Search

Large boulders lying at abnormally high levels are common proxies to determine the occurrence of extreme past geophysical events such as storm or tsunami. Empirical formulas are derived to reconstruct the event conditions associated to the boulder displacement (e.g., Nott, 2003). Up to now, to the best of our knowledge, no one directly identified the individual wave responsible for the movement of a large boulder. In this paper, we report such an observation and analyze the associated hydrodynamic conditions.On the 2/28/2017, a large concrete block of about 50T, protecting the Artha breakwater in Saint Jean de Luz (south west of France) was displaced from the block armor unit to the crest of the structure during a large storm event. The vertical displacement of the block is assessed to a minimum of 5.3m. This type of event is quite rare (last known date is 1951), nevertheless, this time, the scene was by chance captured by a photographer taking pictures of the storm at the same moment. This allowed assessing the approximate time at which the event occurred and analyze the corresponding data available locally at this time namely : waves (integrated parameters and free surface signal 1 km off-shore), water level and the associated photographs.First, the water level is found to be close to its maximum (high tide with high tidal coefficients). During the storm, the significant wave height evolved between 5.61 to 7.47m, and the maximum wave height from 8.53 to 11.11m. Surprisingly, at the time of the block displacement, these parameters were not maximum (i.e., 6.8m and 8.53m, respectively). Nevertheless, the investigation of the free surface signal shows the existence of a large wave of more than 14m high and about 25s long approximately 1min30 before the block displacement, not detected by the zero-crossing wave buoy algorithm. This individual wave satisfies the definition criterion of a rogue wave (i.e., Hwave/Hs>2.2).With 1D phase resolving wave simulations, we first validated the wave propagation time from buoy to breakwater (i.e., 1min30) and assessed the reflection component at the buoy. The latter was considered as sufficiently moderate at the time of observation to consider the 14m recorded wave as an incident one. Next, we used a combination of two models nested : SWASH (1D) and the Navier-Stokes OLA-FLOW (2DV), to simulate the event considering the free surface signal as input. The hydrodynamical conditions obtained with the computations were finally compared to the existing empirical formulas and discussed.The simulations show that the long rogue wave keeps its coherence up to the breakwater and generates a strong and abnormally long duration flow above and within the blocks (i.e. the porous medium). This result strengthens the idea that the force duration (along with force magnitude), is an important factor to consider to predict large boulder or concrete block displacement. This study also shows a possible original effect of rogue waves in the nearshore area.

Subaerial landslides can generate impulsive waves which, in turn, may cause significant damages to the facilities and people on the surrounding coasts. In spite of the numerous studies related to this complex phenomenon in the last decades, there is still a lot to understand, especially physically speaking.The present work aims at better understanding the energy transformation process from the slide initial potential energy to the final wave train energy. In particular, we would like to emphasize the existence of an optimum energy rate of transformation and investigate the reason for this existence.To do so, we rely on a Navier-Stokes two or three phases model (OpenFoam) and perform numerical experiments, fixing a few parameters (slope, density, rheology) and studying the effect of the others. The physics of the phenomenon is highly complex, involving liquid phases interaction, transient wave formation, nonlinear wave processes, dispersion, wave breaking, etc. Such a numerical model, despite its inherent uncertainty, is anyway able to provide a rich information, which may be later completed with experimental results. In particular, the model gives access to all the flow variables which allows to characterize in depth the energy processes. The free surface signal analysis is also valuable for wave celerity, and hence generation zone extent and dispersion analysis.In terms of research strategy, in order to restrict the complexity and allow a better understanding of the phenomenon, the idea is to start with a very simple rheology, the inviscid case, and progressively increase the numbers of rheological parameters (i.e., viscous flows, Bingham and finally Herschell Bulkley).During the conference, we will first illustrate the existence of an optimum in the rate of energy transformation for the inviscid slide by progressively increasing the slide volume. We will try to relate this optimum with the physical processes at stake (liquid mass interaction, wave breaking types, dispersion, etc.). Next, we will show the influence of the slide rheology in the process of energy transfer and in particular how the energy optimum varies with respect to the rheological parameters.