This report concerns Garbuna volcano's first historically witnessed eruption. That occurred in mid-October 2005 after a felt earthquake. This report contains a section by members of the Rabaul Volcano Observatory (RVO) and another by Rodger Wilson, a NOAA meteorologist , who made an unofficial visit in November.

Setting. Garbuna is part of the 23 x 15 km Krummel-Garbuna-Welcker complex (a volcanic field with these major topographic highs located in S-to-N progression; figure 1). The field resides at the S end of New Britain island's Willaumez (Talasea) peninsula, a narrow projection jutting well N from the island's W-central region. The peninsula and some local reefs and islands are known for volcanoes and hydrothermal features (including, from N to S, Dakataua caldera and its large lake, Bola stratovolcano, Garua Harbour volcanic field, and the Garbuna complex). In addition to young volcanics found at the complex's three summit (ridge) centers and their associated domes and craters, there have been prior flank and eccentric eruptions, most notably Numundo Maar. The complex's products are mostly high-SiO₂ andesites to high-SiO₂ dacites with more mafic eruptives from Krummel and Numundo Maar. (McKee and others, 2005). Only 5-6 km to the E and W of the volcanic field are some inhabited and intensively cultivated strips along the coast.

Figure 1. Photograph of Garbuna taken on 19 October 2005 from the SSE. View is northward along the Krummel-Garbuna-Welcker ridge, across the general area of Garbuna with Welcker on skyline; Krummel is behind the camera. The two fuming vents can be seen on the periphery of an old lava dome. The bare geothermal nature of the area is apparent. The incised cone to the left is what locals refer to as Mount Garbuna. Photo by Steve Saunders provided courtesy of RVO.

Garbuna and Welcker volcanoes were thought to have had a latest significant/datable eruption at ~ 1,800 BP. Garbuna in particular was very geothermally active, with the central area containing 4 km² of fumaroles, solfataras, hot and bubbling mud and water springs, and patches of hot ground. Conspicuous from the barren, sulfurous and geothermally altered, clay-rich areas was a timber-covered but undated lava dome (or alternately, a short, thick lava flow to the S, a feature sometimes described as a coulée). This dome shows little geothermal alteration, appears very youthful, and is not obviously mantled by the regional tephras from Witori or Dakataua, all features suggesting a comparatively young age. The low cone hosting the dome stands ~ 500-600 m in diameter.

Events surrounding the eruption. The RVO team reported this section and noted that the complex was not instrumentally monitored. A single, locally felt earthquake occurred around midday on 16 October 2005. Jet-like noises were noticed about 2342 that night, rumbling noises started soon after, and at about midnight ash emissions began. The eruption continued and by morning pale to dark gray ash clouds were being driven forcefully into the sky. By 1000 a 3-4-km-high eruption column was visible; the main plume drifted NW with a thin veil of falling material below it. The eruption began to wane between midnight and the morning of 18 October, reducing to slow pale-gray emissions, with only white vapor by the end of the day. A second vent opened quietly during the night of the 18-19 October, with two white vapor plumes visible at dawn on the 19th.

Aerial inspections on 19, 20, and 26 October showed the two vents situated in the central low area of Garbuna, historically an area of high geothermal activity. Both vents are located on or close to the edge of the youthful lava dome mentioned above. The center of the dome is at 05° 26' 48" S, 150° 01' 36" E, with the active vents aligned SW-NE at across a NW sector of the old cone. Both active vents are 60-75 m in diameter and emitted low to moderate amounts of white fume, the southerly plume being more voluminous. On the first two visits fume billowed out gently.

The SW vent seems to have been the source of the initial October emissions. Before the 19th a small incomplete cone had formed around it. Within a kilometer of this vent several ten's of centimeters of ash/mud had been deposited, which thinned very rapidly away from the vent. Old records did not indicate the existence of the SW vent or other conspicuous feature prior to the onset of this eruption.

The second, or NE, vent became active on the night of 18-19 October. Photographs from 1996 showed that prior to the eruption this vent was a small, wooded, funnel-shaped pit, with some evidence of instability on its western side and visible un-vegetated scars.

Although ashfall was reported to the NW, images of the eruption at first light on the 17th showed the fallout to resemble rain rather than dry ash. Vegetation damaged by the fallout had brown blotches rather than uniform discoloration, leading to the conclusion that the initial column was made up primarily of acidic water and mud.

