Nyiragongo Volcano

Nyiragongo volcano (3470m) is a large basaltic stratovolcano located about 15 km north of the city of Goma in the Democratic Republic of Congo. The volcano is famous for persistent lava lake activity in its huge summit crater. Such lakes are currently only observed at few volcanoes, including Erta Ale in Ethiopia, Ambryn in Vanuatu, and Erebus in the antarctic. The volcano hit the headlines in 2002, when lava flows from an extensive fissure system on its southern flank destroyed much of the eastern part of Goma, resulting in about 150 fatalities and the destruction of thousands of homes. This was the first occasion when lava flows significantly damaged a major city since the eruption of Etna in 1669. The speed of the flows in 2002 and in a similar fissure eruption in 1977 were exceptionally high, making Nyiragongo one of few volcanoes where lava flows present an immediate hazard to human life.

Nyiragongo volcano Nyiragongo volcano, illuminating clouds at night

View to SE side of Nyiragongo from Ruanda

Nyiragongo lava lake illuminating haze / steam over summit

Nyiragongo Volcano Map Nyiragongo Volcano Map

Fig.1 Annotated Satellite Image of Nyiragongo and its Surroundings. Solid white lines: Bay of Bufumbira, Dashed white lines: Kamatembe Rift. Inset Shows East African Rift System in the Area Surrounding Nyiragongo. With kind permission of T. Platz.

Fig.2 Geological Map of Nyiragongo Volcano following the 2002 eruption. With kind permission of T. Platz.

It is noted that compared to other maps, this map appears not to show the full extent of the 2002 lavas, especially insofar as they were erupted over the 1977 products.

Basic Geology

Nyiragongo volcano lies in the Virunga volcanic field which includes 8 volcanic edifices broadly trending in an E-W direction to the north of Lake Kivu. The Virunga field belongs to the western branch of the East African Rift System and is located at the intersection between the N-S trending Albert Rift and Kamatembe and Bay of Bufumbira Rifts (see Fig.1). The wider geodynamical setting is discussed in more detail by Ebinger and Furman (Acta Vulcanol. 14/15 (1-2), p. 9-16 (2002)).

Nyiragongo and nearby Nyamuragira mark the western end of the Virunga volcanic field and together with Visoke are the only volcanoes which have been historically active in the area. Virunga volcanoes erupt extremely fluid low-silica high-alkaline lavas. Several fault systems meet at Nyiragongo, (1) the main N-S fault which was active in the 1977 and 2002 eruptions, (2) a NW-SE trending system linking Nyiragongo to Nyamuragira, (3) a NE-SW trending fault system including the Rushayo chain of scoria cones (formed in 1948), and possibly (4) a further fault approximately parallel but west of the Rushayo chain over which lie the scoria cones of Muhuboli and Gituro (see Fig.2). These appear to have been active in recent times. Further faults are likely to extend from the edifice, yet the dense vegetation and security situation make studies difficult.

Whilst the frequently active Nyamuragira takes the form of a classic shield volcano, Nyiragongo has a classical cone shape with slopes up to 50 degrees near the summit. The main summit crater of Nyiragongo is a 1-1.2 km wide steep-sided pit with terraces marking the former locations of lava lakes. The volcano is famous for the frequent presence of active lava lakes within the summit crater. The inner walls of the edifice down to the first platform reveal numerous alternating layers of lava flow and pyroclastic deposits, intersected by 13 dykes. Prominent craters named Baruta (3148m) and Shaheru (2600m) are located 1.5 km north and 2 km south of the summit crater, respectively, and have been referred to as parasitic cones or coalescing volcanic structures forming a "Nyiragongo complex" in different publications. About 100 further small cones are found on the flanks, with most aligned over fault structures. Lava erupted by Nyiragongo in recent times have had 36.7-41.5% silicate content, which is exceptionally low even for for basaltic lavas (Santo et al. 2003. Acta Vulcanol. 14/15, p.63-66). The lavas can be classified in the nepheline, melilite, and melilite-nepheline series based on the type of crystalline inclusions therein. Studies of such inclusions, attributable to fractionization of magma at different pressures (and thus depths), along with seismic date, lead to the conclusion that the volcano has two main magma chambers, one at a depth of 10-14 km, and a shallower one within the edifice (Demant et al. 1994. Bull. Volcanol. 56, p.47-61; Platz et al. 2004. J. Volc. Geotherm. Res. 136, p.269-295). Collapse of the roof of the shallow magma chamber is a likely mechanism for formation of the summit crater. Magma supply to the volcano is facilitated by the extensional tectonic regime which is also implicated as the trigger mechanism for the 1977 and 2002 eruptions.

