2. Volcanic Lakes

An Inherited Violence

In contrast to the slow-motion violence that forms glacial lakes, volcanic lakes are created by decidedly faster processes—processes literally explosive in nature. Of course the lakes that can be found in the various depressions that volcanoes create are formed long after all the Sturm und Drang, and at a much slower pace than the volcanic eruption itself. Nevertheless, an understanding of the lakes that originate in the craters and calderas and behind the lava dams formed by volcanic activity depends on a general understanding of how volcanoes work in the first place. So, we would do well to begin by exploring these violent spewers of magma and ash.

We are all aware of the existence of volcanoes, and lest we forget, we are periodically reminded of their bad behavior by news of a volcano erupting in Hawaii or Indonesia or Iceland. We see video of orange lava flowing into the sea and local residents fleeing towns and villages by any means necessary. But as devastating as recent volcanic eruptions have been, to get a real sense of what a volcano can do, we need to go back a bit in the historical record, which provides ample evidence of the truly destructive power of a volcanic eruption.

Perhaps the best-known example of a killer volcano is the AD 79 eruption of Vesuvius. The event buried the cities of Pompeii and Herculaneum and killed an estimated 16,000 people. Incredibly, we know some of the specific details of this eruption in part from letters written by Pliny the Younger, whose uncle Pliny the Elder perished while seeking to explore the eruption as well as to rescue those trapped at the base of the mountain. Pompeii was completely covered in ash and volcanic debris, the devastated city not rediscovered for over a thousand years. Extracting the story of the eruption was facilitated by the fact that the hardened volcanic debris left cavities where the encased bodies of Vesuvius’s victims decomposed. These cavities have been exploited by archaeologists who filled the spaces with plaster to create three-dimensional replicas of the dead frozen in their death throes. These statues reveal how quickly a volcano can kill, showing people crouched down, attempting to shelter their children, or grasping bags of silver or gold coins.

But as bad as Vesuvius was, volcanoes can be even more destructive. In 1815, Mount Tambora, a volcano in Indonesia, exploded and sent an enormous quantity of rock into the air. The strength of the explosion is illustrated by the distance these rocks were ejected, some as far as 25 miles away. An estimated 20 cubic miles of ash was sent into the atmosphere during the blast. The repercussions were global; so much of the sun’s light was blocked by the ash that 1816 was sometimes called “the year without a summer,” and harvests failed in Europe and North America. The blast was heard over a thousand miles away, and 10,000 people were killed. In the ensuing months, another 82,000 people died on nearby islands due to famine and disease.

Even more destructive, though with a lower death toll, was the eruption of Krakatau in 1883, which literally impacted the entire planet. It is thought that the Krakatau eruption was the loudest sound in history. Heard thousands of miles away, the shock wave was detected by barographs around the world. The eruption caused the deaths of 36,000 people in the surrounding area (most from the tsunami that followed the eruption), and like Tambora, impacted the entire planet via the sheer volume of material it ejected into the atmosphere. When Krakatau exploded, an entire island was destroyed, the volcano sending over 4 cubic miles of rock ash into the air. Eruptions like Tambora and Krakatau distribute dust and debris throughout the planet and to such an altitude that it takes years for it to fall back to earth. Similar to the effects of Tambora, the eruption of Krakatau reduced average temperatures around the world for the next three years. It is estimated that the sun’s intensity was decreased by 10 percent in Europe, and the appearance of the sun, moon, and sky was altered. Among other things, the material in the atmosphere was responsible for fantastic sunsets; some believe one is included in the background of Norwegian Expressionist artist Edvard Munch’s painting “The Scream”.

