ERUPTION AT NYIRAGONGO VOLCANO, DEMOCRATIC REPUBLIC OF CONGO

17-18 JANUARY 2002

A REPORT ON A FIELD VISIT TO ASSESS THE INITIAL HAZARDS

OF THE ERUPTION

Prepared for the World Health Organization

Dr Peter J Baxter MD

University of Cambridge

18 February 2002


Introduction

1. The writer was a member of a French-English official scientific team sent to evaluate the initial impact of the Nyiragongo eruption on 17-18 January 2002. This report has been written at the request of the World Health Organization. It is a summary of an initial evaluation of the health hazards of the eruption made in the field during 22-26 January. The team over-flew the volcano and Goma in a helicopter from Kigali on 22 January and made field observations in the area from 23-28 January (the writer, 23-26 January). One team member remained to continue studies on Lake Kivu until 13 February.

2. Nyiragongo is a volcano which does not erupt frequently, but it gained its notorious reputation at its last eruption in 1977, when 40-300 villagers were killed when a set of perpendicular fissures suddenly opened on the high flanks to release exceptionally fast-moving lava flows from its lava lake. The longest flow on that occasion travelled 10 km to within 2 km. of Goma airport. No further major activity ensued after that event until the new eruption on 17-18 January 2002, when lava flows entered the city of Goma. This disaster triggered an immediate response from international agencies and governments. From the natural hazard perspective, the expansion of populations in areas of active volcanism exposes large numbers of people to novel eruptive threats, and this is the first time that a large city has been invaded by extensive lava flows. This new disaster scenario required an urgent public health response to evaluate the hazards to life and health. This assessment was made all the more urgent because of the humanitarian crisis in the eastern Democratic Republic of Congo (DRC) that led many of the people who had fled from the lava flows to return to the city within one or two days, despite the danger of further eruptive activity or the possible risks arising from the proximity of the cooling lava flows. On 22 January, about 100 people died in an explosion at a gas station when a petrol leak ignited on the cooling lava, resulting in the largest death toll to date from the eruption.

3. The writer was invited by the French government to join a team of three French volcanologists who were assigned to leave Paris on 21 January with an official delegation of the French and English foreign ministers, which had been previously planned as a diplomatic mission to countries in eastern Africa, including DRC and Rwanda. The assistance of the French embassies in the two countries was invaluable in enabling the scientific team to reach Goma by the afternoon of 23 January, completing the journey by road from Kigali, Rwanda. A French volcanologist, Jacques Durieux and two Italian volcanologists, Dario Tedesco and Paolo Papale, had arrived the day before for the United Nations and we joined with them to form a working group.

4. The official French-English team comprised Patrick Allard (CNRS, France), Jean-Christophe Komorowski (IPGP, France), Michel Halbwachs (Universite de Savoie, France) and the writer, Peter Baxter (University of Cambridge). Allard had studied the volcano some years earlier, and Halbwachs was already working on Lake Kivu to assess the feasibility of extracting methane stored in its depths for power generation in Rwanda. The eruption presented a new and potentially devastating possibility of the huge amounts of carbon dioxide and methane stored at the bottom of the lake being released in a gas burst, resembling the lethal event at Lake Nyos, Cameroon, in 1986. Baxter had been a member of the international scientific response-team to the Lake Nyos disaster, and Halbwachs had recently headed an international team which had successfully begun the operational degassing of the lake to remove its gas hazard. The eventuality of either Nyiragongo or Nyiramuragira volcanoes erupting into Lake Kivu had, rather surprisingly, not been foreseen by scientists, and one of the Goma lava flows had formed a new delta at the lake shore. The danger from the stability of the lake waters being disturbed by the entry of hot lava at depth added substantially to the urgency of the scientific task.

The eruption of 17-18 January 2002, and subsequent eruptive events.

