Un modèle général de d'évolution de l'océan indien

The last global synthesis of the evolution of the Indian ocean which satisfactorily describes the kinematics of the five major plates, up to the initial stages, dates back to Norton and Sclater [1979]. Although the most recent phases of evolution have been largely detailed, in particular in the proximity of the triple junction, the initial stages of evolution have been generally described by models quite speculative and generally only constrained by pairs of plates. The recentmost gravity data based on the altimetric satellites SEASAT, GEOSAT and ERS1 [Sandwell and Smith, 1992] give an homogenoeus image of sea-floor topography. These data, declassified to the south of 30°S, reveal previously unknowkn features of the sea-floor fabric and allow to further detail the geometry of fracture zones. This thesis presents a general model of evolution which takes account of these major physiographic features (fracture zones) as well as all the constraints provided by detailed works and concerning key areas such as the Kerguelen- Broken Ridge system and Mozambique continental shelf. The kinematic reconstructions, here proposed, are substantially based on the assumption that fracture zones are traces of relative motion between plates. Consequently segments of fracture zones used to define a kinematic stage are consistent with sets of small circles. Definition of kinematic stages and choice of the stage rotation poles have been performed by pursuing the objective of internal coherence of the model among the five plates. The approximations generated by such a global "stage by stage" model cannot resolve neither transitions from one stage to another nor the 2nd order kinematic rearrangements, which can be also masked by intraplate deformations. The whole work has been carried out by following four main steps : 1) collection and synthesis of marine and satellite geophysical data (bathymetry, gravity, altimetry, magnetics); 2) digitization of data; 3) computer generated paleogeographic maps based on kinematic reconstruction and 4) interpretation of the resulting maps and proposal of an evolutionary model which takes account of all the geological constraints. Because all the basins created between Africa and Antarctica are located south of 30°S, the reconstructions for each stage were precisely referred to Africa and Antartica plates whereas India and Australia were then included in the puzzle. Basins located NW of Australia present enough evidence to support the hypothesis of a "Greater India Plate" containing Block North India and the India s.s. plate. The adoption of this last hypoyhesis represents a strong constraint to be included into the general evolutionary model. It is possible to obtain, without intraplate deformation a remarkable morphological assemblage of different margins, which provide a beginning of kinematics reconstructions. The Socotra Bank (Arabian sea) and the Mozambique Ridge occupied slightly different position; as a consequence, they moved with respect to the adjacient fixed margins. Similarly micro continental blocks such as Aghulas Plateau, Sri-Lanka and Tasmania had exhibited, for a given timespan, a different evolution with respect to their major plates. Furthermore, a reconstruction gap in the south of Madagascar suggests that the northern part of Madagascar Ridge is of continental origin. The analysis of transform directions suggest the absence of significant changes in relative motion between magnetic anomaly 20 and the Present except for Africa and India. Between Africa and Antartica recent, satellite-derived data reveal major change in plates trajectories between anomalies 32 and 20-24. Another major change occurred little before anomaly 34 at around 90-93 My. The elongated topography of aseismic ridges in the Indian ocean, frequently related to hot spot traces can be associated with triple junction at a moment when one of the branch aded as 2nd order limit. Ninetyeast Ridge and Chagos-Laccadive Ridges coincide to original transform directions on turn associated to triple junction. In summary, the main results of this work can be concluded as following: - At the scale of the whole Indian ocean the kinematic evolution of plates can be described by major phases (each during 20-30 My) which are also recognizable in other oceans. In this context the occurrence of 2nd order rearrangements of regional extent did not significantly affect the global evolution. - The genesis of abnormal structures i.e. aseismic ridges, generally associated with the activity of hot spots, is strictly related to kinematic rearrangements. These structures can be viewed as the consequences and not the causes of kinematic reorganisations. - A major plate reorganisation occurring during middle Cretaceous (90-93 My) and documented all over the other oceans of the Earth is also confirmed in the Indian ocean. - The continental margin located to the west of the Australia correspond to a portion of one plate locked to the India plate and currently mostly disappeared beneath the Asiatic continent. The structures observables on this australian margin have recorded the motion of that ancestral plate, so supporting the hypothesis of a "Greater India" plate.