Nyiragongo volcano Nyiragongo volcano, lava lake at night

Nyiragongo crater with lava lake inside, Jan. 2011

Nighttime view of Nyiragongo crater

Nyiragongo volcano, lava lake Nyiragongo volcano, lava lake

Lava lake sitting in almost filled pit left behind following 2002 eruption

Lava lake with activity along side (back side) and within lake

Nyiragongo volcano, lava lake Nyiragongo volcano, lava lake at night

Evening view of lava lake

Nighttime view of lava lake

Nyiragongo volcano, lava lake Nyiragongo volcano, lava lake degassing

Evening view of lava lake

Intense degassing during phase of high lake activity

Nyiragongo volcano, lava lake illuminating crater at night Nyiragongo volcano, lava lake illuminating crater at night

Nighttime view of western crater wall

Nighttime view of eastern crater wall

Nyiragongo volcano, lava lake activity Nyiragongo volcano, lava lake activity

Lava fountains near pit crater wall

Lava fountains distributed across the lake surface

Nyiragongo volcano, lava lake fountaining activity Nyiragongo volcano, lava lake fountaining activity

Lava fountain lifts up piece of lake crust

Violent lava fountaining

Nyiragongo volcano, lava lake fountaining activity Nyiragongo volcano, lava lake fountaining activity, Pelees hairs

Lava fountaining

Lava fountain draws lava into thin threads - formation of Pele's hair

Nyiragongo volcano, lava lake activity Nyiragongo volcano, lava lake activity

Crust of lake opens at edge allowing degassing

Activity from small fountain in lake and active zone at lake edge

Nyiragongo volcano, lava lake fountaining activity Nyiragongo volcano, lava lake fountaining activity

Fountaining is generally accompanied by vigorous degassing

Lava fountain and associated degassing

Eruptive History & Impact

First reports of Nyiragongo volcano reached the developed world in 1894. Following this date, numerous expeditions reached the volcano until more regular monitoring began in the 1950s. Historical descriptions of pre-2002 activity have been assembled by Durieux (Acta Vulcanol. 14(1-2), 2003, p.137-144). A lava lake was first directly documented within a pit in the summit crater in 1930, although gaseous emissions were already reported to be illuminated red in 1928. Since this time, the volcano has almost continuously hosted a lava lake, although the level and activity have fluctuated and overflows and partial collapses have gradually changed the morphology of the crater and the series of platforms therein. At times the lake was crusted over (e.g. 1965-70, 1996-2001) and only minor fumarolic activity or lava spattering from small cones was observed at the surface. At other times, particularly in the years preceding the 1977 eruption, vigorous fountaining activity could often be observed and the lake reached a level only 155 meters below the summit (compared to e.g. 395 m below the summit in 1959).

Nyiragongo volcano, upper crater wall Nyiragongo volcano, unstable platform Nyiragongo volcano, unstable platform

Steep upper walls of crater

Unstable terrace in crater

Sections of terrace have detached and lean inwards

Nyiragongo volcano, dyke Nyiragongo volcano, crater Nyiragongo volcano, lava channel on flank

Inner crater wall with remains of terraces from earlier lava lake activity. Arrow indicates position of one of the man dykes visible in crater wall.

Lake and W crater wall.

Old lava channel on upper flank. Evidence for high lava levels in past.