5. Local volcanologists informed us that from 4 January the population were aware of frequent felt earthquakes which continued for nearly two weeks until the eruption, which began without an increase in earthquake activity, or any other warning, at 0830 on 17 January. A fissure opened on the east flank high up on the volcano, which drained lava from the lava lake in the crater to form a limited lava flow that destroyed two villages. The area had been affected in 1977, when the lake had emptied its lava over 18 km3 in less than one hour. Villagers were aware of this event, but in any case they had been advised by the Goma Volcano Observatory since 1993 to flee to higher ground if a new eruption occurred, which is what most of them did. As it was a clear day, people could see the lava flows coming. The people who were killed in this and the subsequent fissure flows were amongst the elderly, or groups such as patients left behind in a dispensary, but even fit people were also unable to avoid being engulfed. Two children were known to have died in one village.

6. The population of Goma (about 500,000 people) was not advised on the need to evacuate after this first event. Further lava flows flowed from new fissures, or fractures, which propagated along the southern flanks of the volcano towards Goma later on that day (Figs.1-4). The two main flows extended to the city of Goma (Fig. 5), one traversing part of the airport and the older commercial centre before ending at the shore of Lake Kivu, where it formed a delta and lava tube into the water. The other shorter flow stopped where it cut the main west road out of the city. All buildings in the direct path of the flows were engulfed and destroyed. The high temperature of the flows (> 800 0C) led to the immediate combustion of all flammable materials in their paths and inside concrete or block structures remaining standing but in contact with the flows and subjected to the intense radiant heat of the lava. The lava in the city was fluid enough to flow inside buildings through doors and openings and moved at a steady, walking pace, but the flows also built up against buildings before they collapsed and became engulfed. MONEC personnel (blue helmets) helped to warn the people to flee from the oncoming flows. Over 200,000 escaped, with some people going west further into Congo, but most fled east across the border into Rwanda. The majority of the evacuees started to return next day after the flows had stopped moving, though the lava remained very hot.

7. The homes of about 100,000 people in Goma were destroyed, together with a substantial core number of commercial buildings. About 15-20 % of the city area was covered by lava. Altogether, 47 deaths were reported in the initial event. Those who lost their homes also lost all their possessions, as they had to leave without any warning; they were not insured.

8. Seismicity, as felt earthquakes, increased over 19-20 January, some being felt at Kigali, 150 km away. This led to fears amongst the people of a major earthquake. Some houses collapsed as a result of these “volcanogenic” quakes, about Richter scale 4-5, and up to nine deaths and 150 injuries were reported from this cause in nearby Gisenyi in Rwanda. The seismicity and the felt quakes waned over the next few days.

9. On the night of 21 January, a collapse and explosion occurred inside the crater, which by now had been observed by volcanologists to have lost its lava lake. The absence of the lake implied that most of the stored lava had been released during the eruption. A fall of hot, deep ash (about 10 cm.) occurred on the upper south-west flank of the volcano, damaging vegetation, with a light dusting of fine ash falling in Gisenyi and Goma.

10. By 25 January, the volcanic activity, as measured by seismicity, had declined and there had been no further lava flows since 18 January. Volcanologists agreed that the threat of further activity had, for the time being at least, diminished. Daily life in Goma was rapidly resuming, and the top loose surface rocks of the lava flows were being bulldozed away to open up the main roads to pedestrian and motorized traffic previously made impassable by the flows. The lava flows were still very hot beneath the surface, and heavy rain falling on the lava produced clouds of steam, which also created unfounded anxiety that further activity was occurring. Reports of sporadic, small explosions around the city over the days since the lava flows stopped were received, including during bulldozing over the flows. A pervading smell of methane gas was also reported, strongest in certain places and buildings, including at the airport.

