Mechanisms of slab avalanche release and impact in the Dyatlov Pass incident in 1959

On the night of February 1, 1959, nine Russian hikers died under unexplained circumstances near the Kholat Saykhl slope in the northern Ural Mountains. Although hypothermia was identified as the main cause of death, several victims exhibited severe thoracic and skull injuries, and the campsite showed puzzling features, including a tent cut from the inside and an apparent lack of avalanche debris on what appeared to be a relatively mild slope. Numerous alternative explanations have been suggested, and official inquiries in 2019–2020 concluded that an avalanche was the most probable cause, yet did not provide a quantitative mechanism reconciling the counterarguments (low slope angle, scarce avalanche signs, unclear trigger, unusual injuries). This work addresses that gap by proposing and quantifying a delayed slab-avalanche release mechanism initiated by a man-made cut into a snow slab resting over a weak layer and progressively loaded by wind-transported snow due to strong katabatic winds and local topography. The study aims to determine whether such a mechanism could have occurred at Dyatlov Pass and whether it can explain the observed timing, lack of clear debris, and injury patterns.

Prior discussions of the Dyatlov incident emphasized apparent inconsistencies with a classic avalanche: average surface slope reportedly below 30°, minimal debris, and atypical injuries. Avalanche science, however, recognizes that slab avalanches can initiate on slopes as low as ~20° when dynamic (crack-face) friction is low, particularly in very cold conditions, and that weak layers such as depth hoar can fail under additional loading. Measurements have reported dynamic friction angles in snow as low as ~15°. Recent advances include models of delayed slab release driven by rate-dependent processes or additional loading (e.g., wind transport or snowfall) and detailed understanding of weak-layer fracture and slab mechanics. Numerical methods such as the Material Point Method (MPM) with elastoplastic constitutive laws have been used to simulate snow dynamics and impacts. Field compilations show a range of human-triggered avalanches on slopes around and below 30°, and observations indicate wind transport can create highly nonuniform loading controlled by local terrain features. These strands of literature inform the proposed mechanism coupling terrain-induced variable slab thickness, undercutting by the tent installation, and wind loading over time.

The study combines an analytical model of delayed slab-avalanche release with numerical simulations of slab dynamics and human impact using the Material Point Method (MPM). Analytical model: The slope is represented with an irregular local topography that yields a buried weak layer inclined more steeply (~28°) than the average surface (~23°), producing an upward-thinning slab. Installing the tent required cutting into the slab, reducing tensile support and locally perturbing stresses. Strong katabatic winds interacting with a shoulder above the tent promote wind-driven snow deposition upslope of the cut, thickening the slab over time. The model formulates shear and normal stresses in the weak layer as functions of slab geometry, weak-layer cohesion c and friction angle φ, slab and drifted-snow densities, and a spatially variable thickness profile that evolves with wind deposition. Closed-form criteria are derived for (i) immediate stability after the cut, (ii) bounds on φ allowing delayed release, and (iii) critical wind-deposited thickness at failure. Time-to-failure is computed from the critical wind-deposition thickness using a deposition rate Q inferred from weather conditions. Numerical simulations: A 2D/3D MPM framework with finite-strain elastoplasticity and a cohesive Cam Clay-type constitutive model simulates the failure and downslope motion of the small slab segment impacting occupants lying on the tent floor. Snow blocks of specified volumes (0.125–0.5 m³) and density (~400 kg/m³) are impacted at velocities consistent with the analytic pre-failure dynamics (~2 m/s). Thorax deformation is compared to automotive crash-test data to estimate injury severity (AIS scale). Geometry and parameters representative of the Dyatlov site (weak-layer inclination ~28°, initial slab thickness on the order of decimeters, cut length on the order of several meters, strong cold winds) are used to constrain the scenarios. Wind deposition flux Q is back-calculated to match forensic delay estimates (about 9.5–13.5 h).

