Previous Next

Fire Mountains of the Islands

12. Eruption Alert at Rabaul Caldera: 1971–1994

In a statement issued in Papua New Guinea on Monday [23 January 1984], the principal volcanologist, Dr P. Lowenstein, said that ‘evidence is accumulating to suggest that the volcano has embarked on an irreversible course towards the next eruption and that it is only a matter of time before this occurs … the eruption that was previously only a possibility is now much more likely to occur within the next few months’.

Peter Hastings (1984)

Crisis Build-up and Stage-2 Alert

Villagers living near the south-eastern end of Matupit Island, Rabaul, were by 1970–1971 aware that nearby coastal cliffs had encroached perilously close towards their homes as a result of sea-wave erosion. Their concerns were alleviated after 1971, however, when a new beach began to form at the foot of the pumice cliffs which gradually became stranded inland. The south-eastern end of the island was rising episodically, following the two, major, Solomon Sea earthquakes that had shaken Rabaul in July 1971. Rabaul Volcanological Observatory (RVO) staff led by Rob Cooke began measuring the amount of ground uplift in 1973 using a survey line that ran southwards from Rabaul town to the end of the island. About 60 centimetres of uplift had been detected by 1979, the year of Cooke’s death, and to more than a metre by 1983.1

Matupit is a low-lying island that in 1971 was already well known for its vertical oscillations, most noticeably at times of major earthquakes or tsunamis when a causeway linking the island with the shore might disappear and then reform.2 The island is made up of flat-lying pumice beds, but there is no persuasive geological evidence that it is, or ever was, an eruptive centre. The island rather has risen out of the waters of Rabaul Harbour, perhaps very recently in geological terms, as early European observers of Matupit Island had commented on the island’s youthful vegetation. Furthermore, emergence and occupation by people of the island only a few centuries previously at most, is supported by genealogical evidence.3 The growth of Matupit’s size may indeed have encouraged increases in population. Matupit, by the 1970s–1980s, supported more than 2,000 people who, in common with other Tolai communities — and Melanesian communities in general — had a strong link with their land. They were, however, volcanically vulnerable, being only about two kilometres from the western slopes of Tavurvur volcano across the entrance of Greet Harbour. Collection of eggs from the buried, volcanically heated, nests of megapode birds near Tavurvur was a source of some cash income for the Matupit people.

Uplift of the south-eastern end of Matupit Island in 1971–1983 was accompanied by increased numbers of local earthquakes, some of them felt in Rabaul town, and many taking place in ‘swarms’ that became more pronounced as this 12-year period progressed. Cooke was the first to illustrate the unusual, possibly unique, pattern that these earthquakes formed where plotted as a map of epicentres.4 The pattern he concluded had a broad D-shape, rather like a doughnut flattened straight along its north-western side, and enclosing a mainly submarine area in the harbour that was more-or-less earthquake-free. The straight segment covered the south-eastern part of Matupit and trended directly — and rather menacingly — towards Rabalanakaia volcano. Cooke thought that this linear segment was a possible fault zone, and he identified the complementary curved segment of the ‘D’ as part of the seismically active margin of Rabaul Caldera. The D-pattern, however, was based on relatively few earthquake data obtained from the inferior number and distribution of recording stations that existed in the early 1970s.

The uplift and seismic swarms began to raise concerns about future volcanic activity at Rabaul, and geophysical monitoring was enhanced by RVO. Cooke, a specialist in gravity measurement, introduced surveys of the gravity field to complement the surveying lines, and new recording stations were established to improve monitoring of the earthquake activity. The likelihood of a volcanic eruption similar to those in 1878 and 1937 was discussed, but larger-scale eruptions were possible too. This point was highlighted by timely publication in 1974 of a study of the volcanic geology of Rabaul, which demonstrated for the first time that the volcano was built up in part by major ignimbrites formed by huge, caldera-forming, explosive eruptions, the most recent one only about 1,400 years ago.5 Such a major Krakatau-type eruption taking place in, say, 1980 — the year of a National Census — would impact severely on most of the 100,000 people living in the north-eastern part of the Gazelle Peninsula, and particularly the 15,000 people living in Rabaul town itself.


Figure 97. Villages are not shown in this simplified map of the main features of Blanche Bay and Rabaul town in the 1970s.

Source: Adapted from Johnson & Threlfall (1985, figure on p. 4).

Peter L. Lowenstein, a geochemist and Englishman, took over leadership of RVO after Cooke’s death in 1979. Lowenstein, like his predecessors in this position, had had no training in applied volcanology, but volcanologists Chris McKee and Ben Talai by this time had gained practical experience in monitoring many eruptions in Papua New Guinea. Lowenstein came to Rabaul with natural organisational and management strengths that would prove invaluable in the years ahead as a major volcanic crisis developed in the Rabaul area. He could also use his impeccable English accent and sense of humour to great effect in clearly articulating his views, especially opinions that he held strongly. A new volcanological team was established under Lowenstein’s leadership after French volcanologist Patrice de Saint Ours, Papua New Guinean seismologist Ima Itikarai, American–Japanese seismologist Jim Mori, and surveyor Malcolm Archibald, an Australian, joined the staff at RVO.

A dramatic and exponential increase in volcano unrest at Rabaul Caldera began in late August 1983 and became particularly noticeable on 19 September 1983 when an intense earthquake swarm was felt in Rabaul, accompanied by a sharp increase in the uplift rate at Matupit Island. These events marked the start of what came to be referred as the ‘seismo-deformational crisis’ at Rabaul, and which would demand virtually the sole attention of RVO staff.6 A disaster plan had been prepared for East New Britain Province a few months earlier by a United Nations disaster-management specialist, in consultation with national and provincial government authorities, including RVO, and together with the private sector. The plan assumed eruptions only of the scale of those in 1878 and 1937, and it included a scheme of four stages of volcanic alert — Stage 1 where a volcanic eruption was expected within years to months, up to Stage 4 where one was expected within just days to hours.7

Table 5. Stages of Volcanic Alert at Rabaul

Possible eruptive activity

Summary of meaning


Within years to months

Risk exists. No immediate cause for alarm.


Months to weeks

Increased risk. Still no public action necessary.


Weeks to days

Risk is serious. Precautionary actions required.


Days to hours

Situation is critical. Event imminent. Public Red Alert.

The East New Britain Provincial Disaster Committee (PDC), which was chaired by the Secretary of the Department of East New Britain, Nason Paulius, held its inaugural meeting on 13 April 1983, and government authorities practised the first of several evacuation exercises in late May and early June. A range of other public awareness-raising activities were initiated, including — in the background in Canberra — preparation of my own book, with Rev. Neville Threlfall, on the 1937–1943 eruptions at Rabaul. This publication was sponsored by the Insurance Underwriters’ Association of Papua New Guinea, and was aimed at illustrating how the previous eruptions had affected the Rabaul area.8


Figure 98. Principal features of the north-eastern Blanche Bay area are seen clearly in this computer-enhanced aerial photograph mosaic compiled in the early 1980s, before the uplift of the south-eastern end of Matupit Island became strongly noticeable.

Source: Adapted from Johnson & Threlfall (1985, dust cover).

