Paleoseismic evidence of great surface rupture earthquakes along the Indian Himalaya

Abstract

Toward understanding the relationship between strain accumulation and strain release in the context of the mechanics of the earthquake and mountain building process and quantifying the seismic hazard associated with the globes largest continental thrust system, we describe the late Quaternary expression and paleoseismic evidence for great surface rupture earthquakes at six sites along the Himalayan Frontal Thrust (HFT) system of India. Our observations span a distance of ~250 km along strike of the HFT. Uplifted and truncated fluvial terrace deposits resulting from the Holocene displacements on the HFT are preserved along canyons of the Ghaggar, Markanda, Shajahanpur, and Kosi Rivers. Dividing the elevation of the bedrock straths at each site by their ages yields estimates of the vertical uplift rate of ~4-6 mm/yr, which when assumed to be the result of slip on an underlying thrust dipping at ~20°-45° are equivalent to fault slip rates of ~6-18 mm/yr or shortening rates of ~4-16 mm/yr. Trench exposures reveal the HFT to fold and break late Holocene surface sediments near the cities and villages of Chandigarh, Kala Amb, Rampur Ganda, Lal Dhang, and Ramnagar. Radiocarbon ages of samples obtained from the displaced sediments indicate surface rupture at each site took place after ~A.D. 1200 and before ~A.D. 1700. Uncertainties attendant to the radiocarbon dating currently do not allow an unambiguous definition of the capping bound on the age of the displacement at each site and hence whether or not the displacements at all sites were contemporaneous. Trench exposures and vertical separations measured across scarps at Rampur Ganda, Lal Dhang, and Ramnagar are interpreted to indicate single-event displacements of ~11-38 m. Dividing the observed single-event vertical components of displacement by the estimated longer-term uplift rates indicates ~1330-3250 or more years should be required to accumulate the slip sufficient to produce similar sized displacements. Surface rupture appears to not have occurred during the historical 1905 Kangra (M_w = 7.7), 1934 Bihar-Nepal (M_w = 8.1), and 1950 Assam (M_w = 8.4) earthquakes, which also occurred along the Himalayan front. Yet we observe clear evidence of fault scarps and displacements in young alluvium and progressive and continued offset of fluvial terrace deposits along the HFT. We suggest on this basis and the size and possible synchroneity of displacements recorded in the trenches that there exists the potential for earthquakes larger than recorded in the historical record and with the potential to rupture lengths of the HFT greater than the ~250 km we have studied.