30 December 2000.

Analysis of 6 November 2001 satellite imagery. A routine check of Australia's maritime boundaries in the Southern Ocean by Geoscience Australia showed that the McDonald Islands had doubled in size, and it appears that the separate islands of McDonald Island and Flat Island are now one. Geoscience Australia's Bill Hirst was comparing an aerial photograph of the McDonald Islands taken on 11 March 1980, with satellite imagery from Landsat 7 EGM data acquired on 6 November 2001, when he noticed that the islands had changed shape (figure 6). The islands earlier combined area of 1.13 km² is now thought to have changed to 2.45 km². Some features have disappeared.

Figure 6. Aerial photograph of the McDonald Islands taken on 11 March 1980 from a helicopter (left) and satellite imagery from Landsat 7 EGM data acquired on 6 November 2001 (right). The outline of the islands in 1980 is superimposed on the satellite image. Courtesy of Geoscience Australia.

The senior surveyor onshore during a 6-day visit in 1980 was Geoscience Australia's John Manning, who named many features of the McDonald Islands. He noted that "Thelander Point doesn't appear to be an appropriate name now, Williams Bay seems to be filled in, and The Needle may be gone . . . Windward Point is no longer a point because there are about 400 m of new land in front of it. The tumultuous bay I called Cauldron is now full of rock, and Flat Island is probably joined to McDonald Island by a shingle comprising gravel and pumice." Other new features appear to be a volcanic hill and a spit to the E of the island similar to one on Heard Island. Macaroni Hill was once the highest point.

Observations in late November 2002. Experienced observers noted changes to the McDonald Island group in late November 2002 from on board the Akademic Shokalskiy, which was visiting the Heard Island region on a voyage organized by the New Zealand-based tour company Heritage Expeditions. A comparison of old and new photographs of the area shows that the N part of the island is much higher than before, and 75% of the land area that is now there may be completely new. During the last five years Australian national program vessels that have observed the McDonald group have reported seeing steam issuing from vents at various locations.

Three of the passengers on the Akademic Shokalskiy had worked on Heard Island in the 1950's and 1960's, and one of them, Graham Budd, was one of the first two people to set foot on McDonald Island, in 1971. When the ship was travelling towards Heard Island en route from Crozet early on the morning of 26 November, Budd noticed the changed profile of the McDonald islands and expedition leader Rodney Russ decided to take a closer look after the end of the visit to Heard Island. It was not possible to sail too close to the islands because the water around them is uncharted. Under Australian management plans for McDonald Island, landings cannot be made there without a permit and only then for "compelling scientific reasons."

On the second sail past the island, passengers observed steaming slopes and "two types of lava dome." The highest part of the islands was now at the N end, not in the S at Maxwell Hill as it had been previously. Analysis of enlarged digital photographs taken by passengers indicates that considerable sedimentation has occurred along the coastline, such that the formerly separate Flat Island is now joined to the main island. It also appears that several meters of ash have blanketed the N half of McDonald Island, and Macaroni Hill at its N end has disappeared. A low-lying spit and reef now extend over 1 km E of McDonald Island.

Although it is not certain when the activity occurred, wildlife did not appear to have been affected. Penguins were still nesting up to the top of Maxwell Hill and on ash-covered remnants of the old land inshore of the new spit. The birds appear to have deserted Flat Island. There were a large number of penguins and seals on the beaches, and several dozen fur seals swimming offshore.

The two geologists on the voyage, Australian Jon Stephenson and New Zealander Margaret Bradshaw, believe that a scientific visit should be made so that the sequence of the new volcanic events and the composition of the lavas can be determined. The Australian national program currently plans to conduct a scientific program on Heard Island during the 2003-04 austral summer, but currently has no plans to do land-based research on McDonald Island.

MODVOLC Thermal Alerts. Following the distribution of the above reports via the Volcano Listserv, David Rothery and Diego Coppola (The Open University) searched for "thermal alerts" at McDonald Island using the MODIS Thermal Alerts website (http://modis.higp.hawaii.edu/). This system is the first truly global high-temperature thermal monitoring system. It is capable of detecting and documenting changes in active lava flows, lav domes, lava lakes, strongly incandescent vents, and hot pyroclastic flows. No alert is likely to be triggered by an ash cloud.

As described by Flynn et al. (2001) and Wright et al. (2002), the MODIS Thermal Alerts website provides a series of maps updated every 24 hours to show "thermal alerts" based on night-time (approximately 2230 local time) infrared data from a 1-km-resolution instrument called MODIS that is carried by NASA's Terra and Aqua satellites. Thermal alerts are based on an "alert ratio" (3.9 µm radiance - 12 µm radiance) / (3.9 µm radiance + 12 µm radiance), and an alert is triggered whenever this ratio has a value more positive than -0.8. This threshold value was chosen empirically by inspection of images containing known volcanic sites at high temperature, and is the most negative value that avoids numerous false alarms. There are also some daytime (approximately 1030 local time) alerts that are based on the same algorithm but incorporating a correction for estimated solar reflection and a more stringent threshold whereby the alert ratio is required to be more positive than -0.6 in order to trigger an alert.

Thermal alert data are available for the region including McDonald Island from 13 May 2000 onwards (with a gap 26 May-2 June 2000). No thermal alert occurred at McDonald Island from 13 May 2000 through 30 January 2003. This null result does not prove that the activity must have occurred before 13 May 2000, because MODIS cannot see through cloud, which is common in that region. However, there were multiple thermal alerts for nearby Heard Island during the same period (24 May; 3, 5, and 6 June; 25 September; 29 October; 5, 15, 19, and 24 November; 16, 17, 26, and 30 December 2000; 2 February 2001). Had McDonald been active on the same dates, it is highly likely that this activity would have been detected at least once.

Climate and Biology. The following is taken from the AUSGEO News report. The McDonald Islands are remote, and people have landed on the islands only twice since a British sealer sighted them in November 1833. The islands have cliff-lined coasts and are surrounded by rocky shoals and reefs that are treacherous for boats and landing parties. They lie in stormy seas where temperate water from the Indian Ocean meets icy Antarctic water. Most days are cloudy, making it very difficult to obtain satellite imagery and photographs of the islands. Maximum temperatures average 3°C, and wind gusts can reach 210 km/hour. Two Australian scientists looking for fur seals made the first landing in 1970, a 20-minute visit, by helicopter from the French Antarctic ship Gallieni. The second landing, in March 1980, was from the Cape Pillar, chartered by National Mapping to survey the Heard Island-Kerguelen region. The small shore party, which included a botanist, biologist, geologist, and surveyor, landed by helicopter and amphibious vehicle. They stayed ashore for six days while the ship sailed its survey lines.

The McDonald Islands were designated a World Heritage site in December 1997 because of their pristine sub-Antarctic ecosystems and geological activity. Local waters are teaming with Patagonian toothfish, Mackerel icefish, Grey rockcod, and Unicorn icefish. Colonies of Macaroni and Gentoo penguins breed and feed from these islands.

References. Flynn, L.P., Wright R., Garbeil, H., Harris, A.J.L., and Pilger, E., 2001, A global thermal alert system using MODIS: initial results from 2000-2001: Advances in Environmental Monitoring and Modelling, no. 3, Monitoring volcanic hotspots using thermal remote sensing, edited by Harris, A.J.L., Wooster, M.J. and Rothery, D. A. (Http://www.kcl.ac.uk/ kis/schools/hums/geog/advemm/vol1no3.html).

Wright, R., Flynn, L., Garbeil, H., Harris, A., and Pilger, E., 2002, Automated volcanic eruption detection using MODIS: Remote Sensing of Environment, v. 82, p. 135-155.

Geologic Background. Historical eruptions have greatly modified the morphology of the McDonald Islands, located on the Kerguelen Plateau about 75 km W of Heard Island. The largest island, McDonald, is composed of a layered phonolitic tuff plateau cut by phonolitic dikes and lava domes. A possible nearby active submarine center was inferred from phonolitic pumice that washed up on Heard Island in 1992. Volcanic plumes were observed in December 1996 and January 1997 from McDonald Island. During March of 1997 the crew of a vessel that sailed near the island noted vigorous steaming from a vent on the N side of the island along with possible pyroclastic deposits and lava flows. A satellite image taken in November 2001 showed the island to have more than doubled in area since previous reported observations in November 2000. The high point of the island group had shifted to the McDonald's N end, which had merged with Flat Island.