A Stage Two volcanic alert indicative of a possible volcanic eruption within only weeks to months was declared on 29 October 1983 following further intense earthquake swarms. The crisis period escalated up to April 1984, and by 15 May major seismic swarms had taken place since the previous September.9 The number of recorded earthquakes in April reached a monthly maximum of more than 13,000, and the southern end of Matupit Island had risen a total of about 1.6 metres above its 1973 height. These events caused considerable apprehension amongst the communities of Rabaul town and the surrounding region. There was a partial and voluntary evacuation of the town itself, including businesses.10 Many villagers moved to land outside the caldera area, including perhaps as many as 40 per cent of those people living in the highest risk areas south of the main business district. Furthermore, blocks of government-owned land south of the Warangoi River were made available for settlement, including at Sikut. Other people left the province altogether — for West New Britain, for example — and some expatriates departed for Australia. An inevitable topic of debate in the community and media, as well as at different levels of government, concerned the suitability of Rabaul as a place for a town and provincial capital — a repetition of the debates that had taken place after the 1937 eruption and after the Second World War. Prime Minister Michael Somare entered the debate by announcing in mid-February 1984 that, should an eruption take place, Kokopo would be developed as a new administrative centre and that Rabaul had insufficient land for expansion anyway.11

Disaster-preparedness activities intensified in early 1984. Maps of danger areas in Rabaul were posted on public notice boards together with information about points of assembly and refuge. Roads were cleared of overhanging tree branches, and old airstrips and wharfs outside the caldera were upgraded. Drafting of legislation led to the passing in March of eight Acts of Parliament to cater for disaster preparations, including the Natural Disaster Act, which resulted in the formal establishment of both the National Disaster Centre and Provincial Disaster Committees. This national legislation was indicative of the prominence of the Rabaul area in both the national economy and psyche of Papua New Guinea. RVO and the PDC established a close working relationship, and scores of situation reports and information bulletins on the condition of the volcano were provided by RVO to the PDC as a basis for official decision-making.12 Evacuation rehearsals continued at Rabaul and the crisis began to receive international attention.


Figure 99. The number of harbour earthquakes recorded monthly at Rabaul increased dramatically in 1983–1985.

Source: Mori et al. (1989, Figure 2). Reproduced with the permission of Springer-Verlag.

There was high expectation by early 1984 that an eruption would take place — that, indeed, the volcano had embarked on an ‘irreversible course’ towards a likely eruption.13 Yet the crisis for RVO in practice represented the challenge of management of scientific uncertainty, of there being no scientifically based and precise way of presenting accurate predictions or even definitive forecasts about the expected eruption, much as the public and media would have liked or expected it. Information about the likelihood and scale of the expected eruption was at first restricted, but people in the town and in both the national and international media, increased the demand for the latest information, particularly situation statements from RVO directly. RVO staff, however, were already overworked trying to keep up-to-date with the increased schedules of instrumental monitoring, data collection and interpretation. Reports had to be written for government authorities, particularly the PDC, which was responsible for official information releases rather than RVO itself. Having to deal directly with the visits of many individual media personnel and anxious citizens concerned about Rabaul volcano, was a time-consuming burden.

A group of three, including myself, from the Bureau of Mineral Resources (BMR) in Canberra, and funded by the Australian International Development Assistance Bureau (AIDAB), was sent to Rabaul in January 1984 to assist RVO. I made visits to Manam and Langila volcanoes, and to the Esa’ala area in Milne Bay Province, and New Zealander volcanologist Brad Scott visited Manam, to investigate reports of increased activity that could not be checked by RVO staff because of work commitments in Rabaul.14 My visit to Esa’ala in February 1984 was in response to local reports of increased activity in the thermal areas bordering Dawson Strait area, including coral-reef die-off at Dobu Island, and therefore of eruption fears which were reported in national newspapers. No evidence was found for any impending eruptions, but volcano fears in Milne Bay had clearly been heightened by the prominent reporting of the happenings in Rabaul.

Some news media began to spread misinformation, or sensationalised reports, as well as criticisms of the government and agencies — including RVO — about the management of the crisis. One widely distributed article was written by a journalist for the influential Australian weekly magazine the Bulletin, in which the authorities grappling with the Rabaul situation were portrayed as secretive, indecisive, bungling and stressed, and involved in political infighting.15 The front page of the magazine blazed the words ‘Rabaul’s killer volcano’ and showed a weather cloud above old Kabiu volcano, backlit by the dawn light, which gave the false impression of an actual volcanic eruption. The situation was made even more difficult in RVO by the natural reluctance or anxiety of some scientists — as seen in scientific agencies in other countries — of being misinterpreted or over-interpreted by aggressive, story-seeking media, or publically stepping beyond their areas of expertise and authority, or wanting to avoid perceptions amongst scientific peers of ‘grandstanding’. Science communication today is still a challenge for many scientists in other parts of the world who are employed in observatories that have mandates for early warnings of hazard events.

The public relations situation had become unmanageable, and it was not mitigated until the PDC established a Public Information Unit, and until an Australian geologist, Dr Hugh L. Davies, who had had many years of experience working in Papua New Guinea, was appointed for some months as a volcanological liaison officer. An information newsletter funded by the private sector and called ‘Rabaul Gourier’ — a combination of Post-Courier, a national newspaper, and guria meaning earthquake — was only one of several initiatives developed to keep the citizenry and media informed.

Scientific Responses to the Caldera Unrest

Instrumental monitoring of Rabaul Caldera by RVO had increased dramatically by 1984. Additional funds for the purchase of new equipment were made available by the national government and by the development-assistance agencies of foreign countries, particularly Japan, Australia, New Zealand and the United States. The number of earthquake recorders increased and several, different, ground-survey methods for measuring the type and extent of ground deformation — uplift, tilt and horizontal extension —were deployed.16 These included electronic-distance measurements using laser beams shot to numerous reflectors positioned around the shores of Rabaul Harbour. Tide gauges were used to measure seawater depths, which decreased as parts of the coastline rose. Never before, or since, has such an impressive array of geophysical monitoring instruments been deployed in Rabaul Harbour. Even the heights of emerged and stranded marine barnacles were mapped in an attempt to determine the extent of any longer term, harbour-wide uplift.17

The large amount of volcano-monitoring data collected by RVO before and during the seismo-deformational crisis was used not only for attempts at eruption forecasting, but also for interpretations of what the internal structure of Rabaul Caldera might look like. The major scientific analysis to emerge from this work was later published by Jim Mori and McKee, in association with several co-workers, in numerous peer reviewed papers in the geoscientific literature.18 Their important interpretation was based on the following five observations:

  1. 1. Most earthquake epicentres plotted on a map defined a ‘seismic annulus’, an elongate doughnut shape about ten kilometres long from north-south, rather than the D-shape noted by Cooke. This annulus was thought to mark the ring-like boundary of a block of coherent rock that was presumed to have subsided during the latest caldera-forming event, about 1,400 years ago.
  2. 2. The earthquakes were mostly less than about four kilometres deep, and they defined the top of a postulated, large, underlying magma reservoir.
  3. 3. The sides of the subsided block seemed to dip steeply outwards as if the block were an inverted keystone.
  4. 4. Much of the detected uplift, tilt and horizontal extension could be explained by a smaller magma body that had intruded upwards from the larger magma reservoir into the block to within about two kilometres of the surface. This smaller magma body or intrusion was causing the bulging of the sea floor centred south-east of Matupit Island.
  5. 5. Some of the ground-deformation data hinted at the existence of a second, shallow source of upward intrusion beneath the sea floor east of Vulcan and located over the seismic annulus.

