RUSSIA / Night of 1–2 February 1959
Dyatlov Pass: testing the avalanche reconstruction
Nine hikers died in the northern Urals. This dossier examines what modern snow modelling can explain, which assumptions it needs and why a plausible reconstruction is not a recording of the event.
Northern Urals, Dyatlov Pass area · Regional orientation marker near the pass; not a precise tent position or a location of human remains.
Assessment: A physically supported avalanche explanation exists; the complete historical sequence remains uncertain.
Investigation and inspector’s assessment
Incident brief
The case concerns the deaths of nine hikers after their camp was established on a snowy slope in the northern Urals in February 1959. A 2021 study investigates whether a relatively small slab avalanche could have been released after a delay and caused severe injuries. Its proposed mechanism combines a cut in the slope, wind-driven accumulation and slab failure. The paper’s contribution is a physical feasibility argument, not the discovery of a film or instrument record of the accident. [D1]
What later fieldwork changes
The 2022 follow-up describes expeditions and observations of avalanches in the area. This weakens an absolute claim that slab avalanches cannot happen there. It cannot establish the precise loading, snow structure or timing on the night of the deaths. A regional possibility and a historical attribution remain different propositions. [D3]
Inspector’s assessment
The useful question is not whether a computer can make an avalanche happen. It is whether the model produces relevant outcomes within independently defensible ranges for terrain, snow strength and loading. Separate measured modern values from inferred historical ones. When an input has no 1959 measurement, show how changing it changes the result. Avoid treating the output of a tuned model as a unique solution.
Reconstruction limits
A physical trigger, an evacuation decision, subsequent movement and later exposure form different stages. Establishing one stage does not automatically explain all the others. The dossier avoids graphic images and unsupported allegations about named individuals. Original legal and medical records are not reproduced here, and the English scientific papers should not be represented as certified translations of the complete historical investigation.
Documentary chronology
- 1–2 February 1959
Fatal events occur during the expedition; the detailed sequence is reconstructed retrospectively. [D1]
- 1959 search and investigation
Recovery and inquiry supply the historical material later researchers discuss. [D1]
- 28 January 2021
The slab release and impact study is published. [D1]
- 2022
Follow-up fieldwork and its implications are published. [D3]
Dates distinguish the event, the creation of a record and its later publication. Approximate dates remain approximate.
Evidence ledger
Hypothesis review
Delayed release of a small snow slab
Supported physical scenario
- Supporting considerations
- The published model addresses delayed failure and impact under specified assumptions.
- Limits and objections
- Historical snow conditions and the full sequence are not directly measured.
- Discriminating test
- Repeat the calculation across input ranges and compare with all available site constraints.
A different initiating event followed by exposure
Alternative not specified by this dossier
- Supporting considerations
- Exposure can be a later stage regardless of the original reason for leaving shelter.
- Limits and objections
- A general possibility without a specific supporting record is not a competing reconstruction of equal strength.
- Discriminating test
- Identify dated evidence that discriminates between proposed triggers before adding a new narrative.
Inspector’s next steps
Build an assumption ledger
For each model parameter, identify the source, range and whether it was measured, inferred or selected. Prioritise parameters that change whether a slab releases.
Compare field and historical conditions
Create separate columns for modern surveys and 1959 evidence. Do not transfer a modern weather observation to the historical night.
Check the scope of the conclusion
List what the model directly addresses and what it leaves outside its scope. Keep these limits next to any summary or visual reconstruction.
These are proposed research tasks, not completed fieldwork or requests already sent to an archive.
Documents and source register
Original reports, scans, catalogue records and later analyses are identified separately. Alternate copies of the same document do not add independent evidence. Follow the named archive for full access information.
Gaume and Puzrin (2021): slab avalanche release and impact
Analytical and numerical models of a proposed sequence; includes methods and supplementary material.
Language: en
Open original source ↗2021 research paper — PDF
Reading copy of D1, not an independent source.
Language: en
Open original source ↗Gaume and Puzrin (2022): follow-up expeditions
Reports subsequent field observations bearing on avalanche feasibility.
Language: en
Open original source ↗2022 follow-up — PDF
Reading copy of D3; later field evidence is not an observation of the 1959 night.
Language: en
Open original source ↗Downloadable working files
The pack contains the English dossier, a structured case record, source register, chronology, evidence ledger, hypothesis comparison and inspector task sheet. Downloaded source documents, when available for redistribution, are identified in the pack manifest. Other documents remain direct source links.
- English dossier · Markdown
- Structured case · JSON
- Chronology · CSV
- Evidence ledger · CSV
- Hypothesis comparison · CSV
- Source register · CSV
- Inspector task sheet · Markdown
- File manifest and attribution · JSON
Eklipsis authored this synthesis and its working sheets. The original archives and researchers retain credit for their documents. This is a documentary investigation, not a claim that Eklipsis conducted the historical fieldwork.