Distinguishing deposits from andesitic eruptions

Telling apart different kinds of deposits associated with volcanic eruptions isn't always easy. There are a lot of factors that can affect their appearance: the location and type of eruption, the magma/lava type, where they're emplaced, etc. On Montserrat, volcanologists are lucky to have both ancient and modern deposits; they can look at what's currently being erupted and compare it to the older volcanics on the island. We did quite a bit of this on our field trip, and one of our assignments was to summarize the characteristics of andesitic eruption deposits on Montserrat. (I emphasized that because the characteristics we saw are not necessarily going to be the same for all volcanic eruptions, or even for all andesitic eruptions. I've tried to generalize a bit, but apply these cautiously if you're going into the field to look at other volcanic deposits; things may look very different in your field area.)

Block and ash flows are a kind of pyroclastic density current. "Pyroclastic flow" is a kind of catchall term, but there are more specific ones that better describe the makeup of one of these currents. "Block and ash flow" implies that the flow is composed of blocks (either of denser lava or pumice or both) and ash; "pumice flow" means that the contents are mainly pumice and ash, "ash flow" that there are few blocks and mostly ash in the current.

Faint reverse grading in a block and ash flow deposit at Old Road Bay

Block and ash flows contain a range of clast sizes from mm-sized ash to m-scale boulders. The clasts are usually somewhat angular and tend to be a combination of lithics (lava) and pumice. The matrix of these deposits (what larger clasts are embedded in) is generally ashy and may be crystal-rich. The deposit may be clast or matrix supported (which describes whether there is enough matrix that the clasts are not touching) and poorly sorted (clasts are not separated by size). Oxidation and alteration of clasts as well as fractured blocks (including radially jointed clasts) may occur if the flow is deposited into water and cools quickly. Sometimes you can see grading in the deposits (i.e., a change from small to large clasts, or the reverse), which has to do with the energy of the pyroclastic current and the conditions under which clasts are being deposited from it.

Pointing out matrix in a deposit on the side of the Belham River Valley

Debris avalanches form from older volcanic material that's unstable enough to collapse, either because of erosion or alteration or both. They form "hummocky" deposits that are very obvious in aerial photos, but in cross section they're also fairly easy to identify.

Panorama of a debris avalanche deposit near Jack Boy Hill. Different colors mark chunks of preserved stratigraphy.

Debris avalanche deposits contain mm to m sized clasts (or larger), as well as chunks of material from older deposits that retain their original stratigraphy. These clasts and chunks are poorly sorted, and rounded to angular in shape. Clast compositions are almost always mixed, and the matrix is often clayey and highly altered. Jigsaw jointing is apparent in individual clasts, with clast fragments separated by matrix material but orientation with respect to neighboring clasts preserved. Clasts or deposit blocks may be "smeared" out into trains or show internal faulting.

A "jigsaw" fractured block in a debris avalanche deposit

Lahars form when water mixes with volcanic material and flows downslope. These can have the consistency of soup to concrete, and they're a major concern even when a volcano isn't erupting (especially if the volcano is located in a tropical area that gets lots of rain.)

House buried in lahar deposits in the Belham River Valley

Clasts in a lahar deposit range from mm to multi-meter size. They are matrix supported, poorly sorted, and often contain clay or silt in the matrix. The clast shapes vary from rounded to angular and may be either monolithologic (all one rock type) or mixed rock types, depending on the source material. Sometimes the deposits show bedding features such as cross-bedding. There is one absolutely diagnostic feature for lahars: preserved voids where bubbles of air were trapped in the matrix mud. The voids are spherical and usually pretty tiny, but they don't form in pyroclastic currents. (The lack of these voids does not, however, mean that a deposit is not a lahar, so this feature is only useful if it's present.)

Cross-section of a fairly recent lahar deposit. There's not much in the way of sorting or grading here.

A fourth type of pyroclastic current, called a surge, also shows up on Montserrat, but they're not often preserved because they consist of a blast of ash and hot gases (sometimes derived from block and ash flows), and they leave very thin deposits. I don't have any good photos of them, but they tend to be fine-grained (mm to cm sized clasts at the most), and sometimes show cross-stratification. It's rare that they're preserved, especially in a tropical environment, because the fine material washes away very easily.

Volcano Vocab #4: Lahar

As suggested by a commenter on the last Volcano Vocab post, here's a water-and-volcano-related term for you: Lahar ("lah-haar"). 