1977 Eruption

Since there was no permanent monitoring of the volcano at the time, little details exist about precursory events leading up to the 1977 eruption. The eruption is described in some detail in e.g. Durieux, 2003 (Acta Vulcanol. 14/15, p.145-148) and analysed further by Komorowski et al. 2003 (Acta Vulcanol. 14/15, p.27-62) with comparison to the 2002 event. From September 1976 onwards, several strong tremors were felt, and on Dec. 23, Nyamuragira started an effusive eruption which continued until June 1977. Further earthquakes occurred on Jan. 1 and 6, 1977. At 10:01 on January 10, a series of fissures opened along a N-S axis on both sides of the summit, accompanied by felt seismic activity. The main fissure extended from just south of Shaheru cone for about 6 km towards the village of Bushwaga with the S Shaheru vent at its upper end accounting for about 75% of the lava erupted. Extremely fluid lava flows poured down through the forested flanks and within minutes reached settlements kilometers away. Eyewitness reports suggest flow speeds of between 20 and 60 km/h and the flow passed the forest at such a speed that it did not set fire to the trees and even left some thicker leaves in place covered with a thin glassy layer. Ground deposits were at places only centimeters thick. Numerous people, especially the elderly or children, were unable to escape from the flows. Exact numbers of victims could not be established. Although the official count was 74, it is estimated that up to 400 people may have died. As the flows cooled and reached less steep ground, they slowed and a transition from pahoehoe to a'a morphology occurred. The longest flow stopped after about 9 km, just 1 km short of Goma airport. It took little more than an hour to cover this distance.

Over 20 million cubic meters of relatively degassed magma from the upper conduit / lake were erupted in less than an hour, after which the eruption rapidly ceased. A maximum volumetric eruption rate of up to 6000 cubic meters per second was estimated, which is unprecedented for basaltic magma in historic times. The driving force was primarily the hydrostatic pressure of magma in the upper 1000 m of the conduit which was effectively overhanging the main vents. Lava bombs found in trees around the main vents testify to vigorous fountaining caused by the extreme pressure exerted by the magma column at the onset of the eruption. Minutes after the start of the eruption, the terrace systems in the crater collapsed into the emptying conduit / chamber below, resulting in a powerful phreatomagmatic eruption due to the interaction of meteoric water and pore fluids entrapped in the terraces with the residual magma in the upper conduit. Images suggest that partial collapse of the eruption column resulted in pyroclastic flows descending the upper flanks of the volcano. The ash cloud reached an estimated altitude of nearly 10 km. In the days following the eruption, several powerful earthquakes were felt, and on January 16, an explosion and powerful quake accompanied the collapse of remaining parts of the first platform and the base of the crater, leaving a 900 meter deep pit. Based on the size of the pit, it can be estimated that about 235 million cubic meters of magma was emitted during the eruption, yet only about 22 million are found on the surface. This suggests that only 10% of the magma reached the surface, with the rest presumably being intruded in the fracture system.

Renewed magmatic activity was first observed in the 1977 pit in June 1982. This marked the onset of a period of rapid influx of lava into the crater, with the lava rising to within 400 m of the rim by November 1982, forming a 700 meter wide lake with an estimated volume of 70 cubic meters. In the mid-1990s, the lake rose further, reaching within 250 meters of the summit by Dec. 1995, shortly after which activity suddenly ceased. The surface of the lake remained solidified in the run-up to the 2002 eruption.

2002 Eruption

Due to its unusual nature and social impact, the 2002 eruption has been extensively discussed in the scientific literature. A detailed chronology and analysis is found in Komorowski et al. 2003 (Acta Vulcanol. 14/15, p.27-62), and much of the following summary is based thereon.

Monitoring Nyiragongo is somewhat complicated by the fact that field equipment is often vandalized or stolen in the area. Hence, whilst it was possible to detect seismic events and increases in fumarolic activity in the lead-up to the eruption, no data on ground deformation was available to the Goma Volcanological Observatory (GVO). In December 2000, an increasing number of long-period earthquake swarms commonly followed by volcanic tremors was detected in the region of Nyiragongo and Nyamuragira, yet due to the limited number of instruments, the tremor could not be specifically attributed to either volcano. On Febr. 6, 2001, a 2-week effusive eruption occurred at Nyamuragira. Unusually, after the end of the eruption, seismicity remained at high levels. These persisted until the 2002 Nyiragongo eruption. The general seismic unrest was punctuated by several stronger tectonic events. On Oct. 2 and 7, 2001 and on January 4, 2001, strong earthquakes were felt in the Goma region. Reports suggest that a dark plume rose from the volcano following the Jan. 4 event. Tectonic tremors continued at high levels in the lead-up to the eruption, although in the 8 hours directly before the eruption signals strangely dropped to low levels. Other pre-eruption signals were observed by GVO staff or reported by local people. In particular, fumarolic activity from fissures in the summit area was increasing, as was degassing through the crust of the lava lake. Further, near the rim of the summit crater the ground temperature was measured at 28'C in an area usually only measuring 5'-9'. Further, villagers reported steaming holes and cracks in the ground on the southern flank, together with locally elevated temperatures.