Information Contacts: Bruce Hull, Senior Environment Officer, Environmental Management & Audit Unit, Australian Antarctic Division, Environment Australia, Channel Highway, Kingston, Tasmania 7050, Australia (URL: http://www.antarctica.gov.au/environment); AUSGEO News and National Mapping, Geoscience Australia, GPO Box 378, Canberra, ACT 2601, Australia (URL: http://www.ga.gov.au/); David A. Rothery and Diego Coppola, Department of Earth Sciences, The Open University, Milton Keynes MK 6AA, United Kingdom.

Pinatubo's catastrophic 1991 eruption left the volcano with a 2.5-km-wide summit caldera that eventually came to contain a lake (table 8). During 2001 a crisis occurred as the lake's surface neared the low point on the caldera's rim. PHIVOLCS provided a detailed report on trenching and release of lake water to avoid catastrophic breakout of the crater lake. The report that is summarized here was authored and contributed by Ma. Antonia V. Bornas and the Quick Response Team. The brief version given here omits the lengthy list of Team members as well as several figures and the references.

Table 8. Pinatubo crater-lake-water surface level through time and computed monthly and average lake-rise increments. See the original report for data sources. Courtesy PHIVOLCS.

Mount Pinatubo's summit caldera lake surface rose 40 m between May 1998 and July 2001. By July 2001 lake water approached the caldera rim's lowest point, the Maraunot Notch (~960 m elevation). Its surface then stood at 955 m elevation, 5 m below the notch.

The record of the crater lake's rise implied overtopping of Maraunot Notch in the last quarter of 2001. A breach at Maraunot could lead to rapid escape of lake water into an area of abundant unconsolidated pyroclastic deposits (figure 35). Such an event would threaten upriver towns as well as the larger Botolan, Zambales (population ~40,000).

Figure 35. Digital terrain map of the NW Pinatubo quadrant, showing the Maraunot Notch and the contiguous Maraunot-Balin-Baquero-Bucau river system. Botolan town proper and upriver villages are shown. Digital elevations are from the PHIVOLCS-GIS lab. Sources include USGS (1991), Philippine Bureau of Mines (1983), and Fire and Mud (1996). Courtesy PHIVOLCS.

The beheaded upper Maraunot river sits on the NW flank (figure 36) and flows 15 km NW into the Balin-Baquero river. Lahars have long threatened to inundate Botolan town proper. As with the 1991 pyroclastic flows, lahars obliterated villages in the Balin-Baquero and Bucao valleys (e.g. Villar, Burgos, and Poonbato).

Figure 36. Oblique aerial photograph showing the Pinatubo crater, the Maraunot Notch, and the Maraunot-Bucao river system (looking NW) as seen in 2000. Photo courtesy of S. Suto, PHIVOLCS.

Notch and dam characteristics. The valley of the Maraunot Notch contains 150-m-high walls composed of dome rocks and lithified block-and-ash deposits, cut by steep NW- and E-trending faults. Dome rocks also crop out within the first kilometer-long reach of the Maraunot channel and are inferred to form its bedrock. Less competent deposits fill the valley floor and edge off abruptly at the crater, damming the crater lake. This dam is approximately 85 m wide at the edge or crest but narrows as it slopes 8° down-valley to its toe at a prominence of dome rock 70 m away and 10 m below the crest (the nose).

Comprising the dam are a lower pre-1991 terrace of three boulder-rich breccia units and an upper sequence of 1991 deposits. Pre-1991 breccia units are poorly indurated and contain dense dacite-andesite clasts (median diameter, 10-15 cm) in coarse (B1) or fine (B2) ash or coarse sand (B3) matrix. Exposures of the dam in 1998 indicated that pre-1991 breccia may be as much as 14 m thick at the crest. The units also occur as in-channel terraces along the first 700-m reach of the Maraunot River. An overlying 1991 eruption sequence also occurs. It is unconsolidated and up to several meters thick, but has been gullied down to a meter thick along the channel thalweg, creating a 5 m-wide natural spillway at the dam's axis. Thus, unconsolidated 1991 eruption deposits at the dam's upper part left it vulnerable to rapid erosion and possible catastrophic breach.

A potential breach was expected on the occasion of intense rainfall. Dam failure was thought to be potentially initiated by erosion or headcutting of 1991 deposits where the valley narrows or "noses" and the channel drops. The removal of material would lead to increasing flow perimeter and head, which would increase discharge and weaken the dam. Discharge would escalate into a tremendous rush of water, accelerating erosion headward in a runaway process that culminated in dam failure. This same process has been documented in numerous cases of overtopped natural and man-made dams that have breached.

In the worst case, a 10- to 20-m-depth of the channel dam corresponding to the vertical gap between the crest and shallow channel bedrock could have been breached, releasing lake volumes of 28 x 10⁶ to 55 x 10⁶ m³. For a 10- to 20-m-deep breach, estimated peak discharges at the breach in such a circumstance are 3,000 and 11,000 m³/s. The breakout flow would be expected to erode and incorporate pyroclastic-flow and lahar sediments at the mid- to lower reaches of the Maraunot River, causing it to bulk up 3-6 times. Resulting large lahars could reach 3- to 7-fold larger distances than in previous typhoons (e.g. 1993). Faced with this hazard, PHIVOLCS proposed in early August 2001 to trench across the channel dam. This formed the core element of a rapid mitigation plan that included information drives, evacuation of risk areas, and lahar watches.

Trenching took place during 23 August-5 September 2001. The bulk of the trench was manually dug by an 80-man crew using pick axes and shovels and, later, by sluicing with a portable 50 m-long pressure hose. Excavation followed the channel thalweg or the natural spillway from crest to toe of the dam. The fully-excavated trench was 70 m long, 4 m wide, and nearly 3.5 m deep. It contained a 1-m-wide and 1.5-m-deep inner terrace that resulted from belated prioritization of depth over width (figures 37 and 38). Its bottom was originally graded ~2%. At the mouth it sloped steeply into 5 m-long plug that confined the lake until its release. In the end, about 700 m³ of material was excavated. On 4 September, observers were stationed at four sites. Evacuation of Botolan began the following day in anticipation of potential lahars.

Figure 37. Oblique photo of Pinatubo's Maraunot Trench looking NE, taken the day before the channel was opened. Inset shows the mouth on 1 September 2001, ~ 2 m above the lake level; bottom lefthand inset is the profile of the trench. Courtesy PHIVOLCS.

Figure 38. View showing of the mouth and the terraced inner geometry of the Pinatubo's Maraunot Trench, 6 September 2001. Courtesy PHIVOLCS.

On 6 September, with a 10-cm-head of water, the plug was removed by sluicing. At 0653, after less than 1.25 hours of sluicing, lake water spill into the trench commenced, but discharge remained sluggish in the first four hours (~0.03 m³/s). Political developments led to the trench being left in a state that thwarted rapid, planned breaching.

Monitoring the newly opened trench. From 6 September to 5 November, local rainfall and outflow conditions and changes in configuration of the Maraunot trench were monitored. An estimated 4.4 x 10⁶ m³ (~86,000 m³/day) of rainwater entered the crater between 6 September and 5 November. In response, discharge across the trench fluctuated but rarely exceeded 1 m³/s under a lake head generally under 1 m. The total water output at the trench was roughly 3 x 10⁶ m³ (~59,000 m³/day) for the same period.

Time-series profiles of the trench floor revealed a total 1.5 m of downcutting in the period 8 September-21 October, an average of ~3.5 cm/day. As the terminus lowered close to bedrock and precipitation waned, however, the floor more or less stabilized, as did the trench's mouth-to-terminus elevation drop of 2.2 m. No substantial lateral erosion occurred at the 5-15 reach or in the first 30 m reach between 6 September and 5 November. Nevertheless, there was significant lateral erosion of as much as 2 m at the 55-65 m reaches and beyond. Erosion was attributed largely to the steeper channel and more turbulent flow at the trench's terminal reaches.

