An earthquake-triggered avalanche in Nepal in 2015 was exacerbated by climate variability and snowfall anomalies

Large rock-ice avalanches are fast, long-runout geophysical mass flows composed of rock and ice that can generate destructive air blasts. In earthquake-prone high mountains, climate change appears to exacerbate these hazards by influencing both snow cover and temperature. Prior risk assessments have rarely considered how climate-driven factors affect avalanche runout and flow regime transitions, including the formation of hazardous air blasts. Two climate-related phenomena are central: (1) snowfall anomalies at high elevations that produce thick snow covers, enhancing snow entrainment, lubricating motion, and fostering powder cloud formation; and (2) warming, which alters avalanche thermodynamics, meltwater production, and flow regimes—effects often ignored by models that treat the sliding mass as thermally insulated from ambient air. The 25 April 2015 Langtang rock-ice avalanche (triggered by the Mw 7.9 Gorkha earthquake) provides a striking case: thick anomalous snow cover and warm conditions coincided with severe destruction primarily from the air blast. This study reconstructs the avalanche and associated air blast, then quantifies how varying snow cover depth and ambient air temperature affect runout and flow regime, clarifying the climate controls on destructive potential.

The study builds on evidence that rapid climate change increases instability in glacial and periglacial mountains and may exacerbate rock-ice avalanche hazards. Observations indicate decreasing snowfall duration but increasing snowfall intensity in many high-altitude regions, producing thick snow covers that are easily entrained, increase avalanche volume, reduce friction, and favor the formation of rock-ice-snow powder avalanches capable of powerful air blasts. Previous models often neglect ambient air interactions, treating avalanches as thermally insulated despite clear sources of heat (frictional shearing, entrainment, particle collisions) and the porous, turbulent nature of avalanches that promotes air intake and heat exchange. Prior work on the Langtang event attributed the destruction to snow cover anomalies and documented extensive air-blast impacts, including village destruction and forest blowdown. Studies on tree breakage by air blasts, avalanche thermomechanics, and powder cloud dynamics inform the present modeling approach. The paper addresses a gap in hazard assessment by explicitly quantifying how snowfall anomalies and warming (including diurnal temperature variation) influence avalanche runout, lubrication via meltwater, and air blast dynamics.

Study case and data: The 25 April 2015 Langtang rock-ice avalanche was analyzed using field investigations, satellite imagery, digital surface models, and meteorological data from nearby stations (Kyanjing at 3862 m, Yala Base Camp at 5058 m, and a pluviometer at 4831 m). Observations documented multi-source ice release above 6000 m, a snowline at 4000–4500 m, and thick snow cover due to four anomalous snowfall events in winter 2014–2015. Pre- and post-event DEMs indicated a total involved volume of 14.38×10^6 m³ with 6.95×10^6 m³ deposited in the valley. Air temperatures around the event time indicated warm conditions; a lapse rate of ~0.75 °C per 100 m was applied. Snow depth at Yala (1.5 m) and snowline location were used to estimate and calibrate a snow cover gradient and release-area snow depth. Numerical framework: Simulations were performed with the RAMMS::RockIce model, a depth-averaged two-layer framework representing (i) the dense avalanche core (granular rock-ice-snow-water mixture) and (ii) the powder cloud air blast. The core module includes entrainment of path materials (snow, rock/debris), thermomechanical processes with partitioning of shearing work into heat and granular fluctuation energy, meltwater production from snow/ice melting, and heat exchange with ambient air. The powder cloud module models a turbulent dust-laden air blast (rock and ice components) with momentum input from the core, gravity, air entrainment, turbulence production and decay, and drag. Vertical profiles for velocity and density define dynamic pressure estimates. Tree-breakage was assessed following an established bending-stress criterion. Scenario design and calibration: A baseline reconstruction used a release-area snow depth estimated at 3 m with a snow cover gradient of −0.15 m per 100 m, yielding a simulated initial ice volume of 3.65×10^6 m³ and total entrainment of 11.20×10^6 m³ (total 14.85×10^6 m³), matching observations within 5% error. Sensitivity scenarios varied: (1) snow cover depth at the release area from 0 to 3 m (gradient fixed at −0.15 m per 100 m) while holding air temperature fixed; and (2) ambient air temperature at 3862 m from −1 to 19 °C with a constant lapse rate of 0.75 °C per 100 m while holding snow cover fixed. RAMMS adjusted snow distribution by slope, curvature, and elevation. Modeled outputs included avalanche core runout, velocity, deposit distribution, air blast dynamic pressure fields, meltwater mass and water content within the core, and predicted tree-breakage extent. Key model features and assumptions: The Voellmy-type, process-based rheology allowed friction reduction with increasing fluctuation energy and water content; lubrication effects were parameterized via an ever-decreasing Coulomb friction with water content. Heat transfer between core particles and ambient air used an experimentally based correlation (sphere assumption) sensitive to particle size; representative particle radii were assigned based on prior experience (snow ~7 cm, ice ~10 cm, rock ~30 cm). The cloud model incorporated both laminar and turbulent contributions to air entrainment and drag, with a turbulence decay parameter controlling pressure magnitudes. DEMs from SPOT imagery defined the terrain for simulations.

