1The Himalayan Frontal Thrust of India is not blind Senthil Kumar1, Steven G. Wesnousky1, Thomas K. Rockwell2, Richard W. Briggs1, Vikram C. Thakur3, R. Jayangondaperumal3 1Center for Neotectonic Studies, University of Nevada, Reno NV 89577, USA 2Department of Geological Sciences, San Diego State University, San Diego CA 92182, USA 2Wadia Institute of Himalayan Geology, Dehra Dun, UA 248 001, INDIA ABSTRACT We report evidence of surface rupturing earthquakes at Chandigarh, Kala Amb, Rampur Ganda, Dehradun, Lal Dhang, and Ramnagar covering a distance of ~400 km along strike of the Himalayan Frontal Thrust (HFT) of India. Trench exposures across the HFT and radiocarbon ages place limits on the timing of the last surface rupture at each site, but Dehradun, during the last ~600 years: Chandigarh (1404 – ~1600 A.D.), Kala Amb (1424 – 1950 A.D.), Rampur Ganda (1222 – 1422 A.D.), Lal Dhang (1282 – 1632 A.D.), and Ramnagar (1263 – 1433 A.D.). The overlapping ages allow the interpretation that all sites ruptured simultaneously at about 1413 + 9 A.D. (or, between 1404 – 1422 A.D.). The interpretation is consistent with the large coseismic displacements observable at Rampur Ganda, Lal Dhang, and Ramnagar. At these sites, trench exposures and vertical separations of ~9 – 13 m are interpreted to indicate ~18 – 26 m of coseismic slip during the last surface rupture earthquake, assuming an average fault dip of 30o. The sites at Chandigarh and Kala Amb may also preserve evidence of a penultimate earthquake. Long-term displacement on the HFT has resulted in the occurrence of uplifted and truncated fluvial strath terraces along canyons of the Ghaggar, Markanda, Shajahanpur, Kosi and Nandaur Rivers. Minimum vertical uplift rates of the terraces are estimated at 5 to 11 mm/yr by dividing the elevation of2the terraces by the maximum radiocarbon age of the terrace abandonment. Dividing observed single event vertical displacements at Rampur Ganda, Lal Dhang, and Ramnagar by the estimated longer-term uplift rates indicates 1 – 2 thousand years would be needed to accumulate the slip released during the most recent surface rupture earthquake. Assuming an average dip of 30o for the HFT, the long-term uplift rates of 5 to 11 mm/yr equates to fault slip rates of 10 to 22 mm/yr and shortening rates of 9 to 19 mm/yr. Given that we observe clear evidence of surface rupture earthquakes and its long-term geologic expression in progressive and continued offset of fluvial terraces, the question arises as to why surface rupture has not been recorded in the well-documented major historical events of the 1905 Kangra earthquake, 1934 Bihar-Nepal earthquake, and 1950 Assam earthquake that occurred during the last century. We suggest on the basis of size and possible synchroneity of the most recent displacement recorded between 1404 and 1422 A.D. (or, 1413 + 9 A.D.) that the earthquake recorded in the trenches is larger than historical earthquakes and indicate a potential for sections of the HFT to rupture simultaneously along lengths of the HFT greater than the ~400 km we have studied. INTRODUCTION The ongoing collision of India into Eurasia has resulted in three major earthquakes along the Himalayan front during the past ~100 years (Seeber and Armbruster, 1981) (Figure 1A). From east to west, the sequence includes the 1905 Kangra earthquake (Ms ~7.8), the 1934 Bihar-Nepal earthquake (M = 7.7 + 0.2), and the 1950 Assam earthquake (Mw ~8.6) (Pandey and Molnar, 1988; Ambraseys and Bilham, 2000; Bilham, 2001).3Although none of the earthquakes are reported to have produced primary surface rupture (Seeber and Armbruster, 1981), it has generally been assumed on the basis of isoseismals and location that the earthquakes are the result of slip on the Himalayan Frontal Thrust (HFT). Lack of primary surface rupture during the major historical earthquakes has led previous workers to attribute their occurrence to a blind thrust, whereby strain release is expressed as anticline growth rather than primary surface rupture or coseismic surface rupture (Stein and Yeats, 1989; Yeats et al., 1992; Yeats and Thakur, 1998). In this paper, we describe the late Quaternary expression of the HFT at seven sites and demonstrate that the HFT is not blind but, rather, an emergent fault system. We then discuss the mechanical implications that arise if the major historical earthquakes along the HFT have been the result of slip on the HFT but have not produced coseismic surface rupture. Finally, we discuss observations that suggest the HFT has and will produce earthquakes of size greater than those observed historically, perhaps as large as the greatest thrust earthquakes observed along the major convergent oceanic plate boundaries. REGIONAL TECTONICS The Himalayan mountain belt formed as a result of the collision of India into Eurasia (Figure 1A) and has accommodated ~2000-3000 km of convergence along the ~2500 km length of plate boundary since the Eocene (Molnar and Tapponnier, 1977). The collision has produced three major crustal-scale south verging thrust faults that strike the length of the Himalayan arc (Figure 1A). The northernmost is the structurally highest and oldest Main Central Thrust (MCT) system, which dips 30o to 45o northward and marks the contact4between the High Himalaya and the Lesser Himalaya (Gansser, 1964). The age of the youngest deformation in the MCT shear zone is unknown (Hodges, 2000). Seeber and Gornitz (1983) suggested that at least some segments of the MCT might still be active based on a general correlation of the MCT trace with distinctive knickpoints in the gradients of antecedent rivers that drain the southern flank of the Himalaya. However, the MCT has not been observed to cut Quaternary deposits and is generally considered inactive (Nakata, 1989). South of the MCT, the south verging Main Boundary Thrust (MBT) forms a series of north dipping thrust faults that mark the contact between the predominantly pre-Tertiary Lesser Himalayan sediments and the Tertiary and Quaternary sub-Himalayan sediments. The MBT is clearly expressed as a fault in bedrock along nearly its entire length, and in places transports pre-Tertiary to Quaternary Lesser Himalayan and sub-Himalayan sediments over younger Quaternary deposits (Nakata, 1972; Nakata, 1989; Valdiya, 1992). The MBT along the majority of the arc is expressed by south facing scarps and is locally expressed by north facing