But if these examples fail to impress, one need only consider supervolcanoes, a less-discussed type of volcano, but one whose explosive power is so large they may be a threat to our very existence. The magma chamber for one supervolcano lies beneath Yellowstone National Park and is responsible for the active geysers and steam and hydrothermal vents one finds there. The last time this supervolcano exploded was 630,000 years ago, which is of some comfort. Considerably less comforting is what it did when it did erupt, which was to cover virtually all of the United States west of the Mississippi in a layer of ash. It left behind a caldera over 40 miles in diameter (it currently comprises virtually all of the park). The massive explosion is estimated to have been a thousand times more powerful than that of Mount St. Helens. Should it erupt today, it is difficult to imagine what life would be like in North America, or to what degree life would exist at all. And, we should be nervous. This supervolcano hasn’t erupted for 630,000 years, but erupts roughly every 600,000 years. As Bill Bryson aptly writes in his book A Short History of Nearly Everything, “Yellowstone, it appears, is due.”

But what does all of this have to do with lakes? How does the explosive force of volcanoes manifest itself in lake formation? The answer is simply that anything capable of creating a depression or hole in the ground can make a lake, and volcanoes are adept at nothing if not creating holes in the ground.

Nor is it just the holes at the peaks of volcanoes, the craters and calderas, that result in lakes. While impressive in size, these are actually small in number. Volcanoes generate lakes in large numbers via several other mechanisms including, for example, the formation of maars and pit craters, located all along the slope of a volcano and even in the flat land some distance from a volcano. The lava flows emanating from a volcano can also create lakes. River valleys are dammed, creating sometimes enormous lakes. Lakes can also form in the lava flows themselves, resulting, as we shall see, in numerous strange depressions that pockmark those solidified flows. Volcanoes create depressions and holes all over the planet, and wherever there is sufficient rain or groundwater or runoff, these holes will become lakes.

Volcanoes are not evenly distributed. In the continental United States volcanic lakes are found primarily in the west, particularly in the northwest where Oregon’s Crater Lake and Hole-in-the-Ground Lake are just two examples. But, as we saw in the last chapter, lakes in North America, particularly in the northern part of the continent, tend to be of glacial, not volcanic origin. In other parts of the world, however, volcanoes dominate the limnological landscape. For example, in Japan the study of limnology itself is focused on volcanic lakes. Most of what is known about tropical limnology is due to the study of volcanic lakes in the volcanically active areas of Java, Bali, Sumatra, and Central America. Volcanic lakes are common throughout New Zealand, Iceland, the Auvergne district in France, and in the Eifel district of Germany. If we focus on volcanoes themselves, about three-quarters of all the active volcanoes on Earth can be found in what is referred to as the Circle of Fire, roughly coinciding with the edges of the Pacific Ocean, running along the Aleutian Islands, down the Kamchatka Peninsula, through Japan, the Philippines, Indonesia and New Zealand and then across the Pacific to the western edge of South America and on up through central America to the West coast of the United States and Canada. But this is only where recent volcanic activity exists. It is almost certainly true that there is no part of the planet that has not been subject to volcanism at some point in the geological record. So, while volcanic lakes are much smaller in number than those caused by glaciers, volcanic lakes can be found over a much wider geographical region.

The study of volcanic lakes lies at the intersection of limnology and volcanology and the resulting terminology and classifications is not entirely consistent between these fields (or even within them). Here, we classify volcanic lakes into two groups. The first consists of those lakes that form in the depressions on the volcano itself, either right at its peak or along its slope. The second group is lakes that evolve from the basins created by the flow of lava (or sometimes mud, and sometimes both) once it has solidified.

The first category includes lakes that are probably foremost in the minds of the popular imagination—a body of water occupying the crater located at the very peak of a volcano. The depressions formed at these peaks are sometimes referred to as craters and sometimes as calderas, a distinction that is far from agreed-upon. Most authors agree calderas are bigger than craters, with a dividing line at about a diameter of one mile. Many researchers typically refer to craters as depressions formed by the volcanic activity proper, that is, the buildup of ash, lava, rock, and other ejecta at the edge of the vent where the explosion emanates. This ejecta builds up to form a bowl-shaped depression that can turn into a lake once the volcanic eruption has ceased (this, as we will see, is never a guarantee). For explosions on the smaller end of the scale, such craters may be the end of the story. The volcano may erupt multiple times at the same location, or at nearby locations, creating a cluster of craters. But if these eruptions are relatively small, the resulting craters will be the defining depression created by the volcano. Once the eruption has ceased, lakes can form in all of these craters.

