Emplacement dynamics of debris avalanches and submarine landslides at Soufrière Hills volcano, Montserrat

Lead Research Organisation: University of Southampton
Department Name: School of Ocean and Earth Science


Volcanoes are inherently unstable geological features, and their collapse produces some of the largest landslides on Earth's surface. One of the best known examples of this type of landslide is that which accompanied the 1980 eruption of Mount St. Helens. This event generated a rapidly moving mixture of rock, known as a debris avalanche, with a volume of nearly three cubic kilometres. Similar landslides have occurred around many volcanic islands, such as Hawaii and Tenerife, and in these settings can be over 100 times larger than the Mount St. Helens debris avalanche. Debris avalanches on volcanic islands may potentially generate very large tsunamis. However, the size of the tsunami depends on how the landslide moves. For example, a landslide that occurs in several distinct phases, or which involves material from the deep, surrounding seafloor, will produce much smaller tsunamis than a single landslide of the same volume that all derives from the volcano.

No large-scale collapses of volcanic islands have been observed in modern times. However, several submarine deposits from past landslides indicate their widespread occurrence. The same deposits provide the key to understanding how these landslides moved, and for therefore accurately modelling their associated tsunami hazard. Understanding the formation and development of landslides around volcanic islands is the major aim of this research project. Specifically, the research will show: whether debris avalanches on volcanic islands typically occur in single or multiple stages; what controls how far the landslide moves underwater; and whether debris avalanches trigger further landslides once they land on the seafloor.

This research will focus on two landslide deposits off Montserrat, in the West Indies. Each deposit is a similar size to the Mount St. Helens debris avalanche. Geophysical data from the area shows that the two deposits have very different shapes and structures. One of the landslides travelled much further and also appears to have involved a large volume of material from the seafloor, rather than from the volcano. Examples similar to both the Montserrat deposits can be found around other volcanic islands. The Montserrat deposits therefore provide general information for understanding the processes of landslide development around volcanic islands.

Fully understanding how the Montserrat landslide deposits formed requires combining the geophysical data with samples of the actual deposits. These samples will show what different parts of the landslide deposits are made of, and will reveal what structures seen in the geophysical data mean in terms of landslide movement. Samples from the deposits will be collected in March 2012 by the International Ocean Drilling Program (IODP). This will be the first time that deposits of this type have been directly sampled. This unique sample set therefore provides an opportunity for significantly advancing our knowledge of landslides around volcanic islands.

This research project involves four stages. The first will reinterpret the geophysical data in light of the IODP samples. This will help show how the landslide moved underwater. The second stage will investigate the sediment deposited around the margins of the landslide deposits, using physical and chemical measurements to show if the landslides occurred in multiple phases. The third stage will determine the constituents of the landslide deposits, to measure how much of the deposit was formed by material from the seafloor. The final stage will directly date the younger landslide deposit, and then further investigate its constituents and stages of collapse, using samples and data that have already been collected. Together, this work will provide a detailed understanding of the Montserrat landslide deposits, and will provide important results for a better understanding of landslide processes and tsunami hazards around other volcanic islands.

Planned Impact

The proposed research is of direct interest to organisations working to understand, mitigate and respond to landslide hazards. As such, the findings will also be of interest to the wider public community. An important aspect of this research is its interdisciplinary nature. Academic beneficiaries include those working across engineering and the geosciences, and this project will lead to further collaborative efforts in tsunami modelling, slope stability and submarine landslide processes. This research and its associated academic outcomes aim towards a more accurate ability to: determine the typical progression of volcanic collapse; accurately model the tsunami hazard associated with volcanic collapse and linked submarine landslides; and understand the failure and propagation of submarine landslides in unconsolidated sediment on very low gradients. All of these outcomes are important in the fields of natural hazards, and there are therefore beneficiaries well beyond the academic community.