The January 17 2002 eruption of Nyiragongo volcano started suddenly at 8:25 local time. The eruption effectively involved a reopening and extension of the 1977 fault system and a renewed drainage of the lava lake, followed by the eruption of some magma from a deeper source. The eruption started with opening of a vent (V1) north of Shaheru crater at an altitude of 2800 meters. The vent aligned with the 1977 eruptive fissure. A further vent (V2) at an altitude of 3100 m below the western rim also emitted a short lava flow, presumably during the early stages of the eruption. Several minutes after the onset of powerful fountaining activity at V1, two sets of parallel fissures, about 300 meters apart, opened at an altitude of 2250 meters south of Shaheru.

Nyiragongo volcano, Shaheru crater Nyiragongo volcano, Shaheru crater

Shaheru crater viewed from the climbing track

Nighttime view over Shaheru crater showing the city lights of Goma

Nyiragongo volcano, Shaheru crater Nyiragongo volcano, 2002 fissure

Shaheru crater viewed from summit. Note: 2002 fissure below left side of crater and lava field in crater from 1977 eruption.

Upper section of 2002 fissure with summit area in background.

Nyiragongo volcano, lava nests in trees after 2002 eruption Nyiragongo volcano, lava nests in trees after 2002 eruption Nyiragongo volcano, lava nests in trees after 2002 eruption

Lava nests in trees resulting from vigorous fountaining at onset of 2002 eruption.

Nyiragongo volcano, lava tree mould Nyiragongo volcano, lava collar around tree trunk Nyiragongo volcano, lava tree mould

Tree mould with remains of tree behind.

Lava collar around tree trunk.

Tree mould left behind after tree stuck in lava decayed.

Lava began to spew out of the twin eruptive vents (collectively V3) formed near the top of these fissures. The fractures continued to propagate downhill in a southerly direction in the following hours with further vents opening at 10:00 (V5), 10:30 (V4 at 2000 m) and 14:10-16:20 (V8 at 1580 m), with V8 being the most southerly vent involved in the eruption. A further vent (V7) opened at 15:30 at an altitude of 1950 m, about 1.5 km to the west of the main eruptive fault. As in the 1977 eruption, extremely fluid lavas were erupted from the vents, especially in the early stages of the eruption. The most prolific vents in terms of erupted lava volumes were V3, V7 and V8. Lavas from V3 burned houses in several villages as they rapidly flowed downhill, reaching an eventual distance of 9 km. Lava flows from vents V3, V4 and V5 partially joined each other and a number of elderly people and children in an area south of V5, unable to escape to higher ground such as cinder cones, were overwhelmed by the lava and killed. The fissure extension was accompanied by strong local tremors and was clearly visible on the surface, not only at the eruptive vents but also in the form of the opening of a number of visible graben (i.e. trenches), some 10-20 meters wide and up to 10 m deep. In places, the cumulative extension of the fractures was as high as 39 meters (i.e. the ground either side of the fracture system moved nearly 40 meters apart).

Vents V7 and particularly V8 were responsible for most of the destruction in the town of Goma. V7 fed what ultimately became the longest flow of the eruption (10 km). The flow was fortunately relatively narrow and of slow-moving a'a morphology as it entered the western districts of town, damaging largely residential property. V8 was effectively a collection of vents along a length of about 1 km, with the most southerly vent being only 1.5 km north of the airport. Initial spattering and pahoehoe flows from this vent were soon followed by more substantial a'a flows which covered a third of the runway and proceeded through much of the commercial center of town, destroying many businesses, 45 schools and 3 hospitals, before continuing to flow into Lake Kivu, where a 80 m wide lava delta formed which extended up to 120 meters into the lake. The eruption lasted for about 24 hours, although lava continued to flow for many hours thereafter. Volumes of lava erupted by the main vents (in millions of cubic meters) were 1.16 (V1), 0.96 (V2), 5.8 (V3), 7.57 (V7) and 8.78 (V8). In total, about 25 million cubic meters were erupted. The missing volume of the crater resulting from the eruption was however at least 60 million cubic meters, suggesting that most of the magma therefrom was intruded into the underground fissure system.

