Anne Le Friant

Pedestrian tsunami evacuation depends on wave arrival time, warning dissemination, the spatial configuration of the exposed area and population response. These factors, studied by distinct disciplines, are rarely integrated within a single framework, limiting the use of evacuation models for local planning. We present TsunaWAVE (Web interface for Agent-based modelling and Visualisation of Evacuation), an open web interface that links tsunami hazard scenarios with agent-based evacuation scenarios through a one-way sequential coupling, in which simulated arrival times define the time budget against which precomputed evacuation outputs are read, making both explorable by decision-makers. The agent-based model simulates warning dissemination, pre-evacuation delay and movement towards safe areas, while the published surrogate model (Landslide-Tsurrogate v1.0) provides first-wave arrival times for submarine landslide scenarios. The framework is applied to the Mamoudzou waterfront (Mayotte, France), where lead times are 10 to 16 minutes in two reference scenarios and 759 exposed individuals were counted using drone imagery, video line-crossing and manual counts. Ninety scenarios, each replicated 42 times, rank three operational levers (warning, safe-area accessibility and social contagion). Warning reliability is the first-order control. Under degraded warning without social contagion and a 10 min lead time, fewer than 8% of agents reach safety, whatever the safe-area configuration. Social contagion, implemented as neighbourhood-based imitation, ranks second; it cannot substitute for a reliable warning but partially compensates for a degraded one when time allows, raising the share reaching safety at 16 min from 19.4% to 43.8%. TsunaWAVE makes these findings openly available to local authorities at https://geodata.univ-montp3.fr/web-app/irima-mayotte/.

Submarine gravity-driven sliding of sediments are common processes in the vicinity of volcanic islands. In the Lesser Antilles arc, the Montagne Pelée volcano on Martinique Island underwent several flank-collapse events during its long-term eruptive history, resulting in debris avalanches. When the debris avalanches entered into the seawater, they were emplaced over the unstable slope of the volcano, triggering a seafloor sediment failure and massive landslides downslope. Using a laboratory modeling approach, we simulated the gravity-driven sliding of a sand layer lying above a silicone layer. The experiments were performed using various slope geometries (slope lengths and number of slope breaks separating the slopes with different angles), under both dry and aqueous conditions, and while varying the amount of additional sand inputs upslope. The resulting deformations were characterized in each experiment in order to compare the obtained structures with those shown by the seismic lines offshore to the west of Martinique Island. During all the experiments, a compressional frontal deformation zone made of several reverse faults formed downslope, often near the slope breaks. Downslope, a portion of the sediments was mostly displaced and poorly deformed in a damping zone, while an extensional deformation zone formed upslope. The displacements of the surficial markers were measured through time to characterize the sliding dynamics. Our study demonstrates that the slope geometry and additional sand inputs primarily favor and increase the sliding deformation, whereas the hydrostatic pressure plays a secondary catalytic role over time. These results provide new constraints on the driving factors and their consequences on gravity-driven sliding in terms of deformations and runout distance over time. This may have a significant impact on the associated hazard assessment related to offshore infrastructures, in a region known for its seismic and volcanic risks.

Since May 2018 till the end of 2021, Mayotte island has been the locus of a major submarine volcanic eruption characterized by the offshore emission of more than 6.5 km 3 of basanitic magma. The eruption occurred along a WNW–ESE trending submarine ridge on the east flank of the island where, in addition, several seemingly recent phonolitic bodies were also identified close to the island. To define realistic scenarios of magma ascent and potentially predict the style of an upcoming event, it is crucial to have a precise understanding on the plumbing system operating below volcanoes. The putative relationships between basanites emitted by the new volcano and these recent phonolites have been experimentally explored by performing crystallization experiments on a representative basanite over a large range of pressures (up to 400 MPa). The results show that the crystallization of basanite at crustal levels ( ≤ 12–15 km) yields a phonolitic residual liquid containing up to 3–4 wt% H 2 O after ≥ 65 wt% of an assemblage of olivine + plagioclase + amphibole + clinopyroxene + biotite + magnetite + ilmenite + apatite. The final iron content of the residual phonolitic liquids is strongly controlled by the depth/pressure of fractionation. Fe-rich phonolites from the submarine ridge are produced at 6–8 km depth, while a shallower differentiation ( ≤ 4–5 km) results in the production of liquids with trachyte–benmoreite affinities. If the fractionation process occurs at depths higher than 8 km, the resulting phonolitic melts are progressively enriched in SiO 2 – Al 2 O 3 but depleted in FeO*, ie unlike those erupted. We therefore conclude that phonolitic magma production and storage at Mayotte is a rather shallow process.

The “Fani Maoré” eruption off the coasts of Mayotte has been intensively monitored by applying methods similar to those used for subaerial eruptions. Repeated high-resolution bathymetric surveys and dredging, coupled with petrological analyses of time-constrained samples, allowed tracking the evolution of magma over the whole submarine eruptive sequence. Indeed, after one year of direct ascent (Phase 1), basanitic magma switched to a different pathway that sampled a tephri-phonolitic subcrustal reservoir (Phase 2). Later, the magma pathway shifted again in the crust resulting in a new eruption site located 6 km northwest of the main edifice (Phase 3). The petrological signature of lava flows reveals both an evolution by fractional crystallization and syn-eruptive mixing with a tephri-phonolitic magma. We demonstrate that high-flux eruption of large volumes of basanitic magma from a deep-seated reservoir can interact with shallower reservoirs and remobilize eruptible magma. This has significant hazards implications with respect to the capacity of such large eruptions to reactivate shallow-seated inactive reservoirs from a transcrustal magmatic system that could be located potentially at a distance from the high-flux eruptive site.

Since May 2018, Mayotte Island has been experiencing seismo-volcanic activities that could trigger submarine landslides and, in turn, tsunamis. To address these hazards, we use the HySEA numerical model to simulate granular flow dynamics and the Boussinesq FUNWAVE-TVD numerical model to simulate wave propagation and subsequent inundations. We investigate 8 landslide scenarios (volumes from 11 . 25 × 10 6 m 3 to 800 × 10 6 m 3 ). The scenario posing the greatest threat involves destabilization on the eastern side of Mayotte’s lagoon at a shallow depth and can generate sea-surface deformations of up to 2 m. We show that the barrier reef surrounding Mayotte plays a prominent role in controlling water-wave propagation and in protecting the island. The tsunami travel time to the coast is very short (a few minutes) and the tsunami is not necessarily preceded by a sea withdrawal. Our simulation results provide a key to establishing hazard maps and evacuation plans and improving early-warning systems.

The development of overpressure in continental margins is typically evaluated with hydrogeological models. Such approaches are used to both identify fluid flow patterns and to evaluate the development of high pore pressures within layers with particular physical properties that may promote slope instability. In some instances, these models are defined with sediment properties based on facies characterization and proxy values of porosity; permeability or compressibility are derived from the existing literature as direct measurements are rarely available. This study uses finite-element models to quantify the differences in computed overpressure generated by fine-grained hemipelagic sediments from the Gulf of Cadiz, offshore Martinique and the Gulf of Mexico, and their consequences in terms of submarine slope stability. By comparing our simulation results with in situ pore pressure data measured in the Gulf of Mexico, we demonstrate that physical properties measured on volcanic-influenced hemipelagic sediments underestimate the computed stability of a submarine slope. Physical properties measured on sediments from the study area are key to improving the reliability and accuracy of overpressure models, and when that information is unavailable, literature data from samples with similar lithologies, composition and depositional settings enable better assessment of the overpressure role as a pre-conditioning factor in submarine landslide initiation.