Lahar is an Indonesian word for a mudflow of volcanic material - that is, a mass movement of volcanic debris that contains some amount of water. (A dry flow of volcanic material would usually just be called a debris flow or debris avalanche.) The key thing that distinguishes a lahar from a "regular" mudflow is the presence of volcanic material in the flow, which can include tephra, ash, hydrothermal alteration products, blocks of lava flows, and other pyroclastic materials. Some descriptions liken this mixture to a flow of cement, and it's capable of moving house-sized boulders huge distances from their source.

Lahar deposits characteristically show poor sorting (lots of different sizes of material, from boulders to sand), multiple rock types, rounded clasts, and muddy matrix supporting the clasts. (Sometimes in deposit matrices you can find rounded voids where bubbles of air were trapped as the deposit hardened around them!) Lahars are most common on stratovolcanoes, but (as we've seen in Iceland), other types of volcanoes can also create the correct conditions to form a lahar. Lahar formation depends on having lots of loose material, and the addition of lots of water over a short period of time (such as from a melting glacier, a hurricane or storm, or a breached crater lake, among other things). They do not require the volcano to be actively erupting, which is one reason why they are so dangerous.

Here's a video of a lahar from Mount Ruapehu in New Zealand (March 2007):

Lahars are an especially dangerous volcanic hazard because they appear and disappear so quickly. Lahar debris that's deposited in a riverbed can easily be eroded by normal river flow, and lahars that have spread beyond valleys and drainages can easily become reclaimed by vegetation. In fact, one of the most devastating lahars to have come from Mount Rainier in Washington State (the Osceola Mudflow)  is now covered with small towns. Because the people in the area are now aware of the danger a repeat lahar would present, they are required to hold evacuation drills. Lahars are somewhat easier to monitor and avoid; acoustic flow monitors (specially calibrated seismometers) can be placed in source zones, and a timely warning can be sent downstream when lahar signals are detected. Sometimes evacuation can be simply a matter of climbing to a higher elevation, although on a floodplain it could be necessary to travel much greater distances to safety, but it does require advance warning.

Armero, Colombia, destroyed by lahar on November 13, 1985. Photo from the USGS CVO website.

Lahar warnings aren't always heeded, however. It's impossible to mention lahars without also mentioning the town of Armero in Columbia. On November 13, 1985, a small eruption of the nearby volcano Nevado del Ruiz melted part of the snow and ice capping the volcano's summit, and produced a lahar. Volcanologists knew that river valleys on the volcano's flanks could channel lahars toward populated areas, and sent warning to towns in the lahar's path; unfortunately, local officials either received incomplete or conflicting information, and/or decided not to listen to the scientists' warnings. As a result, more than 23,000 people were killed in Armero and nearby villagees, when they could have reached safety by climbing only a short distance up the slopes on the sides of their valley. This tragedy drove a USGS scientist to develop the Acoustic Flow Monitors mentioned above, in hopes that more deaths could be avoided; the system is now used at lahar-prone volcanoes worldwide.

Looking backward: Past eruptions at Volcán Santa Maria

On our way to visit the Santiaguito Volcano Observatory, Gustavo Chigna of INSIVUMEH (the Guatemalan equivalent of the USGS) was kind enough to take an afternoon off and show us some of the older deposits near Santiaguito. Our first stops were at an exposure of the air-fall deposit from the October 24, 1902 eruption of Volcán Santa Maria. This eruption was a devastating one, stripping the land for more than 50 km around the volcano, burying villages and fincas (plantations) in more than 3 meters of ash, mud and rock, and killing more than 7,000 people (the exact number will probably never be known). The area had already experienced months of earthquakes prior to the eruption, and activity at the crater formed in Santa Maria continued for weeks afterward.

This location is about 4 km from the volcano. The 1902 air-fall deposit here is more than 2 meters thick (the photo is only showing the top bit, and Gustavo is standing a few meters behind the outcrop, so the sense of scale is a bit wonky. The big clast in the center is about fist-sized, though.). This fall deposit contains a mix of lithics (old and new lava and country rock) and pumice, and what's really impressive is that many of the lithics are the same size as the pumice (up to 15 cm). Generally, in a fall deposit the material will have had the same terminal velocity, which means the mass of the pumice and lithics should be roughly equal. Because pumice is much less dense than lithics, this usually means that the lithics are much smaller than the pumice. Possibly the lack of difference in clast sizes here has to do with the proximity to the volcano; in other places that we stopped, the pumices were definitely larger than the lithics.