It appears that the NE vent is predominantly a collapse feature, surrounded by a small apron of brownish-gray mud, with a jagged edge and bright red/yellow walls. Following a locally felt earthquake, summit observations on the 20th showed that the NE vent had increased ~ 10-20 m by concentric collapse since the previous day and was 50-60 m deeper. At this time it contained a boiling mud lake ~ 60-70 m below the rim.

Observations on the 26th showed the ash cone around the SW vent had all but disappeared as it had increased in size by collapse. The resulting pit was irregular in shape. Quite vigorous steam emissions occurred and some ash was visibly mixed with the fume and dropping out as fine droplets of dilute mud. Impact craters from projectiles were evident around the vent, especially to the SW. Near the vent these small projectiles were visible and block-like. Up to 500-600 m to the SW small impact craters could still be seen but the projectiles themselves were not apparent.

Close study of the old dome, on whose periphery the vents have opened, suggests that it has undergone little or no movement during the onset of this eruption. Foliage-stripped trees are mainly up-right, and no fresh cracks or heaved boulders are evident. Thermal imagery also showed no hot cracks in the dome, suggesting that the eruption was not preceded by intense surface deformation, and that the vents are now enlarging by concentric collapse.

A large area of grass N of the vents gave the impression of having been flattened in a uniform direction. Trees and shrubs, although stripped of leaves, did not show this flattening. This may suggest that floods of water were responsible for the flattened grass, rather than blast effects. To the S, flooding is also suggested by the apparently recent incising of the valley floor (headwaters of the Garu-Haella sulfur stream). The edge of the jungle also shows undercutting with trees having fallen toward the vents.

Since the start of the eruption changes in the amount of discharge, along with unusual discoloration and dying fish were seen in the streams draining from Garbuna. On 26 October, aerial observers followed one of the Garu thermal streams from the Plantation-Garu village road on its ascent to the summit. At higher elevations the stream's water level dropped markedly, until within a few kilometers of the summit, it dried out completely. The stream was not blocked or dammed, and the falling levels of streams and drying of springs appeared to be related to the drying of the summit, or fracturing, allowing water to percolate into the mountain rather than flow off the geothermally produced clay-rich area. At first light on the morning of 29 October, the watercourse commonly known as the Walindi river was milky white with a blue tinge. It was odorless, of normal temperature, and tasted simply of clayey water with no bitterness. This is the first recorded case of one of the eastern drainage systems exhibiting this behavior, although it is common in the W and SW regions.

A few locally felt earthquakes and sulfur odors were reported from areas not traditionally affected by the complex's sulfatara emissions. On several occasions explosions or booming noises from the Garbuna area were heard at Garu Plantation.

Two seismic stations were installed on 18 October. A 3-component digital recorder was located at Garu Plantation ~ 5.5 km SW of the active vents. An analog recorder was installed at Sisi near Walindi, ~ 5.6 km E of the vents. Notable seismicity was recorded on 20, 21, and 22 October. On the 20th a ML ~ 2.5 local volcano-tectonic earthquake was felt, which was followed by a dozen smaller VT earthquakes between 1300 and 1500. Starting about 0500 on 21 October many very small (ML < 0.5) VT earthquakes began to occur. These events continued throughout 21 October and ceased about 1600 on 22 October. Random VT earthquakes numbering 1-4 per day were recorded on 23, 25, and 28 October. Other volcano-related earthquakes recorded included some small low-frequency earthquakes on 22 October by the Sisi station. Continuous tremor was recorded immediately when a new telemetered seismometer was installed about 0.9 km NE of the active vents. At the end of the month the tremors were continuing.

The West New Britain Provincial Disaster Committee has ensured a smooth civic response to this unforeseen event with public education and preparations for a possible evacuation being well advanced.

Observations during mid-November 2005. Rodger Wilson submitted the following report of his visit to Garbuna with John Seach.

"We climbed Garbuna the first time on 14 November. We smelled H₂S(?) from at least three locations at lower elevations along the trek. Wind at the time was to the NE, so I don't believe we were sensing the summit gas plume. Also, there was an area that I jokingly referred to as, 'The Valley of Death,' where we all (four) felt nauseated (on both climbs at the same location, both coming and going), but did not detect any odor of gas. On the second climb, we found an immature parrot on the ground at the (SE?) edge of this area (we removed him from that area, and he seemed fine afterward). Again, we were unable to get a GPS fix there, but it occurs along a (NW to SE-running?) depression just prior to a steeper climb to the summit.