For larger volcanic eruptions, calderas can form. All volcanic explosions result from the existence of pressurized gas located beneath a vent. This gas may emanate from the magma where it was dissolved and subsequently came out of solution to form a gas. The gas can also be superheated steam, created when the magma encounters water (“phreatic” explosions, described in greater detail below). But in either case, when the gas pressure becomes large enough it explodes through a vent, often pushing out an enormous amount of lava in the process. This lava originates in a magma chamber at some depth beneath the vent and once the volcanic explosion is complete (which can take from minutes to days), the magma chamber is significantly changed. Its pressure has decreased, and, for large explosions, the volume of magma has decreased significantly, resulting in a large, unoccupied volume beneath the overlaying crust.

This process often results in the downward collapse of the crust, resulting in a massive depression called a caldera. Because the magma chamber can be much larger than the cone formed above it, the resulting depression, the caldera, may be far larger than the cone that formed during the actual eruption. Crater Lake in Oregon is an example of such a caldera. Crater Lake is huge (and, really, it should be called Caldera Lake), having a diameter of about six miles and a depth of almost 2000 feet at its deepest point; the distance from the rim of the caldera to the bottom of the lake is about 3900 feet. It is the second deepest lake in North America and one of the ten deepest in the world. Subsequent and less-violent activity has resulted in smaller craters on the floor of Crater Lake, one of which created a small island within the lake, Wizard Island, a visual illustration of the range in size of the structures that volcanoes can create.

Not all volcanic lakes are found at a volcano’s peak. They can also form in depressions on the slopes of a volcano or on the flatter regions nearby. One example of such craters and one that often results in lakes are craters formed by phreatic explosions. These occur when magma rises from below and encounters groundwater. The groundwater flashes into superheated steam, which bursts up through the crust, creating a hole in the ground that may be away from the volcanic peak. Because the cause of such explosions is the groundwater already existing above the magma chamber, the resulting crater is typically filled with water, forming a lake fed by the water table.

Such lakes are called maars and are found throughout the world. They are especially prevalent in the Eifel region of Germany, where they have been well-studied. Maars are typically circular in shape, less than a half a mile in diameter, and less than 500 feet deep. In some cases, the superheated steam released during the explosion results in little lava release and so the debris cone around the lake is minimal, consisting only of the fragmented crust that was ejected during the explosion. This can make the lake difficult to identify, since it lacks the characteristic volcanic rock that surrounds crater lakes. Indeed, simply identifying the debris cone itself can be tricky, resulting in a lake that seems somewhat out of place, a hole in the ground with no apparent origin. Indeed Hole-in-the-Ground Lake, referred to earlier, is such a maar, a bit larger than typical, having a diameter of about one mile.

A lack of a debris cone can be even more pronounced for pit craters, craters that can range in diameter from tens of feet to about a mile and their depth from a few tens of feet to more than a thousand feet. Pit craters are not formed by explosions of any type, but rather by the sinking in of a part of the volcanic surface. The details are not entirely understood, but it is thought that pit craters form when underlying magma recedes downward, perhaps through release at a lower side vent or an opening on the slope of a volcano, thus draining magma from above, the surface collapsing when the magma leaves. It is also possible that pit craters are formed by vertical cracks formed as a volcano settles and that cause a collapse at the surface, or when a magma conduit eats away at the rock above, melting it so that there is only a thin layer of solid crust at the top, which subsequently collapses, forming a crater. Excellent examples of pit craters can be found along the Chain of Craters Road in Hawaii’s Volcanoes National Park. Though a common feature in the study of volcanoes, pit craters are absent in limnology textbooks. It seems likely that many of these have filled with water to form lakes. Indeed, so long as the floor of the pit is below the water table, lake formation seems almost guaranteed. It is possible that since these lakes also lack a debris cone of any kind, they are simply classified as maars in the limnological literature.

