Pyroclastic Flows


Pryoclastic flows are probably the second best-known aspect of volcanoes after lava. What exactly are pyroclastic flows, though? What’s the real definition? How do they form and why do they travel so far? What kind of devastation do they leave? How? What can be done in the future to save lives? These are all questions that will be dealt with in the upcoming article.

So what is the definition of a pyroclastic flow? Well, unfortunately the field of volcanology gets muddled down in terminology around the subject. Let’s start slashing our way through this jungle of words. Probably the oldest term on the subject would be “Nuee Ardente” which means “incandescent cloud” in French. This is a cloud of hot debris that sweeps down the side of the volcano and devours anything in it’s path. This is virtually the same thing as the more modern word “Pyroclastic Surge.” A “Pyroclastic Flow” is technically a ground-based debris flow (mostly without the cloud) composed of extremely hot material. A debris flow can describe Pyroclastic Flows as well, but usually refers to cold debris cascading down the side of the volcano. Of course that’s a lot of vocab to remember…especially when it’s not really that necessary. Personally, I like the pop-culture definition of Pyroclastic Flows: a volcanic debris flow/density current following topography. That definition includes nuee ardentes, pyroclastic surges, and true pyroclastic flows. In short: burning hot gas, ash, and boulders come cascading down the side of the volcano to envelop anything in its path.

As you should have already realized, these types of eruptions are extremely dangerous. The surges and flows can extend miles from the summit and can be instigated by a number of different mechanisms. They can be unpredictable in their formation and behavior. Let’s leave this introduction and look to the backbone of the subject: the mechanics of how pyroclastic flows form and behave.



Pyroclastic flows can form in a number of different ways, which include: deformation-induced landslides, flank collapse/slope failure, lateral blasts, and ash column collapse. Deformation-induced landslides and flank collapse/slope failure are intimately linked methods of formation in the production of pyroclastic flows, so these will be discussed in tandem. Deformation of volcanic slopes is fairly common in active volcanoes, especially among those awakening from a long period of dormancy. The primary cause of this is the growth of a magma body within the volcano itself. The eruption of Mt. St. Helens is the perfect example of this, where a second small magma chamber, often referred to as a cryptodome, formed near the summit of the volcano. This body of magma grew in size, causing a large bulge to form on the side of the mountain. This bulge continued to grow, with large cracks forming in it as the fairly brittle earth attempted to deal with the new formation. Finally, the ground gave way in a massive landslide; the bulge had grown to a point at which it was no longer stable, and the whole thing collapsed. As this enormous landslide cascaded downslope, it removed the earth that had covered the cryptodome. The rapid depressurization of this magma body resulted in its literal explosion, and the formation of an enormous lateral blast. In some pictures, it appears that the entire side of the volcano completely exploded and this is not far from what happened. Another example of this collapse would be the growth of lava domes. When an eruption has progressed for some time, the lava composition evolves and so, consequently, does the eruption style. Eventually the degassed magma will rise from the vent much like toothpaste squeezed from the bottle. The magma will be extremely viscous and will rise and grow in potentially unstable arrangements. As the dome grows, it may form large pillars rising from the top or may out grow the crater and begin to overlap the slopes. In situations where a pillar forms (more often called pinnacles), the needle-like projection eventually breaks, depressurizing the interior which may explode, and then flow down slope. Much more catastrophic, when the dome outgrows its crater and begins to flow over the edges of the slope, a piece of the dome may break off and flow downslope. Like the breaking of the pinnacle, this shattering of debris results in the rapid depressurization of the interior gas-rich magma which, in turn, explodes. This style of dome eruption is usually much more dangerous and larger than the pinnacle collapse.

These aren’t the only style of landslide-style pyroclastic flow. There is another method of formation, which isn’t as impressive, but is still important. Here, debris that has broken off from the summit non-explosively cascades downslope, causing a landslide of hot ash from the slope to join it and form a fairly small, but none-the-less dangerous, pyroclastic flow. This doesn’t involve any sort of blast, however, so the amount of energy it incorporates is naturally lower than most types. As you’ve seen from the previous descriptions, lateral blasts can occur together with other events. It can happen on its own, but it more often occurs as a result of some sort of slope failure, as described above.

There is another method of formation that is extremely dangerous, difficult to predict, and far from obvious for those who are not familiar with large eruptions. This style is called ash column collapse. In truly massive plinian eruptions, the rising ash column is supported and raised by the convecting currents of air surrounding it. Occasionally, however, the column can become “overloaded” and will collapse in on itself. I won’t repeat the entire section from the “characteristics of ash” article, but the column does not rise by its own force, but through the supplied energy of convecting hot air that engulfs the column. If the column becomes overladen with ash, and thus too dense/heavy for the convecting air to carry, a section of it may collapse catastrophically back to the earth. It will move directly down from the column and down the slopes of the volcano. This style of pyroclastic flow contains few, if any, boulders, but abundant hot ash and gas. It can be very difficult, if not impossible, to predict this type of event before an eruption begins. In fact, this style of formation is very poorly understood at present: its presence was only first confirmed during the eruption of Mt. Pinatubo during the mid-nineties. On-going studies are being conducted through use of computer simulations; because these events only occur during very large eruptions, they are rare occurrences and consequently difficult to observe first hand.

