Climate & Environment

    Natural Catastrophes

Post-Doctoral Fellowships


Predicting the most unpredictable volcanic hazards

Some volcanoes are like pressure cookers, but without a safety valve. The hydrothermal activity within can result in dangerously unpredictable hazards, like devastating laterally directed explosions or collapse of the volcanic edifice. The 2014 non-magmatic (phreatic) eruption of the Japanese volcano Mount Ontake, which showed no precursor signs and claimed the lives of 63 people, is a stark example of how sudden and unforeseeable such events can be. Dr. Marina Rosas-Carbajal is leading a project to investigate the still opaque subsurface dynamics at work in these so-called 'wet’ volcanoes. A post-doctoral researcher at the Institut de Physique du Globe in Paris, she chose to focus her study on the La Soufrière de Guadeloupe volcano (Lesser Antilles), an archetype of explosive subduction volcanoes, which holds one of the most hazardous volcanic hydrothermal systems in the world.
 In simple terms, hydrothermal eruptions occur when an underground reservoir of water is heated by the volcano and violently explodes. « In most cases of volcanic eruptions, we are able to provide some forecast for the event, albeit uncertain, from months to days in advance, thanks to precursor signs such as earthquakes, ground deformation or gas emissions, explains Dr. Marina Rosas-Carbajal. But with this type of explosions, there are generally no warning signs.» While quite limited in magnitude, and non-magmatic, these explosions can generate a diversity of hazards, such as ballistic showers, small pyroclastic flows, partial edifice collapse and debris flows. In addition,they sometimes precede more devastating magmatic eruptions. « This was, for example, the case with the infamous 1902 Montagne Pelée eruption, in Martinique ( which killed as many as 30,000 people), and was preceded with a few small phreatic explosions ». « On the other hand, the last phreatic eruption of La Soufrière, in July 1976, resulted in complex crisis response that included a 6-month evacuation of 70,000 people that was perceived as 'unnecessary’ by the population and engendered severe socio-economical consequences for months to years thereafter». The controversial management of this non-magmatic eruption reflects the need for adequate data and models, not only to better forecast hydrothermal explosions, but also to interpret them in terms of other hazardous outcomes. «An in-depth understanding of the physical processes that control volcanic hydrothermal systems is thus critical for developing efficient monitoring strategies and forecasting models,» the Argentinian researcher concludes.



Institut de Physique du Globe de Paris





ORCID Open Researcher and Contributor ID, a unique and persistent identifier to researchers

 What’s cooking in the shallower parts of a volcano

 To contribute to this end, her project aims at characterizing and modeling the changes in the internal state of the La Soufrière de Guadeloupe hydrothermal system, currently in a phase of marked unrest and their relationship with the magmatic source. Among the questions addressed are: is the increasing unrest of La Soufrière de Guadeloupe related to changes in the magmatic source? What is the influence of climate in the cooling and general dynamics of the system? What are the typical pore-pressure changes that occur in the dome, and what is the relation with changes in fumarolic activity, flank instability? To answer these interrogations, Dr. Rosas-Carbajal and her team are collecting new data on the field on the temporal and spatial density variations. These will be inferred from novel tomographic imagining techniques, called muon tomographies. This technology uses cosmic ray muons, an elementary particle, to generate three-dimensional images of a geological object. «The data obtained will be combined with multi-disciplinary surface observations and numerical modelling to be able to characterize the internal physical processes of the hydrothermal system,» the researcher adds.

The outcome of the project will be a simple functional model that characterises the changes in the internal physical state of the volcano’s hydrothermal system. Once the model is developed and tested against extreme conditions scenarios (overpressurization of the magmatic source, deep-sourced magmatic heating, low rainfall levels), the project’s ultimate objective will be to test which type of monitoring is the most efficient, thus paving the way for a better assessment and mitigation of the unpredictable hazards related to volcanic hydrothermal systems.