- A delayed-release slab avalanche is physically plausible at the Dyatlov tent site when accounting for local terrain (shoulder above the tent), an upward-thinning slab over a weak layer, the undercut made to pitch the tent, and wind-transported snow loading the slab above the cut. - Effective weak-layer inclination is estimated around 28°, within the range of human-triggered avalanches when dynamic friction angles are low (reported as low as ~15° in very cold conditions), enabling failure on slopes below the commonly cited 30° threshold. - The analytical framework yields a delay to failure consistent with forensic timing: approximately 9.5–13.5 hours after the cut, for reasonable weak-layer strength parameters and wind deposition rates. - Back-calculated wind deposition fluxes correspond to average wind speeds on the order of a few meters per second (about 2–12 m/s), consistent with nearby station data under katabatic flow that night. - The failed slab size is limited by the local geometry; analytical estimates and MPM simulations predict a tensile failure length of about 5 m and a small runout, consistent with the scant avalanche signs reported 26 days later. - Simulated slab impact velocities around 2 m/s and block masses corresponding to snow blocks of 0.125–0.5 m³ at 400 kg/m³ density produce thorax deflections of approximately 28–34%, implying mostly non-fatal but severe-to-moderate thoracic injuries per AIS, aligning with autopsy findings (severe thorax/skull trauma without typical high-speed avalanche impact patterns). - The scenario explains the coexistence of severe injuries with an absence of extensive debris or long runout: victims were pinned between a small, dense slab and the tent floor in a confined space, rather than being swept into obstacles downstream.

The results reconcile the avalanche hypothesis with previously cited inconsistencies by introducing a physically quantified delayed-release mechanism. The key insight is that the average surface slope understated the effective weak-layer inclination due to irregular topography; the undercut to install the tent removed tensile support; and subsequent wind-driven deposition above the cut provided the additional load required to overcome weak-layer cohesion after several hours. This produces a small slab with limited runout and scarce observable debris weeks later, addressing the lack of classic avalanche signs. The MPM impact simulations demonstrate that a slow, dense slab impacting occupants in a confined tent space can produce significant thoracic injury without being fatal, matching the autopsy record and explaining the atypical injury pattern for avalanche victims. Beyond the Dyatlov case, the mechanism highlights how variable slab thickness from local terrain and time-dependent wind (or snowfall) loading can control delayed failures, suggesting broader relevance for storm- and wind-triggered slab avalanches. The analytical criteria and timing bounds offer a framework for assessing delayed-release potential when detailed site geometry, weak-layer properties, and meteorological data are available.

This study proposes and quantifies a delayed slab-avalanche release mechanism driven by wind-transported snow loading of an undercut, upward-thinning slab over a weak layer. The mechanism explains the Dyatlov Pass observations—timing of the event, minimal avalanche traces, and severe but largely non-fatal injuries—within a coherent physical model. The combined analytical framework and MPM simulations constrain the required wind deposition rates and slab geometry, and predict a small failed slab with limited runout and impact-induced injuries consistent with the forensic record. Future work should include field-scale validation of the delayed-release criteria under wind loading, improved in situ characterization of weak-layer friction and cohesion at very low temperatures, refined coupling of deposition models with high-resolution terrain and atmospheric flows, and systematic application of the framework to other wind-exposed sites to assess hazard under similar conditions.

The analysis relies on reconstructed geometry, weather conditions, and weak-layer properties rather than contemporaneous in situ measurements at the time of the event. Several parameters (e.g., weak-layer cohesion and friction, exact wind deposition flux, effectiveness of a possible parapet above the cut) are uncertain and were constrained indirectly. The terrain and loading are simplified (e.g., parabolic thickness profiles), and 2D/3D MPM simulations represent idealized blocks and material behavior. Observational constraints (e.g., lack of fresh avalanche signs due to a 26-day delay before site inspection and subsequent snowfall/wind) limit direct validation of slab dimensions and runout. Despite these uncertainties, converging analytical and numerical predictions support the plausibility of the proposed mechanism.