"Just before reaching the clearing at the summit, along a more N-S-running depression, we encountered an area where the trees appeared to have been 'sprayed' horizontally as evidenced by ash being 'plastered' to their N (crater) sides. Bark on many of the larger trees appeared to be at least lightly abraded, but not removed. Numerous smaller trees of approximately 6 cm or less in diameter had been neatly 'clipped' or sharply bent over at just less than 2 m height. There was no evidence of high temperature in connection with the physical damage. Our visit was restricted to along the S edge of the summit, bounded by the hydrothermal area on the W and the two old phreatic craters to the E. This area of damaged trees was, as far as we could see, the only significant damage to the surrounding forest (by a base surge or a cold density current?) in contrast to the more complete devastation suffered by the fewer trees and lower vegetation at the summit. Interestingly, the trees still standing at the summit, appeared to be stripped solely by vertically falling, not horizontally moving, debris.

"We exited the forest at the summit at about 1100, along a N-S bare ridge (old crater rim?), that is clearly visible in aerial photos of the area. Copious fume emanated soundlessly from the two new craters. White fume exited the western crater, with yellow-tinged fume rising from the eastern one. There was a fairly strong smell of H₂S, but not the eye-stinging or choking sensation I've felt with SO₂ at Etna and Stromboli. As we rested there, we noted the water in our bottles was in constant motion and once we made our way to the thermal area, we clearly felt frequent (several per minute) small shocks while we took temperatures at several spots there. The highest temperatures were all at 100°C. During the first couple of hours at the summit, we had two brief bands of rain showers pass overhead, but by about 1330 the rain became sustained and heavy. Run-off in most of the surrounding gullies had increased to several inches deep. We . . . were picking our way back toward the forest when, just as we left the southern edge of the thermal area, we heard a loud roar and witnessed a lahar issue from the gully draining the crater . . ..

"Changes we noted during the second visit 3 days later were [as follows]: a low rumble or rushing noise associated with the summit vents which was heard through most of the journey to the summit, although [they observed] a complete lack of detectable seismicity while at the summit. The interval between "huffs" of fume was shorter, on the order of maybe 4-5 minutes rather than the 8-10 minutes observed during the first visit. Fume leaving the western vent remained white, while ash was clearly visible as it fell over the E flank of the volcano from the eastern plume. That plume also had a more yellow cast as it issued from its source, compared with a few days before.

"We heard low booming rumbles from near Walindi at around 0530 on 17 November and loud roaring the next morning at about 0700 from the same location. The latter was preceded (by as much as 10 minutes) by dogs at our location being agitated and barking, simultaneously with others in the distance.

Reference. McKee, C.O., Patia, H., Kuduon, J., and Torrence, R., 2005, Volcanic Hazard Assessment of the Krummel-Garbuna-Welcker Volcanic Complex, Southern Willaumez Peninsula, WNB, Papua New Guinea: Geological Survey of Papua New Guinea—Report 2005/4.

Geological Summary. The basaltic-to-dacitic Krummel-Garbuna-Welcker Volcanic Complex consists of three volcanic peaks located along a 7-km N-S line above a shield-like foundation at the southern end of the Willaumez Peninsula. The central and lower peaks of the centrally located Garbuna contain a large vegetation-free area that is probably the most extensive thermal field in Papua New Guinea. A prominent lava dome and blocky lava flow in the center of thermal area have resisted destruction by thermal activity, and may be of Holocene age. Krummel volcano at the south end of the group contains a summit crater, breached to the NW. The highest peak of the group is Welcker volcano, which has fed blocky lava flows that extend to the eastern coast of the peninsula. The last major eruption from both it and Garbuna volcanoes took place about 1800 years ago. The first historical eruption took place at Garbuna in October 2005.

Information Contacts: Steve Saunders, Ima Itikarai, and Herman Patia, Rabaul Volcano Observatory (RVO), P.O. Box 386, Rabaul, Papua New Guinea; Rodger Wilson, Meteorological Technician, US National Oceanic and Atmospheric Administration (NOAA) and National Weather Service (NWS), WSO, P.O. Box 1685, Valdez, AK 99686, USA.