Acute volcanic hazard assessment

11. The hazards of lava flow eruptions are governed by geological considerations, such as the chemical composition of the magma for a particular volcano. The lava at Nyiragongo is mafic, i.e., it has a silica content less than 55%, and so has a lower viscosity than the silicic lavas. Mafic lava flows occur at such familiar volcanoes as Kilauea, Hawaii and Mount Etna, Italy, but the lava at Nyiragongo differs from these and most other volcanoes around the world in having a much lower viscosity and being capable of flowing under gravity in an unusually fluid form. The flows are potentially much more hazardous in being difficult or impossible to out-run, with speeds of 30 km/hr on average, and an initial velocity of 100 km/hr at the fractures. The eruptive history of Nyiragongo is not well described, as the scientific studies have not been undertaken. The edifice of the volcano is comprised of numerous layers of old lava flows, but the maximum run-outs of these in the largest foreseeable eruptions of the volcano are not known.

12. The main hazards of lava flows that have been previously recognized at volcanoes include:

· Mechanical destruction and fires (lava lake temperature > 1000oC)

· Fall of heavy fragments from lava fountaining and forceful spattering

· Explosions when lava flows advance over vegetation (forming flammable organic gases)

· Explosions when hot lava enters water expanses: steam explosions inside lava tubes

· Gases from fissures

· Opening of fissures and shifting ground, with risk of subsidence

· Ground instability from lava loading on steep slopes leading to landslides triggered by heavy rain, earthquakes

· Collapse events at shorelines where “new land” of lava forms on steep shore gradients

All of these hazards have arisen at this eruption.

13. Nyiragongo’s northern neighbour, Nyamuragira, erupts more frequently with lava flows (it had a small eruption about a year ago). In 1938 lava from this volcano went as far as Lake Kivu. In 1977 both volcanoes erupted at the same time, and a concern of volcanologists has been that Nyramuragira would erupt again on this occasion, but there has been no evidence of such a threat so far. The ash from Nyiramuragira is believed to be toxic; in an eruption in 1995 thousands of cattle grazing in the ash fall area died. The ash fell in needle rather than round-particle form, which may have interfered with the feeding or digestive processes of the cattle, or the ash could have had a toxic coating of fluoride. I am not aware of any studies performed on this ash in 1995. Nyamuragira last erupted in February-March 2001.

14. The summits of the two volcanoes are about 15 km. apart and separated by a valley area. Vast ground emissions of carbon dioxide occur here, the lowest depressions containing the skeletons of many wild animals that have been lured to their deaths by the lush grass. The area is also known as the “Cemetery of the Elephants.” Mofettes, or small vents through which volcanic gases (carbon dioxide) outflow independently of eruptive activity, are also found in this active volcanic area (e.g., at Masuko). The ground carbon dioxide emissions are known to be a hazard to local people. As recently as last December four people died from this cause whilst working in a field.

15. The smell of gas, in particular methane, was reported in the city after the eruption. The possibility of soil gas emissions being a hazard from the eruption causing fracturing of the ground in and around Goma required assessment. Minor explosions were also reported in the days after the eruption (Fig.6).

16. A gas plume was visible from the helicopter arising from the Nyiragongo crater, but the altitude of the summit (3,489 m., or11, 385 ft.) is probably too high for the plume to pose an air pollution hazard to Goma. No measurements of the gas flux from the crater have been undertaken. A major plume from the crater was visible on satellite photos during the lava flow eruptions. Volcanic gases include sulphur dioxide, hydrogen sulphide and hydrogen chloride and hydrogen fluoride, as well as carbon dioxide. A more relevant consideration is the emission of gases from the fracture sites opened in the eruption from where lava has, or has not, been emitted. The fractures are lengthy and pass through, or close to, villages (Fig. 7). Lava fountains were observed from the vents during lava emission, so the high temperatures would have driven gases upwards in plumes, which would be unlikely to have grounded and been hazardous to local people at the time. This type of lava rapidly degases, usually as it emerges from the vents, but some residual degassing can occur from flows themselves, though the high temperature also results in strong upward convection of any such emissions.