T31A-2845
Late Pleistocene to Historical Activity of the Hovd Fault (Mongolian Altay) from Tectonic Geomorphology and Paleoseismology

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Matthieu Alexis Ferry1, Davaasambuu Battogtokh2, Jean-Francois Ritz1, Robin Kurtz1, Regis Braucher3, Yann Klinger4, Munkhuu Ulzibat2, Odonbaatar Chimed2 and Sodnomsambuu Demberel2, (1)Géosciences Montpellier, Montpellier Cedex 05, France, (2)Mongolian Academy of Sciences, Institute of Astronomy and Geophysics, Ulaanbaatar, Mongolia, (3)CEREGE, Aix-en-Provence Cedex, France, (4)Institut de Physique du Globe de Paris, Paris, France
Abstract:
Active tectonics of western Mongolia is dominated by large strike-slip fault systems that produced great historical earthquakes: the Bulnay fault (Mw 8.1 and 8.4 in 1905), the Fu-Yun fault (Mw 8.0 in 1931) and the Bogd fault (Mw 8.1 in 1957). Central to these faults is the Altay Range that accommodates ~4 mm/yr of right-lateral motion. An earthquake of similar magnitude occurred in 1761 and has been attributed to the Hovd fault were seemingly fresh surface rupture was reported in 1985.

Here, we study the Ar-Hötöl section of the Hovd fault where surface rupture was described over a length of ~200 km. Detailed mapping of stream gullies from high-resolution Pleiades satellite images show a consistent pattern of right-lateral offsets from a few meters to ~500 m. At Climbing Rock, we surveyed a gully offset by 75 ± 5 m. The associated surface was sampled for 10Be profile which yields an exposure age of 154 ± 20 ka. The resulting minimal right-lateral slip rate ranges 0.4-0.6 mm/yr. However, drainage reconstruction suggests this surface may have recorded as much as 400 ± 20 m of cumulative offset. This implies the Hovd fault may accommodate as much as 2.6 ± 0.4 mm/yr, which would make it the main active fault of the Altay.

At a smaller scale, TLS topography documents offsets in the order of 2.5-5 m that likely correspond to the most recent surface-rupturing event with Mw ~8. A value of 2.8-3.0 m is reconstructed from a Uiger grave dated AD 750-840. At Marmot Creek and Small Creek, short drainages flow across the fault and form ponds against the main scarp. Two paleoseimic trenches reveal similar stratigraphy with numerous peat layers that developed over alluvial sands. The fault exhibits near vertical strands affecting pre-ponding units as well as a well-developed peat unit radiocarbon-dated AD 1465-1635. This unit likely corresponds to the ground surface at the time of the last rupture. It is overlain with a sandy pond unit on top of which a second continuous peat unit developed after AD 1795-1895. According to OxCal modelling of the whole stratigraphic sequence, the most recent event occurred between AD 1515 and AD 1865 with a favored time of occurrence at the end of the earlier peat development, i.e. after AD 1635. Considering known seismicity for that time window, our findings strongly support that the Hovd fault produced the AD 1761 earthquake.