Thousands of people and animals died of suffocation when carbon dioxide emissions from the supersaturated waters of volcanic Lake Nyos flowed down nearby valleys.
On the evening of August 21, 1986, a mysterious gas emission from volcanic Lake Nyos in the African nation of Cameroon flowed down nearby valleys, killing about seventeen hundred people and eight thousand head of livestock. Many survivors reported that they had heard rumbling sounds. One witness told of hearing bubbling sounds from the lake and seeing a white cloud rising from the lake and a large surge of water. Many survivors who had been overtaken by the cloud reported smelling a sulfurous odor and feeling a warm sensation before they rapidly lost consciousness; some remained unconscious for as long as thirty-six hours. Cattle were killed up to one hundred meters above the valley floor, indicating that the cloud was about one hundred meters deep. Small animals, birds, and insects were also killed, but vegetation was unaffected, and there was virtually no property damage.
All evidence from the disaster was consistent with exposure to a gas cloud composed of carbon dioxide. Carbon dioxide concentrations of 10 percent lead to rapid unconsciousness and to coma and death in fifteen minutes or less; 40 percent concentrations are almost immediately lethal. Many of the victims at Lake Nyos seemed to have died very quickly, suggesting very high carbon dioxide concentrations.
Pockets of gas lingered in low and sheltered areas and caused a few fatalities the next day, when people unwittingly walked into the gas pockets. The warm sensation reported by survivors is an effect of high carbon dioxide concentration; the gas itself was not unusually hot, and none of the victims was burned by it. The sulfurous odor might have been attributable to small amounts of hydrogen sulfide, but high levels of carbon dioxide are known to cause sensory hallucinations that might account for the unusual odors and sensations of warmth. High carbon dioxide concentrations also cause irritation of the mucous membranes, a symptom reported by many survivors for a few days after the event.
Lake Nyos is slightly more than one kilometer in diameter and two hundred meters deep, surrounded mostly by steep bluffs up to one hundred meters high. It does not lie in a volcanic cone, as do many crater lakes; rather, it occupies a maar, a crater formed by gas-rich explosive eruptions. Maar eruptions blast out craters but do not erupt enough material to build cones. Lake Nyos is surrounded by volcanic ash deposits up to thirty meters thick that were deposited by the eruptions that formed its crater only a few hundred years ago. The deeper levels of the lake are so highly supersaturated with carbon dioxide that water samples effervesce when brought to the surface.
The gas release itself was not a volcanic eruption; the deep waters of the lake were not unusually warm after the event, bottom sediments were undisturbed, and there was very little sulfur present in the lake water, contrary to what is found following normal volcanic eruptions. Instead, the magma source that fed past volcanism in the Lake Nyos area continued to seep carbon dioxide into the deep waters of the lake, where it dissolved. Eventually, the carbon dioxide content of the water became so great that carbon dioxide escaped to the atmosphere. The final event that triggered the release may have been trivial, possibly only a slight breeze. Once gas began escaping, the gas content of the water dropped, permitting still more gas to escape. About one cubic kilometer of gas escaped, enough to lower the lake level by more than one meter.
The release of gas was abrupt enough to create large surges in the lake, sending water as high as eighty meters up the shoreline in places. The escaping gas carried oxygen-poor deep lake water with it. When this water reached the surface, dissolved iron reacted with oxygen in the air, creating iron hydroxide that stained the lake yellow for several weeks.
The Lake Nyos disaster was not the first of its type. On August 15, 1984, a gas cloud from Lake Monoun, about ninety-five kilometers south of Lake Nyos, killed thirty-seven people. This earlier incident came at a time of severe political instability in Cameroon and did not attract widespread attention. The occurrence of both disasters in the month of August has suggested to some researchers that the gas releases were triggered by seasonal overturn of the lakes, which usually occurs when surface waters become cooler than deep waters and sink because of their greater density. The hypothesis remains plausible but unproven, because very few data have been gathered concerning the seasonal stability of lakes in equatorial Africa.
