<p>River-blocking snow avalanches pose a significant compound hazard in high-mountain environments by obstructing river channels, creating temporary lakes, and triggering downstream floods upon dam failure. This study presents a computational analysis of the 2024 avalanche event near Tholang village in the Lahaul–Spiti region of the Indian Himalayas, which temporarily dammed the Chenab River. A depth-averaged model incorporating Mohr–Coulomb, Voellmy–Salm and Pouliquen–Forterre rheologies was solved using an open-source finite volume code. The model was first validated against the Chowkibal–Tangdhar (CT) avalanche site in India. Adaptive mesh refinement reduced runtime by about 90% while retaining near fine-grid accuracy. For the Tholang event, mobility was controlled primarily by basal resistance and release-location uncertainty. Within the Mohr–Coulomb closure, increasing the bed friction angle from <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(15^\circ \)</EquationSource> </InlineEquation> to <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(27.5^\circ \)</EquationSource> </InlineEquation> shortened runout distance by <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(17\%\)</EquationSource> </InlineEquation>, reduced peak mean velocity by <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(40\%\)</EquationSource> </InlineEquation>, and increased maximum deposition depth by <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(50\%\)</EquationSource> </InlineEquation>. Release position also significantly affected avalanche behavior, with runout distance decreasing by up to 14% and peak velocity reducing by 15–25% depending on the friction angle. Voellmy–Salm simulations showed the expected opposing roles of dry friction and turbulent drag: increasing <InlineEquation ID="IEq6"> <EquationSource Format="TEX">\(\mu \)</EquationSource> </InlineEquation> lowered peak mean velocity by up to <InlineEquation ID="IEq7"> <EquationSource Format="TEX">\(40\%\)</EquationSource> </InlineEquation> and increased peak depth by <InlineEquation ID="IEq8"> <EquationSource Format="TEX">\( 12\%\)</EquationSource> </InlineEquation>, whereas increasing <InlineEquation ID="IEq9"> <EquationSource Format="TEX">\(\xi \)</EquationSource> </InlineEquation> from 1000 to 2000 <InlineEquation ID="IEq10"> <EquationSource Format="TEX">\(\mathrm{m\,s^{-2}}\)</EquationSource> </InlineEquation> increased peak mean velocity by <InlineEquation ID="IEq11"> <EquationSource Format="TEX">\( 45\%\)</EquationSource> </InlineEquation> and reduced peak depth by <InlineEquation ID="IEq12"> <EquationSource Format="TEX">\( 18\%\)</EquationSource> </InlineEquation>. Pouliquen–Forterre results were most sensitive to <InlineEquation ID="IEq13"> <EquationSource Format="TEX">\(L\)</EquationSource> </InlineEquation> and <InlineEquation ID="IEq14"> <EquationSource Format="TEX">\(\phi _1\)</EquationSource> </InlineEquation>, but remain less constrained for dry snow due to limited calibration data. The reconstructed Mohr–Coulomb scenario reproduced deposition into the Chenab River consistent with the reported temporary blockage. These results provide valuable insights for hazard mapping, mitigation planning, and the analysis of avalanche-induced flood risk in data-scarce Himalayan regions.</p>

Computational modeling of river-blocking snow avalanches: a case study in the Indian Himalayas

期刊论文
发表日期：

2026年4月2日

查看全文

Vishal Sharma,
Gaurav Bhutani

摘要

River-blocking snow avalanches pose a significant compound hazard in high-mountain environments by obstructing river channels, creating temporary lakes, and triggering downstream floods upon dam failure. This study presents a computational analysis of the 2024 avalanche event near Tholang village in the Lahaul–Spiti region of the Indian Himalayas, which temporarily dammed the Chenab River. A depth-averaged model incorporating Mohr–Coulomb, Voellmy–Salm and Pouliquen–Forterre rheologies was solved using an open-source finite volume code. The model was first validated against the Chowkibal–Tangdhar (CT) avalanche site in India. Adaptive mesh refinement reduced runtime by about 90% while retaining near fine-grid accuracy. For the Tholang event, mobility was controlled primarily by basal resistance and release-location uncertainty. Within the Mohr–Coulomb closure, increasing the bed friction angle from \(15^\circ \) to \(27.5^\circ \) shortened runout distance by \(17\%\) , reduced peak mean velocity by \(40\%\) , and increased maximum deposition depth by \(50\%\) . Release position also significantly affected avalanche behavior, with runout distance decreasing by up to 14% and peak velocity reducing by 15–25% depending on the friction angle. Voellmy–Salm simulations showed the expected opposing roles of dry friction and turbulent drag: increasing \(\mu \) lowered peak mean velocity by up to \(40\%\) and increased peak depth by \( 12\%\) , whereas increasing \(\xi \) from 1000 to 2000 \(\mathrm{m\,s^{-2}}\) increased peak mean velocity by \( 45\%\) and reduced peak depth by \( 18\%\) . Pouliquen–Forterre results were most sensitive to \(L\) and \(\phi _1\) , but remain less constrained for dry snow due to limited calibration data. The reconstructed Mohr–Coulomb scenario reproduced deposition into the Chenab River consistent with the reported temporary blockage. These results provide valuable insights for hazard mapping, mitigation planning, and the analysis of avalanche-induced flood risk in data-scarce Himalayan regions.