The trouble with lahars

by Dr Rebecca Williams



What I'm thinking about...a word cloud of my 2008 paper validating Titan2D

This summer I plan on thinking about lahars as a refreshing change to troubling over pyroclastic density currents, which are ‘my thing’. Pyroclastic density currents (PDCs) are mixtures of hot gas, ash, pumice and blocks of rock that flow down the sides of volcanoes during catastrophic eruptions. They are what I’ve spent most of the last 7 years thinking about.



A boulder carried by a lahar from Volcán Cotopaxi in 1877. A much younger me for scale.

Video of the Vazcun Valley Lahar, Volcán Tungurahua, 2005. See how it impacts the bridge and take a look at the photos below to see the size of the boulders that were piled up against the central pillar. More videos on my youtube channel.

 Lahars, on the other hand, are flows made of water, volcanic material, and whatever else they can erode including soil and trees.  The word lahar is Indonesian and roughly translates to volcanic mudflow. They range from dirty water to flows so laden with material that they have the consistency of concrete and can carry boulders the size of a house!

Between 1600 and 2010 AD lahars caused 44,252 deaths, so they are pretty deadly. This is because they can travel great distances (The Chillos Valley Lahar flowed from Volcán Cotopaxi to the Pacific Ocean – that’s 326 km!) and at great speeds (>85 km/hr).   They can form during a volcanic eruption if hot ash is erupted onto snow and ice at the top of a volcano. But they can also form when the volcano hasn’t erupted for a long time if heavy rainfall erodes the loose volcanic deposits on the sides of the volcano. This makes it very difficult to know when one might happen - they can take communities by surprise. Lahars can be big, or small, fast or slow and contain lots of debris, or be very watery. If they are so different, how can we begin to try to figure out where they might flow or how fast?



The Vazcún Valley with Baños in the foreground.

The El Salado baths are circled.

Volcán Tungurahua’s summit is hiding in the clouds

I last thought about this problem for my Masters dissertation, which I did at the University of Buffalo. Then, I was working amongst a large collaborative effort to create computer simulations of hazardous volcanic flows. When I joined UB Titan2D (their computer model for simulating PDCs) was well established and they were just starting to adapt the mathematical equations to use it for modelling lahars. I was given the task to see if it would work, and to see if it could be used to predict where lahars might travel and how fast they might flow. 

On the 12th of February, 2005, a small ash-rich lahar travelled down the Vazcún Valley which drains Volcán Tungurahua through the town of Baños, Ecuador. This lahar was triggered by heavy rainfall eroding recent ashfall deposits. The lahar swept past the El Salado baths, much to the surprise of the bathers enjoying the volcanic springs here. It travelled 10 km at speeds of up to 7 m/s. In the summer of 2005, me and some colleagues mapped the deposits left behind by this lahar. Back in the office I used Titan2D to simulate a lahar travelling down the valley and compared where the model said it would go and at what speeds to what we knew from the field survey. Turns out, the model did a pretty good job. The area that the model predicted would be covered matched pretty closely with what we mapped and the predicted thickness of the flow was within 4% of the measured flow thickness. But, the velocity of the simulated flow was only within 17–67% of the velocities recorded during the event, depending on what we told the model to simulate. So something wasn’t quite right.

(A) Photograph of the El Salado baths in the Vazcún Valley. At the weekends, it is estimated that more than 300 people visit the baths per day. (B) The 2005 Vazcún Valley Lahar came to within a few tens of centimeters of inundating the El Salado baths. Photograph shows lahar at peak stage height. Photograph courtesy of Defensa Civil de Baños. (C) Looking downstream towards the new footbridge replacing the one destroyed by the lahar. Foreground shows area outside of channel inundated by the lahar. (D) Meter-scale boulders piled up at the foot of the central support pier for the main road out of Baños. Note also splash/tide marks on both bridge piers. From Williams et al., 2008.

In Titan2D you can change the percentage of water in your lahar, and the volume of material in your lahar, and differences in these can vastly affect the output of the model. What it can’t do, is change the amount of clay, silt, sand, gravel or boulders in your lahar. And this has got me thinking...how does changing the composition of a lahar change the distance it can travel, the speed at which it travels and how much it stays within the confines of its valley? The answer? I’m going to think about how to answer that this summer.

Video of a Titan2D simulation of the Vazcun Valley lahar.

Williams et al., 2008. Evaluation of the Titan2D two-phase flow model using an actual event: Case study of the 2005 Vazcún Valley Lahar. Journal of Volcanology and Geothermal Research 177 (2008) 760--766 
This paper can be found here or contact me for a copy. Tw: @volcanologist 

The trouble with lahars is that they are very erosive so they cut valleys very quickly. The valley here, the Quebrada Achupasha, suffered heavily from lahars washing away roads and bridges, isolating communities on the west side of Volcán Tungurahua.

Gratuitous Volcán Tunguragua shot. Taken from the volcano observatory during a small ash eruption.