The pre-1991 breccia matrix eroded with vertical scour experienced uniformly across the entire floor and lateral scour (sidecutting) confined to the terminal reaches. Matrix erosion resulted in armoring of the trench floor with dense boulders. This partly accounted for restrained vertical scouring.

Trenching impacts to the lake breakout problem. Although the trench did not trigger a rapid breach as PHIVOLCS originally intended, the monitoring determined that the armoring provided by coarse pre-1991 breccia limited vertical scouring of the dam. Lateral matrix erosion and bank collapse were considered to deliver even further armor to the trench bed, as well as some sideways expansion of the channel.

Trenching by itself had significantly reduced the breakout hazard. The lake was averted from growing an extra 11 x 10⁶ m³ and relieved of another 3 x 10⁶ m³ with a trench now draining it. This minimized the magnitude of lake breakout. Had natural overtopping been allowed to occur under sustained intense rainfall, initial outflow could have easily scoured a wider channel across the loose 1991 deposits, attaining discharge rates possibly too high for pre-1991 breccia to counteract with armoring.

Geologic Background. Prior to 1991 Pinatubo volcano was a relatively unknown, heavily forested lava dome complex located 100 km NW of Manila with no records of historical eruptions. The 1991 eruption, one of the world's largest of the 20th century, ejected massive amounts of tephra and produced voluminous pyroclastic flows, forming a small, 2.5-km-wide summit caldera whose floor is now covered by a lake. Caldera formation lowered the height of the summit by more than 300 m. Although the eruption caused hundreds of fatalities and major damage with severe social and economic impact, successful monitoring efforts greatly reduced the number of fatalities. Widespread lahars that redistributed products of the 1991 eruption have continued to cause severe disruption. Previous major eruptive periods, interrupted by lengthy quiescent periods, have produced pyroclastic flows and lahars that were even more extensive than in 1991.

Information Contacts: Ma. Antonia V. Bornas and theQuick Response Team, Geology and Geophysics Research and Development Division, Philippine Institute of Volcanology and Seismology, C.P. Garcia Ave., University of the Philippines Campus, Diliman 1101, Quezon City, Philippines.

Higher-than-normal seismic and explosive activity occurred at Semeru during June-September 2002 (BGVN 27:09). During 9 September-29 December, activity continued to be higher than normal. Seismicity was dominated by explosions and avalanche earthquakes (table 10). Throughout the report period, a white-gray ash plume rose 400-500 m high above the Jonggring Seloko crater rim. There were eight explosions on 23 December, one explosion on 25 December, seven explosions on 26 December, eight explosions on 27 December, and another seven explosions on 29 December.

Table 10. Earthquakes recorded at Semeru during 9 September 2002-1 January 2003. "*" indicates that the report was part of a special report issued by VSI and may break the sequence of weekly reports. Courtesy VSI.

On 25 December, a pyroclastic flow traveled 2.5 km and entered the Besuk Kembar river. On 27 December lava avalanches traveled 250 m toward Besuk Kembar. On 29 December a 5 km pyroclastic flow occurred. The same day during 1700-2015 a lahar flowed along Besuk Kembar closer to Supit village. Early on the morning of 30 December residents of Supit village were evacuated. The same day at 0720 a pyroclastic flow traveled 2.0 km toward Besuk Kembar and at 1000 a pyroclastic flow traveled 4.0 km, approaching Supit village. Semeru remained at Alert Level 2 (on a scale of 1-4).

Geologic Background. Semeru, the highest volcano on Java, and one of its most active, lies at the southern end of a volcanic massif extending north to the Tengger caldera. The steep-sided volcano, also referred to as Mahameru (Great Mountain), rises above coastal plains to the south. Gunung Semeru was constructed south of the overlapping Ajek-ajek and Jambangan calderas. A line of lake-filled maars was constructed along a N-S trend cutting through the summit, and cinder cones and lava domes occupy the eastern and NE flanks. Summit topography is complicated by the shifting of craters from NW to SE. Frequent 19th and 20th century eruptions were dominated by small-to-moderate explosions from the summit crater, with occasional lava flows and larger explosive eruptions accompanied by pyroclastic flows that have reached the lower flanks of the volcano.

Information Contacts: Volcanological Survey of Indonesia (VSI), Jalan Diponegoro No. 57, Bandung 40122, Indonesia (URL: http://www.vsi.esdm.go.id/).

Following heightened seismicity during June-July 2002 that culminated in an explosion on 24 July (BGVN 27:07), major activity lessened until late December.

On 28 December, an effusive eruption started at the base of Crater 1 of the NE Crater in the summit area. This eruption ended on 29 December and a helicopter-borne thermal camera survey that day revealed three lava flows that had spread in the eastern Sciara del Fuoco and had reached the sea. Along the coast, the joined flows were ~300 m wide, but were no longer being fed.

Visibility improved on 30 December, when a new survey found an eruptive fissure running NE. The fissure started from the base of Crater 1 at ~700 m elevation and spread down to ~600 m elevation, along a length of ~200 m. On 30 December observers saw a ~200-m-long lava flow emitted from the base of the fissure, spreading in the upper Sciara del Fuoco into a small depression.

Landslides and tsunami. On 30 December at 1315 and 1322 two landslides formed along the Sciara del Fuoco. They reached the sea accompanied by fine (0.1 mm grain-size) wet dust falling on the SE flank of the island (from rock collisions during the landslides). The volume of the first landslide was estimated at ~6 x 10⁶ m³ of rock while the second was smaller at ~5 x 10⁶ m³ of rock. These landslides detached the lava from the 28 December eruption along the slope together with a large portion of the ground below.

The large volume of rock crashing into the sea caused two tsunamis, each with waves several meters high. The waves spread onto the villages of Stromboli and Ginostra damaging buildings and boats and injuring several people (according to news reports, six people were evacuated by helicopter and taken to two hospitals on Sicily). Large waves were reported on the northern coast of Sicily, 60 km S of Stromboli. The two separate landslides were formed from two distinct bodies of rock, and left a ridge on the Sciara del Fuoco wall between them. This ridge may collapse in the future; its volume is estimated to be similar to that of the first landslide.

As of 6 January 2003, the effusive eruption and thin lava flows continued along the Sciara del Fuoco. Two vents located at ~500 m and ~300 m elevation in the middle of the Sciara del Fuoco were feeding two narrow flows that merged and reached the sea. Occasional small landslides from the unstable walls of the Sciara covered the lava flows with a thin talus. Concern over another major landslide had diminished due to several small-volume rockfalls from the walls of the depression. The summit craters had not shown any explosive activity since the start of the eruption on 28 December, and no earthquakes were recorded by the indigenous seismic network. Two shocks recorded by INGV seismic stations were directly related to the spreading of the two landslides on the Sciara del Fuoco.

Previous tsunamis at Stromboli occurred in 1930, 1944, and 1954. These were related either to paroxysmal eruptive activity, to landslides along the Sciara del Fuoco, or to pyroclastic flows, but not associated with lava flow venting.

Geologic Background. Spectacular incandescent nighttime explosions at this volcano have long attracted visitors to the "Lighthouse of the Mediterranean." Stromboli, the NE-most of the Aeolian Islands, has lent its name to the frequent mild explosive activity that has characterized its eruptions throughout much of historical time. The small island is the emergent summit of a volcano that grew in two main eruptive cycles, the last of which formed the western portion of the island. The Neostromboli eruptive period took place between about 13,000 and 5,000 years ago. The active summit vents are located at the head of the Sciara del Fuoco, a prominent horseshoe-shaped scarp formed about 5,000 years ago due to a series of slope failures that extend to below sea level. The modern volcano has been constructed within this scarp, which funnels pyroclastic ejecta and lava flows to the NW. Essentially continuous mild Strombolian explosions, sometimes accompanied by lava flows, have been recorded for more than a millennium.

Information Contacts: Sonia Calvari, Instituto Nazionale di Geofisica e Vulcanologia (INGV); Sezione di Catania (URL: http://www.ct.ingv.it/); Stromboli On-Line (URL: http://www.stromboli.net/).