The conclusion that a ‘bulge’ existed close to Matupit Island led to some speculation that the doming there represented the site of possible future submarine eruptions which might build a new volcano above the sea, rather than the anticipated eruptions taking place at either Tavurvur or Vulcan, as in 1878 and 1937. The final hypothesis favoured by the RVO volcanologists, however, involved lateral, underground migration of magma from the proposed shallow magma body to beneath both of these well established volcanoes. This interpretation has some characteristics similar to those for Glen Coe-type calderas, particularly existence of a coherent, near-cylindrical block rather than the disintegrated and highly fractured rocks envisaged for Krakataua-type calderas. Speculation also centred on an especially threatening aspect of the caldera block. Could it subside again, perhaps by a reduction of magmatic pressure, and the ‘inverted keystone’ drop catastrophically into the large, underlying magma reservoir, generating another major eruption the size of the one 1,400 years ago?

Attention was paid not only to the monitoring data but to improving fundamental knowledge about the structure and evolution of Rabaul Caldera and how its volcanoes behave, both now and in the past. A marine geophysical survey of the Rabaul Harbour floor was undertaken by the United States Geological Survey (USGS) in 1982, including the area of the ‘bulge’ south-east of Matupit Island. Up-domed and folded sediments, slumps, and steep faults were discovered on seismic-reflection profiles, adding support to the RVO theory.19 Questions that arose, however, concerned both the age and full extent of these structures and the age of the bulge itself. Did they all start forming in 1971 or were they much older in origin, including the bulge, which might represent only the latest episode of the long, oscillating, uplift history of Matupit Island? Another question was why there was no cluster of earthquake epicentres marking the area of the bulging and fracturing? Such a cluster might not be expected if the rocks above the inferred magma body were soft and weak, but in this case upward streaming of heat, and even magma, might have been expected to reach the surface more easily, particularly if the bulge was a long-lived feature. One highly respected volcanologist known for his bold thinking even suggested to me in the early 1980s that small volcanic eruptions may have taken place already on the deeper parts of the sea floor.


Figure 100. Two different patterns for the ‘seismic annulus’ were obtained by mapping the epicentres of earthquakes in the Blanche Bay area. The D-shape on the left was produced by Cooke (1977, adapted from Figure 2) and encompasses the epicentres of earthquakes recorded in the early 1970s. The pattern on the right is for the epicentres of more than a thousand earthquakes recorded from 1971 to 1983 (Mori et al. 1989, Figure 3).

Source: Both figures were redrawn from the originals and presented together by Johnson et al. (2010, Figure 22).

I arranged for a Remotely Operated Vehicle equipped with a video camera and temperature sensor to undertake dives at 11 different sites over the bulge in 1985, but we found no evidence of elevated water temperatures, hot springs, upward streaming of gas bubbles, or submarine craters or eruptions.20 This inspection of the sea floor, however, in turn raised the need for systematic surveys of the discharge or flow of heat from the sea floor of Rabaul Harbour, as a means of understanding how volcanic heat is distributed in relation to the submarine centres of uplift and to the much deeper, larger, magma body. A series of AIDAB-funded heat-flow surveys — the main one in 1992 — provided no evidence that the main ‘bulge’ south-east of Matupit Island coincided with an area of elevated heat flow.21


Figure 101. The four-kilometre-deep magma reservoir beneath Rabaul is shown in this cartoon of a vertical ‘slice’ through Rabaul Caldera, as envisaged by RVO volcanologists. The magma reservoir presses up on the inverted-keystone block, helping to keep it in place. Magma from the smaller, shallower body moves obliquely through the block and its boundary — as shown by the lines of question marks — in order to feed Tavurvur and Vulcan volcanoes on the surface.

Source: Mori et al. (1989, adapted from Figure 22 by Johnson et al, 2010, Figure 26).

Major advances in understanding the land geology of Rabaul Caldera were also made as a result of geological mapping in 1979 and 1984.22 One important conclusion of these studies was that as many of five, but possibly nine, major ignimbrite eruptions of Krakatau-type may have taken place at Rabaul during the previous 20,000 years or so, meaning that such catastrophic eruptions take place once every 2,000 to 3,300 years, more or less. The latest of these major eruptions was the one about 1,400 years ago, corresponding to an eighth century AD age. It deposited the so-called Rabaul Pyroclastics, the pumiceous pyroclastic flows of which were particularly energetic. These extended out from Rabaul to at least 50 kilometres, flowing over the sea and reaching Watom Island. There were erroneous speculations that the eruption was responsible for major atmospheric effects in the northern hemisphere and for dateable acidity layers in polar ice cores.23 The Rabaul Pyroclastics are of dacite composition, and evidently are a sample of the large magma reservoir beneath Rabaul Harbour. Furthermore, the large number of ignimbrites at Rabaul can also be taken as evidence that the caldera, like the one at Long Island, probably originated by a series of collapses, and therefore that a high ‘ancestral mountain’ may never have existed, at least above sea level.

Worldwide Volcanic Crises and Developments in Risk Awareness

The Rabaul crisis achieved considerable prominence internationally, even at this time — in the early to mid-1980s — when disastrous eruptions that might otherwise have diverted attention were taking place at other volcanoes worldwide, and when there were threatening signs of unrest at calderas in other countries. This was a period when important contributions by volcanologists in the United States achieved worldwide prominence and impact, triggered particularly by the highly publicised volcanic eruption at Mount St Helens, Washington State, on 18 May 1980. Government volcanologists in the USGS had accumulated many years of experience monitoring the relatively ‘passive’ volcanoes of Hawaii, but the explosive activity at Mount St Helens — a subduction-zone type volcano on the continental mainland — was a new experience for them. Managing the 1980 disaster raised similar challenges to those facing RVO staff at Rabaul during the Stage Two period. These included handling the critical but sensitive relationship between volcanologists, who were dealing with scientific uncertainty and ambiguity in their monitoring of volcanic unrest, and the authorities, public and media, who had their own need for clear and decisive information.24 There were also conflicts with some US academic volcanologists who wanted to study the unrest and eruption themselves and who believed that any restrictions imposed of them were a violation of their right to unfettered scientific freedom.

A further issue for USGS volcanologists was being able to continue gaining experience in monitoring explosive-type volcanoes in the United States during those long periods when those volcanoes are dormant. The USGS therefore created the Volcanic Crisis Assistance Team (VCAT), which was to be partly funded by the US Agency for International Development (USAID), and would provide a rapid-response capability for those foreign countries who invited its assistance during a volcanic crisis.25 The response included donation of volcano-monitoring equipment to the country concerned, as well as temporary secondment of USGS staff for in-country training, help in instrument installation, and advice on volcanic hazards and eruption forecasting, if requested. USGS volcanologist Norman G. Banks — founder of VCAT — came to Rabaul in 1984 accompanied by technicians as part of this USAID-funded arrangement. They helped install the laser-beam, electronic-distance measurement equipment for use by RVO, and provided training.