The second group of volcanic lakes are those formed some distance from the volcano itself and as a consequence of a lava flow emanating from the volcano. We should note that while an erupting volcano is a pretty incredible thing with its fountain of glowing molten rock that can rise thousands of feet into the sky, lava flows are pretty amazing, too. Though they may not be able to compete with the drama of the volcano they emanated from, lava flows make up for what they lack in drama with their sheer scale, for the lateral size of these flows can be simply enormous. The geological record contains many examples of lava flows that extend for dozens of miles and some for over 100 miles in distance and having thicknesses of over 100 feet. In terms of volume, the material comprising a lava flow can be mind-boggling. Some volcanoes in the Hawaiian Islands are estimated to have produced more than 16 billion cubic feet of lava. In Iceland, the lava flow that was emitted just from the Laki fissure in 1783 alone is estimated to have been 2.8 cubic miles in volume. Indeed, Iceland is so dominated by volcanoes that its bedrock is 90 percent igneous (solidified magma)—the entire island nation is essentially an enormous patchwork quilt of lava flows.

Perhaps the areal extent of lava flows on the planet should be no surprise. The beginning of the rock cycle starts with igneous rock (rock formed from magma), from which all other rock, sedimentary and metamorphic, are subsequently formed. Pick up a rock, and you can safely claim that its constituents were all once some kind of lava. But still, knowing that a vast swath of the planet’s surface was created by flowing rivers of liquid rock can be hard to wrap our minds around. Geological maps of the northwestern portion of the United States (as well as many other parts of the world) reveal vast areas covered with basaltic lava flows. Indeed, it is estimated that the Roza lava flow in eastern Washington State covers 20,000 square miles and comprises 600 cubic miles of rock. It is difficult to imagine a single cubic mile of hot, orange, molten rock, let alone 600 of them.

Given the prodigious amount of material that flows from a volcano, it should be no surprise that these flows can dam up river valleys, thereby forming lakes in a way that is much faster than, but not dissimilar from, the lakes resulting from the buildup of glacial moraine at the end of a valley described in the previous chapter. One example of such a lake is Lake Kivu, which straddles the border of Rwanda and the Democratic Republic of the Congo. This is a particularly interesting lake because what is now Lake Kivu was once a tributary of the Nile River. Lava flows during the Pleistocene dammed up that river, causing the lake to form. Once filled, Kivu no longer drained into the Nile and now drains south toward Lake Tanganyika (itself the second deepest lake in the world at 4826 feet) and ultimately into the Congo River basin, entering the ocean on the far western side of the continent rather than the north as it originally did. Lakes of this form can be found all over the world and include Lac d’Aydat in France, Lake Bunyonyi in Uganda, and Lakes Pankeko and Penkeko in Japan, both of which were actually dammed up by mudslides formed when an eruption occurred in the side of Bandaisan in 1888.

Perhaps more interesting than the lakes formed when lava dams a valley are the lake basins formed within lava flows themselves. Surfaces and edges of lava flows can cool and partially or completely solidify, even as molten lava continues to flow beneath. Thus, when the eruption generating the lava flow ends, some parts of the lava flow may have solidified surfaces, with liquid lava still flowing inside, creating what is essentially a flattish rock tube within which liquid flows. The hotter lava located on the upper slope may drain through these tubes, emptying and spreading out at a lower location and solidifying. Once lava from the upper slopes has completely drained, the inside of these tubes empty, leaving large flat hollow structures. These vacant tunnels may be stable enough to maintain their shape once completely cooled, resulting in tunnels or lava tubes that are sometimes large enough to be explored on foot. A hikeable mile-long lava tube can be found at Lava River Cave in Coconino National Forest about ten miles northwest of Flagstaff, Arizona. In other cases, the draining of lava beneath the partially hardened surface of the lava flow will slump as the relatively soft walls and ceiling of these tunnels collapse. These depressions can become lakes.

