The same swirly, red-blue image as is shown at the top of the article, but this version has depth measurements added.

The unique makeup of Africa’s Lake Kivu prevents the mixing typically seen in other deep lakes, leading to unusual stratification of the waters. There are distinct density differences between each layer. The sharp transition between two of those layers is shown here, with the lower, warmer, saltier water below (red) and the cooler, fresher water on top (blue). The boundary between the two layers is just a few centimeters thick.

Lake Kivu is one of Africa’s strangest bodies of water. An unusual set of properties make it an intriguing subject for scientists, as well as a potential source of peril and prosperity for the millions of people living nearby.

Kivu doesn’t behave like most deep lakes. Typically, when water at the surface of a lake is cooled — by winter air temperatures or rivers carrying spring snowmelt, for example — that cold, dense water sinks, and warmer, less dense water rises up from deeper in the lake. This process, known as convection, generally keeps the surfaces of deep lakes warmer than their depths.

But at Lake Kivu, circumstances have conspired to block this mixing, giving the lake unexpected qualities — and surprising consequences.

Straddling the border between Rwanda and the Democratic Republic of the Congo, Kivu is one of a string of lakes lining the East African Rift Valley where the African continent is being slowly pulled apart by tectonic forces. The resulting stresses thin the Earth’s crust and trigger volcanic activity, creating hot springs below Kivu that feed hot water, carbon dioxide and methane into the lake’s bottom layers. Microorganisms use some of the carbon dioxide, as well as organic matter sinking from above, to create energy, producing additional methane as a byproduct. Kivu’s great depth — more than 1,500 feet at its deepest point — creates so much pressure that these gases remain dissolved.

Lake Kivu straddles the border between the Democratic Republic of the Congo and Rwanda. This animation, compiled with data from the Landsat-5 satellite’s Thematic Mapper sensor, zooms in on the lake’s location in the volcanically active East African Rift Valley.


This mixture of water and dissolved gases is denser than water alone, which discourages it from rising. The deeper water is also saltier due to sediment raining down from the upper layers of the lake and from minerals in the hot springs, which further increases the density. The result, says limnologist Sergei Katsev of the University of Minnesota Duluth, is a lake with several distinct layers of water of sharply different densities, with only thin transition layers between.

The layers can be separated roughly into two regions: one of less-dense surface water above a depth of about 200 feet and, below that, a region of dense saline water that is itself further stratified, says Alfred Wüest, an aquatic physicist at the Swiss Federal Institute of Technology in Lausanne. There is mixing within each layer, but they don’t interact with each other. “Just think of the entire water mass sitting there for thousands of years and doing nothing,” says Wüest, author of a 2019 article in the Annual Review of Fluid Mechanics surveying convection in various lakes of the world, including weird outliers like Lake Kivu.

But Lake Kivu is more than just a scientific curiosity. Its unusual stratification and the carbon dioxide and methane trapped in its deeper layers have researchers worried that it could be a disaster waiting to happen.

Lurking danger

About 1,400 miles northwest of Kivu, a crater lake in Cameroon known as Lake Nyos similarly accumulates and traps large amounts of dissolved gas — in this case carbon dioxide — from a volcanic vent at the bottom of the lake. On August 21, 1986, the lethal potential of that gas reservoir was demonstrated in spectacular fashion. Possibly due to a landslide, a large amount of water was suddenly displaced, causing the dissolved carbon dioxide to rapidly mix with upper layers of the lake and release into the air. A large, deadly cloud of the gas asphyxiated about 1,800 people in nearby villages.

Events like this are called limnic eruptions, and scientists fear that Kivu may be ripe for a similar, even deadlier event. Nyos is a relatively small lake, measuring a little more than a mile long, just under a mile wide and less than 700 feet deep. Kivu is 55 miles long, 30 miles across at its widest point and more than twice as deep as Nyos. Because of its size, Katsev says, Kivu “has the potential for a major, catastrophic limnic eruption where many cubic miles of gas would be released.”

About 14,000 people lived near Nyos at the time of the eruption; more than 2 million live in the vicinity of Lake Kivu today, including roughly 1 million residents of the city of Bukavu, in the Democratic Republic of the Congo. If Kivu were to experience a limnic eruption, says limnologist Sally MacIntyre of the University of California, Santa Barbara, “it would be completely catastrophic.”