Reconstruction of the Langtang event: - Release and entrainment: Simulated released ice volume 3.65×10^6 m³; entrained snow and rock 11.20×10^6 m³; total 14.85×10^6 m³ (within ~5% of observed initial 3.50×10^6 m³ and total 14.38×10^6 m³). - Deposits and velocities: Two main deposit areas—platform at ~4500 m a.s.l. and Langtang Valley—with deposit depths >30 m in the valley, consistent with observations. Maximum core velocity >90 m/s at 5000–5500 m a.s.l.; velocity ~57 m/s passing the valley before striking the opposite mountain toe, consistent with run-up estimates (~63 m/s). - Air blast pressures and impacts: Modeled mean dynamic pressure over Langtang village >15 kPa with maxima up to 28 kPa at a representative point; on the opposite mountainside mean ~10 kPa, max ~18 kPa at the toe, decreasing upslope. Tree-breakage modeling (Abies and Rhododendron, average diameter ~0.16 m) predicted ~0.8 km² of damage, extending ~1 km along the valley and ~550 m up the slope, matching observations. Effect of snow entrainment (snow depth at release 0–3 m): - Mobility and runout increase markedly with thicker snow cover; with 3 m, modeled dynamics match observed runout and damage. - Without snow entrainment (0 m), the avalanche stops before the opposite mountain; limited deposition reaches the valley. Air blast impact area and pressures shrink substantially: mean pressure at the village ~2.5 kPa with minimal tree damage—far less destructive than observed. - Conclusion: Entrained snow is a primary driver of dispersed powder cloud formation, increased avalanche volume, and enhanced destructive air blast in Langtang. Effect of ambient temperature (−1, 9, 19 °C at 3862 m, fixed lapse rate): - Cold scenario (−1 °C): Heat loss to cold air restricts melting; final meltwater ~74,000 t; maximum core water content ~600 mm m⁻³; comparatively smaller valley deposition than observed. - Warm scenario (19 °C): Meltwater ~170,000 t (>2× cold); maximum water content >1800 mm m⁻³; strong lubrication at the front leads to a fluid-like regime in the valley and a long runout (>1.5 km downstream of the valley). - Timing: Meltwater appears ~20 s after initiation and accumulates in the frontal lobe of deposits; warm daytime conditions intensify heat exchange and lubrication compared with cold/nighttime conditions. Overall: Thick anomalous snow cover and warm ambient temperatures jointly amplified the avalanche’s mobility, runout, and air blast destructiveness, explaining the extensive damage despite the village not being struck by the dense core.

The findings show that climate-linked factors substantially modulate rock-ice avalanche dynamics and air blast hazard. Snow entrainment thickens the flow, raises normal and shear stresses, and increases shearing work, producing more fluctuation energy and heat. Combined with snow’s low friction, this reduces effective resistance and increases mobility. Entrained snow also supplies mass and momentum to the powder cloud, enlarging impact area and dynamic pressure; without snow entrainment, destructive air-blast effects at Langtang would have been minor. Ambient temperature governs heat exchange between the porous, turbulent avalanche core and entrained air. Warm air enhances meltwater generation, strongly lubricating the flow and promoting fluid-like behavior and longer runout; cold air hinders melting, shortening runout and reducing damage potential. The results indicate that not only long-term warming trends but also short-term diurnal temperature variations can materially alter destructive potential. For hazard assessment in high-altitude regions, incorporating snow cover state and ambient temperature into dynamic modeling is essential to realistically predict runout, powder cloud formation, and air blast pressures, thereby improving risk management for settlements and infrastructure.

At noon on April 25, 2015, the Mw 7.9 Gorkha earthquake triggered a large rock-ice avalanche in Nepal’s Langtang Valley. Although the dense core did not directly strike the village, the resultant air blasts devastated the settlement and nearby forest, causing over 350 fatalities. Through field investigations and advanced numerical modeling, the study identifies two principal amplifiers of the disaster: (1) anomalously thick snow cover that promoted substantial snow entrainment, increased avalanche volume, reduced friction, and fostered a dispersed powder avalanche and powerful air blast; and (2) warm ambient air that intensified heat exchange, increased meltwater production, and lubricated the flowing mass, extending runout. These mechanisms highlight the critical need to include snow cover and air temperature—both seasonally and climatically variable—in hazard analyses of high-altitude rock-ice avalanches. Small variations in environmental conditions can markedly alter mobility and destructive potential, suggesting that climate variability can either mitigate or exacerbate outcomes in similar future events.

- Model simplifications: Heat transfer between particles and ambient air is based on correlations for spheres, neglecting particle shape and porosity effects. Representative particle sizes (snow, ice, rock) were assigned due to lack of detailed size distributions. - Thermodynamic assumptions: The mean core temperature is assumed not to exceed the melting point until all snow/ice melt; parameterizations of friction reduction with water content are simplified. - Case specificity: Results are calibrated to a single, well-documented event (Langtang 2015); generalization to other settings requires caution. - Environmental inputs: Ambient temperature fields use a constant lapse rate and station-based records; spatial-temporal variability and microclimate effects are simplified. - Scope: The study focuses on changes in runout and flow regimes under varying snow cover and temperature, not on climate-driven changes to event occurrence or triggering probabilities.