This report presents a summary of activity throughout 2002. During 2002 several episodes of intense seismic activity occurred that shared certain characteristics: clusters of long-period (LP) earthquakes, tremor related to ash emissions, and an increase in VT events on some occasions. Magmatic intrusions during January-March 2002, were generally preceded by LP clusters with dominate frequencies of 3.8 Hz with some oscillating around 1.5-1.6 Hz. Following these clusters, increased tremor occurred, some related to the emission of gas and ash. Eruptive activity included explosions and Strombolian blasts.

In April, activity changed, LP clusters ceased including events with a dominant frequency of 3.8 Hz and began to contain frequencies of ~6 Hz. Since June, VT events seemed to precede LP events or tremor episodes. Precursors of magmatic activity changed slightly. In almost every case, fewer precursory events were registered. Instituto Geofisica (IG) stated that the present eruptive process could be more uncertain than before. In September, the acceleration of processes seemed to indicate variations in internal conditions, such as changes in magma within the conduit, increased temperatures, diminishing percentages of crystals, lower SiO₂, and addition of new gases.

During October-November there was none of the intense tremor activity that usually accompanies new magma injections. Energy remained at very low levels. IG stated that a large number of VT events and their decreased influence on volcanic activity could indicate a low contribution of magmatic gases that could be mobilized and released outside the volcano by means of explosions, continuous ash emissions, or Strombolian activity as previously observed. Further details of 2002 activity follow.

Detailed activity. During the first 2 weeks of January 2002 a high number of low-energy LP earthquakes took place. Some of the LP's were associated with emissions of mainly steam with a moderate magmatic gas concentration. During the last 2 weeks of the month the number of LP's increased remarkably. The LP's occurred in clusters, most of which were preceded by VT events at depths of 4-11 km beneath the summit. Beginning on 15 January it was possible to see a glow coming from the crater, accompanied by the emission of gases. While the emissions diminished during the last week of January, explosions increased in number and magnitude. By the end of January sporadic episodes of tremor and light ashfall occurred in Ambato and Baños. These seismic characteristics, along with frequent roaring noises that occurred with the explosions, indicated possible degassing of a small volume of magma that entered the conduit beginning on 15 January.

During February magma injection apparently disturbed the system, and new gases ascended. Steam and ash emissions occurred, as well as the possible formation of a lava lake. Strombolian activity during 4-18 February was so strong that pyroclastic flows (PF's) descended the WNW flank along the Juive and Cusua valleys. Seismicity was characterized by LP's, tremor related to emissions, a few volcano-tectonic events (VT's), and small explosions.

During the first 3 weeks of March there was Strombolian activity with emissions of lava, gas, and ash, and almost-continuous roaring noises. During the third week of March, activity diminished in intensity until it disappeared almost completely by the last week of the month. Although incandescence was observed at night, it was not as intense as that observed in previous months. Ashfall occurred in Ambato, Quero, Latacunga, Cusua, Chacauco, Penipe, Peula, Patate, Pelileo, Cotaló, and Pillate.

Most of the LP's registered during April were small and rather sporadic, but frequency content changed on 17 April from 4-4.8 Hz to 6-8 Hz. On 22 and 23 April, VT events at 6-8 km depths were followed by strong gas-and-ash emissions. These became quite intense during 24-30 April.

Activity was quite intense during 12-13 and 28-30 May. On 13 May a total of 8 explosions took place, preceded by an increase in the number of LP events. The same day ashfall occurred in Ambato and Baños. On 24 May VT activity took place just before an increase in explosive activity. During 17-26 May explosions were preceded by VT events, and by 30 and 31 May were preceded by LP events. As of the second week of May Strombolian activity, roaring noises, and incandescence in the crater was intense and almost constant. Lava was present in the crater, accompanied by tremor and ongoing emissions. During the last week of the month a continuous gas-ash column drifted mainly W.

During the last week of June intense tremor registered. The tremor occurred for 3 days and contained dominant frequencies of 2.2-2.7 and 1.5 Hz. Tremor lasted up to an hour with an amplitude that saturated seismographs. Many LP's and explosions accompanied the tremor. During June VT events (4-7 km deep) occurred just before tremor and LP events. Several LP's and tremor episodes preceded explosive events. On average the LP's and tremor occurred 2-4 hours before an explosion.

Explosions occurred during the first week of July. During the first 2 weeks, deep VT earthquakes (5-10 km deep) occurred at a rate of ~1 per day and there was an increase in the number of LP's and hybrid earthquakes. VT and LP events preceded new cycles of explosions, not immediately as had previously been noticed, but in this case by about 15 days. After the new cycle of explosive activity began, most of the LP events had frequencies of 1.5-2.5 Hz. Some VT's preceded the LP's and had frequencies of 3.8 and 1.5 Hz. During the second week intense roars were heard, and increasing ash emissions mainly drifted W. There was strong persistent incandescence, and frequent explosions produced loud noises and ash columns 2-4 km above the crater.

During the first 2 weeks of July, several episodes of Strombolian activity were observed, along with continuous but light ash emissions that were accompanied by roaring noises. Ash was deposited in a thin N-S strip between Hualcango and San Pedro de Sabañag (S of Quero), extending toward the W and Igualata. Ash accumulated up to 2.5 mm thick in "El Mirador" at Cerro Arrayán. Activity decreased toward the end of the month, when small plumes were emitted.

During 5-13 September, 8-10 VT earthquakes registered. These preceded the harmonic tremor seen during 13-21 September. Strong explosions and ash emissions also occurred. Ashfalls were noted in distant cities such as Píllaro and Riobamba, located ~30 km NW and SW, respectively.

During the first week of October explosions with reduced displacements greater than 10 cm² took place and ashfall occurred in Pillate, Ambato, Cusua, Penipe, Altar, Bayusig, Matus Alto, and Matus Bajo. During the second and last week of the month VT events preceded explosions. During the last week of the month incandescence and roaring noises were heard. Three ashfalls were noted, two in Guadalupe and one (on 29 October) in Baños (up to 1 mm), Runtún, Pondoa, and Pintitin.

On 10 and 26 November, two peaks of LP activity occurred that were very close to the peaks of VT activity. The first LP peak preceded the first VT peak by two days. This was unusual because the VT peak normally preceded the LP peak. The second LP peak took place around the same time as the VT peak, indicating that the circulation of fluids was almost simultaneous. Incandescence was observed before the VT activity on 26 November. An increase of LP activity seemed to be correlated with the increase of sounds emitted by the volcano. Frequent incandescence in the crater preceded a VT peak.

Magmatic intrusions during 2002. Five magmatic intrusions (figure 18) apparently occurred during (1) 15-29 January, (2) 15-30 April, 12-13, 24-30 May, (3) 28-30 June, (4) 3-13 July, and (5) 5-13 September. Two periods of intense activity also occurred during 8-13 and 21-27 October, and on 10 and 26 November. During April-June magmatic intrusions did not occur along with a peak of seismic activity, but VT's, hybrids, and emissions all occurred, though in smaller numbers than registered in previous years.

Figure 18. Monthly earthquakes at Tungurahua during January 1999-November 2002. Peaks indicated with arrows correspond to periods of inferred magmatic intrusion. Courtesy IG.

Tremor activity was an essential indicator of these magmatic intrusions (figure 19). Later peaks of tremor activity were always during periods of seismicity related to magmatic intrusions, although it was not clear whether the June peak was related to a possible intrusion. Tremor energy was quite variable.

Figure 19. Tremor energy at Tungurahua, 14 September 1999 through 14 November 2002. Many of these tremor episodes were related to small emissions of gas or ash. Arrows indicate 2002 peaks. Courtesy IG.

Deformation measurements. During 2002 EDM measurements on the N flank showed a slight tendency of inflation. This inflation was first noticed during the first half of 2000. During 2002 a shortening of the distance occurred between prisms and reference bases, between -2 and -6 cm with respect to values observed before the reactivation of the volcano. Although there were variations in measurements taken during the year, the overall tendency has been inflation of 4 to 6 cm with respect to that during 1998-2000.

Data from inclinometers RETU and JUIV show a positive drift of the radial axis of station RETU (elevation 4,000 m). The drift would mean a deflation in the NW sector. During September 2002, when numerous explosions occurred, inclinometer movements changed.

