Smithsonian Institution
Bulletin of the Global Volcanism Network, v. 20, no. 11/12, November-December 1995
Rabaul (Papua New Guinea)  Small ash-bearing emissions from
     Tavurvur

Rabaul
New Britain Island, Papua New Guinea
4.27 S, 152.20 E; summit elev. 688 m
All times are local (=3D GMT + 10 hours)

Throughout most of November 1995 the two recently active centers
remained quiet, with Tavurvur emitting only steam and Vulcan not
emitting any visible vapor (figure 12). Then on 28 November,
Tavurvur suddenly began erupting, creating a parasitic crater.
Vulcan continued to remain quiet throughout December.

The volume of Tavurvur's faint blue vapor emissions seemed to
increase in the weeks prior to 28 November. On the morning of the
eruption an impressive white steam cloud stood several hundred
meters above Tavurvur's summit. The new eruption, which was
preceded by weak roaring sounds, started at about 1020, and
initially consisted of forceful emissions of gas and dark ash at
2-6 minute intervals. Those emissions lacked explosion sounds; they
rose 400-800 m above the crater rim and blew over a broad arc
between the SE and SW, resulting in fine ashfall both onshore and
over the sea. No ashfall reached Kokopo, 25 km SE. The next day, 29
November, two intervals of stronger emission took place (at
1200-1300 and 1415-1430), sending columns ~1 km above the summit.

An aerial inspection on 30 November revealed a new crater on the
1994-95 crater's SSE rim. Although the 1994-95 crater displayed no
new activity, fumaroles were particularly active along its E walls.
An old explosion crater along the base of Tavurvur's S flank, in
which 6 people were killed in 1990 by inhalation of carbon dioxide,
was releasing weak-to-moderate emissions of white vapor from its N
to E walls. Directly downslope and immediately offshore of this
explosion crater a spring had become considerably more active since
the 1994 eruption; during the 30 November aerial inspection it was
prominent, giving off a strong stream of rusty brown water. During
November and December, ground deformation remained low.

Tavurvur discharged dark ash clouds in December, typically at 3-6
minute intervals, that rose 400-1,000 m above the summit. On 2
December two ash clouds rose to 1.5-2 km. The second brought
intense lightning causing minor damage to home appliances in Rabaul
Town (figure 12). On 5 December, a particularly loud explosion,
heard 30-40 km away, accompanied the discharge of an ash cloud that
rose to 1.2 km. Additional loud explosions accompanied dense ash
clouds that rose to 1-1.2 km; these took place during December as
follows: 11th (1 time), 13th (1), 14th (4), 18th (1), 23rd (1),
24th (1), and 29th (2). Moderate-sized clouds blew SE, and very
fine ash occasionally fell both in Kokopo and, due to shifting
winds, in Rabaul Town. On December nights, observers saw
incandescent fragments and during the second half of the month they
heard occasional deep roaring noises.

Seismicity. November seismicity generally remained low, but was
punctuated by 11 high- and 42 low-frequency events. Eight of the
high-frequency events were located. Five occurred within the
caldera's seismically active elliptical fault zone, in the NE (1
event), W (1), and S (3) quadrants. Although one of the
extra-caldera events was centered S of the caldera, two events were
located immediately to the caldera's NE, an area where the bulk of
the high-frequency earthquakes have occurred in the past few
months. One of these two events, ML 3.0 on 24 November, produced a
felt intensity of MM III at Rabaul Town.

Of the 42 low-frequency earthquakes during November, 17 came from
around Tavurvur volcano. Two of these occurred in late October, and
9 others in November prior to the 28 November eruption. The last
time such events appeared was during the eruptive activity in March
1995. The other 25 low-frequency earthquakes not centered around
Tavurvur were more difficult to locate accurately due to emergent
waveforms and fewer stations outside the caldera. Many may have
originated immediately N of the caldera. On 10 November a
low-frequency earthquake centered 7-8 km outside of the caldera was
strong enough to trigger aftershocks.

During December, seismic instruments detected 30 high-frequency
earthquakes, 684 low-frequency earthquakes, and 488 explosion
events. Instruments also recorded occasional discontinuous
non-harmonic tremors. About 70% of the high frequency earthquakes
occurred during 4-6 December. The five located events had
epicenters in either the S part of the caldera's seismically active
zone (the largest one, M 2.7), NE of the caldera (two events), or
within the caldera. All of the seismic explosions and most
low-frequency earthquakes originated at Tavurvur; the 20 exceptions
originated farther NW and took place at the end of the month.

Fumarole and soil sampling. During 21-27 November, rainwater, water
from hot springs, and gases from subaerial and submarine fumaroles
were sampled at 13 sites (table 2). Compared to Vulcan, fumaroles
at Tavurur displayed relatively high temperature, low pH, and high
conductivity. Hot springs sampled near the shore of Greet Harbor
were slightly acidic and comparatively conductive. All samples were
more acid than those assessed prior to the 1994 eruption episode.

A soil CO2 survey E of Simpson Harbor (figure 12) showed that CO2
concentrations varied widely, 0.4-20% (figure 13). As reported by
Mori and McKee in 1987, the CO2 concentrations peaked along the
seismically active fault zone (near the old airport), some distance
from either Tavurvur or Vulcan. Other anomalously high
concentrations were seen at the Matupit causeway and Sulphur Creek.
Low concentrations were seen at other places, including Matupit
Island.

Isotopic analysis of six selected samples along the profile found
that ^13C ranged from -29.8 to -18.4 per mil suggesting chiefly
biogenic contributions. A mixing process with a minor contribution
of volcanogenic CO2 might also account for the wide range of ratios
seen. High soil CO2 levels could be related to the effects of a
higher thermal gradient along active fractures and faults.

Information Contacts: Same as for Manam; N. M. Perez and H. Wakita;
University of Tokyo, Earth Chemistry, Bunkyo-ku, Tokyo 113 Japan.

Figure 12. Index map of Rabaul and detail of soil CO2 transect.
Elevation contours given in meters; base map after Johnson (1995).

Figure 13. Soil CO2 concentrations at Rabaul Caldera along transect
A-A'. Courtesy of RVO.

Table 2. Summary of fumarole and hot spring sampling at Rabaul
Caldera, 21-27 November 1995. Courtesy of RVO.

Location      Number of     Temp        pH        Electrical=20
             samples/type  (deg C)               conductivity
                                                   (mS/cm)
Tavurvur      3/fumarole    202-98.9   1.21-3.53   0.327-10.4
Vulcan        1/fumarole    99.8       5.28        0.0758
Rabalanakaia  1/fumarole    99.3       3.20        0.444
Vulcan        1/hot spring  99.2       5.98        73.3
Greet Harbor  4/hot spring  62.6-84.8  5.89-6.66   52.9-53.8
  shore
Sulphur Creek 1/hot spring  N.D.       6.21        4.29
Rabaul        1/rainwater   N.D.       6.06        N.D.