Reports / sciencemag.org/content/early/recent / 17 December 2015 / Page 1 / 10.1126/science.aac9865 The Mw 7.8 Gorkha earthquake that struck Nepal in April 2015 confirmed high seismic risk projections for the Himalayas (1, 2), inferred mostly from paleoseismological proxies of past events (3–6). Strong ground shaking caused the collapse of more than half a million homes, killing more than 8,500 people, and injuring more than 20,000. Landslides buried villages, roads, and river channels, consistent with coseismic impacts reported from other active mountain belts (7). Detailed sediment budgets show that landslides triggered by strong seismic ground shaking may rapidly detach millions to billions of cubic meters of rock, soil, and biomass, providing this material for subsequent entrainment by surface runoff and river flows (8). The resulting sediment pulses may fill even rapidly incising bedrock rivers by up to several tens of meters (9), cause protracted channel instability, impede access to emergency areas, destroy hydropower facilities, and compromise post-disaster rehabilitation efforts. Detailed mass balances of recent large earthquakes in China (10), Taiwan (11), Japan (12), and New Zealand (13) offer blueprints of how river networks recover from sudden input of excess sediment over several years to centuries. The high sediment transport rates in many mountain rivers, however, rarely sustain evidence of prehistoric earthquake- induced sedimentation pulses. In these cases depositional rec- ords of catastrophic aggradation in forelands are more instructive (14). Both of these lines of evidence for understanding river network recovery have remained elusive in the Himalayas. We present exceptionally well- preserved sedimentary archives that connect landscape-scale disturbance around Pokhara, Nepal, to at least three documented medieval mega- thrust earthquakes (3). The city of Pokhara (28°13´N, 83°59´E, 870 m asl) is located at the foot of the >8,000-m peaks of the Annapurna Massif in the Seti Khola valley (Fig. 1). This steep orographic gradient receives monsoonal rainfall of ~4,000 mm yr –1 (15). Pokhara’s geology features primarily Precambrian metamorphic sand- stones, shales, and dolomites. Other rocks include Paleozoic phyllites and schists of the Lesser Himalayan Series (LHS). Higher Himalayan Crystalline (HHC) rocks north of the Main Central Thrust (MCT) are Precambrian high-grade metamorphic quartzites, schists, and gneisses (16, 17). Marine calcareous metasediments of the Tethyan Sedimentary Series (TSS) prevail north of the South Tibetan Detachment Zone (18, 19). Pokhara sits on a large sediment fan built by the upper Seti Khola that drains the partly glaciated and debris-filled Sabche Cirque in the Annapurna Massif. The fanhead near the MCT grades into a ~60-km long flight of prominent terraces downstream that rise up to 140 m above the river bed, and envelop several LHS bedrock hills (20). The fan has three stratigraphic units called the Tallakot, Ghachok, and Pokhara Formations. We focus on the youngest Pokhara Formation composed of extensive coarse gravel sheets, numerous boulders >10 m in diameter, and thick debris-flow deposits. Digital topographic data (21) confirm damming of the more than a dozen tributaries along Seti Khola’s course of >60 km by Pokhara Formation Repeated catastrophic valley infill following medieval earthquakes in the Nepal Himalaya Wolfgang Schwanghart, 1 * Anne Bernhardt, 1 Amelie Stolle, 1 Philipp Hoelzmann, 2 Basanta R. Adhikari, 3 Christoff Andermann, 4 Stefanie Tofelde, 1 Silke Merchel, 5 Georg Rugel, 5 Monique Fort, 6 Oliver Korup 1 1 Institute of Earth and Environmental Science, University of Potsdam, Potsdam, Germany. 2 Institute of Geographical Sciences, Freie Universität Berlin, Berlin, Germany. 3 Institute of Engineering, Tribhuvan University, Kathmandu, Nepal. 4 Helmholtz-Zentrum Potsdam, German Centre for Geosciences GFZ, Germany. 5 Helmholtz-Zentrum Dresden-Rossendorf, Helmholtz Institute Freiberg for Resource Technology, Dresden, Germany. 6 CNRS UMR 8586 Prodig, Département de Géographie, Université Paris-Diderot-SPC, Paris, France. *Corresponding author. E-mail: w.schwanghart@geo.uni-potsdam.de Geomorphic footprints of past large Himalayan earthquakes are elusive, though urgently needed for gauging and predicting recovery times of seismically perturbed mountain landscapes. We present evidence of catastrophic valley infill following at least three medieval earthquakes in the Nepal Himalayas. Radiocarbon dates from peat beds, plant macrofossils, and humic silts in fine-grained tributary sediments near Pokhara, Nepal’s second largest city, match the timing of nearby M > 8 earthquakes in ~1100, 1255, and 1344 C.E. The upstream dip of tributary valley fills and X-ray fluorescence spectrometry of their provenance rule out local sources. Instead, geomorphic and sedimentary evidence is consistent with catastrophic fluvial aggradation and debris flows that had plugged several tributaries with tens of meters of calcareous sediment from a Higher Himalayan source >60 km away.