Volcanic eruptions are largely powered by gases, and the two most abundant gases in volcanic emissions are relatively benign: water vapor and carbon dioxide. Although hazards from explosions and hot gas clouds are well documented, and global climatic effects of volcanic carbon dioxide emissions have been suspected, the gases themselves are rarely dangerous. Toxic hazards from volcanic eruptions are normally associated with the less abundant components of volcanic emissions: sulfur dioxide, hydrogen sulfide, hydrogen chloride, and hydrogen fluoride. Cases of lethal carbon dioxide emissions during volcanic eruptions have been reported, but the Lake Monoun and Lake Nyos incidents were the first recorded cases of lethal gas emissions from volcanic lakes.
Lake Nyos in Cameroon a few days after the deadly gas was emitted from the lake, killing more than seventeen hundred people.
The conditions required for a lethal gas discharge of the Lake Nyos type are unusual. Among the immediate impacts of the Lake Nyos incident was an awareness of the need to investigate this previously unrecognized type of disaster. Researchers launched three lines of investigation: a search for the mechanism that accumulated the gas and released it, a search for evidence of past gas releases, and a search for other potentially dangerous lakes.
A variety of lines of evidence point clearly to slow seepage of gas from deep volcanic sources into the lake waters, which ultimately become supersaturated with carbon dioxide and become so unstable that even slight disturbances may trigger a gas release—thus the exact triggering mechanism in a given case may never be known. Evidence for past gas emissions may exist in local legends, which tell of explosive events that could have been maar eruptions, gas outbursts, or both. Physical evidence for past emissions may exist in the form of eroded ash deposits around the lake, which may have been washed away by waves from earlier gas emissions. It is possible that some mass deaths preserved in the fossil record may have resulted from events like the one that took place at Lake Nyos in 1986, so identification of physical evidence for gas outbursts may be of interest to scientists interpreting the geologic past in other regions.
Identification of other potentially dangerous lakes is not difficult. The numerous volcanic crater lakes of Cameroon are candidates, as are crater lakes elsewhere. Sampling of lake-bottom waters for dissolved gases is enough to identify hazardous lakes and rule out nonhazardous ones. To accumulate hazardous gas levels, a lake must be deep, both to have a large enough water volume to contain the gas and to allow stratification so that bottom water is prevented from rising to the surface and releasing gas harmlessly. Shallow lakes frequently emit methane, a gas created by decaying plant matter, but are not deep enough to accumulate dangerous amounts of gas.
Relatively small lakes are more likely sites for gas accumulation than are large lakes. Large lakes can be kilometers deep, but seasonal overturn causes mixing of deep and surface waters in most large lakes. Another cause of mixing in large lakes is upwelling of deep water, which may occur as winds push the surface waters to one side of the lake and deep water rises to replace the surface water. Perhaps the most gas-charged large lake in the world is Lake Kivu, between Rwanda and the Democratic Republic of the Congo, which is estimated to contain about sixty cubic kilometers of dissolved methane
The abundance of gas-charged lakes in Africa is closely related to the geology of Africa. The volcanoes of Africa are noted for their unusual chemistry and richness in carbon dioxide. One of the most famous and most unusual, Ol Doinyo Lengai
In hot-spot volcanoes such as those in Hawaii or Yellowstone National Park, a long-lived upwelling stream of hot mantle rock causes volcanism in the interior of a plate. Whenever mantle rock melts, some elements tend to concentrate in the melt, typically large atoms such as potassium or those with large ionic charges, such as uranium. These atoms are too large or of the wrong electric charge to fit in the crystal structures of most minerals in the mantle. Over the history of the earth, the crust has been enriched in these materials and the mantle has been depleted. Volatile components such as water and carbon dioxide also escape during melting. The mantle has thus become more chemically homogeneous over time.