The 1980 eruption at Mount St Helens was of moderate size and was assigned a VEI of five. VEI stands for ‘Volcanic Explosivity Index’, a concept proposed by two US-based volcanologists in 1982.26 VEI is a semi-quantitative measure of eruption ‘size’ by which a score, or index, is assigned largely on the basis of the volume of magma erupted by a particular explosive volcanic eruption. Non-explosive eruptions, even if large in volume, are arbitrarily assigned a VEI of zero. Use of the VEI system increased greatly during the 1980s, especially when scientists of the Smithsonian Institution, Washington D.C., adopted it in a global database of volcanic eruptions that they had been developing.27 The Mount St Helens eruption of 1980 has a higher VEI than does, for example, the Mount Lamington activity of 1951 — as shown in the following short list of selected historical eruptions.28 Its VEI, however, is less than those for the caldera-forming eruptions at Rabaul in the eighth century and at Long Island in the seventeenth century. Furthermore, all of these examples have smaller VEI than does the great eruption at Tambora in 1815. It, impressively, has a VEI of seven, the only historical eruption rated so highly.

Table 6. VEI Values for Major Eruptions

Tambora, 1812–1815, Indonesia


Rabaul, eighth century, Papau New Guinea


Long Island, seventeenth century, Papau New Guinea


Krakatau, 1883, Indonesia


Tarawera, 1886, New Zealand


Mount St Helens, 1980–1985, United States


El Chichón, 1982, Mexico


Pelée, 1902–1905, Caribbean


Rabaul, 1937, Papau New Guinea


Lamington, 1951–1956, Papau New Guinea


Taal, 1965, Philippines


Galunggung, 1982–1983, Indonesia


Rabaul, 1878, Papau New Guinea


Manam, 1956–1958, Papau New Guinea


Nevado del Ruiz, 1985–1991, Colombia


The above list includes two other eruptions of relevance from 1980–1985. The first of these was at Galunggung, Indonesia, in 1982 when Boeing 747 aircraft — belonging to two different international airlines — flew, at separate times, into drifting ash from the volcano, causing all four engines to stall on each aircraft.29 These ash/aircraft encounters mark the start of considerable international effort to mitigate the threat of high-rising ash clouds to both international and domestic aviation. Many such ash clouds could be seen on images taken from satellites, but not where regions were covered by extensive weather clouds, such as those in monsoonal tropical areas like Indonesia and Papua New Guinea. There had been, however, considerable advances by the mid-1980s in the ways that eruption clouds could be observed from polar-orbiting and geostationary satellites. These included development of multispectral scanners that allowed for the satellite detection and tracking of the sulphur-dioxide gas in volcanic clouds. Furthermore, CSIRO scientists in Australia in the early 1980s began developing a multispectral-scanner technique for discriminating volcanic-ash clouds — initially those from Galunggung in 1982 — from clouds of meteorological origin, by using the contrasting infra-red characteristics of each type of cloud.30 Volcanic aerosols injected into the stratosphere could also be identified and tracked using ‘lidar’— meaning Light Detection and Ranging — instruments on board satellites and at ground-based stations. Aerosols from the 1980 eruption at Ulawun, for example, were detected in this way.31

The second important eruption in the above list was that at Nevado del Ruiz, Colombia, in 1985–1991 when a small VEI 3 eruption caused melting of the snow and ice cap on the summit of the Andean volcano and produced lahars. Highly destructive lahars swept down river valleys late at night on 13 November 1985, killing more than 22,000 people, and resulting in the second largest volcanic disaster of the twentieth century after Mont Pelée in 1902, in terms of numbers of human fatalities.32 A major disaster-management question to emerge from the aftermath of the Ruiz disaster was why people had not evacuated before the lahars overwhelmed them. A volcanic-hazard map for Ruiz had been made available to local authorities a month earlier, and an early warning had been issued by scientists when the summit eruption began. The USGS Volcanic Disaster Assistance Program (VDAP), the successor to VCAT, came into formal existence in 1986 prompted by the 1985 Ruiz eruption and its disastrous outcome. VDAP continues to operate internationally, including in Papua New Guinea.

Pioneering techniques for mapping volcanic hazards had been developed by USGS volcanologists at Mount St Helens long before the 1980 eruption33 and applied elsewhere, but the importance of such maps seems to have been underestimated by the Colombian authorities at Ruiz volcano. Volcanic-hazard maps are neither ‘geological’ nor ‘evacuation’ maps. Rather, they portray those areas most likely to be affected by particular kinds of volcanic processes — pyroclastic flows, lahars, ash falls, and so on — and, ideally, they are developed also for eruptions having different VEI. Geological maps of volcanoes, in contrast, show areas where volcanic rocks of different types and ages are exposed. Evacuation maps simply provide designation of escape routes, such as roads, tracks, airstrips and wharfs for shipping, on or near active volcanoes, without explicit — though commonly inferred — reference to volcanic hazards.

Alarming earthquake activity, ground deformation, and new gas emissions in one case, were also being reported in the early 1980s from the large, active calderas at Long Valley, eastern California, United States, and at Phlegrean Fields, west of Naples in Italy. The concurrent restlessness of these two calderas, together with Rabaul Caldera, in three different parts of the world, was coincidental, but were compared with considerable interest by volcanologists internationally. The three calderas also featured prominently in a major scientific literature review in the 1980s that was undertaken by two USGS volcanologists. These authors summarised almost 1,300 episodes of historical ‘unrest’ at 138 large calderas worldwide. Their conclusions included the following:

The remarkable unrest at Rabaul Caldera that began in 1971 is perhaps the most threatening example in this compilation.34

A fundamentally important point that also emerged internationally in the 1980s was the distinction that must be made in volcanic-disaster management between ‘risk’ and ‘hazard’.35 Volcanic hazard refers simply to the size and frequency of a volcanic event and carries no implications about its impact on people and their settlements. Risk, on the other hand, is a more complex concept because it involves definition of what is being threatened by a particular volcanic hazard — people, schools, businesses, critical infrastructure, lifelines, and agricultural land. All of these different ‘at-risk’ categories, or ‘elements’, may in total be called ‘exposure’ or ‘value’, and defining risk depends on knowing and ideally measuring the vulnerability of each one of them. Thus, the small volcanic eruption at Ruiz in 1985 had high risk because of the large number of people and villages who would be, and were, affected by it, rather than because of the size of the eruption. Conversely, a large volcanic eruption in an unpopulated area may have low risk, but might have higher risk to aircraft flying overhead. Risk can be summarised by means of the following disarmingly simple formula:

Risk = The hazard (their magnitudes and frequencies) x Elements exposed to risk x Their separate vulnerabilities

An obvious corollary of this formula is that large populations of vulnerable people, together with agricultural lands and built environments that are also at-risk, have the potential to be affected by the greatest disasters. Reducing such disasters became a worldwide theme for the United Nations in the late 1980s, and the 1990s were declared an International Decade for Natural Disaster Reduction (IDNDR), a worthy venture that was successful particularly in those countries where funding could be identified for significant, national, IDNDR activities. The late 1980s and 1990s were, for me, a time of career transition when I greatly reduced my scientific interests in volcanic petrology, geochemistry and the tectonic settling of volcanoes, and became more involved in work on disaster-risk reduction in both Papua New Guinea and Australia.

