Small or huge, with a color capable of covering practically all shades of turquoise. Crater lakes are among the most fascinating phenomena that volcanoes can boast. Evocative mirrors of water, which extend as far as the eye can see in vast depressions or which fill perfectly circular basins, on the hills or on the peaks of the highest volcanoes.
But are these lakes really only formed on volcanoes? And are they dangerous? To answer these and other questions, we had a chat with Dmitri Rouwet, an INGV volcanologist and a great fan of these evocative events, who guided us to discover the most famous crater lakes in the world.
Dmitri, what is meant by "crater lakes"?
As can be easily guessed from the name, these are lakes that form inside craters. Despite what one might think, however, not all crater lakes are volcanic lakes (think of the lakes that form in craters produced by meteorite impacts, for example), and vice versa.
How are they formed, then?
Leaving aside the example of the meteorite I just gave, these lakes usually form from volcanic eruptions that create a perfect crater, a holewhich is then filled with rainwater. We often find these lakes on top of volcanoes, filling the main crater of the volcanic edifice, but even in this case it is not a golden rule, there are exceptions. Think of Lake Albano, on the outskirts of Rome: it's a volcanic lake, but it's not exactly on the top of the volcano. However, it is considered a crater lake because the basin in which it was formed is actually one of the volcano's craters.
Non-crater volcanic lakes, on the other hand, can also form on the sides of a volcano, or in any case in a volcanic area which is not necessarily the crater: an example is that of the lava flows which 'close' the valleys (as if they were real dams) allowing the formation of a body of water.
Let us think, then, of the lakes found in the maar, i.e. inside cavities originating from eruptions in areas such as monogenetic volcanic fields. These environments are particularly suitable for the formation of lakes since the eruptions that give rise to the craters do not generate volcanic cones, but only depressions, holes in the ground that reach the underlying aquifer which, in turn, gives shape to the lake basin.
In any case, it should be remembered, however trivial it may seem, that these lakes are formed only where there is ample availability of water, whether it is rain or coming from an aquifer: in areas where neither of these two conditions occurs, the craters fail to fill and, therefore, there will be no crater lake.
What information can crater lakes tell us about the underlying volcanoes?
I would say many, and it is precisely this aspect that makes them particularly important and interesting. Some of them, in fact, allow us to look in the aquifer below a volcano, giving us a 'window' into the volcano itself. And this is true both for active volcanoes (think of hydrothermal systems), and for i maar. Therefore, instead of drilling a well, researchers simply take water samples from the lake: logistically it is a huge advantage!
There are also various possible scenarios: you can analyze, for example, crater lakes located on top of active volcanoes that degas, and in this case the lake waters will absorb almost everything that is emitted by the volcano. Monitoring and studying samples of these waters over time allows us researchers to constantly follow the behavior and state of the volcano under the lake.
In general, then, we can say that the more active a volcano is, the more frequently sampling and analysis of the waters of the crater lake located on its summit should be done.
Isn't the presence of crater lakes directly connected to the state of activity of the volcano?
No, these lakes can form both on active volcanoes and on extinct or quiescent volcanoes. It all depends, as I said before, on the amount of water available. Think of the crater of the Fossa di Vulcano, in the Aeolian Islands: it would be a perfect basin for the formation of a crater lake, however the absence of consistent rainfall (or underlying aquifers) does not allow it. If it rained more and if a lake could be formed, the gases emitted by the fumaroles could be absorbed by the water and, instead of sampling the fumaroles, we could take water samples: the story that the samples would tell us would be almost the same, with advantage that the water preserves a state of outgassing of the volcano for a variable period depending on how large the lake is and how 'pushed' the outgassing.
Are crater lakes dangerous?
They can be, yes. This is because the combination of magma and water generates potentially explosive, so-called phreato-magmatic eruptions. In the absence of magma, on the other hand, steam eruptions can be generated, real phreatic eruptions, particularly common in the case of crater lakes located on rather 'lively' volcanic systems. The latter are very dangerous and, above all, unpredictable, very difficult to monitor.
Still another case is that of eruptions so powerful as to "expel" the entire lake water outside the crater: this water is dispersed along the sides of the volcano and, if it is in large quantities, it can generate lava, or potentially destructive mudflows.