During 2002 measurements of the inclinometer at station JUIV5 were stable until October 2002, when there were disturbances in the radial axis and to a greater degree in the tangential axis. Since 10 November both axes showed significant changes of up to 200 µrad. The negative tendency indicated a progressive inflation. This change agreed exactly with the first LP peak on 10 November. The change lasted until 20 November and included the greater peak of VT activity during 2002. After 20 November, both axes became stabilized. The oscillations seen in this slope between September and October occurred simultaneously with other activity, possibly representing slow but continuous magma movement in the lower parts of the volcano.

Geochemistry. SO₂ flux measurements determined by COSPEC during 1999-2002 were generally less than 2,000 tons/day (figure 20). The peaks took place during March and October, with values reaching 3,000-5,000 tons/day. These high values seemed to correspond with the magma injections of December 2001and January and September 2002. Other episodes of seismic activity related to magmatic injection seemed to precede the peaks in SO₂ emission. The high point in August ("3 y 4" on figure 14), followed increased seismicity during June and July.

Figure 20. COSPEC-measured SO2 emissions at Tungurahua during 1999-2002. The arrows indicate the peaks of SO2 that occurred during May and August 2002.

Thermal waters generally increased in temperature ~0.5°C. A small reduction in pH occurred, with a tendency toward alkaline values. During 1998-99, when the seismicity increased, pH also increased, probably because of the magmatic unrest at the time. Conductivity did not change, and neither did geochemical characteristics such as abundances of sulfates, chlorides, and bicarbonates. IG stated that it could not yet be explained how an increase in seismicity seemed to shift the pH of thermal waters (figure 21).

Figure 21. Temperature and pH of thermal waters at Tungurahua during 1994-2002. Courtesy IG.

Future scenarios. Since 1999 Tungurahua has shown frequent, moderate volcanism with occasional lava emissions. This period can be divided into 13 magmatic intrusions of similar characteristics, although the last three injections displayed slight differences. Starting in 1916 Tungurahua displayed intermittent activity until 1918, with periods of tranquility and greater activity than at present.

The present process has been characterized by LP clusters just before and during eruptions. During October and November 2002, VT events usually preceded cycles of increased activity. Strong incandescence on 2 December was not accompanied by strong explosions, Strombolian activity, or lava emissions.

Geologic Background. Tungurahua, a steep-sided andesitic-dacitic stratovolcano that towers more than 3 km above its northern base, is one of Ecuador's most active volcanoes. Three major edifices have been sequentially constructed since the mid-Pleistocene over a basement of metamorphic rocks. Tungurahua II was built within the past 14,000 years following the collapse of the initial edifice. Tungurahua II itself collapsed about 3000 years ago and produced a large debris-avalanche deposit and a horseshoe-shaped caldera open to the west, inside which the modern glacier-capped stratovolcano (Tungurahua III) was constructed. Historical eruptions have all originated from the summit crater, accompanied by strong explosions and sometimes by pyroclastic flows and lava flows that reached populated areas at the volcano's base. Prior to a long-term eruption beginning in 1999 that caused the temporary evacuation of the city of Baños at the foot of the volcano, the last major eruption had occurred from 1916 to 1918, although minor activity continued until 1925.

Information Contacts: Patty Mothes and Indira Molina, Geophysical Institute (Instituto Geofísico, IG), Escuela Politécnica Nacional, Apartado 17-01-2759, Quito, Ecuador.

Additional information about Mt. Pago's recent eruption (BGVN 27:07-27:09) has been provided by members of the U.S. Geological Survey's Volcano Disaster Assistance Program (VDAP). The team donated to the GVP archives an extensive suite of digital photographs (still and video) taken during August-October 2002. The photographers included the helicopter pilot Alan Cameron (Heli Niugini), and VDAP members Andy Lockhart, Jeff Marso, and Elliot Endo.

In terms of the basic distribution of eruptive products, the August-October 2002 photos (figures 7-16) appeared similar to those shown in earlier reports (BGVN 27:07-27:09). All photos were taken from a helicopter, often during routine observation flights provided by the West New Britain Provincial Government. For scale on some of the photos, Cameron estimated that tree heights ranged from 5-30 m, with the taller trees in the low-lying areas and most of the ones in the photos at the shorter end of that range.

Figure 7. A false-color Landsat satellite image labeling some key features at Mt. Pago and its vicinity. N is upwards (parallel to the grid lines) and, for scale, Pago lies ~20 km S of the coast at Cape Hoskins. Although the settlement at Hoskins is labeled, several others also lie along the coast, including some E of Lolo volcano. Taken by LANDSAT 7 on 26 May 2002 (path 94, row 64) and provided courtesy of USGS-VDAP.

Figure 8. An overview of Pago's N sector taken on 7 October 2002 and showing middle to lower flanks and caldera. The shot was taken from the NW, sighting cross-wise to the aligned chain of recent eruptive vents. Freshly erupted lavas have thus far remained confined within the caldera. The extruded massive dacitic lavas include two lava tongues flowing towards the viewer and a larger lava flow ponded in the distance, banked up against older (1911-18) intra-caldera lavas and the caldera's topographic margins. The wide zone of discolored vegetation continues well beyond both the caldera's topographic margin and the photo's left-hand edge. This and several other features such as a zone of deformation and faulting (lower center) appear less distinct here but are highlighted on later figures. Courtesy of USGS-VDAP.

Figure 9. Upper NE flanks of Pago highlighting the broad zone of denuded and knocked-down vegetation there. Most of the trees have been laid flat, and there exist occasional cleared-out gullies resembling avalanche chutes, washouts, and lahar paths. Courtesy of the USGS-VDAP.

Figure 10. A 16 September 2002 view of Pago, as seen looking SSE towards the summit along the aligned, radial-trending chain of vents. Massive lava flows lie in the foreground. Their extrusive vent sits along the main fissure below the lowest cone, in an area of local degassing and conspicuous yellow deposits. Provided courtesy of USGS-VDAP.

Figure 11. A 13 September 2002 photo of Pago's middle-to-upper flanks, including the summit crater and the higher-elevation radial-vent areas. This photo was taken from the NW; in many other photos taken during August-October 2002 white steam plumes tended to obscure the ground. Note the sub-linear swaths of denuded vegetation, particularly two swaths in the left foreground, and the broad area of discolored vegetation in the background behind the fresh lava. The swaths denote the surface traces of recent faults with significant offset, places where existing trees had fallen over. Observation flights in mid- to late September disclosed still further visible, meter-length deformations in this area. Observers inferred that these features reflected a graben formed in the upper portion of a cryptodome. Courtesy of USGS-VDAP.

Figure 12. A close-up photo of Pago's ravaged summit crater taken from the N on 16 September 2002. Despite their proximity to the crater, some portions of the cone's flanks appear relatively undisturbed. Although difficult to see at the limited scale and resolution of this rendition, the original image clearly shows that a band of denuded trees remained standing within the highly disturbed zone along the breach. Many trees in a zone farther downslope were knocked flat. Courtesy of USGS-VDAP.

Figure 13. A closer view of a portion of Pago's NW outer flanks (seen in figure 3 and part of figure 5) centered on Pago's zone of intense deformation and faulting. The traces of two sub-parallel faults offset the intervening area (D) downward, forming a graben, which crosses the steep sides of older, tree-covered lavas. Farther upslope, the two faults intersect the steaming, lowermost cone (C) at several points (D'' and D'''). Downslope, the two faults join a larger system, which seems to curve back towards the massive lavas (E and E'). The massive lavas (A) discharge at the surface at a point just below A'. Courtesy of USGS-VDAP.

Figure 14. Preliminary structural interpretation by Elliot Endo of Pago's zone of intense faulting and deformation. In this interpretation, the upslope area contains a graben; the downslope area a thrust or a region of mass wasting. Courtesy of Elliot Endo, USGS-VDAP.

Figure 15. A closer view showing Pago's graben deformation feature. Earliest photographs available (~ August 15) show this feature in the early stage of development. The photo was taken looking E on 16 September 2002. For scale, mature trees midway along the fault are 10-15 m in length. Courtesy of the USGS-VDAP.