Numbers of casualties are not exactly known, although less than 50 are thought to have died as a direct result of the eruption. Between 60 and 100 people died, and many more were injured, when a main petrol station surrounded by lava exploded. It was suggested in the media at the time that looters trying to steal petrol may have triggered the explosion. Several isolated fatal looting incidents were also reported.

Nyiragongo vents from 2002 eruption Nyiragongo vents from 2002 eruption

Cone and lava flow field along V8 vent system of 2002

Most southerly prominent cone on V8 vent system of 2002

In the 5 days following the eruption, around 100 tectonic events with a magnitude of over 3.5 were measured in the Goma area. Just after midnight on January 20, a magnitude 5 event took place. Building collapses caused a number of deaths in Goma and the adjoining Rwandan town of Gisenyi. Unlike in 1977, the floor of the crater lake did not collapse near the start of the eruption. However, the structure started to fail at about 20:51 on January 22. The collapse triggered violent phreatomagmatic activity, which persisted for about 4 hours. Intense seismic activity accompanied the event, powerful blasts could be heard and incandescence could be seen over the crater. Hot ash and scoria formed up to 10 cm thick deposits in the Rusayo area (8 km SW of summit). Surveys performed in the following weeks revealed the full extent of the explosive activity. The flanks were peppered with up to 50 cm large lava bombs down to an altitude of about 2900 meters, numerous scoria clasts were found down to about 2700 m on the south flank and the forest on the N and NE flanks had been flattened for up to several hundred meters from the crater (presumably due to a column collapse event). Part of the crater rim also collapsed and the crater was now once again a huge gaping hole with a depth of 900 meters, similar to after the 1977 eruption.

In Goma, in the days following the eruption, extremely high levels of methane were detectable at various locations in town resulting from seepage of methane and carbon dioxide-rich gases through tiny fissures. Locally, small explosions due to methane combustion were observed, and levels in the airport area were measured before flights were restarted due to concerns that levels in the area could be approaching the 5% flammability threshold in air.

Detailed morphological studies, including radioisotope analysis, have been performed on lavas erupted from different portions of the fissure system at different times during the eruption. Together with gas analyses, the results were used to analyse the eruption mechanism (Komorowski et al. 2003; Tedesco et al. 2007. J. Geophys. Res. 112, 12 p.). Initial lavas emanating from the upper part of the fracture system (i.e. vents V1 and V2) were largely degassed and highly fluid. These properties indicate the rapidly draining lava lake as their source. In the conduit, magma then rose and entered the developing fissure system as a result of depressurization and gas exsolution induced by the reduced hydrostatic load of the draining lava lake. It appears that the magma that filled the laterally propagating dike and was erupted from vents V3-V8 was more gas-rich and at least partially sourced from greater depth than initial V1-V2 eruptates. Indeed it seems that this magma may have bypassed the main conduit and directly fed into the dike system, possibly playing a role in its propagation. The southernmost system of vents (V8) erupted particularly vesicular material and formed a chain of scoria and spatter cones.

Tedesco et al. 2007 attributes the seismic activity in the year leading up to the eruption, with its pattern of tectonic > long-period > volcanic tremor, to subterranean fracture formation and associated magmatic intrusions. Hence, the deep-sourced feeder dike may have already extended to near the surface before the onset of the actual eruption. Whether magma movement only occurred as a result of tectonic rifting in the area or whether, and to what extent, magmatic pressure was involved in inducing the rifting cannot be conclusively resolved.

The predominance of a rift-trigger mechanism is supported by the fact that (i) the fractures formed during the eruption lie along a N-S axis which corresponds to the normal rifting orientation in the Kivu area, (ii) seismicity continued and indeed increased in the days following the eruption, in contrast to the normal situation following volcanic eruptions, (iii) the tip of the dike is visible at places in the eruptive fissure without having actually erupted, suggesting little pressure thrusting it upwards and that it may have risen upwards before draining relatively passively to lower parts of the fracture system as it extended. This lack of pressure suggests that rising magma was not pressurized sufficiently to drive the rifting mechanism. Indeed, (iv) the small magnitude of the eruption suggests it is unlikely that magmatic pressure could have triggered the large-scale fracturing of the edifice. However, it is noted that the fracture on the NW flank of the volcano does not obviously support the rift-trigger theory, since it is aligned more or less perpendicular to the rift.