Finally, a final danger scenario for these lakes concerns those located on non-active but degassing volcanoes (that is, they constantly release gases in variable quantities, in particular carbon dioxide, CO2): in these cases, if the lake is deep enough, CO2 can accumulate on the bottom of the lake until it reaches a level of supersaturation and generates a higher pressure than that exerted by the overlying water column. This can lead, in fact, to an explosion that "releases" the gas accumulated on the bottom.
The levels of gas accumulated on the seabed can only be investigated by introducing probes into the water and taking samples at depth. The interesting aspect is that these gas levels on the seabed can be very high even in the case of lakes that are "normal", silent on the surface. Two examples are Lake Albano, near Rome, and Lake Averno, in Pozzuoli: in the case of Lake Albano, in fact, there are well-known carbon dioxide outgassing events, often associated with seismic swarms. Fortunately, however, in the modern age these accumulations of CO2 have never been such as to reach the explosion point.
An event which, however, occurred in 1986 at Lake Nyos, in Cameroon. In that case there was a sudden release of carbon dioxide which, being denser than air, stopped at ground level sliding into the valleys around the lake and killing about 1.800 people by asphyxiation during the night. The gaseous upheaval in the waters of the lake led the basin to turn red (due to the iron deposited on the seabed which was suddenly brought back to the surface by the rising CO2), but apart from this there was no evident sign of a real volcanic eruption own: only the victims remained to bear witness to the danger that a volcanic lake can have.
The Lake Nyos event effectively inaugurated a new season of study of these phenomena which, as we have seen, can even be lethal, with the aim of getting to know them better and of implementing monitoring and prevention actions that can protect the life of those who live near these lakes. In doing so it was discovered that Lake Albano is also subject to similar dynamics, with the enormous advantage, however, of being located in a climatic zone such that during the winter the surface waters of the lake cool down, becoming denser and descending in depth, so as to mix with the deeper ones and dilute the accumulations of carbon dioxide. In this way the CO2 never reaches saturation levels such as to trigger a violent explosion.
What are the most famous crater lakes in the world?
Those located in tropical areas, where rainfall is sufficient to fill the craters of active volcanoes. In Indonesia, for example, there is the Kawah Ijen, which is the largest acid lake (ie located on an active volcano) in the world; in New Zealand, Ruapehu and White Island, which erupted in 2019; in Central America the Santa Ana in El Salvador, El Chichón in Mexico, the Poás in Costa Rica (very active between 2006 and 2017 with phreatic eruptions); in South America some lakes on the border between Argentina and Chile, the Copahue and the Peteroa.
We must also mention those present in Japan, the Yugama lake of the Kusatsu-Shirane volcano and the Yudamari lake of the Aso volcano: we are not in the tropics but, conversely, here there is snow which is in any case an excellent meteoric recharge for the formation of crater lakes.
…and in Italy?
Definitely Lake Albano, which we have mentioned several times. Then, if you wish, the Fangaia, at the Solfatara of Pozzuoli, the two lakes of Monticchio, in Basilicata, and the lake of Averno at the Campi Flegrei. In our country, due to the low humidity of the climate (which, therefore, does not generate sufficiently copious rainfall) we do not have crater lakes located on active volcanoes (for example in the Aeolian Islands...).
What is, for you, the most scientifically interesting crater lake among those you have visited?
I worked on Poás, in Costa Rica, during the last cycle of phreatic eruptions, and I must say that among those active it was the one that struck me the most. The work on the volcano with local colleagues was very interesting: we identified precursors in the chemistry of water, accumulating a great deal of data and information which will be fundamental for elaborating the next developments in research in this area.
The very first one I worked on during my PhD, on the other hand, was El Chichón in Mexico, which has a less 'lively' activity than Poás but which has other peculiarities such as the springs on its side that make it unique in its own way.
…and the most suggestive?
Kawah Ijen (in Indonesia) and Ruapehu (in New Zealand) are truly beautiful lakes: huge, majestic, evocative. Real icons for a fan like me!
In-depth link on the INGVvulcani blog:
Cover: The world's largest hyperacid crater lake, Kawah Ijen, Java, Indonesia (September 2014, photo by Dmitri Rouwet).