Research results relating to volcanic collapse are relevant in volcanically active regions across the world, in continental and island settings, since flank collapse is a near-ubiquitous volcanic hazard. The impact of these hazards is significant in human and economic terms, and the broad beneficiaries of an improved understanding of flank collapse processes include governmental and non-governmental organisations working to identify volcanic hazards. More specifically, an improved understanding of the processes controlling the magnitude and evolution of volcanic flank collapse will allow more precise constraints to be used in the preparation of hazard maps in risk prone regions, which are vital to the safe development of infrastructure in potentially affected regions, and important to minimise human and economic impacts from volcanic landslides.

In terms of reducing risk associated with tsunamis, the beneficiaries are broadly the same as those identified above. Vulnerable coastal sites are likely to be more reliably identifiable from a better knowledge of tsunami generating source conditions, which will be the outcome of this research. For example, landslides generated on the island of Montserrat may have damaging impacts on the coastlines of surrounding islands, but their magnitude is unlikely to generate waves capable of propagating across the wider Caribbean region. I will use the uniquely detailed data set resulting from this research to work with academic collaborators and workers at Montserrat Volcano Observatory. We will produce tsunami hazard assessments for this site, taking into account the complexity of landslide processes. Being able to make comparable assessments in other settings is important in the case of future volcanic island instabilities. In such instances, the example of hazard assessment from this research will benefit organisations working towards hazard forecasting elsewhere. It is important to emphasise, however, that scientists working in this field must provide accurate predictions that do not sensationalise the potential risks and thereby cause unnecessary alarm to local populations and damage to regional economies, which may be critically dependent on tourism.

Research into slope stability has a different set of beneficiaries than that with a focus on disaster mitigation around volcanic regions. This project will advance our knowledge the causes and dynamics of submarine slope failure on low gradients. This is relevant in volcanic and non-volcanic settings, and of importance to those in industry (e.g., oil exploration) and those involved in development of other marine infrastructure (e.g., pipelines, data cables, shipping construction). Suitability of sites, likelihood of slope failure, and initiation of slope failure are all important assessments during the development of such sites, and understanding how soft sediment can fail on very low slopes is vital to ensure that stable sites are selected.


10 25 50

publication icon
Cassidy M (2015) Origin of Basalts by Hybridization in Andesite-dominated Arcs in Journal of Petrology

Description The volcanoes of Montserrat have collapsed several times during their history, generating large landslides that enter the surrounding sea. These landslides have the potential to generate damaging tsunamis, but understanding this potential relies on an accurate understanding of landslide processes. In this work, we aim to use recent samples collected on the IODP340 expedition to provide a clearer understanding of how these landslides were emplaced (the origin of landslide material, and whether landslides were formed in one or multiple phases) and their relative timing (how frequent and how large are these events). We also aim to understand specifically how the volcano responded to a large collapse, in terms of its subsequent explosive activity. Early results suggest an unusual period of explosive volcanism followed collapse. The results will help us understand how and why the style of hazards posed by this volcano has changed through time.
Exploitation Route The results relating to landslide processes can be used as an input to tsunami models. We aim to use these to generate more accurate models of past events, and thereby inform those those involved in tsunami monitoring and mitigation strategies in the Eastern Caribbean region. The detailed findings on landslide processes coming out of this work will be used to generate more accurate tsunami generation models for landslides around ocean islands, in collaboration with other researchers and in conjunction with ongoing projects at other sites (volcanic and non-volcanic) to understand landslide-generated tsunamis. Other results on how the style of activity changes following collapse will feed into ongoing research projects regarding the long-term evolution of volcanoes. Part of this research is also investigating more generally how samples collected in marine cores should be interpreted (i.e. how can we differentiate samples deposited via different mechanisms or volcanic processes), and this will be relevant to a wide range of research projects that use such samples.
Sectors Environment,Other

Description The results and interest of this research overlap with those of NE/I02044X/2, which can be referred to for further details. This research has fed into results shared with the volcano observatory on Montserrat, enhancing our understanding of past processes on the volcano, of what drives these processes, of the subsurface magma storage conditions (relevant to monitoring of the current system) and of the range of potential eruption styles present at this volcano. Our research has also resulted in the development of a new protocol for the interpretation and assessment of volcanic event records preserved in marine sediment cores, which we hope will be taken up widely at other locations and will generally assist the more accurate development of past eruption histories.
First Year Of Impact 2014
Sector Environment,Other