Abstract
Several major earthquakes have struck the 2500 km long Himalayan range during the past century. However, both the maximum size of such earthquakes and the probability of occurrence of a magnitude 9 megaquake are still a mater of debate. In this framework it is of key importance to investigate the lateral segmentation of the Himalayan arc in order to deepen our understanding of the mechanisms that control such events as well as their extents. Besides the remarkably uniform tectonic set-up and the fact that most studies on the Himalayas have focused on Central Nepal, several indications of along strike changes can be gathered.This thesis brings new constraints on lateral variations of crustal structure and of deformation between western Nepal (~80°E) and eastern Bhutan (~92°E).The first part of this thesis focuses on the lateral variations of crustal structure. In the past three years we have carried out four land gravity campaigns. We have established 366 new gravity measurement points to fill the data gaps in central and western Nepal as well as in Bhutan. This new dataset, together with available seismological data, is used to constrain thermo-mechanical models of the bending of the India plate underneath the Himalayas. While the inferred crustal geometry does not exhibit major along-strike variations over central and eastern Nepal, the shorter wavelength flexure of the lithosphere in Bhutan is associated with a weaker upper mantle rheology.The second part focuses on the lateral variations of recent deformation in the Himalayas. Fifteen years of paleoseismological investigations have extended the catalogue of major Himalayan events over the last millennium. Combining these information with secular loading, we assess the spatial and temporal stress changes on the Main Himalayan Thrust along the orogen over the last nine centuries. Our calculations indicate that inter-seismic loading has now nearly overcome the Coulomb stress decrease caused by the great ~1100 medieval earthquake. Our results also point out that the 1897 Shillong plateau earthquake does not have a major influence on the stress accumulated on the Main Himalayan Thrust since then. In order to better characterize active tectonics in the Bhutan Himalayas where no studies were done so far, we carried out a morphotectonic analysis in the south-central part of the kingdom. We show that the same amount of Holocene deformation is accommodated on the frontal thrust in Bhutan as on the neighbouring portions of the Himalayan arc. We also find evidences for two major (M>8) earthquakes on this thrust in the Bhutan Himalayas during the last millennium. Our results therefore show that Bhutan cannot be considered as a seismic gap. They also challenge the interpretation of one single ~1100 medieval mega-event that would have ruptured the Main Frontal Thrust from central Nepal to eastern Bhutan.