‘Natural disaster’, even in the IDNDR title, carries an implication that disasters are caused in the first place by natural hazards, just as technological disasters are ‘caused’ by hazardous events such as industrial fires, oil-depot explosions, or chemical leaks. The affected communities are portrayed as unfortunate victims unable to defend themselves adequately against ferocious and uncontrollable external forces. ‘Acts of God’, the phraseology still enshrined in legalese and insurance policy, is a similar quasi-fatalistic expression. Both terms belie the fact that disasters follow only where the communities are ill-prepared for the impact of, say, a volcanic eruption, have not organised themselves to be appropriately resilient, and are vulnerable because of basic sociological characteristics such as weak economies, poverty, ineffective political leadership, and poor governance. The hazards themselves are not to blame, being without harmful intentions towards societies.

This international shift towards comprehensive risk and hazard assessments was reflected in an array of disaster-management studies in Rabaul. RVO volcanologists and others produced the first volcanic hazard maps for Rabaul,36 and Lowenstein and Talai determined a risk rating for the numbers of at-risk people on or near the young volcanoes of Papua New Guinea, noting that the top five volcanoes — in order of decreasing risk — were Rabaul, Lamington, Manam, Karkar, and Ulawun.37 Two geoscientists from the New Zealand Department of Scientific and Industrial Research (DSIR), undertook an assessment of volcanic, earthquake and tsunami hazards at Rabaul and included the first consideration of the numbers of people at-risk from the different volcanic hazards in the Rabaul area.38 Furthermore, geographer Russell Blong and economist Colin Aislabie collaborated in assessing what would be lost in economic terms by considering the impacts of three different kinds of eruption on the north-eastern Gazelle Peninsula.39 Finally, an Australian geographer, Ken Granger, demonstrated how digital geo-referenced information — where loaded onto computer-based geographic information systems (GIS) — could be used for practical crisis-management purposes in the Rabaul area — and elsewhere.40 GIS was evolving at this time as a valuable tool for risk-assessment purposes.


Figure 102. The north-eastern Gazelle Peninsula would be affected by airfall-ash deposits thicker than 20 centimetres and by far-reaching pyroclastic flows in the case of a large eruption of about VEI 6, as shown in this simple volcanic hazard map for the area.

Source: Adapted from McKee et al. (1985, Figure 13).

Costs, Benefits and Crisis Decline: 1985–1994

Rabaul in April 1984 was prepared for an imminent volcanic eruption. Then, unexpectedly, the number of earthquake swarms and their intensities began to decline. Next, in November 1984, the Stage Two alert was called off, reverting to Stage One, and the seismicity by mid-1985 returned to its pre-1983 levels. Perceptions grew amongst parts of the Rabaul community that the crisis had been a false alarm or ‘failed eruption’ and that RVO had not delivered an accurate prognosis. There were concerns at RVO itself that community complacency might set in, leading to a more blasé approach towards any further crisis periods.41 The two USGS volcanologists who had undertaken their review of geophysical unrest at active calderas worldwide, noted that residents and scientists in such circumstances are faced with a difficult problem:

Unrest [at calderas] is abundantly clear but hard to interpret and more likely than not to subside rather than lead directly to an eruption. On the other hand, the potential always exists for a truly devastating eruption to occur with little additional warning. Worrisome changes that do not lead to an eruption are likely, and the public must be willing to accept false alarms if they wish to receive timely warning of an eruption.42

The total cost of the emergency preparations during the Rabaul crisis and incurred by the national government and private enterprise, may well have exceeded 20 million PNG kina.43 There were substantial losses of business revenue. Insurance premiums during and after the crisis increased greatly for volcano, earthquake and tsunami cover, or cover was not offered at all. This meant that banks would not approve loans, and so a development stand-off gripped the high-risk town. School closures had disrupted the education of children. Some shortages of imported foodstuffs were caused by stocks being purchased for emergency supplies, which, in the end, were not needed. There were fire sales. Agricultural productivity also declined in the province during the crisis, although this was offset, fortuitously, by higher world prices for copra and cocoa. Many people thought, overall, that the emergency was a waste of money, given that no eruption had eventuated.44

There was some criticism, too, about the 1983 provincial disaster plan and its effectiveness. Granger, for example, in promoting the future use of GIS-based data for crisis management, criticised the way in which information had been collected and managed at Rabaul:

In some instances key information was unavailable and had to be collected. In other instances information was available but in a form that required considerable manipulation to make it useable, or was so out-dated that it was of dubious value. [Furthermore] … there was no guarantee that the information being used by one planner was the most current or most accurate available. Nor was it clear that it was the same information as that being used by another planner working on the same, or an associated, aspect in another agency. Under such circumstances it is not surprising that the 1983 Plan contains gaps in essential information and errors or inconsistencies in the data provided.45

Interest in future volcanic activity at Rabaul retained some momentum after the end of the Stage Two alert. The 50th anniversary of the 1937 eruption was commemorated at a well-attended ceremony,46 and some new instrumental monitoring was introduced by RVO. In particular, the University of Queensland, funded by AIDAB, in 1987 installed new tide gauges in Rabaul Harbour. Gauges were deployed on metal frames constructed on each of the two centres of sea floor uplift — the main one south-east of Matupit Island and the possibly secondary one east of Vulcan. These instruments, importantly, were to be linked by satellite communications to RVO and to the university in Brisbane, so providing the potential for near ‘real time’ monitoring of ground-height changes. The momentum of such volcano-focused activities, however, fell away after about 1987, including ongoing funding for the real-time tide-gauge system. In addition, there was reduced attention from the media which, understandably, was motivated more by the excitement and uncertainty of the Stage 2 crisis early in 1984, than by the ‘let down’ of its aftermath. Attention from international volcanologists was also diverted by other events, including particularly the major, VEI 6, eruption at Pinatubo volcano, Philippines, in June 1991 and the persistent threat there from lahars in the following years.

The amount of financial support that could be provided to RVO by the national government declined following the 1983–1985 crisis. Lowenstein resigned from RVO in January 1989, largely because of this reduced support. McKee took over as head of RVO, assisted by Talai as deputy. Leslie Topue retired in 1989, after 38 years of service, and died in 1992. The observatory, by the late 1980s, experienced difficulties in keeping all of the monitoring instruments up-to-date and operational. Some initiatives could not be started, including the high-priority construction of an emergency observatory in case RVO headquarters became inoperable during an eruption. Similarly, proposals to install telemetry links for the automatic measurement of temperature at key points in the caldera could not be funded. The routine, manual measurement of temperatures on the volcanoes was labour intensive, and it declined sharply during and after the crisis period, mainly because of other work demands. Neither was monitoring of changes in volcanic-gas emissions at Tavurvur possible. Six villagers were killed by gas asphyxiation in a crater on the southern side of the volcano on 24 June 1990 when three villagers, who had been collecting megapode eggs in the crater, were overcome by carbon dioxide, and another three villagers were killed the day after when they tried to rescue them.47


Figure 103. Dignitaries lay wreaths during a matamatanai or commemorative ceremony near Tavana village, west of Vulcan, on the morning of 29 May 1987, the 50th anniversary of the 1937 eruption. Villagers also staged a balabalaguan, a mortuary ceremony, in honour of the victims who years before had been memorialised in the old monument shown here.