Lava flows can be complex. Their braided streams bifurcate and rejoin, changing their geometry as the volumetric flowrate of the lava source changes and as the underlying topography over which they flow changes. So, it is not surprising that the lakes that form from these lava flows have complicated shapes as well. A wonderful example of such a lake system is Lake Myvatn in northern Iceland. This highly irregular lake has numerous islands within it that are also highly irregular in shape. Both the islands and the shore of Myvatn are pockmarked with circular depressions, themselves formed by the slumping process described above. Many of these depressions are filled with water, creating small lakes on the islands within the lake proper, as well as along its shores. It is a complex and magical-looking place, as is characteristic of lakes formed in lava flows.

•  •  •

As we have seen, the process of making a volcanic lake can be a violent one. But can the lakes formed by volcanoes themselves exhibit violence? This is certainly the case when volcanic lakes form in the craters or calderas of active volcanoes. If the volcano becomes active again, all of the water collected in the crater can be suddenly ejected. More often the water in such a lake will drain at a high rate of speed out of one or more cracks that develop in the side of the cone. This results in a hot mud flow called a “lahar” that can be extremely dangerous. There is one crater lake of especial note in this regard, the crater lake in the caldera of the Kelut volcano in Java, Indonesia. Water accumulated in this crater lake has been discharged in 1771, 1811, 1826, 1835, 1848, 1851, 1859, 1864, 1901, and 1919, the 1919 lahar implicated in 5110 deaths. These lahars occurred so frequently that engineers were brought in to design and build a pair of mitigating tunnels, the Ampera Tunnels. These successfully keep the lake drained to a low level, thereby minimizing the death and destruction caused by this particularly lethal volcano.

But what if the volcano is dormant? When viewing a placid lake that occupies the crater of a long-dead volcano, such as Crater Lake in Oregon, it is perhaps tempting to see this as the denouement of this story of volcanic destruction. Such lakes appear calm enough, nothing like the glaring red lava lakes that once occupied these water-filled depressions. And indeed, catastrophes related to lakes in dormant volcanoes pale in comparison to a full-on volcanic eruption. But still, as if inheriting the ways of the magmatic furies that formed their basins, some volcanic lakes seem to have a penchant for death—a tendency to destroy, even without any volcanic activity. And there is no better or stranger example of this than that of Lake Nyos, a volcanic lake located in the West African nation of Cameroon.

•  •  •

On the night of August 21, 1986, Lake Nyos erupted. During that night, the lake emitted not lava, not ash, not hot mud, but instead a massive cloud of cool carbon dioxide gas that silently raced down the slope, killing almost everything below. About a quarter of a cubic mile of carbon dioxide was released from Lake Nyos that night, traveling downhill at close to 45 miles an hour. In the nearby villages 1746 people died, most as they slept. In the town of Nyos itself, virtually every soul died.

Unlike many volcanic disasters, the Nyos event did not occur a thousand or even a hundred years ago. Occurring as it did in 1986, scientists were able to travel to the site within days. Accordingly, we have a detailed picture of what happened at Lake Nyos—one that is both terrifying and strange.

For the few survivors of the disaster, the situation they woke to must have extended beyond terror and into the horribly surreal. Some of the survivors did not wake for two days, and when they did, everyone around them had been killed—their families were dead and their neighbors were dead. Stumbling out of their houses, they could be forgiven for thinking that some otherworldly force had descended upon them and that the entire world had come to an end. Every living thing had died. Their chickens lay dead in the streets. Their livestock lay dead in the fields. The corpses of birds lay scattered randomly about. Even the insects were dead; rescue workers who arrived later noted the silence, the absence of insectile cacophony so common to equatorial Africa. It was not until days later that insect life reappeared, arriving at about the same time as the vultures that came from adjacent areas to feast on the bodies.

For some time the cause of the disaster was unclear. Other than the bodies, everything was normal. The sun was shining. The fields were green. Buildings were not knocked down. Nothing was burned. Initially some suspected a virulent epidemic that left only the few with natural immunity to live. But none of the outsiders and government officials who trickled into the villages became ill. It quickly became clear that something else was at play, and that something else turned out to be carbon dioxide.

