Erosional dam breach modelling for multi-glacial lake outburst process and downstream flood exposure assessment in the Eastern Himalayas

Abstract

Accelerating global warming has led to the retreat of glaciers and the concurrent expansion of glacial lakes in the Himalayas, increasing the risk of glacial lake outburst floods (GLOFs) with potential to destroy infrastructure and lives in the downstream river valley. In the Sikkim Himalayas, the potentially hazardous Gurudongmar lake complex (GLC) consists of four lakes containing approximately 148 x 10(6) m3 of water with an enlargement rate of 74 +/- 3 %. Therefore, evaluating possible GLOF hazards from GLC under the current scenario is very important. We have presented in this paper an integrated and robust GLOF model that combines the physical moraine- dam breach using TELEMAC 2D and SISYPHE with 1-D inundation modelling using HEC-RAS for the multi-lake complex to estimate the peak flood (m3/s), flood depth (m), and flow velocity (m/sec) and finally assess the downstream inundation exposure. The present study also incorporates different breaching scenarios calculated based on remote sensing and field survey data of surrounding moraine-dam parameters of the GLC. In the largest- cases of GLOFs, with an 80 % overtopping dam breach, the flood peak would reach 5833.7, 7684.0, and 8882.0 m3/s for scenarios 1C, 4C, and 5C, respectively. These floods would release a total volume of 39.9, 48.5, and 59.4 x 10(6) m3 of water for the three scenarios, respectively. The downstream flood exposure assessment is based on different land use types and flood heights observed across nine scenarios (labelled 1A-5C) for each settlement using a 15 x 15 m fishnet method. The results for all 19 settlement sites along the main Tista River channel show a logical trend of increasing potential exposures. However, when the water volumes from GL-2 and GL-1 or GL-3 and GL-1 are combined, the risk of a potential flood rises. An increase in the potential flood depth directly results in higher exposure to infrastructure. The GLOF exposure levels for each scenario at 19 settlement sites differed. However, the Thangu Valley and Chungthang town have the highest inundation exposure due to their size and potential losses. This research article contributes to developing moraine-dam erosional breach modelling for multi-glacial lake outburst scenarios and downstream flood exposure assessment in the Hindu Kush-Karakoram- Himalayan region.