This isn’t just a theoretical concern. Scientists have found what may be evidence of at least one previous limnic eruption at Kivu that likely occurred between 3,500 and 5,000 years ago, and possibly several more recent ones. Sediment cores taken from the bottom of the lake have revealed features known as brown layers that are unlike the surrounding sediments. These sediment bands are “very unusual, organic-rich layers,” Katsev says, that may be the result of eruptions.

Limnic eruptions can occur for two reasons. If the water becomes completely saturated with dissolved gases, any additional carbon dioxide or methane injected into the lake will be forced to bubble out of solution, rise and be released into the air. Eruptions can also be caused when something forces the deep water with its dissolved gases to mix with the layers above, reducing the pressure on the gases and allowing them to quickly come out of solution and escape, similar to the effect of shaking a can of soda and then opening it.

An offshore biogas plant floating in the waters of Lake Kivu

The KivuWatt project, launched at the end of 2015, seeks to pump gases trapped at the bottom of the lake to harvest methane for energy production. Once fully online, the plant will produce more than 100 megawatts of power for Rwanda, helping the country to reach its ambitious goal of providing electricity to the entire population by 2024.


While a landslide of the scale suspected in the Nyos eruption might not cause enough mixing at Kivu, due to the lake’s size and depth, there are several other possible triggers. Kivu is in a seismically active area, so an earthquake could generate waves in the lake that would mix the layers enough to release the trapped gases. Climate is also a potential culprit. At least one past eruption discovered in the sediment record appears to have been caused by drought that evaporated enough water from the top of the lake to reduce the pressure at the lower levels and release the dissolved gases. Lowered water levels during dry periods could also leave Kivu more vulnerable to disruption from particularly heavy rain events. They could flush enough built-up sediment from the dozens of streams feeding into the lake to cause the layers to mix, MacIntyre says.

The chances of such a sequence of events may go up as the planet warms, MacIntyre says. Climate change will bring more rain to East Africa, and “it’s going to come in the form of more extreme rain events with bigger intervals of drought in between.”

Another possible trigger is volcanic activity under the lake or from surrounding volcanoes, but scientists think the risk of that is low. A 2002 eruption of nearby Mount Nyiragongo didn’t bring in enough material to disrupt Kivu’s bottom layers. And modeling studies have shown that volcanism below the lake wouldn’t cause a big enough disruption either, MacIntyre says.

Whatever the culprit, the effect would be the same: Accumulated gases are released from their dissolved state, creating dense clouds of carbon dioxide and methane that, as happened with the carbon dioxide at Nyos, could displace oxygen and asphyxiate people and animals alike. And if enough methane is released into the air at Kivu, there is the additional risk that it could ignite.

Katsev says that the lake is monitored regularly for signs of upticks in gas concentration, so a sudden upwelling “will not catch us by surprise.” More than a dozen seismic stations measure the activity near the lake in real time as well. And in 2001, an effort began to reduce the risk of another disaster at Nyos by siphoning water from the bottom of the lake through a pipe to the surface, where the carbon dioxide is released into the air at a safe rate. Similar efforts are underway at Kivu.

Mining the depths

As gas concentrations rise in Kivu’s depths, so does the risk. Wüest and colleagues found that from 1974 to 2004 the concentration of carbon dioxide increased by 10 percent, but the bigger concern at Kivu is the methane concentration, which rose 15 to 20 percent during the same period.

There may be a way to turn Kivu’s risk into reward, though. The same gas that could fuel a deadly natural disaster has potential as a renewable energy source for the region. In 2008, Rwanda launched a pilot program taking methane from the lake to burn as a natural gas and last year signed a contract to export bottled methane. A much larger program, called KivuWatt, came online in 2015.

The projects pump water from the deep layers of the lake, and as the pressure is reduced on that water, the gases are released. The methane is extracted to be used as fuel, and the carbon dioxide is pumped back down to the bottom of the lake. “They take this gas, ship it via pipeline onshore and burn it the way you would burn fossil fuels to generate electricity,” Katsev says.

This harvesting might help to reduce the risk from accumulated gas in the lake, though it won’t eliminate it. Still, for a lake with that much danger lurking below, anything helps. And for the region around the lake, it could be an important source of energy. Once KivuWatt is fully online, the 100 megawatts of power produced by that project alone will make a significant difference for Rwanda, a developing country that is aiming for universal access to electricity.