Development of an integrated approach to reduce the risks associated with explosive volcanism, from hazards research to crisis management tools: Martinique case study – RAVEX

Complementary methods are used by RAVEX project participants. The field work consists of studying the eruption deposits and collecting samples for laboratory analysis; surveys are also conducted among the population of the island of Martinique to assess the perception of risk by the inhabitants. Analogue laboratory experiments investigate (i) the collapse dynamics of eruptive fountains and the formation of resulting pyroclastic flows, and (ii) the mechanisms of interaction between pyroclastic flows and a body of water, which can lead to the formation of a tsunami. Finally, numerical models are used to simulate volcanic phenomena at the scale of Martinique and to evaluate the areas that may be affected by the products.

A theoretical 1D eruptive column model, taking into account the effect of pyroclasts in gas trapping and sedimentation, was developed to specify the conditions leading to the collapse of a volcanic plume (Task 1). Additional laboratory experiments documented the process of particle accumulation at the impact zone and revealed the occurrence of interstitial fluid pressure in the emerging flow (Tasks 1 and 2). Flow propagation was studied theoretically by developing a high performance computational code that simulates dense biphasic flows, in particular by capturing rigid zones in the flow as well as the interface with the ambient fluid (task 2). Another experimental study identified scaling laws that relate the degree of turbulence of a diluted flow to the concentration of particles; these are applicable to plumes and the diluted portion of pyroclastic flows. Work in Task 3 focused on the construction of experimental devices to study the initiation of tsunamis by pyroclastic flows. A 2D prototype was used to release liquid flows into water and to study the characteristics of the wave; the results were compared to Navier-Stokes simulations. A second large scale device (7 m) was built to study during the second part of the project the interaction mechanisms between a fluidized granular flow and a mass of water. Field work in Martinique (Task 4) clarified the rhythm and characteristics of past plinian eruptions and identified two previously unknown events. Surveys and coordination and communication meetings were organized to assess the representations and issues of volcanic risk in Martinique.

Several additional studies are planned for the second part of the RAVEX project. The effect of wind on column collapse mechanisms will be considered in the eruptive fountain model. A particle concentration measurement method using acoustic waves will be applied in laboratory experiments on dense or dilute biphasic mixtures. Tsunami formation experiments will be conducted to investigate in detail the mechanisms that operate when a granular flow rapidly comes into contact with a body of water; the scaling laws that link the characteristics of the waves created to the properties of the flow will be identified. Field work in Martinique will be continued in order to clarify the characteristics of past eruptions based on the analysis of their products, and information activities with stakeholders likely to take part in the risk chain will be continued.

Peer-reviewed journals:
- Weit A., O. Roche, T. Dubois, M. Manga (2018). Experimental measurement of the solid particle concentration in geophysical turbulent gas-particle mixtures. Journal of Geophysical Research-Solid Earth, doi: 10.1029/2018JB015530.
- Chupin L., T. Dubois (2016). A bi-projection method for Bingham type flows. Computers and Mathematics with Applications, 72: 1263-1286.
- Michaud-Dubuy A., G. Carazzo, E. Kaminski, F. Girault. A revisit of the role of gas entrapment on the stability conditions of explosive volcanic columns. Journal of Volcanology and Geothermal Research, revised.

The RAVEX project aims at developing an integrated approach to reduce the risks associated with explosive volcanism. It associates researchers specialized in the physics of explosive volcanic eruptions and researchers specialized in the study of public policy. It will be run by a consortium of laboratories whose fields of expertise are renowned at the international level and which encompass disciplines of Earth and environmental Sciences and of human and social Sciences. The project will be led by the Laboratoire Magmas et Volcans (Univ. Blaise Pascal-IRD, Clermont-Ferrand), with the following partners: the Laboratoire de Mathématiques (Univ. Blaise Pascal), the Institut de Physique du Globe (Paris), the laboratoire des Sciences de l'Ingénieur Appliquées à la Mécanique et au génie Electrique (Univ. Pau), Sciences Politiques (Paris) and the laboratoire Intégration et Coopération dans l'Espace Européen (Univ. Paris 3). The project is structured in three tasks dedicated to the study of volcanic hazards and a fourth task on the analysis of representations, vulnerabilities and risks, which is transverse to the whole project.

We will investigate, through laboratory experiments and theoretical models, phenomena that can occur in chain during most explosive volcanic eruptions generating highly hazardous hot gas-particles mixtures: the collapse of volcanic fountains and the resulting pyroclastic density currents, and we will investigate how these currents can generate tsunamis in coastal environments.
(1) We will study how the complex coupling between the ascending and descending parts of an eruptive fountain and the particles in the jet can control the entrainment coefficient of the ambient fluid, the key parameter for determining the fountain stability. A new model will be developed in order to predict the characteristics of the gas-particles mixture formed at the vent and it will be validated by scaled laboratory experiments.
(2) Considering a logical succession of the phenomena, we will perform laboratory experiments to study the mechanisms of particles accumulation at the impact zone and of segregation in the emerging pyroclastic density currents. The particle concentration in the flows will be measured using new acoustic methods. We will develop a complementary mathematical model of dense gas-particles flows for which a new method for resolution of the equations will be proposed.
(3) A third task will deal with the triggering mechanisms of tsunamis by pyroclastic flows. We will perform laboratory experiments that will consist of impacting of a mass of water by a fluidized granular flow simulating a pyroclastic flow, a novelty compared to earlier studies. Numerical simulations with a multi-fluid code will take into account the thermal exchanges and the phase changes and will complement the experiments whose results will in turn validate the models.
(4) The results of these works will serve as a base for establishing a model case. We will consider the island of Martinique because it is a key area for studying explosive volcanism and related tsunamis, and the island and Caribbean context provides an interesting ecosystem for studying risk governance schemes. Numerical simulations of flows and tsunamis in real configuration will be done to create hazards and risks maps, and they will be complemented by new field data. New procedures will be developed to allow the appropriation of this knowledge by the various actors and strengthen the strategy of disaster risk reduction. This requires mapping the actor networks and analyzing their different levels and modalities of interactions. We also plan to investigate the representations different actors may have of the hazards, vulnerabilities, risks and the role of science in order to identify possible disharmonies and to propose methods to overpass them. Finally, we will make general recommendations for improving the scientific and policy strategies with respect to disaster risk reduction.