Source: R.W. Johnson. Geoscience Australia (no registered number).

Some benefits, however, emerged from the 1983–1985 crisis.48 The extensive emergency planning in particular was a positive result — not only the eight Acts of Parliament concerned with national and provincial disaster-management arrangements, but also the general raising of community awareness of the volcanic hazards that threaten Rabaul, and what should be done in responding to the different stages of alert. The community in effect was made more risk-aware and therefore more resilient. Nevertheless, some parts of the community — for example, the European expatriates — may have been much better informed of what was happening than, say, Sepiks and other non-Tolai people in the settlements. Disaster preparations in the province that were funded by the national and foreign governments also resulted in improvements to infrastructure, including airstrips, wharves, roads, bridges, and water and power supplies. For example, an old airstrip at Tokua, 12 kilometres east of Kokopo, was developed in view of the considerable vulnerability of the Lakunai Airport within the caldera just to the north-west of Tavurvur volcano.

Benefits also accrued directly to RVO. The observatory had for a time received considerable and increased attention and resources from foreign governments, and from the national government, including increases to its annual recurrent budget. RVO also benefited from the practical experiences of volcanic-crisis management and of enhanced instrumental monitoring, especially the use of new geophysical instruments and techniques. Furthermore, its management of the crisis in hindsight can be judged as successful — despite the criticisms levelled at it at the time — in that it kept the community and authorities informed about the state of the volcano and what its expectations were for a possible eruption, despite the uncertainty under which it was operating. The pressure to raise the alert to Stage Three, or even Stage Four, had been considerable, but Lowenstein and his team resisted this.

Authorities at Rabaul, to their credit, continued to rehearse their emergency-management responsibilities. The rehearsals by April 1994, however, did not involve the communities themselves, which, in retrospect, can be regarded as an unfortunate disengagement from the population at large. Memories of the 1983–1985 crisis and its intense earthquake activity faded. Instrumental monitoring continued at Rabaul, although the volcano seemed now to be relatively ‘quiet’. McKee and some other RVO staff spent several periods away from Rabaul, monitoring eruptions at Manam in 1992 and Long Island in 1993, and establishing monitoring networks at Ulawun and Karkar. McKee and others were also occupied in mapping the geology of Lolobau and Hargy volcanoes in West New Britain, work that brought out McKee’s particular strengths as a careful scientific observer in the field. RVO in 1993 also hosted a group of overseas volcanologists who had been attending a major international volcanological conference in Canberra and who visited Rabaul on a field excursion. RVO life almost seemed back to ‘normal’.

Yet, the level of earthquake activity in the seismic annulus at Rabaul in 1992–1994 had not declined to pre-1971 levels. Indeed, there were indications of slight increases in the number of earthquakes, although not nearly to the levels seen in 1983–1985. Furthermore, some earthquakes in 1992 were detected beneath Namanula Hill and, later, further out into St Georges Channel — that is, well outside of the seismic annulus. These formed a zone that came to be referred to as the ‘northeast earthquakes’. In addition, the south-eastern end of Matupit Island continued to rise slightly, and increasingly so after 1991. In other words, there had still been no deflation of the ‘bulge’. All of these slight changes were insufficient to justify RVO raising the alert level at Rabaul, but they would later, in hindsight, become more significant for what would follow in 1994.

Volcanic Alert on Simbo Island

Interest in the volcanic crisis at Rabaul, Papua New Guinea, may have declined by 1993, but the opposite was true for people living on remote Simbo Island, across the international border in the western Solomon Islands, about 400 kilometres west-north-west of the capital Honiara. Simbo is a geologically youthful volcanic island on which, in the south, there are active hot springs, fumaroles, and solfataras, including at Pakeru or Ove Crater and Lake at the south-western end of the island.49 There are no confirmed reports of historical eruptive activity from any of the three main eruptive centres on Simbo, but radio reports from the island on 11 February 1993 were sufficiently dramatic to suggest to expatriate government geologists, R. Addison and M.G. Petterson, from the Department of Natural Resources in Honiara, that an eruption may have started at Ove Crater.

The radio reports were of ‘volumes of black smoke, a strong smell of sulphur, explosions and fire’ and, later on the same day, of ‘villagers complaining … of increased smoke and fumes causing nausea, chest pains and stringing eyes. The ground was reported to have been shaken by seismic shocks’ and there was ‘a white glow around the vent’.50 Advice from officials in Honiara was provided to the villagers on how to cope with volcanic eruptions and there were discussions on evacuation procedures that may need to be adopted. Honiara officials, including Addison and Petterson, flew to the island on both 12 and 13 February. Observations were made from helicopter on 12 February:

The circuits of the crater revealed … a number of active fumarolic vents (fissures between rocks) from which steam and smoke were issuing to be carried in low billows downwind in a northeasterly direction. The area from which most of smoke or steam appeared to be issuing was black, as if discoloured by volcanogenic sublimates … . A moderately strong smell of sulphur dioxide was detected from within the helicopter … a second series of circuits of the crater were [sic] made, during which the smoke was perceived to have thickened in consistency and to be flowing over the crater rim, down the slope of the cone, to be carried away as a ground-hugging cloud across the ridge NE of Ove Lake.51

A ground inspection revealed that the discolouration seen from the air appeared to be the result of burning caused by heat or acid attack. Trees and leaves were discoloured within the crater and trees there had shed large amounts of dead leaves. A key observation on the northern side of the crater wall was of sulphur that seemed ‘to have been burnt and melted. Patches and pools of black and yellow sulphur appear to have flowed and solidified in small pools and lobate flows in hollows of the wall.’52

The conclusion reached was that combustible sulphur had been ignited as a result of a bush or grass fire lit, evidently by tourists, on the top of the crater rim, and the burning had produced the clouds of ‘smoke’, sulphur dioxide and vapour. The ‘white glow’ may refer to the flames of the sulphur burning. Seismologist A.K. Papabatu had installed a seismometer near Ove on 14 February and had maintained recording there for four weeks, but no local, volcano-related earthquakes were recorded on the seismometer. The emissions had in fact died away after rainfalls on 12 and 13 February, which evidently extinguished any further burning. There had been no volcanic eruption, and the Simbo event was yet another, although unusual, example of a volcanic false alarm.


Archbold, M.J., C.O. McKee, B. Talai, J. Mori & P. de Saint Ours, 1988. ‘Electronic Distance Measuring Network Monitoring during the Rabaul Seismicity/Deformational Crisis of 1983–1985’, Journal of Geophysical Research, 93, no. B10, pp. 12, 123­–136.