In June 2002, the crater left by the eruption began to refill. The process was associated with intense degassing (Carn 2003. Acta Vulcanol. 13/14, p.75-86) sand at times violent lava fountaining and mild phreatomagmatic activity throwing material hundreds of meters in the air. An active lava lake has remained in the crater since then until the present day and its level has gradually risen from 2002-2010, approaching the level attained before the 2002 eruption. Hence, the system is essentially primed for another catastrophic lava lake drainage.

Hazard Assessments

Detailed analysis of the lava flow hazard presented by Nyiragongo to the adjoining urban regions of Goma and Gisenyi has been made, together with an assessment of the possibility of diverting possible new flows emanating from the fracture system which was associated with the 1977 and 2002 eruptions (Favalli et al. 2009. Bull. Volcanol. 71, p.363-374; Chirico et al. 2009. Bull. Volcanol. 71, p.375-387). It was concluded that the eastern part of Goma, most affected by the 2002 flows, remained most at risk and could not effectively be protected, whilst large barriers positioned at two locations could divert lava away from the rest of the town as long as the lava was emitted from a fissure on the flank of the volcano. The suggested abandonment of eastern Goma is however clearly not practical and building barriers for diverting lava away from the rest of town and into the eastern suburbs and the border zone between Goma and the adjoining Rwandan town of Gisenyi would clearly meet resistance from inhabitants of the affected areas. Further, a risk would remain that further extension of the N-S trending fissures as seen in the 2002 eruption (compared to 1977) would result in disruption of the barriers and / or eruption of lava directly into the urban areas. Indeed, minor N-S oriented fissuring could be detected around the airport building following the 2002 eruption and minor emissive fractures were later found 500 m nearer to the airport than the southernmost part of vent 8. Further vents may have been hidden by overflowing lavas from V8. Hence, the fracture system already appears to extend into town, although significant opening of the fracture has yet to occur in the area.

The risk of explosive events at Nyiragongo also needs to be considered. The risk of a catastrophic explosive eruption from the summit crater appears small whilst the system is essentially open and able to degas easily, as in historical times. However, the interaction of meteoric water with magma / lava can result in violent phreatomagmatic activity. For example, partial collapses of the intra-crater terraces or lava lake crust into the conduit following drainage of the upper section of the conduit led to violent phreatomagmatic explosions in 1977 and 2002, shortly after the effusive events. These events led to a peppering of areas of the crater-rim with large ballistics and to the flattening of some areas of trees, probably as a result of small column-collapse pyroclastic flows in the vicinity of the crater rim. Phreatomagmatic or hydromagmatic activity may also occur when lava flowing laterally through fissures encounters water. For example, studies of the volcanic cones southwest of the volcano, and in Goma itself, shows that cones located within about 1.5 km of Lake Kivu are largely tuff cones, the formation of which involved explosive activity, whilst cones located further inland were largely effusive cinder or spatter cones (Capaccioni et al. 2003. Acta Vulcanol. 14/15, p.129-136). Mt. Goma, rising 200 meters above the harbour of Goma city and currently site of the Volcano Observatory is an example of a lakeside tuff cone. If it had been formed during the 2002 eruption, violent surges from it could have devastated the surrounding town.

Further hazards resulting from Nyiragongo have been considered. Lake Kivu to the south of the edifice entraps large quantities of carbon dioxide and methane in its depths by virtue of the hydrostatic pressure of the overlying water body. In 2002, Lake Kivu was estimated to contain 55 and 250 cubic km of methane and carbon dioxide, respectively. Disruption of the water-column in a lake holding such high concentrations of dissolved gases can lead to rapid exsolution and release thereof. The fatal effects of such events could be observed at the Nyos crater lake in Cameroon, where a landslide into the lake in 1986 destabilized the water column and exsolved carbon dioxide flowed down the flanks leading to 1800 fatalities by suffocation (Baxter et al. 1989. Brit. Med. J. 298, p.1437-1441). Were an equivalent event to occur at Lake Kivu then the large urban centers at the lakeside, such as Goma and Gisenyi, would be at risk. Indeed, geological studies suggest that a catastrophic lake-turnover has occurred in the last 10,000 years (Haberyan and Hecky, 1987. Paleogeogr. Paleoclim. Paleoec. 61, p.169-197). A detailed risk assessment was made following the 2002 eruption (Schmid et al. 2003, Acta Vulcanol. 14/15, p.115-122). It was concluded that it was unlikely that heat released from the lava flows entering the lake would be able to cause sufficient convection at depth to trigger a catastrophic gas release. Only a large eruption within the lake was considered to have the potential to release the gases. Commercial exploitation of the methane reservoir could further reduce the risk in coming years. Small-scale extraction has been performed from 1962 to 2004 by the lakeside Bralirwa Brewery to power its generators and heat the boilers, and the KivuWatt project is extracting methane for power generation since 2010.