Beneath long-stable continents, however, no processes have acted to remove gases and incompatible chemical elements. Geologic mapping shows that Africa was assembled by the collision of numerous blocks of crust between 900 and 600 million years ago. Since that time, the mantle beneath the interior of Africa has remained largely undisturbed. When the crust of a long-stable continent is rifted apart, the volatile materials and incompatible elements in the underlying mantle can escape to the surface. The initial rifting of a continent is thus often marked by gas-charged volcanism of unusual chemistry.
There are two principal zones of rifting in Africa. By far the most important zone is the East African Rift,
When a continent begins to break apart, the first stage of the breakup appears as local centers of uplift, volcanism, and crustal fracturing. Fractures radiate away from each center in a three-armed pattern, and eventually some fractures link to form a complete network that cuts the continent into two or more pieces. As the pieces spread apart and become separate continents, bulges on one piece match reentrants on the formerly adjacent piece. The pieces can be joined like pieces of a jigsaw puzzle; the fit between Africa and South America is one of the most dramatic pieces of evidence for plate tectonics.
Each reentrant is outlined by two arms of the initial fractures. The third arm extends into the continent, marked by a rift valley and perhaps volcanic activity. Because this third arm never develops into a full-fledged break, it is sometimes called a “failed arm.” The Benue Trough is such a failed arm, a sediment-filled rift extending into the interior of Africa. Much more recently, the zone of weak and faulted crust along the Benue Trough allowed magma to penetrate to the surface, creating the Cameroon Volcanic Line. The East African Rift may also be a failed arm, the other two arms being the Red Sea and Gulf of Aden. The long stability of Africa and its recent rifting account for both the unusual chemistry of African volcanoes and the continent’s abundance of dangerously gas-charged lakes.
-
Freeth, Samuel J., and R. L. F. Kay. “Volcanic Geology: The Lake Nyos Gas Disaster.” Nature 325 (January 8, 1987): 104-105. Provides a moderately technical description of the geologic setting of Lake Nyos and the probable sources of gas. -
Freeth, Samuel J., George W. Kling, and Minoru Kusakabe. “Conclusions from Lake Nyos Disaster.” Nature 348 (November 15, 1990): 201. Presents a short and not overly technical summary of the state of Lake Nyos and offers suggestions for preventing future outbursts by piping deep lake water to the surface. -
Kling, George W. “Seasonal Mixing and Catastrophic Degassing in Tropical Lakes, Cameroon, West Africa.” Science 237 (August 28, 1987): 1022-1024. Somewhat technical presentation of the evidence that seasonal variations in water temperature and density may have helped destabilize lakes Monoun and Nyos. -
_______, et al. “The 1986 Lake Nyos Gas Disaster in Cameroon, West Africa.” Science 236 (April 10, 1987): 169-175. Presents a thorough, moderately technical discussion of the Lake Nyos disaster. Discusses various possible sources of the gas and accounts for the symptoms reported by survivors. -
Sigurdsson, Haraldur. “A Dead Chief’s Revenge? Scientists Now Understand the Mechanics of the Deadly Cameroon Gas Burst One Year Ago, but the Trigger Is Still a Mystery.” Natural History 96 (August, 1987): 44-49. Presents a nontechnical summary of the disaster and includes good photographs. Title refers to a local folk explanation for the disaster. -
Smith, Keith. Environmental Hazards: Assessing Risk and Reducing Disaster. 4th ed. New York: Routledge, 2004. Comprehensive volume covers all varieties of hazardous conditions in the environment. Includes discussion of the Lake Nyos disaster in chapter 7. Features photographs and index. -
Stager, Curt. “Silent Death from Cameroon’s Killer Lake.” National Geographic, September, 1987, 404-420. A nontechnical account of the disaster and its aftermath from the personal viewpoint of a scientist who had studied the lake not long before the disaster. Includes excellent photographs of the lake and diagrams of the flow of the gas.
Mount St. Helens Erupts
Toxic Gas Leaks from a Union Carbide Plant in Bhopal, India
Eruption of Mount Pinatubo