Blong, R. & C. Aislabie, 1988. ‘The Impact of Volcanic Hazards at Rabaul, Papua New Guinea’, Institute of National Affairs Discussion Paper 33, Port Moresby.

Cooke, R.J.S., 1977. ‘Rabaul Volcanological Observatory and Geophysical Surveillance of the Rabaul Volcano’, Australian Physicist, February, pp. 27–30.

Crandell, D.R. & D.R. Mullineaux, 1978. ‘Potential Hazards from Future Eruptions of Mount St. Helens Volcano’, Washington, United States Geological Survey Bulletin 1383-C.

Darius, W., 1984. ‘Move to Kokopo is On’, Papua New Guinea Post-Courier, 16 February, p. 1.

De Saint Ours, P., 1993. ‘Lessons Learned by Rabaul Volcano Observatory in Volcano Crisis Management’, World Organisation of Volcano Observatories Workshop, Guadaloupe, French West Indies, Program and Abstracts (unpaginated).

De Saint Ours, P., B. Talai, J. Mori, C. McKee & I. Itikarai, 1991. ‘Coastal and Seafloor Changes at an Active Volcano: Example of Rabaul Caldera, Papua New Guinea’, Workshop on Coastal Processes in the South Pacific Islands Nations, Lae, Papua New Guinea. SOPAC Technical Bulletin, 7, pp. 1–13.

Dunkley, P.N., 1986. ‘The Geology of the New Georgia Group, Western Solomon Islands’. British Geological Survey Overseas Directorate, British Technical Cooperation Report, Western Solomon Islands Geological Mapping Project, 21, Report MP/86/6.

East New Britain Provincial Disaster Plan, 1983. East New Britain Provincial Government, Rabaul.

Ewert, J., C.D. Miller, J.W. Hendley II & P.H. Stauffer, 1998. ‘Mobile Response Team Saves Lives in Volcano Crises’, United States Geological Survey Fact Sheet 064-97.

Fisher, N.H., 1939. Geology and Vulcanology of Blanche Bay, and the Surrounding Area, New Britain. Territory of New Guinea Geological Bulletin 1.

Fournier d’Albe, E.M., 1979. ‘Objectives of Volcanic Monitoring and Prediction’, Journal of the Geological Society of London, 136, pp. 321–26.

Graham, T.L., M.G. Swift, R.W. Johnson, J. Pittar, P. Musunamasi & I. Kari, 1993. Rabaul Harbour Heat Flow Project 1993 Papua New Guinea: Final Report. Australian International Development Assistance Bureau.

Granger, K.J., 1988. ‘The Rabaul Volcanoes: An Application of Geographical Information Systems to Crisis Management’. Master of Arts thesis, The Australian National University, Canberra.

——, 1990. ‘Process Modelling and Geographic Information Systems: Breathing Life into Spatial Analysis’, Mathematics and Computers in Simulation, 32, pp. 243–47.

Greene, H.G., D.L. Tiffin & C.O. McKee, 1986. ‘Structural Deformation and Sedimentation in an Active Caldera, Rabaul, Papua New Guinea’, Journal of Volcanology and Geothermal Research, 30, pp. 327–56.

Grover, J.C., 1955. ‘Simbo Volcano’, in Geology, Mineral Deposits and Prospects of Mining Development in the British Solomon Islands Protectorate. Interim Geological Survey of the British Solomon Islands Memoir, 1, pp. 46–48.

Guppy, H.B., 1887. The Solomon Islands: Their Geology, General Features, and Suitability for Colonisation. Swan Sonnenschein, Lowrey, London.

Hastings, P., 1984. ‘Volcano Set to Blow: 20,000 Plan Island Escape’, Sydney Morning Herald, 26 January, p. 1.

Heming, R.F., 1974. ‘Geology and Petrology of Rabaul Caldera, Papua New Guinea’, Geological Society of America Bulletin, 85, pp. 1253–64.

Honey, F.R., 1991. ‘Passive, Two-Channel, Thermal-Infrared Imaging Systems for Discrimination of Volcanic Ash Clouds’, in T.J. Casadevall (ed.), Volcanic Ash and Aviation Safety: Proceedings of the First International Symposium on Volcanic Ash and Aviation Safety. United States Geological Survey Bulletin, 2047, pp. 347–50.

Johnson, R.W., 1984. ‘Volcanological Investigations in Papua New Guinea, February 1984’. Geological Survey of Papua New Guinea Report 84/4.

——, 1986. ‘Underwater Video Survey of the Volcanic Bulge on the Floor of Rabaul Harbour, Papua New Guinea, December 1985’, CCOP/SOPAC Proceedings of 15th Session, Rarotonga, Cook Islands, pp. 138–39.

Johnson, R.W. & T.J. Casadevall, 1991. ‘Aviation Safety and Volcanic Ash Clouds in the Indonesia–Australia Region, in T.J. Casadevall (ed.), Volcanic Ash and Aviation Safety: Proceedings of the First International Symposium on Volcanic Ash and Aviation Safety. United States Geological Survey Bulletin, 2047, pp. 191–97.

Johnson, R.W., I. Itikarai, H. Patia & C.O. McKee, 2010. ‘Volcanic Systems of the Northeastern Gazelle Peninsula, Papua New Guinea: Synopsis, Evaluation, and a Model for Rabaul Volcano’, Rabaul Volcano Workshop Report, Papua New Guinea Department of Mineral Policy and Geohazards Management, and Australian Agency for International Development, Port Moresby.

Johnson, R.W., & N.A. Threlfall, 1985. Volcano Town: The 1937–43 Rabaul Eruptions. Robert Brown and Associates, Bathurst.

Latter, J.H., 1988. ‘Quantitative Volcanic Risk in Asia and the Pacific’, in Urban Geology in Asia and the Pacific. United Nations Economic and Social Commission for Asia and the Pacific Atlas of Urban Geology, 2, pp. 16–23.

Latter, J.H. & A.W. Hurst, 1987. ‘An Assessment of Volcanic, Seismic, and Tsunami Hazard at Rabaul and Neighbouring Areas of New Britain, Papua New Guinea’, New Zealand Department of Scientific and Industrial Research, Wellington, Contract Report 25.

Lowenstein, P.L., 1988. ‘Rabaul Seismo-Deformational Crisis of 1983–85: Monitoring, Emergency Planning and Interaction with the Authorities, the Media and the Public’. Geological Survey of Papua New Guinea Report 88/32.

Lowenstein, P.L. & B. Talai, 1985. ‘Volcanoes and Volcanic Hazards in Papua New Guinea’, Geological Survey of Japan Report, 263, pp. 315–31.

McCormick, M.P., 1985. ‘Aerosol Observations for Climate Studies’, Advances in Space Research, 5, no. 6, pp. 67–73.

McKee, C.O., P.L. Lowenstein, P. de Saint Ours, B. Talai, I. Itikarai & J. Mori, 1984. ‘Seismic and Ground Deformation Crises at Rabaul Caldera: Prelude to an Eruption?’, Bulletin of Volcanology, 47, pp. 397–11.

