Non renseigné / Most mega-earthquakes (i.e. earthquakes with Mw ≥ 8.5) occur along subduction mega-thrusts, the interfaces between the subducting - and the overriding plates in convergent margins. These events may have catastrophic impact on human societies due to their destructive potential. For this reason being able to predict the timing and size of these earthquakes became one goal of the international scientific community. The subduction seismic cycle is influenced by many different parameters. The interplay between these parameters governing the frequency and size of megathrust earthquakes still remains unclear, mainly due to the short (i.e. limited to the last century) seismic record.The seismogenic part of the subduction thrust fault spans between depths of 11±4 and ± 51 km (Heuret et al. 2011). In this zone a combination of temperature, pressure and rocks characteristics creates conditions favourable for seismic behaviour. Whether a specific area in the subduction thrust fault has the ability to trigger mega-earthquakes can be expressed using the degree of seismic coupling, i.e. the amount of slip that occurs with respect to the total amount of plate convergence (e.g. Scholz 1998; Scholz & Campos 2012). When a fault is fully coupled, all of the fault slip occurs during earthquakes instead of also during aseismic behaviour (e.g. slow slip events). The internal structure of the interplate fault zone mainly determines whether an area within a subduction zone behaves seismic or aseismic (Wang & Bilek 2011). This is influenced by the topography of the plate interface (e.g. subducting seamounts; Wang & Bilek 2014), but also subducted sediments and fluids in the subduction channel may play an important role.The main goal of this project is to understand which parameters affect the behaviour of mega-earthquake ruptures. This will be done by comparing natural data (e.g. seafloor roughness, sediment thickness and fluid content in the subduction channel) to rupture characteristics of major recent earthquakes. With this analysis also more knowledge can be gained on the triggering of slow earthquakes instead of mega-earthquakes. These are slow slip events with lower frequencies and longer durations than ‘regular’ earthquakes (Saffer & Wallace 2015).The database of natural data, implemented by the long-term scientific joint venture between the Univ. Montpellier and the LET (Roma Tre) will be used for the analysis. Ongoing work is done on determining a method for estimating the seafloor roughness, i.e. the distribution of high, low and smooth areas (by Michel Peyret in collaboration with Serge Lallemand, Univ. Montpellier). Also data is available on the trench sediment thickness around the world (Heuret et al. 2011). In the frame of this project, information on the roughness of the seafloor will be added to the database. In addition the rupture characteristics of major recent earthquakes will be collected. By performing a multiparametric statistical analysis of the database, a conceptual model will be realized, exploring the possible link between all the different parameters. The aim is to validate this model in the lab using scaled 3D analogue models. This will be done both at the LET and at Univ. Montpellier by using a broad range of geometries and contact materials with different rheologies (e.g. gelatin, foam rubber and a new analogue material; Caniven et al. 2015; Corbi et al. 2013). This jointed experimental approach with both the Univ. Montpellier and the LET involved creates a rich environment where differences and similarities of the two different approaches can be used to validate the results.

La Mer d’Iroise est observée depuis 2006, par des radars à haute fréquence (HF) qui estiment les courants de surface. Ces mesures ont une finesse temporelle et spatiale pour permettre de capturer la dynamique fine du domaine côtier. Ce travail de thèse vise à la conception et l’application d’une méthode d’assimilation de ces données dans un modèle numérique réaliste pour optimiser le frottement sur le fond et corriger l’état du modèle afin de mieux représenter la circulation résiduelle de marée et les positions des fronts d’Ouessant en mer d’Iroise. La méthode d’assimilation de données utilisée est le Filtre de Kalman d’Ensemble dont l’originalité est l’utilisation d’une modélisation stochastique pour estimer l’erreur du modèle. Premièrement, des simulations d’ensemble ont été réalisées à partir de la perturbation de différents paramètres du modèle considérés comme sources d’erreur : le forçage météo, la rugosité de fond, la fermeture turbulente horizontale et la rugosité de surface. Ces ensembles ont été explorés en termes de dispersion et de corrélation d’ensemble. Un Lisseur de Kalman d’Ensemble a ensuite été utilisé pour optimiser la rugosité de fond (z0) à partir des données de courant de surface et d’un ensemble modèle réalisé à partir d’un z0 perturbé et spatialisé. La méthode a d’abord été testée en expérience jumelle puis avec des observations réelles. Les cartes du paramètre z0, optimisés, réalisées avec des observations réelles, ont ensuite été utilisées dans le modèle sur une autre période et les résultats ont été comparés avec des observations sur la zone. Enfin, des expériences jumelles ont été mises en place pour corriger l’état modèle. Deux méthodes ont été comparées, une prenant en compte la basse fréquence en filtrant la marée des données et du modèle pour réaliser l’analyse ; l’autre prenant en compte tout le signal. Avec ces expériences, on a tenté d’évaluer la capacité du filtre à contrôler à la fois la partie observée du vecteur d’état (courant de surface) et la partie non-observée du système (température de surface). / The Iroise Sea has been observed since 2006 by High Frequency (HF) radars, which estimate surface currents. These measurements offer high resolution and high frequency to capture the dynamics of the coastal domain. This thesis aims at designing and applying a method of assimilation of these data in a realistic numerical model to optimize the bottom friction and to correct the model state in order to improve the representation of the residual tidal circulation and the positions of the Ushant fronts in the Iroise Sea. The method of data assimilation used is the Ensemble Kalman Filter. The originality of this method is the use of a stochastic modeling to estimate the model error. First, ensemble simulations were carried out from the perturbation of various model parameters which are the model error sources: meteorological forcing, bottom friction, horizontal turbulent closure and surface roughness. These ensembles have been explored in terms of dispersion and correlation. An Ensemble Kalman smoother was used to optimize the bottom friction (z0) from the surface current data and from an ensemble produced from a perturbed and spatialized z0. The method is tested with a twin experiment and then with real observations. The optimized maps of parameter z0, produced with the real currents, were used in the model over another period and the results were compared with independent observations. Finally, twin experiments were conducted to test the model state correction. Two approaches were compared; first, only the low frequency, by filtering the tide in the data and in the model, is used to perform the analysis. The other approach takes the whole signal into account. With these experiments, we assess the filter's ability to control both the observed part of the state vector (currents) and the unobserved part of the system (Sea surface Temperature).