The key to understanding the Lake Nyos explosion is to understand how carbon dioxide dissolves in water. All gasses have a certain solubility in water, a limit beyond which no more of the gas can be added. For carbon dioxide that limit is about one liter of the gas per liter of water at atmospheric temperature and pressure. Further attempts to add more gas to the water, for example by bubbling gas into the water, will have no effect. The mass of gas in each of the bubbles will stay the same as they rise and ultimately exit the surface.

The solubility of a gas in water can be increased by increasing the pressure and/or decreasing the temperature. And, high pressure and low temperature is exactly what one finds at the bottom of a deep lake. Like many crater lakes, Nyos is quite deep. At 682 feet, its bottom lies over two football fields in length below the surface. At this depth, the pressure is intense: 20 times larger than that at the surface, a pressure where the solubility of carbon dioxide is 20 times larger than at the surface. Since water can hold a liter of carbon dioxide per liter of liquid at atmospheric pressure, at the bottom of Nyos a single liter of water can hold an incredible 20 liters of carbon dioxide. This is rather a lot of carbon dioxide, and since virtually no light penetrates the nearly 700 feet to Nyos’ bottom, it is also quite cold, likely dissolving yet more carbon dioxide.

But being able to hold a lot of carbon dioxide gas doesn’t mean that a lake will hold that much gas. There must be a source for that carbon dioxide, and for most lakes that source is the air at the lake surface. So, for most deep lakes, the potential to dissolve carbon dioxide at depth is high, but the actual amount of dissolved carbon dioxide will be quite low since there is no effective way for the surface carbon dioxide to diffuse down to the lake bottom. But in the case of Lake Nyos, the carbon dioxide did not come from the surface. Though long dormant, some volcanic activity does exist deep beneath the floor of Lake Nyos, resulting in the formation of carbon dioxide. The gas seeps up through cracks and fissures in the rock, ultimately bubbling up through the lake floor, introducing it precisely where its solubility is highest.

Imagine these carbon dioxide bubbles, rising from the bed of Lake Nyos. As they rise, they encounter water that is very cold and at a very high pressure—water starved for carbon dioxide. And so, if you were at the bottom of Lake Nyos and could somehow see in the murky depths, what you might observe emerging from some crack in the lake floor would be a plume of bubbles rising upward and getting progressively smaller as they rose, the gas dissolving into the water until the bubbles simply disappeared. Imagine being on the floor of Lake Nyos, surrounded by numerous columns of bubbles rising upward and gradually disappearing, much the way the steam above your coffee cup disappears as it rises through the air above. What would be especially interesting about watching these bubbles disappear would be the dreadful knowledge that as each bubble dissolved, the lake would be getting closer and closer to disaster.

The process of dissolution of carbon dioxide into deep, cold water can go on for a very long time, perhaps a hundred years, a time when nothing untoward may happen. But bit by bit, the lake will become unstable. Year after year, decade after decade, the continuous dissolution increases the potential for disaster. Let’s once again imagine we are at the bottom of Lake Nyos, this time with a bottle. If we fill that bottle with water at the bottom, seal it, bring it to the surface and then open it, the water will violently effervesce, foaming explosively out of the bottle. This is because water that is saturated at the pressures found at the bottom of Nyos becomes supersaturated at the far lower pressure of the lake surface. The carbon dioxide in the water at the lake bottom wants to be in gas form when suddenly exposed to the lower pressure at the lake surface, causing rapid degassing of the water in the same way that a bottle of soda may effervesce when opened. Now, what if instead of a person going to the bottom of the lake and bringing the bottom water to the surface, something else destabilized Lake Nyos, causing some of its deep water to move upward into the warmer, lower-pressure layers of the lake? Should this happen, the upwelling water would release its dissolved carbon dioxide in the form of bubbles. This sounds harmless, but in fact it would be catastrophic.

