As geodetic data is becoming more and more common there is a need for some simple models relating crustal deformation and seismicity that would provide some physical basis to help assess the frequency and size of major earthquakes. The Himalaya, which is the most active intracontinental mountain range on earth, is a most appropriate case study to address this question. In collaboration with Nepali colleagues from the Department of Mines and Geology and many French colleagues, we have carried on various investigations in the Himalaya of Nepal. Some idea of the seismic cycle could be derived that relate mountain building in the Himalaya and the large M>8 Himalayan earthquakes.
The seismic permanent network, initially settled around Katmandu was extended in 1994 and now consists of 23 short period stations. This network has revealed a belt of intense microseismic activity which follows the front of the high range [Pandey et al., 1994; 1999].
In the sub-Himalaya, abandoned Holocene terraces have recorded active folding with uplift rates up to 1.5 cm/yr. It shows that as much 21.5+/-1.5 mm/yr of horizontal shortening is accommodated by localized slip along the MFT [Lavé and Avouac, 2000]. Structural geology suggests that this fault emerges from a decollement at the top of the Indian basement, at 5-6 km depth, extends northwards beneath the Lesser Himalaya and roots into a mid-crustal sub-horizontal shear zone beneath the Higher Himalaya and southern Tibet that could be imaged from INDEPTH seismic experiments in southern Tibet. Incision rates along the major rivers across the Himalaya of Nepal are keeping with this geometry : little incision in the Lesser Himalaya and Southern Tibet, is observed and up to 5-10mm/yr in the Higher Himalaya, a pattern consistent consistent with 20-23mm/yr of slip along the ramp-and-flat geometry of the MHT [Lavé and Avouac, 2001]. By contrast, geodetic data shows that, during the interseismic period, horizontal shortening is mainly absorbed within a 100 km wide zone that spans over the Higher Himalaya, while deformation in the sub-Himalaya and southern part of the Lesser Himalaya seems negligible. It indicates that the fault is fully locked with horizontal shortening being mainly absorbed within a 100 km wide zone that spans over the Higher Himalaya. This information is reconciled from a mechanical FEM model in which we consider a lithospheric section with a realistic rheology, submitted to horizontal shortening [Cattin and Avouac, 2000]. The model accounts for surface processes and the thermal structure of the lithosphere. In this model, on the long term, provided that friction on the MHT is low (less than 0.2-0.3) shortening across the Himalaya is essentially accommodated by localized frictional slip along the MHT in the brittle upper crust and by ductile flow in the lower crust beneath the high range and southern Tibet . During the interseismic period, the MHT is fully locked from the sub-Himalaya to beneath the Higher Himalaya. Microseismic activity is enhanced in the zone of increasing Coulomb stress.
Interseismic stress build-up by elastic straining of the upper crust isprobably the main process responsible for the observed belt of microseismicity that can be traced along the front of the high range all along the Himalayas of Nepal. This is consistent with the geodetic data that also suggest that the MHT is locked everywhere including the seismic gap between the rupture areas of 1905 and 1934. It seems highly probable that this portion of the Himalayan arc also produces large recurrent earthquakes similar to the 1934 and 1905 events. Motion along the MHT is probably stick-slip as a result of recurring large earthquakes similar to the 1934 Bihar-Nepal or 1905 Kangra events.
Additional information comes from the result of a MT sounding experiment also carried on across the Himalaya of central Nepal. This experiment has revealed a deep conductor, with a resistivity of the order of 30 W.m, (resistivity across the Himalaya) that roughly coincides with the zone of intense microseismic activity but apparently extends to greater depth [Lemmonnier et al., 1999]. This pattern suggests that the high conductivity results from a well interconnected fluid phase that would be fed from metamorphic reactions as the Indian crust is thrusted under the mid-crustal ramp, channeled upward in the zone of interseismic straining.