Figure 16. Closeup showing the extreme surface roughness of the recent Pago dacite extrusions appearing in an area near the lower vent. Large fractures sub-parallel to the vent developed during extrusion. Offsets along fractures were estimated to be as much as 5-7 m and the height of numerous adjacent points on the lava flow's surface easily varied by a meter. Courtesy of the USGS-VDAP.

Movie 1. Digital movie of Pago filmed from a helicopter on 6 October 2002 showing the zone of deformation and faulting followed by a views of the lava flows and vents with the summit crater in the distance towards the SSE. Courtesy of the USGS-VDAP. (30 seconds, 10.7 MB MPEG)

During all or part of this August-October 2002 interval, lavas erupted at high rates: 10-20 m³/s. The crystal-poor dacitic lavas were roughly the same as those produced during the ancestral caldera-forming eruption. The same composition had also been consistent for the intervening lavas. By or before the end of October the current eruption had emitted ~60 x 10⁶ m³ to ~100 x 10⁶ m³ of magma. There was some evidence of magma mixing. Available evidence suggested that the magma rose in a dike from source depths of 6-8 km. A vital question was whether a gas-rich eruptive phase might start.

Highlighted in the August-October photos were recent faults and associated surface deformation. These had been documented by Chris McKee (Geophysical Observatory, PNG) who found that these features covered an area on Pago's mid-to-lower NW flanks. In many cases the faults left conspicuous trails marked by swaths of fallen trees across the rainforest (figures 5 and 8). Despite their clear expressions and documentation, a thermal-imaging device found that the faults and adjacent areas generally lacked anomalous high-temperature signals (Steve Saunders, RVO). The obvious exceptions to this occurred where faults cut across either vent areas and their cones or across massive lava flows in the caldera (figure 7). The inferred cause of the faulting and associated deformation was a shallow magmatic intrusion.

The USGS contributors expressed gratitude to their colleagues affiliated with Rabaul Volcano Observatory in Papua New Guinea and the West New Britain Provincial Government who had helped them with field and logistical support.

At the close of 2002 Alan Cameron (Heli Niugini) wrote Endo the following brief note. "Since you left, interest in Mt. Pago seems to have diminished; I have not flown over it for some time. Yesterday I flew a [medical evacution] past it, and smoke, etc. was still rising but the weather was bad and I did not get closer than about a half mile [(~1 km)], so I don't know what it is doing. Hoskins [airport] is still closed to aircraft, and the Talasea [air]strip is often closed due to water over it and the soft surface, so air travel is somewhat unreliable from here."

In the first week of February, Cameron sent another message. "The last time I had a close look at Pago was about a month ago. It still looked to be fairly active in most respects, however there is not much emission of ash now and the lava seems to have slowed, but I think this is on account of the flow being restricted in its exit to the [S]. To my eye it seems that the lava deposit may be increasing in height due to that restriction . . . . I do recall that there is still a great deal of heat from the lava ( I could feel its effect on the helicopter), which supports my feeling that it is building vertically and the lava is still flowing."

Reference. Cooke, R.J.S., 1981, Eruptions at Pago volcano, 1911-1933 (Compiled by R.W. Johnson), in Cooke-Ravian Volume of Volcanological Papers (editor, R.W. Johnson) Geological Survey of Papua New Guinea Memoir 10, 135-46; Printed in Hong Kong by Libra Press Ltd.

Geologic Background. The 5.5 x 7.5 km Witori caldera on the northern coast of central New Britain contains the young historically active cone of Pago. The Buru caldera cuts the SW flank of Witori volcano. The gently sloping outer flanks of Witori volcano consist primarily of dacitic pyroclastic-flow and airfall deposits produced during a series of five major explosive eruptions from about 5600 to 1200 years ago, many of which may have been associated with caldera formation. The post-caldera Pago cone may have formed less than 350 years ago. Pago has grown to a height above that of the Witori caldera rim, and a series of ten dacitic lava flows from it covers much of the caldera floor. The youngest of these was erupted during 2002-2003 from vents extending from the summit nearly to the NW caldera wall.

Information Contacts: Elliot Endo, John Ewert, C. Dan Miller, Andy Lockhart, Jeff Marso, and Chris Newhall, U.S. Geological Survey, David A. Johnston Cascades Volcano Observatory, Volcano Disaster Assistance Program (VDAP), 1300 SE Cardinal Ct, Building 10, Suite 100, Vancouver, WA 98683, USA; Alan Cameron, Chief Pilot, Heli Niugini Kimbe, Box 404, Kimbe WNB, Papua New Guinea; Ima Itikarai and Steve Saunders, Rabaul Volcano Observatory (RVO), Papua New Guinea; Chris Mckee, Port Moresby Geophysical Observatory, PO Box 323, Port Moresby NCD, Papua New Guinea; Hugh Davies, Earth Sciences, University of Papua New Guinea, PO Box 414, University Post Office NCD, Papua New Guinea.

In discussing the week ending on 12 September, "Earthweek" (Newman, 1997) incorrectly claimed that a volcano named "Mount Pinukis" had erupted. Widely read in the US, the dramatic Earthweek report described terrified farmers and a black mushroom cloud that resembled a nuclear explosion. The mountain's location was given as "200 km E of Zamboanga City," a spot well into the sea. The purported eruption had received mention in a Manila Bulletin newspaper report nine days earlier, on 4 September. Their comparatively understated report said that a local police director had disclosed that residents had seen a dormant volcano showing signs of activity.

In response to these news reports Emmanuel Ramos of the Philippine Institute of Volcanology and Seismology (PHIVOLCS) sent a reply on 17 September. PHIVOLCS staff had initially heard that there were some 12 alleged families who fled the mountain and sought shelter in the lowlands. A PHIVOLCS investigation team later found that the reported "families" were actually individuals seeking respite from some politically motivated harassment. The story seems to have stemmed from a local gold rush and an influential politician who wanted to use volcanism as a ploy to exclude residents. PHIVOLCS concluded that no volcanic activity had occurred. They also added that this finding disappointed local politicians but was much welcomed by the residents.

PHIVOLCS spelled the mountain's name as "Pinokis" and from their report it seems that it might be an inactive volcano. There is no known Holocene volcano with a similar name (Simkin and Siebert, 1994). No similar names (Pinokis, Pinukis, Pinakis, etc.) were found listed in the National Imagery and Mapping Agency GEOnet Names Server (http://geonames.nga.mil/gns/html/index.html), a searchable database of 3.3 million non-US geographic-feature names.

The Manila Bulletin report suggested that Pinokis resides on the Zamboanga Peninsula. The Peninsula lies on Mindanao Island's extreme W side where it bounds the Moro Gulf, an arm of the Celebes Sea. The mountainous Peninsula trends NNE-SSW and contains peaks with summit elevations near 1,300 m. Zamboanga City sits at the extreme end of the Peninsula and operates both a major seaport and an international airport.

[Later investigation found that Mt. Pinokis is located in the Lison Valley on the Zamboanga Peninsula, about 170 km NE of Zamboanga City and 30 km NW of Pagadian City. It is adjacent to the two peaks of the Susong Dalaga (Maiden's Breast) and near Mt. Sugarloaf.]

References. Newman, S., 1997, Earthweek, a diary of the planet (week ending 12 September): syndicated newspaper column (URL: http://www.earthweek.com/).

Manila Bulletin, 4 Sept. 1997, Dante's Peak (URL: http://www.mb.com.ph/).

Simkin, T., and Siebert, L., 1994, Volcanoes of the world, 2nd edition: Geoscience Press in association with the Smithsonian Institution Global Volcanism Program, Tucson AZ, 368 p.

Information Contacts: Emmanuel G. Ramos, Deputy Director, Philippine Institute of Volcanology and Seismology, Department of Science and Technology, PHIVOLCS Building, C. P. Garcia Ave., University of the Philippines, Diliman campus, Quezon City, Philippines.