Nobody is sure quite what caused the bottom water to move upward in Lake Nyos. Some have hypothesized that small seismic tremors were the cause, or perhaps a sloshing of the water in the lake due to a wind of just the right speed. Others have posited that steam explosions deep beneath the floor of the lake (a form of phreatic explosion, described earlier) were the cause, or perhaps some kind of underwater landslide. Whatever the cause, some of the cold water from the depths of Lake Nyos did move upward on the night of August 21, 1986, and once this initial perturbation occurred, disaster was only seconds away.

As the cold lake water moved upward, its pressure dropped, and its temperature increased as it encountered the warmer water residing above it. This caused some carbon dioxide to come out of solution and form bubbles. By itself, this wouldn’t have resulted in a disaster. However, the rising bubbles entrained behind them some of the liquid beneath, which in turn became supersaturated as it rose, releasing more bubbles. These bubbles entrained yet more deep water generating yet more bubbles. Thus a feedback cycle formed, each bubble resulting in the formation of other bubbles, each in turn creating more. Such processes go by a variety of names: self-reinforcing processes, positive feedback cycles, exponential growth, and others. Another term for this type of process is explosion. The cold, saturated water that had rested comfortably at the bottom of Lake Nyos for untold years rose to the surface at an explosive rate, causing massive quantities of carbon dioxide to come out of solution all at once. Though deadly, the gas release was relatively quiet. Those who survived reported little more than a rumbling, the sound of something like a distant explosion or rockslide. Most probably heard nothing at all.

Compared to the death tolls that are associated with volcanic explosions, those found at Nyos may seem small. But considering the absence of lava and fire and the darkening of the sun by ash—the fact that these deaths were due to nothing more than an invisible gas, the death toll was staggering. According to the United Nations, 1746 people died. At least 300 people ended up in the hospital, 3000 people were displaced, and 3500 head of cattle were killed. In the town of Nyos itself, there were only four survivors. To give a sense of the volume of gas released, some of the dead were found in villages as far as 12 miles from the lake; dead cattle were found at elevations as high as 330 feet above the crater rim.

That anybody survived at all is a miracle. Perhaps it was the oddities of wind and air flow that enabled some to receive enough oxygen to survive. Perhaps others were located in dwellings with minimal ventilation so that the carbon dioxide didn’t come inside. Perhaps the survivors were in a high location in their dwelling where, like a person trapped in a sinking automobile, there was a bubble of air that allowed them to survive until the carbon dioxide dissipated.

Lake Nyos is now actively studied, and a degassing strategy has been implemented, aiming to prevent a repeat disaster. However, there are other crater lakes in Cameroon. It is estimated that there are 44 of these in Cameroon’s Northwest Province alone, where carbon dioxide may be building in the depths. A similar explosion occurred at Lake Monoun, also in Cameroon, in 1984, where 37 people died. It is unclear if there are more carbon-dioxide lake explosions in Cameroon’s future or whether such conditions may exist in crater lakes in other parts of the world.

But as bad as the Nyos explosion was, it should be noted that carbon dioxide, while lethal, is not combustible. The same cannot be said for Lake Kivu, the lava-dammed lake described earlier in this chapter. Kivu lies on the opposite side of Africa from Cameroon, bordering Rwanda and the Democratic Republic of the Congo. Like Nyos, gas is released from the floor of Lake Kivu, but in this case, that gas is not just carbon dioxide, but also methane, basically natural gas, like what is used in stoves and furnaces. Exactly what the consequences and risks this has for the people who live around Kivu is unclear. But one should probably take note of two facts: natural gas burns, and Lake Kivu is more than a thousand times larger than Lake Nyos. Due to Kivu’s size and the high population density along the shores of the lake, some have stated that a Nyos-like event at Kivu would result in one of the largest natural disasters in human history. Volcanoes, it seems, present a threat to humans long after their eruptions have echoed into history, continuing to threaten lives along the cool shores of the lakes they create.