Modeling the glacial lake outburst flood process chain in the Nepal Himalaya: reassessing Imja Tsho's hazard
Abstract. The Himalayas of South Asia are home to many glaciers that are retreating due to climate change and causing the formation of large glacial lakes in their absence. These lakes are held in place by naturally deposited moraine dams that are potentially unstable. Specifically, an impulse wave generated by an avalanche or landslide entering the lake can destabilize the moraine dam, thereby causing a catastrophic failure of the moraine and a glacial lake outburst flood (GLOF). Imja-Lhotse Shar Glacier is amongst the glaciers experiencing the highest rate of mass loss in the Mount Everest region, in part due to the expansion of Imja Tsho. A GLOF from this lake may have the potential to cause catastrophic damage to downstream villages, threatening both property and human life, which prompted the Nepali government to construct outlet works to lower the lake level. Therefore, it is essential to understand the processes that could trigger a flood and quantify the potential downstream impacts. The avalanche-induced GLOF process chain was modeled using the output of one component of the chain as input to the next. First, the volume and momentum of various avalanches entering the lake were calculated using Rapid Mass Movement Simulation (RAMMS). Next, the avalanche-induced waves were simulated using the Basic Simulation Environment for Computation of Environmental Flow and Natural Hazard Simulation (BASEMENT) model and validated with empirical equations to ensure the proper transfer of momentum from the avalanche to the lake. With BASEMENT, the ensuing moraine erosion and downstream flooding was modeled, which was used to generate hazard maps downstream. Moraine erosion was calculated for two geomorphologic models: one site-specific using field data and another worst-case based on past literature that is applicable to lakes in the greater region. Neither case resulted in flooding outside the river channel at downstream villages. The worst-case model resulted in some moraine erosion and increased channelization of the lake outlet, which yielded greater discharge downstream but no catastrophic collapse. The site-specific model generated similar results, but with very little erosion and a smaller downstream discharge. These results indicated that Imja Tsho is unlikely to produce a catastrophic GLOF due to an avalanche in the near future, although some hazard exists within the downstream river channel, necessitating continued monitoring of the lake. Furthermore, these models were designed for ease and flexibility such that local or national agency staff with reasonable training can apply them to model the GLOF process chain for other lakes in the region.