Xinhua News Agency filed a news report on 27 February under the headline "Volcano erupts in Somalia" but the veracity of the story now appears doubtful. The report disclosed the volcano's location as on the W side of the Gedo region, an area along the Ethiopian border just NE of Kenya. The report had relied on the commissioner of the town of Bohol Garas (a settlement described as 40 km NE of the main Al-Itihad headquarters of Luq town) and some or all of the information was relayed by journalists through VHF radio. The report claimed the disaster "wounded six herdsmen" and "claimed the lives of 290 goats grazing near the mountain when the incident took place." Further descriptions included such statements as "the volcano which erupted two days ago [25 February] has melted down the rocks and sand and spread . . . ."

Giday WoldeGabriel returned from three weeks of geological fieldwork in SW Ethiopia, near the Kenyan border, on 25 August. During his time there he inquired of many people, including geologists, if they had heard of a Somalian eruption in the Gedo area; no one had heard of the event. WoldeGabriel stated that he felt the news report could have described an old mine or bomb exploding. Heavy fighting took place in the Gedo region during the Ethio-Somalian war of 1977. Somalia lacks an embassy in Washington DC; when asked during late August, Ayalaw Yiman, an Ethiopian embassy staff member in Washington DC also lacked any knowledge of a Somalian eruption.

A Somalian eruption would be significant since the closest known Holocene volcanoes occur in the central Ethiopian segment of the East African rift system S of Addis Ababa, ~500 km NW of the Gedo area. These Ethiopian rift volcanoes include volcanic fields, shield volcanoes, cinder cones, and stratovolcanoes.

Information Contacts: Xinhua News Agency, 5 Sharp Street West, Wanchai, Hong Kong; Giday WoldeGabriel, EES-1/MS D462, Geology-Geochemistry Group, Los Alamos National Laboratory, Los Alamos, NM 87545; Ayalaw Yiman, Ethiopian Embassy, 2134 Kalorama Rd. NW, Washington DC 20008.

Following the Ms 7.8 earthquake in Turkey on 17 August (BGVN 24:08) an Email message originating in Turkey was circulated, claiming that volcanic activity was observed coincident with the earthquake and suggesting a new (magmatic) volcano in the Sea of Marmara. For reasons outlined below, and in the absence of further evidence, editors of the Bulletin consider this a false report.

The report stated that fishermen near the village of Cinarcik, at the E end of the Sea of Marmara "saw the sea turned red with fireballs" shortly after the onset of the earthquake. They later found dead fish that appeared "fried." Their nets were "burned" while under water and contained samples of rocks alleged to look "magmatic."

No samples of the fish were preserved. A tectonic scientist in Istanbul speculated that hot water released by the earthquake from the many hot springs along the coast in that area may have killed some fish (although they would be boiled rather than fried).

The phenomenon called earthquake lights could explain the "fireballs" reportedly seen by the fishermen. Such effects have been reasonably established associated with large earthquakes, although their origin remains poorly understood. In addition to deformation-triggered piezoelectric effects, earthquake lights have sometimes been explained as due to the release of methane gas in areas of mass wasting (even under water). Omlin and others (1999), for example, found gas hydrate and methane releases associated with mud volcanoes in coastal submarine environments.

The astronomer and author Thomas Gold (Gold, 1998) has a website (Gold, 2000) where he presents a series of alleged quotes from witnesses of earthquakes. We include three such quotes here (along with Gold's dates, attributions, and other comments):

(A) Lima, 30 March 1828. "Water in the bay 'hissed as if hot iron was immersed in it,' bubbles and dead fish rose to the surface, and the anchor chain of HMS Volage was partially fused while lying in the mud on the bottom." (Attributed to Bagnold, 1829; the anchor chain is reported to be on display in the London Navy Museum.)

(B) Romania, 10 November 1940. ". . . a thick layer like a translucid gas above the surface of the soil . . . irregular gas fires . . . flames in rhythm with the movements of the soil . . . flashes like lightning from the floor to the summit of Mt Tampa . . . flames issuing from rocks, which crumbled, with flashes also issuing from non-wooded mountainsides." (Phrases used in eyewitness accounts collected by Demetrescu and Petrescu, 1941).

(C) Sungpan-Pingwu (China), 16, 22, and 23 August 1976. "From March of 1976, various large anomalies were observed over a broad region. . . . At the Wanchia commune of Chungching County, outbursts of natural gas from rock fissures ignited and were difficult to extinguish even by dumping dirt over the fissures. . . . Chu Chieh Cho, of the Provincial Seismological Bureau, related personally seeing a fireball 75 km from the epicenter on the night of 21 July while in the company of three professional seismologists."

Yalciner and others (1999) made a study of coastal areas along the Sea of Marmara after the Izmet earthquake. They found evidence for one or more tsunamis with maximum runups of 2.0-2.5 m. Preliminary modeling of the earthquake's response failed to reproduce the observed runups; the areas of maximum runup instead appeared to correspond most closely with several local mass-failure events. This observation together with the magnitude of the earthquake, and bottom soundings from marine geophysical teams, suggested mass wasting may have been fairly common on the floor of the Sea of Marmara.

Despite a wide range of poorly understood, dramatic processes associated with earthquakes (Izmet 1999 apparently included), there remains little evidence for volcanism around the time of the earthquake. The nearest Holocene volcano lies ~200 km SW of the report location. Neither Turkish geologists nor scientists from other countries in Turkey to study the 17 August earthquake reported any volcanism. The report said the fisherman found "magmatic" rocks; it is unlikely they would be familiar with this term.

The motivation and credibility of the report's originator, Erol Erkmen, are unknown. Certainly, the difficulty in translating from Turkish to English may have caused some problems in understanding. Erkmen is associated with a website devoted to reporting UFO activity in Turkey. Photographs of a "magmatic rock" sample were sent to the Bulletin, but they only showed dark rocks photographed devoid of a scale on a featureless background. The rocks shown did not appear to be vesicular or glassy. What was most significant to Bulletin editors was the report author's progressive reluctance to provide samples or encourage follow-up investigation with local scientists. Without the collaboration of trained scientists on the scene this report cannot be validated.

References. Omlin, A, Damm, E., Mienert, J., and Lukas, D., 1999, In-situ detection of methane releases adjacent to gas hydrate fields on the Norwegian margin: (Abstract) Fall AGU meeting 1999, Eos, American Geophysical Union.

Yalciner, A.C., Borrero, J., Kukano, U., Watts, P., Synolakis, C. E., and Imamura, F., 1999, Field survey of 1999 Izmit tsunami and modeling effort of new tsunami generation mechanism: (Abstract) Fall AGU meeting 1999, Eos, American Geophysical Union.

Gold, T., 1998, The deep hot biosphere: Springer Verlag, 256 p., ISBN: 0387985468.

Gold, T., 2000, Eye-witness accounts of several major earthquakes (URL: http://www.people.cornell.edu/ pages/tg21/eyewit.html).

Information Contacts: Erol Erkmen, Tuvpo Project Alp.

In December 2002 information appeared in Mongolian and Russian newspapers and on national TV that a volcano in Central Mongolia, the Har-Togoo volcano, was producing white vapors and constant acoustic noise. Because of the potential hazard posed to two nearby settlements, mainly with regard to potential blocking of rivers, the Director of the Research Center of Astronomy and Geophysics of the Mongolian Academy of Sciences, Dr. Bekhtur, organized a scientific expedition to the volcano on 19-20 March 2003. The scientific team also included M. Ulziibat, seismologist from the same Research Center, M. Ganzorig, the Director of the Institute of Informatics, and A. Ivanov from the Institute of the Earth's Crust, Siberian Branch of the Russian Academy of Sciences.

Geological setting. The Miocene Har-Togoo shield volcano is situated on top of a vast volcanic plateau (figure 1). The 5,000-year-old Khorog (Horog) cone in the Taryatu-Chulutu volcanic field is located 135 km SW and the Quaternary Urun-Dush cone in the Khanuy Gol (Hanuy Gol) volcanic field is 95 km ENE. Pliocene and Quaternary volcanic rocks are also abundant in the vicinity of the Holocene volcanoes (Devyatkin and Smelov, 1979; Logatchev and others, 1982). Analysis of seismic activity recorded by a network of seismic stations across Mongolia shows that earthquakes of magnitude 2-3.5 are scattered around the Har-Togoo volcano at a distance of 10-15 km.

