Last month, prompted by VC-reader VerboselyLaconic’s observation about the volcanoes of Colombia and Ecuador seemingly exhibiting a trend to suffer flank collapses in a preferred direction, we invited our readers to search worldwide in order to establish whether or not this was a general observation or if it only applied to the volcanoes mentioned by VerboselyLaconic.
The first result was a nice discussion, a sort of barn-storming session, as to what could be the causes of the observed phenomenon. Cbus proposed gravity in combination with structural weakness: “When stuff collapses, it will typically topple towards the lowest lying area and area of greater structural weakness. Given the greater relief on the western side of these volcanoes, it’s only natural for them to collapse in that direction.”
This is a sound observation, not only because non-volcanic landslides follow this rule, but also because volcanoes, strato- (or layered) volcanoes in particular, are not homogenous, strong bodies of rock, but rather a pile of debris held in place by loosely welded pyroclastic deposits and lava flows. If the solid bedrock upon which the volcanic edifice is built tilts significantly so that in one direction it is significantly higher, say 300-500 m, than in the opposite, there will be so much more loose material on the downslope that a structural weakness in that direction is inevitable. If this is true, volcanoes on such a slope ought to display a preferred direction of collapse while volcanoes on level bedrock would tend to collapse whichever way.
DownUnder replied that: “But gravity only helps with the west side already being structurally less stable or more prone to collapse. If the volcano has more instability on its east flank, that’s were it will collapse if it shall collapse.” This too is a valid observation and led to a discussion on hidden, interior weaknesses caused by magmatic intrusions, so-called “cryptodomes” or “failed eruptions”, or by chemical alteration caused by volcanic gasses in combination with hydrothermal activity.
At this point, Edmdas asked us if we had considered weathering? “Water flow (infiltration) can dissolve any minerals which would be holding the grains, ash, together weakening the cementation process from the pressure of ash layers above. This would make one side weaker than the other. There is also freeze-thaw action and also solifluction to consider which could all be playing a part in some way.” Biologique replied that “In general, the highest rainfall is on the upside of prevailing winds and a rain “shadow” tends to be present on the down side of prevailing winds… …In northern Latin America the prevailing winds are from the east, therefore the eastern side of the Andes in the northern LA region are very wet. The western side tends to be much dryer.”
Here VerboselyLaconic intervened: “Wet side doesn’t sell well, though. This is what got me thinking: Look at the Cascades from BC to CA. All can agree that the west side of the Cascades are the wet side: 3/4s of Oregon and Washington are rain-shadow deserts. But, St Helens, Rainier, Baker, Shasta, and Lassen all have had major flank collapses to the north. So, wet vs dry can’t be the primary driver… …Washington State appears to be primarily northerly driven failures, Oregon is primarily southerly or easterly driven failures, and California switches back to north trending. Maybe there is no rhyme nor reason….”
Thus in no time at all we had identified two main underlying (pun intended) causes as well as a possible minor outside influence in the form of weather. By the time Lucas added: “Anyway, could the alignment of a volcano on a fault influence where it collapses?”, something definitely to keep in mind, we were still confused albeit at a higher level of understanding. Thanks to Irishzombieman, who located a paper bearing on the subject, we could read the following:
On Kamchatka, detailed geologic and geomorphologic mapping of young volcanic terrains and observations on historical eruptions reveal that landslides of various scales, from small (0.001 km3) to catastrophic (up to 20–30 km3), are widespread. Moreover, these processes are among the most effective and most rapid geomorphic agents. Of 30 recently active Kamchatka volcanoes, at least 18 have experienced sector collapses, some of them repetitively. The largest sector collapses identified so far on Kamchatka volcanoes, with volumes of 20–30 km3 of resulting debris-avalanche deposits, occurred at Shiveluch and Avachinsky volcanoes in the Late Pleistocene. During the last 10,000 yr the most voluminous sector collapses have occurred on extinct Kamen (4–6km3) and active Kambalny (5–10 km3) volcanoes. The largest number of repetitive debris avalanches (>10 during just the Holocene) has occurred at Shiveluch volcano. Landslides from the volcanoes cut by ring-faults of the large collapse calderas were ubiquitous. Large failures have happened on both mafic and silicic volcanoes, mostly related to volcanic activity. Orientation of collapse craters is controlled by local tectonic stress fields rather than regional fault systems. Specific features of some debris avalanche deposits are toreva blocks — huge almost intact fragments of volcanic edifices involved in the failure; some have been erroneously mapped as individual volcanoes. One of the largest toreva blocks is Mt. Monastyr — a 2 km3 piece of Avachinsky Somma involved in a major sector collapse 30–40 ka BP. (Ponomareva & al: 2006)
A second look at the volcanoes of Equador listed by VerboselyLaconic yields a slightly different picture. First of all, right through the middle of the Andes is the Guayllabamba river plain, a depression that separates the Andes into two distinct mountain ranges, each of which has erosion gullies caused by rivers flowing down into this plain. While there are exceptions such as Volcán Corazón, which indeed has suffered a flank collapse towards the WNW (even if the terrain to the east of Corazón shows unmistakable evidence of debris avalanches such as toreva blocks), most volcanoes seem to follow the principle suggested by Cbus, that is, collapse in the direction of lowest altitude where one is present.
To the WSW of Corazón lies Volcán Ruminahui, almost in the middle of the plain, and the topography suggests that it has suffered flank collapses to the west, the NNE, east, SE and SW. Even Cotopaxi itself, 15 km to the SE of Ruminahui, has feautures suggestive of toreva blocks and debris avalanches in several directions. Because volcanoes perpetually change the landscape, obliterating what was before, it is indeed very difficult to say anything definitive from the study of maps alone. Yet, kudos to VerboselyLaconic because he got us all thinking!
So where does this leave our project? Well, as we all soon realised, the subject is so complex and the search area so vast that I am afraid that there is little chance of us ever establishing a definitive data base, discover and quantify the underlying causes in every single case (or at least enough cases) in order to ascertain a set of guiding principles. Nevertheless, there is always our own, personal Oddysey of exploration and gaining of knowledge to consider, and the hours I spent familiarising myself with Kamchatcka and particularly the Kuriles gave me a lot of pleasure and a far greater familiarity with that region. So instead of feeling despondent, I feel a sense of satisfaction and personal accomplishment. I hope you share this feeling and thank you for your participation!
Ponomareva, Melekestsev & Dirksen: “Sector collapses and large landslides on Late Pleistocene – Holocene volcanoes in Kamchatka, Russia”, Journal of Volcanology and Geothermal Research 2006
Featured Image – Avachinsky from EOS
Categories – Science, Volcano
Tags – Debris Avalanche, Edifice Collapse, Flank Collapse, Lateral Blast, Reader Participation, Research Project