The legacy of megafloods in the eastern Himalaya: from erosion to deposition

Abstract

Glacial lake outburst megafloods are infrequent, high magnitude (discharge ≥10^6 m^3/s) events that are uniquely efficient agents of landscape change, despite being short lived. Much of our understanding of megaflood hydraulics and their subsequent erosion and deposition comes from the study of megafloods in low relief systems. However, recent research suggests that outburst floods will behave differently in steep, mountainous terrain than they would in relatively flat landscapes. The Yarlung-Siang River (YSR) in the eastern Himalaya provides an excellent case study to examine the impact of megaflooding in high-relief topography because of its history of extreme megafloods throughout the Quaternary. In this dissertation, I investigate outstanding questions regarding the legacy of eastern Himalayan megafloods using interdisciplinary methodologies. Hydraulic simulations of an ancient megaflood sourced from the Tibetan Plateau show that megaflood erosional patterns will differ from those of the modern river due to dynamic interactions between flood hydraulics and the rugged mountain landscape which is far above the inundation of the modern or paleo YSR river channel, even in extreme meteoric flooding. If megafloods are a significant contributor to erosion in this region, as many have suggested in prior studies, then simply using annual stream power estimates of erosion is insufficient to capture the impact of megaflooding. Beyond erosion, these simulations revealed an extensive potential for deposition of sediment, both fine-grained (sand and silt) and coarse-grained (decameter scale boulders). Additionally, hydraulic simulations predict that the Siyom River, a large tributary of the YSR, experienced up to 60 km of with megaflood backflooding. Remote topographic analysis and field geomorphology, sedimentology, and geochronology in fluvial terraces in the valley reveal evidence of extensive deep stagnant and turbulent water, which I interpret as megaflood backflooding deposits. Radiocarbon geochronology reveals megaflood deposit ages ~9-11 ka in the Siyom River valley, which coincide with the timing of a paleolake on the Tibetan Plateau thought to be a source of megaflooding down the YSR. The inundation of a tributary during a megaflood introduces complexities not typically considered in predictive models of river evolution in mountain landscapes. Finally, hydraulic simulations indicate a potential for the deposition of ≥4m diameter boulders in boulder bars in the YSR channel during a megaflood, a finding that is consistent with field and remote sensing observations. Using a 1D landscape evolution model that incorporates the impact of immobile boulders on channel evolution, I determined that megaflood-deposited boulders create hundreds of meter-scale knickpoints which can persist for tens of thousands of years along the longitudinal profile of a river that experiences a megaflood. If that river experiences a sequence of megafloods, as is likely at the end of a glacial cycle, then those knickpoints will be refreshed and the change to the channel will compound, leading to a stepped longitudinal river profile—a potential signature of megaflooding in the landscape. Altogether, this dissertation advances understanding of how megafloods function in steep mountain landscapes and outlines the complex legacy of these immense floods, which includes both extensive erosion and deposition.