Bralirwa Brewery, lake Kivu, Ruanda KivuWatt project rig, Lake Kivu Methane from lake Kivu

Bralirwa brewery aside lake Kivu

Contour Globals KivuWatt project rig for gas extraction in Lake Kivu

Methane from lake was used at the brewery from 1962 to 2004

Indirect hazards relating to eruptions may include the outbreak of epidemics if large numbers of people need to be moved into refugee camps. Indeed, during the Rwanda crisis in 1994, 50,000 refugees dies of cholera in the Goma area. However, following the 2002 eruption, no major outbreaks of infectious diseases occurred, probably in part due to the rapid establishment of temporary water chlorination plants along the shore of Lake Kivu (Baxter et al. 2003. Acta Vulcanol. 14/15, p.109-114). Civil unrest is another potential problem in such a politically unstable region. Following the 2002 eruption, there were isolated fatal looting incidents and reestablishment of the electricity supply was hampered by removal of cables and poles by looters. Mains water supply was also effected, since many of the pumping stations that had remained unscathed during the eruption were without power. However, on the whole, the situation remained relatively calm.

Degassing from both the central crater and from fumaroles in the surrounding areas, including Goma, also can present an ongoing hazard to the population. The plume from the lava lake was estimated to include 60000 tonnes of sulphur dioxide per day in May 2002, with significant amounts of HF, HCl and carbon dioxide (Vaselli et al. 2006. Chinese J. Geochem. 25(Suppl.), p.71-72). The acidic gases and associated acid rain falling downwind had a visible effect on vegetation downwind. High fluoride levels could be measured in rainwater and are cause for concern due to their toxic effects on humans and livestock. In certain areas around the volcano, in particular near Lake Kivu to the SW of Nyiragongo, dry gas vents known locally as Mazukus (= Evil Winds) pose a threat to humans and livestock (Vaselli et al. 2003. Acta Vulcanol. 14/15, p.123-128). These vents release high concentrations of carbon dioxide of presumably magmatic origin, which depending on meteorological conditions may accumulate at dangerous levels in depressions near the source vent. Animals and occasionally humans may suffocate as a result, with humans being particularly threatened at night when they lie close to the ground. Interestingly, there are accounts that on April 22, 2001, a climbing party was temporarily incapacitated by gas emissions at an altitude of 3000 meters on the flank of the volcano, showing that such hazards may not be confined to low-lying areas.

Visitor Information

The region surrounding Nyiragongo remains politically volatile with some nearby areas under control of rebel / criminal groups. Reliable up-to-date information should be obtained before travelling. Most visitors reach Goma via the Goma / Gisenyi border crossing after flying to Kigali in Rwanda. Whilst Nyiragongo continually draws small numbers of tourists and guides and porters can be arranged for the approximately 1600 meter ascent from the National Park Headquarters, the adjacent Nyamuragira shield volcano is only rarely visited.

Nyiragongo National Park Sign Bullet Holes Nyiragongo park entrance

Bullet-ridden National Park sign

Start of path to Nyiragongo at park entrance

Nyiragongo crater rim Nyiragongo summit camp site

Large sign on crater rim

Camp site on terrace at crater rim. Wooden huts will be placed here in mid-2011

In Goma, photography was hindered by corrupt police claiming that a permit was necessary and demanding that one pay a fine (without receipt of course). However, obtaining a permit was essentially impossible since there is no obviously responsible authority. Goma does have modern Supermarkets (serving largely the UN troops) and a couple of modern hotels, yet much of the town remains in a desolate state. Visitors may visit the Gorillas in Virunga National Park. Permits (as for Nyiragongo) must be arranged in town. Whilst the drive to Nyiragongo is easy, the roads leading to the Rumangabo Ranger station where Gorilla trecks start require use of a 4WD vehicle.

Gorillas, Virunga National Park, Congo

Gorillas in Virunga National Park

Photovolcanica Full Index