The view from this location was pretty spectacular, and it highlighted the deposits from the November 2, 1929 dome collapse. Here you can just see the summit of Santa Maria, and the El Brujo and other domes in front of it. Brujo is the dark green dome farthest to the left, and the hummocky low hills in front of it are may be the result of that 1929 collapse. Hummocks are a classic sign of a debris avalanche of some sort, and can be seen in places like Mount St. Helens and Mount Shasta. ***Note: After speaking with Rudiger Escobar, a Santiaguito expert and fellow volcanologist (see the comments section), I wanted to note that the hills pictured here probably have more to do with lava flows coming off of Brujo than older collapse deposits, although some portion of them may be collapse remnants. It's also possible that I have faulty notes, as the conversation was mostly being conducted in Spanish at the time, and my Spanish skills are limited. 

After a bit of backroad driving around the fincas, we reached the Rio Tambor, which was the site of many deaths in the 1929 dome collapse. The 1929 event, in which more than 3 million cubic meters of the lava dome collapsed, was mostly described by foreign geologists who interviewed survivors. Their accounts tell of glowing rains of ash and rock; boiling mudflows in the rivers covered with rafts of glowing rock; a blast that rushed back toward the volcano after it blew down the river valleys, scouring the south side of tree trunks; and people suffocating on gases and hot air. More than 3,000 people were killed, mainly because they were on the fincas for a religious holiday rather than in their villages. Sadly enough, this was a very  minor dome collapse by volcanology standards (a large one might be in the hundreds of millions of cubic meters, such as those at Soufriere Hills on Montserrat).

The coffee plantations that were destroyed in 1929 are no longer here, but the new ones are beginning to encroach. Frankly, I'm not sure I'd want to spend my days working in an area where there are outcrops like this:

These photos are looking east across the Tambor at a mudplain created by lahars, and at the 1902 deposit (the white strip at the base of the cliff) overlain by block-and-ash flow and lahar (volcanic mudflow) deposits. Here the 1902 deposit is about 2 meters thick, and overlain by a thinner gray ashy deposit (perhaps from the 1929 collapse, although it hasn't been well studied and it wasn't clear if that guess was right). The upper blocky deposits are probably post-1929, and consist of a mix of lahar and block-and-ash (pyroclastic flow) leavings.

It was a bit difficult to get a scale into this picture, but the cliff is about 10 meters high, and the bigger boulders at the base are roughly the size of me. This is an interesting photo because it shows a very distinct division between two deposits. What we ended up discussing at this site was what kind of deposits these were, and how to tell them apart. Both block-and-ash flow deposits and lahar deposits look similar at first glance: very poorly sorted, with a clast size range from ash (< 2 mm, in the matrix) to boulders (> 25 cm, supported by the matrix). Both deposits are matrix-supported and both have a bit of a mix of angular and rounded clasts.

So how to tell them apart? One clue might be the angularity of the clasts; block-and-ash flows tend to be a bit more violent in terms of knocking rocks around, and they're dry, which means no muddy cushion for the rocks like a lahar might provide. But what happens if the block-and-ash flow goes down a river (like here)? It could become a lahar, which brings up all sorts of messy discussion about naming conventions. A better indicator of a block-and-ash flow would be the lithology of its clasts; lahars will pick up anything in their path and tend to have a big mix of rock types, while block-and-ash flows tend to be monolithological. Again, however, if a block-and-ash flow travels far enough, it may pick up enough clasts from other places to confuse this.

As it turns out, there is one sure indicator that you've got a lahar deposit, and it's not always easy to find (or present). Because a lahar is a wet flow, it will contain bubbles. When the lahar stops, sometimes those bubbles are unable to rise through the muddy liquid and are trapped; when the deposit dries out, the bubbles leave little spherical cavities in the fine-grained matrix. These are really hard to see and it's easy to debate what you're seeing if you're not sure.

So what was the verdict for these deposits? Based on a bit of arguing discussion, we decided that we were fairly sure we saw bubbles in the lower deposit, but that the upper one had more indicators of a block-and-ash flow deposit. So at this outcrop, we have about 4-5 meters of visible lahar deposits overlain by about 5-6 meters of block-and-ash flow deposits. Needless to say, this would not have been a fun place to have been standing for either event. In fact, Gustavo mentioned that very few people will come down to this area after dark, because some of the victims of the 1929 collapse are thought to haunt the riverbed. Having been down there in the evening, I can attest that it does get a little creepy when it's dark, and it would certainly be easy to turn the jungle sounds into the moaning or wailing of ghosts. Better not to stay there too long, ghosts or no ghosts; it is a channel filled with lahar deposits, after all.