Transportation, Devastation, and Mitigation

Nothing in life is simple, and so follow volcanoes. The transportation of pyroclastic flows is not such a simple thing as “hey, that gas moves down until it stops.” Yeah, it’s a bit more than that. We often like to think of things like a solid block moving down a slope. The base has friction and slows the whole thing. Gasses don’t behave like that, however, and neither do pyroclastic flows. Pyroclastic flows are gravity driven and as they move down the slope of a volcano, the base is slowed by friction with the ground. Rather than slowing the whole body, however, the rest of the flow rides atop this layer like an air cushion. Likewise, the top of the flow is slowed through contact with the surrounding stagnant air. Thus, the flow travels fastest toward the middle and slower on the outside edges. This allows the flow to travel much farther than it otherwise could, resulting in greater devastation along the volcano’s flanks. If a pyroclastic surge reaches water, however, the effect is made more dramatic through the flash-boiling of the surface water. As the surface contacts the pyroclastic surge, a layer of vaporized water forms, allowing the surge to flow overtop of it much like a puck across an air-hockey table. Krakatau’s eruption, in 1883, created pyroclastic flows that traveled as much as 80 km over the sea!

The transportation of pyroclastic flows, however, is by no means fully understood. While we know that larger particles drop out first, much like the energy conserving behavior of lahars, some pyroclastic flows move like a fluid…even when they are composed entirely of rocks and debris (no gas and little ash). The reason for this is still being actively investigated, but we know that vibrations of certain frequencies can liquefy loose particles. Also, while we can theorize all day about what the interior of a pyroclastic flow or surge looks like, there is no way to check at this time due to the extreme conditions present.

Devastation         Warning: this section contains graphic descriptions and is not for the faint of heart

The power of pyroclastic flows can be awe-inspiring. The momentum of the gas and debris can level trees and buildings alike. The extremely high temperatures add to the destruction by incinerating anything flammable in its path. For specific instances in which this hazard has struck, see the article on historic disasters.

The pyroclastic flow (though here, we’ll be talking about surges more) loses energy as it moves down slope. The flow is slowed through friction, loses energy as the slope levels (remember, the flow is gravity-driven), and the surge is diluted by the surrounding air. Thus, the further a city is from the source of pyroclastic flows, the safer it is; this does not necessarily make it safe, however.

Most people killed in pyroclastic flows die either by being crushed by debris or burned by the hot ash and gas. Trees, cars, and parts of buildings may be picked up and thrown like toys during the intense passing of a pyroclastic flow. People outside may be crushed by such debris while people indoors may be buried under a collapsing wall or roof. As terrible as it sounds, these people may be considered the lucky ones. Those not crushed usually succumb to the intense heat. Gas and ash, as hot as 660 degrees Celsius, can instantly ignite clothes, hair, and exposed skin…only to be extinguished by the lack of oxygen. These are later reignited when oxygen reenters the area. The burning gas and ash are breathed in by victims, most of whom die after a few short breaths. The lungs are either seared or are choked to death by massive quantities of ash. Some victims in the city of Saint Pierre, which was devastated in 1902, withstood days of agony before dieing due to burning of the lungs. Bodies were found in contorted positions, desperately gasping for oxygen. Once the cloud passes, however, the trouble is not over for survivors. The freshly laid deposit of ash and debris is extraordinarily hot and will burn through shoes for people trying to escape. Moreover, with the return of oxygen and the presence of burning ash, buildings are soon ignited and raging fires may cause further devastation.

While all of this sounds, and quite frankly is, horrible, there remains plenty of room for survival if adequate measures are taken. The following section contains quotes taken from the article Hazards from Pyroclastic Flows and Surges, by Setsuya Nekada of University of Tokyo.

A pyroclastic flow descended on Unzen volcano, in Japan, on June 3, 1991. The following are accounts regarding survivors. “Another person opened his door soon after the surge hit his house. He suffered burns on his arms where he had no sleeves. Then glass in upstairs windows was broken and a mass of ash invaded the house. The hot ash caused burns to his stomach. Houses nearby were burning as he fled. He left the house without shoes, and felt intense heat as he stepped on the ash deposits. His house was still burning the next morning.”

“A man took refuge in a car as he was engulfed by the surge. He saw that small pebbles hit the windscreen, crackling and glowing in the darkness and he felt intense heat in the darkness. He suffered some burns on his neck and back and it was later found that the surges had melted the vinyl cover of the car seat.”


As you might imagine, attempting to prevent the tragedies from happening is a huge priority. Predicting a pyroclastic flow may be difficult in some situations, but taking proper preventative measures beforehand can reduce the need for a perfect prediction. Where a dormant volcano is reawakening, evacuating people to a distance thought safe will usually be effective enough. Excellent examples of this can be found at the most recent eruptions of Pinatubo and Montserrat where thousands of lives were saved.

By hazard mapping low lying areas and river valleys, where pryoclastic flows are likely to occur, we can evacuate people more effectively when the need arises. Likewise, by keeping an eye on growing lava domes, we can predict where these flows are going to occur and evacuate people accordingly. With the number of people living around active volcanoes, however, the logistics of evacuating everyone and the likelihood of developing a 100% accurate forecasting system don’t look too good. While forecasting these events is becoming much more effective, growing populations mean it’s more difficult to get everyone out of the way. In some situations, where threatened towns lie a greater distance from the volcano, construction of flame and impact resistant houses together with shatter-proof windows may dramatically improve the survivability of these disasters if one strikes without an evacuation. While we’ll probably never see a perfect forecasting method or a completely survivable pyroclastic flow, we can continue to improve our safety record with a hazard that is going to be here for a long, long, time.


By reading this article, you hopefully have a better understanding of the way pyroclastic flows and surges behave. We’ve looked at the numerous ways through which they are formed, as well as the manners in which they travel. We’ve discussed the dangers posed and steps that can be taken to alleviate the danger. While pyroclastic flows remain one of the most dangerous volcanic hazards, it is important to note that lives can be saved through education and vigilant monitoring. This hazard is far from unavoidable. For information on related subjects, I would encourage you to continue reading articles on this site and published articles elsewhere.