Déformation de la lithosphère océanique : Approche numérique par éléments finis, application à la plaque Indo-Australienne.

Abstract : Plate tectonics postulate that plates are rigid and that deformations focus on their boundaries, generally narrow in oceanic areas. Some oceanic plates appear to derogate to these tenets. The Indo-Australian plate, for instance, displays a high level of intraplate seismicity expressing a high level of deformation of the oceanic lithosphere. Earthquake focal mechanisms express compressional deformation in the Eastern part of the plate and extensional deformation East of the Central Indian Ridge and North of the South-East Indian Ridge. Inversion of kinematic data, assuming rigid plates, lead to the concept of composite plates comprising rigid components separated by wide and diffuse boundaries. Using the example of the Indo-australian plate, this dissertation investigates how the oceanic lithosphere behaves when submitted to horizontal stresses and the role of preexisting features such as submarine plateaus or fossil fracture zones on the deformation pattern. The first part of this work focuses on the distribution of the low magnitude seismic activity in the deformed areas using hydroacoustic records from 3 autonomous hydrophones in the Southern Indian Ocean. The objective was to identify and localize active structures in the deformed areas. Due the geometry of this temporary array, little new information could be obtained, however these data confirmed the reactivation of fossil fracture zones in the extensional areas. The discovery of seismic foreshocks preceding high magnitude earthquakes along active transforms of the East Pacific and Juan de Fuca ridges lead us to search for such precursors in our data. A foreshock is defined as an event occurring few hours before a major event within a 100 km radius around it. No such foreshocks were detected in our partial analysis of data from the Comprehensive Nuclear-Test-Ban Treaty (CTBT) hydroacoustic array. From the data of our temporary array, 13 events were identified as foreshocks few hours before major events on fracture zones ; however, with only three hydrophones, localization errors can not be estimated. Thus pursuing our investigations on these two issues would require to combine data from the two arrays, both to extend the data coverage and to improve the event localization. The second and main part of this study is devoted to model the deformation within the Indo-Australian lithosphere, both to test the validity of the composite plate concept and to evaluate the importance of pre-existing structures, such as submarine plateaus and fossil fracture zones, in the deformation pattern. To address this problem, we tested two numerical codes ADELI (Hassani et al., 1997) and SHELLS (Bird, 1989; Bird et Kong, 1994; Kong et Bird, 1995; Bird, 1999). The latter proved to be better adapted to our purpose. The plate is partitioned into a mesh of spherical triangular finite elements, and the numerical code calculates the deformation of each element. This code considers the age of the lithosphere and the presence of fracture zones or topography. Using the plate velocities as boundary conditions and some assumptions on the rheology of the lithosphere, SHELLS then calculates stress and strain on the lithosphere. The Indo-australian plate comprising several disjoint areas of deformation, we chose to model them separately : the first extensional area, between the Indian and Capricorn plates, comprises the Chagos Bank and a series of fracture zones oblique to the general N-S trend of the Central Indian Ridge ; the second extensional area, between the Capricorn and Australian plates, comprises the Amsterdam and St Paul plateau and a series of fracture zone perpendicular to the general WNW-ESE trend of the Southeast Indian Ridge ; the third area is dominated by a compressional deformation between the Indian, Capricorn and Australian plates, is bounded to the NE by the Sunda Trench and comprises the Ninetyeast Ridge and N-S oriented fracture zones of Paleocene to Eocene age in the Wharton and Central Indian basins. The boundary conditions of the models are derived from kinematic models assuming rigid plates. However, to account for the diffuse plate boundaries, we let one side of our four-sided meshes free and by doing so, we were able to reproduce the observed deformation pattern in the three modeled areas. Finally, we combined the three meshes into a single one to model the whole Indo-australian plate, where each of its sides is constrained by plate velocities. All our models predict deformation patterns consistent with the observed ones. Small discrepancies may occur and are discussed. All the models also emphasize the importance of preexisting structures of the lithosphere. For instance, the Chagos Bank focuses the deformation in the diffuse boundary between the Indian and Capricorn plates where magnitude-eight earthquakes occurred. Fossil fracture zones turn out to be weaknesses that accommodate a large fraction of the intraplate deformation. For instance, the Ninetyeast Ridge and the fracture zones immediately east of it explain the different behavior between the Central Indian Basin, where the deformation is essentially compressional in a N-S direction, and the Wharton Basin, where the deformation includes left-lateral shear and NW-SE compression. Our models also predicts rates of extension or shortening consistent with previous studies : east of the Central Indian Ridge, India moves away from Capricorn at ~2 mm/yr in a N-S direction ; west of the Sunda Trench, Australia moves towards India at ~11.5 mm/yr in a NWSE direction. To the first order, the concept of composite plate thus seems valid for the Indo-australian plate, however the shape of the diffuse boundaries may be more complex than anticipated.
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00680190
Contributor : Vincent Brandon <>
Submitted on : Sunday, March 18, 2012 - 9:28:53 AM
Last modification on : Friday, April 5, 2019 - 8:16:55 PM
Long-term archiving on : Monday, November 26, 2012 - 11:20:11 AM

Identifiers

  • HAL Id : tel-00680190, version 1

Collections

Citation

Vincent Brandon. Déformation de la lithosphère océanique : Approche numérique par éléments finis, application à la plaque Indo-Australienne.. Géophysique [physics.geo-ph]. Université de Bretagne occidentale - Brest, 2010. Français. ⟨tel-00680190⟩

Share

Metrics

Record views

914

Files downloads

3224