Figure 1. Photograph of the Har-Togoo volcano viewed from west, March 2003. Courtesy of Alexei Ivanov.

Observations during March 2003. The name of the volcano in the Mongolian language means "black-pot" and through questioning of the local inhabitants, it was learned that there is a local myth that a dragon lived in the volcano. The local inhabitants also mentioned that marmots, previously abundant in the area, began to migrate westwards five years ago; they are now practically absent from the area.

Acoustic noise and venting of colorless warm gas from a small hole near the summit were noticed in October 2002 by local residents. In December 2002, while snow lay on the ground, the hole was clearly visible to local visitors, and a second hole could be seen a few meters away; it is unclear whether or not white vapors were noticed on this occasion. During the inspection in March 2003 a third hole was seen. The second hole is located within a 3 x 3 m outcrop of cinder and pumice (figure 2) whereas the first and the third holes are located within massive basalts. When close to the holes, constant noise resembled a rapid river heard from afar. The second hole was covered with plastic sheeting fixed at the margins, but the plastic was blown off within 2-3 seconds. Gas from the second hole was sampled in a mechanically pumped glass sampler. Analysis by gas chromatography, performed a week later at the Institute of the Earth's Crust, showed that nitrogen and atmospheric air were the major constituents.

Figure 2. Photograph of the second hole sampled at Har-Togoo, with hammer for scale, March 2003. Courtesy of Alexei Ivanov.

The temperature of the gas at the first, second, and third holes was +1.1, +1.4, and +2.7°C, respectively, while air temperature was -4.6 to -4.7°C (measured on 19 March 2003). Repeated measurements of the temperatures on the next day gave values of +1.1, +0.8, and -6.0°C at the first, second, and third holes, respectively. Air temperature was -9.4°C. To avoid bias due to direct heating from sunlight the measurements were performed under shadow. All measurements were done with Chechtemp2 digital thermometer with precision of ± 0.1°C and accuracy ± 0.3°C.

Inside the mouth of the first hole was 4-10-cm-thick ice with suspended gas bubbles (figure 5). The ice and snow were sampled in plastic bottles, melted, and tested for pH and Eh with digital meters. The pH-meter was calibrated by Horiba Ltd (Kyoto, Japan) standard solutions 4 and 7. Water from melted ice appeared to be slightly acidic (pH 6.52) in comparison to water of melted snow (pH 7.04). Both pH values were within neutral solution values. No prominent difference in Eh (108 and 117 for ice and snow, respectively) was revealed.

Two digital short-period three-component stations were installed on top of Har-Togoo, one 50 m from the degassing holes and one in a remote area on basement rocks, for monitoring during 19-20 March 2003. Every hour 1-3 microseismic events with magnitude <2 were recorded. All seismic events were virtually identical and resembled A-type volcano-tectonic earthquakes (figure 6). Arrival difference between S and P waves were around 0.06-0.3 seconds for the Har-Togoo station and 0.1-1.5 seconds for the remote station. Assuming that the Har-Togoo station was located in the epicentral zone, the events were located at ~1-3 km depth. Seismic episodes similar to volcanic tremors were also recorded (figure 3).

Figure 3. Examples of an A-type volcano-tectonic earthquake and volcanic tremor episodes recorded at the Har-Togoo station on 19 March 2003. Courtesy of Alexei Ivanov.

Conclusions. The abnormal thermal and seismic activities could be the result of either hydrothermal or volcanic processes. This activity could have started in the fall of 2002 when they were directly observed for the first time, or possibly up to five years earlier when marmots started migrating from the area. Further studies are planned to investigate the cause of the fumarolic and seismic activities.

At the end of a second visit in early July, gas venting had stopped, but seismicity was continuing. In August there will be a workshop on Russian-Mongolian cooperation between Institutions of the Russian and Mongolian Academies of Sciences (held in Ulan-Bator, Mongolia), where the work being done on this volcano will be presented.

References. Devyatkin, E.V. and Smelov, S.B., 1979, Position of basalts in sequence of Cenozoic sediments of Mongolia: Izvestiya USSR Academy of Sciences, geological series, no. 1, p. 16-29. (In Russian).

Logatchev, N.A., Devyatkin, E.V., Malaeva, E.M., and others, 1982, Cenozoic deposits of Taryat basin and Chulutu river valley (Central Hangai): Izvestiya USSR Academy of Sciences, geological series, no. 8, p. 76-86. (In Russian).

Geologic Background. The Miocene Har-Togoo shield volcano, also known as Togoo Tologoy, is situated on top of a vast volcanic plateau. The 5,000-year-old Khorog (Horog) cone in the Taryatu-Chulutu volcanic field is located 135 km SW and the Quaternary Urun-Dush cone in the Khanuy Gol (Hanuy Gol) volcanic field is 95 km ENE. Analysis of seismic activity recorded by a network of seismic stations across Mongolia shows that earthquakes of magnitude 2-3.5 are scattered around the Har-Togoo volcano at a distance of 10-15 km.

Information Contacts: Alexei V. Ivanov, Institute of the Earth Crust SB, Russian Academy of Sciences, Irkutsk, Russia; Bekhtur andM. Ulziibat, Research Center of Astronomy and Geophysics, Mongolian Academy of Sciences, Ulan-Bator, Mongolia; M. Ganzorig, Institute of Informatics MAS, Ulan-Bator, Mongolia.

An eruption at Mount Elgon was mistakenly inferred when fumes escaped from this otherwise quiet volcano. The fumes were eventually traced to dung burning in a lava-tube cave. The cave is home to, or visited by, wildlife ranging from bats to elephants. Mt. Elgon (Ol Doinyo Ilgoon) is a stratovolcano on the SW margin of a 13 x 16 km caldera that straddles the Uganda-Kenya border 140 km NE of the N shore of Lake Victoria. No eruptions are known in the historical record or in the Holocene.

On 7 September 2004 the web site of the Kenyan newspaper The Daily Nation reported that villagers sighted and smelled noxious fumes from a cave on the flank of Mt. Elgon during August 2005. The villagers' concerns were taken quite seriously by both nations, to the extent that evacuation of nearby villages was considered.

The Daily Nation article added that shortly after the villagers' reports, Moses Masibo, Kenya's Western Province geology officer visited the cave, confirmed the villagers observations, and added that the temperature in the cave was 170°C. He recommended that nearby villagers move to safer locations. Masibo and Silas Simiyu of KenGens geothermal department collected ashes from the cave for testing.

Gerald Ernst reported on 19 September 2004 that he spoke with two local geologists involved with the Elgon crisis from the Geology Department of the University of Nairobi (Jiromo campus): Professor Nyambok and Zacharia Kuria (the former is a senior scientist who was unable to go in the field; the latter is a junior scientist who visited the site). According to Ernst their interpretation is that somebody set fire to bat guano in one of the caves. The fire was intense and probably explains the vigorous fuming, high temperatures, and suffocated animals. The event was also accompanied by emissions of gases with an ammonia odor. Ernst noted that this was not surprising considering the high nitrogen content of guano—ammonia is highly toxic and can also explain the animal deaths. The intense fumes initially caused substantial panic in the area.

It was Ernst's understanding that the authorities ordered evacuations while awaiting a report from local scientists, but that people returned before the report reached the authorities. The fire presumably prompted the response of local authorities who then urged the University geologists to analyze the situation. By the time geologists arrived, the fuming had ceased, or nearly so. The residue left by the fire and other observations led them to conclude that nothing remotely related to a volcanic eruption had occurred.

However, the incident emphasized the problem due to lack of a seismic station to monitor tectonic activity related to a local triple junction associated with the rift valley or volcanic seismicity. In response, one seismic station was moved from S Kenya to the area of Mt. Elgon so that local seismicity can be monitored in the future.

Information Contacts: Gerald Ernst, Univ. of Ghent, Krijgslaan 281/S8, B-9000, Belgium; Chris Newhall, USGS, Univ. of Washington, Dept. of Earth & Space Sciences, Box 351310, Seattle, WA 98195-1310, USA; The Daily Nation (URL: http://www.nationmedia.com/dailynation/); Uganda Tourist Board (URL: http://www.visituganda.com/).