McKee, C.O., R.W. Johnson, P.L. Lowenstein, S.J. Riley, R.J. Blong, P. de Saint Ours & B. Talai, 1985. ‘Rabaul Caldera, Papua New Guinea: Volcanic Hazards, Surveillance, and Eruption Contingency Planning’, Journal of Volcanology and Geothermal Research, 23, pp. 195–37.

McKee, C., J. Mori & B. Talai, 1989. ‘Microgravity Changes and Ground Deformation at Rabaul Caldera, 1973–1985’, in J.H. Latter (ed.), Volcanic Hazards: Assessment and Monitoring. IAVCEI Proceedings in Volcanology, Springer-Verlag, Berlin, 1, pp. 399–428.

Mori, J., C. McKee, I. Itikarai, P. Lowenstein, P. de Saint Ours & B. Talai, 1989. ‘Earthquakes of the Rabaul Seismo-Deformational Crisis September 1983 to July 1985: Seismicity on a Ring Fault’, in J.H. Latter (ed.), Volcanic Hazards: Assessment and Monitoring. IAVCEI Proceedings in Volcanology, Springer-Verlag, Berlin, 1, pp. 429–62.

Nairn, I.A., C.O. McKee, B. Talai & C.P. Wood, 1995. ‘Geology and Eruptive History of the Rabaul Caldera Area, Papua New Guinea’, Journal of Volcanology and Geothermal Research, 69, pp. 255–84.

Neumann, K., 1996. Rabaul Yu Swit Moa Yet: Surviving the 1994 Volcanic Eruption. Oxford University Press.

Newhall, C.G., & D. Dzurisin, 1988. Historical Unrest at Large Calderas of the World, United States Geological Survey Bulletin 1855.

Newhall, C.G. & S. Self, 1982. ‘The Volcanic Explosivity Index (VEI): An Estimate of Explosive Magnitude for Historical Volcanism’, Journal of Geophysical Research, 87, no. C2, 1231–238.

Peterson, D.W., 1988. ‘Volcanic Hazards and Public Response’, Journal of Geophysical Research, 93, no. B5, pp. 4161–170.

Petterson, M.G., D. Tolia, S.J. Cronin & R. Addison, 2008. ‘Communicating Geosciences to Indigenous People: Examples from the Solomon Islands’, in D.G.E. Liverman, C.P.G. Pereira & B. Marker (eds), Communicating Environmental Geoscience. Geological Society, London, Special Publications, 305, pp. 141–61.

Rabaul Volcanological Observatory, 1990. ‘Rabaul Caldera’, Global Volcanism Network Bulletin, 15, no. 6, pp. 8–9.

Sack, P., 1987. ‘The Emergence and Settlement of Matupit Island’, Bikmaus, Journal of Papua New Guineas Affairs, Ideas and the Arts, 7, pp. 1–14.

Siebert, L., T. Simkin, & P. Kimberley, 2010. Volcanoes of the World. 3rd edn. Smithsonian Institution, Washington D.C., University of California, Berkeley.

Stannard, B., 1984. ‘Rabaul Trembles as Fears of Big Bang Grow’, Bulletin, 104, no. 5405, 28 February, pp. 44­–49.

Stothers, R.B., 1984. ‘Mystery Cloud of AD 536’, Nature, 307, pp. 344­–45.

Thompson, D., 2000. Volcano Cowboys: The Rocky Evolution of a Dangerous Science. St. Martin’s Griffin, New York.

Tilling, R.I. & R.S. Punongbayan, 1989. ‘Scientific and Public Response’, in R.I. Tilling (ed.), Volcanic Hazards, Short Course in Geology, 1, American Geophysical Union, Washington D.C., pp. 103–06.

Voight, B., 1990. ‘The 1985 Nevado del Ruiz Catastrophe: Anatomy and Retrospection’, Journal of Volcanology and Geothermal Research, 44, pp. 349–86.

Walker, G.P.L., R.F. Heming, T.J. Sprod & H.R. Walker, 1981. ‘Latest Major Eruptions of Rabaul Volcano’, in R.W. Johnson (ed.), Cooke-Ravian Volume of Volcanological Papers. Geological Survey of Papua New Guinea Memoir, 10, pp. 181–93.

1 McKee et al. (1985).

2 Fisher (1939, p. 18), for example, reported the causeway forming after an earthquake in 1919, and continuing to rise slightly up to the time of the 1937 eruption. The causeway and its road were destroyed in 1971 as a result of tsunamis generated by the 1971 regional earthquakes, but were subsequently reconstructed by civil works.

3 Sack (1987).

4 Cooke (1977).

5 Heming (1974).

6 Lowenstein (1988).

7 East New Britain Provincial Disaster Plan (1983).

8 Johnson & Threlfall (1985).

9 McKee et al. (1984).

10 Lowenstein (1988), Blong & Aislabie (1988) and Neumann (1996).

11 Darius (1994).

12 Lowenstein (1988).

13 Hastings (1984).

14 Johnson (1984).

15 Stannard (1984).

16 Lowenstein (1988).

17 De Saint Ours et al. (1991).

18 See, in particular, Archbold et al. (1988), McKee et al. (1989) and Mori et al. (1989).

19 Greene et al. (1986).

20 Johnson (1986).

21 Graham et al. (1993).

22 Walker et al. (1981) and Nairn et al. (1995).

23 Stothers (1984). The erroneous correlation was based on uncalibrated radiocarbon dates — a calibrated age of AD 720–750 ± 20 is regarded as the best current estimate for the ‘1400 BP’ eruption (C.O. McKee personal communication, 2012).

24 Peterson (1988).

25 Tilling & Punongbayan (1989), Ewert et al. (1998) and Thompson (2000).

26 Newhall & Self (1982).

27 Siebert et al. (2010).

28 Uncertainties in the VEI assignments made by the Smithsonian Institution for some of these eruptions are not shown in this list.

29 Johnson & Casadevall (1991).

30 Honey (1991).

31 McCormick (1985).

32 See, for example, Voight (1990).

33 Crandell & Mullineaux (1978).

34 Newhall & Dzurisin (1988), p. 227.

35 Fournier d’Albe (1979).

36 McKee et al. (1985).

37 Lowenstein & Talai (1985).

38 Latter & Hurst (1987). J.H. Latter, a volcanologist at RVO in the late 1960s, also published on the numbers of at-risk people on active volcanoes in the member countries of the Economic and Social Commission for Asia and the Pacific, or ESCAP (Latter, 1988).

39 Blong & Aislabie (1988).

40 Granger (1988, 1990).

41 Lowenstein (1988) and De Saint Ours (1993).

42 Newhall & Dzurisin (1988), p. 26.

43 Blong & Aislabie (1988) and Lowenstein (1988).

44 K. Neumann (personal communication, 2008).

45 Granger (1988), p. 6.

46 See, for example, Neumann (1996).

47 Rabaul Volcanological Observatory (1990).

48 Lowenstein (1988).

49 Guppy (1887, Chapter 4), Grover (1955, Chapter 13) and Dunkley (1986).

50 Petterson et al. (2008), pp. 149–51.

51 Petterson et al. (2008), p. 151.

52 Petterson et al. (2008), p. 153.

Previous Next