Description: Modern erosion of the Himalaya, the world's largest mountain range, transfers huge dissolved and particulate loads to the ocean. It plays an important role in the long-term global carbon cycle, mostly through enhanced organic carbon burial in the Bengal Fan. To understand the role of past Himalayan erosion, the influence of changing climate and tectonic on erosion must be determined. Here we use a 12 Myr sedimentary record from the distal Bengal Fan (Deep Sea Drilling Project Site 218) to reconstruct the Mio-Pliocene history of Himalayan erosion. We use carbon stable isotopes (d13C) of bulk organic matter as paleo-environmental proxy and stratigraphic tool. Multi-isotopic - Sr, Nd and Os - data are used as proxies for the source of the sediments deposited in the Bengal Fan over time. d13C values of bulk organic matter shift dramatically towards less depleted values, revealing the widespread Late Miocene (ca. 7.4 Ma) expansion of C4 plants in the basin. Sr, Nd and Os isotopic compositions indicate a rather stable erosion pattern in the Himalaya range during the past 12 Myr. This supports the existence of a strong connection between the southern Tibetan plateau and the Bengal Fan. The tectonic evolution of the Himalaya range and Southern Tibet seems to have been unable to produce large re-organisation of the drainage system. Moreover, our data do not suggest a rapid change of the altitude of the southern Tibetan plateau during the past 12 Myr. Variations in Sr and Nd isotopic compositions around the late Miocene expansion of C4 plants are suggestive of a relative increase in the erosion of High Himalaya Crystalline rock (i.e. a simultaneous reduction of both Transhimalayan batholiths and Lesser Himalaya relative contributions). This could be related to an increase in aridity as suggested by the ecological and sedimentological changes at that time. A reversed trend in Sr and Nd isotopic compositions is observed at the Plio-Pleistocene transition that is likely related to higher precipitation and the development of glaciers in the Himalaya. These almost synchronous moderate changes in erosion pattern and climate changes during the late Miocene and at the Plio-Pleistocene transition support the notion of a dominant control of climate on Himalayan erosion during this time period. However, stable erosion regime during the Pleistocene is suggestive of a limited influence of the glacier development on Himalayan erosion.

Description: Soft-sediment deformation structures have been analyzed at six sites of the Kathmandu valley. Microgranulometric study (this Supplement and Fig. 3B of Mugnier et al., Tectonophysics, 2011) reveals that silty levels (60 to 80% silt) favor the development of soft-sediment deformation structures, while sandy levels (60 to 80% sand) are passively deformed. Nonetheless well sorted sand levels (more than 80% sand) generate over-fluid pressure during compaction if located beneath a silty cap, leading to fluidization and dike development. 3-D geometry of seismites indicates a very strong horizontal shearing during their development. Using a physical approach based on soil liquefaction during horizontal acceleration, we show that the fluidization zone progressively grows down-section during the shaking, but does not exactly begin at the surface. The comparison of bed-thickness and strength/depth evolution indicates three cases: i) no soft-sediment deformation occurs for thin (few centimeters) silty beds; ii) the thickness of soft-sediment deformation above sandy beds is controlled by the lithological contrast; iii) the thickness of soft-sediment deformation depends on the shaking intensity for very thick silty beds. These 3 cases are evidenced in the Kathmandu basin. We use the 30 cm-thick soft-sediment deformation level formed during the 1833 earthquake as a reference: the 1833 earthquake rupture zone extended very close to Kathmandu, inducing there MMI IX-X damages. A 90 cm-thick sediment deformation has therefore to be induced by an event greater than MMI X. From a compilation of paleo and historic seismology studies, it is found that the great (M ~ 8.1) historical earthquakes are not characteristic of the greatest earthquakes of Himalaya; hence earthquakes greater than M ~ 8.6 occurred. Kathmandu is located above one of the asperities that laterally limits the extent of mega-earthquake ruptures and two successive catastrophic events already affected Kathmandu, in 1255 located to the west of this asperity and in ~ 1100 to the east.