Prime Rogue Media seal
Prime Rogue Media · Investigation

CFB Wainwright fire: our reporting and open archives, indexed

Open investigation By , Prime Rogue Media · Index · Fire · Index published

What the work has established

As of 1:10 a.m. MT (08:10 UTC), Tuesday, November 17

  • Fire was on the range before the time National Defence gave. Satellite detections put fire on the range hours before DND's stated ignition time; the military later said the fire had been known for hours. Detections · The changed account
  • The weather was the mechanism. A storm line crossed the area and the wind shifted; working model v5 puts the cold front at the fire at 3:28 p.m. MDT, and the flank became the front. Model v5 · Flank to front
  • Two explanations are ruled out by the records. Gas wells played no part, and the seismographs show no explosion. Gas wells · Seismic check
  • The record of the day is thin. There is no GPS trail; the board of inquiry works from range control logs. Details
  • The paper trail is public. Our access-to-information requests to National Defence, and their status, are tracked openly. ATIP tracker
  • Everything is checkable. Every raw file in the archives is kept unaltered, hashed and timestamped on the Bitcoin blockchain through OpenTimestamps, so anyone can verify a file existed in that form at that time. The archives

Reconstruction and analysis

Newest first. Ten pieces, Oct. 18 to Nov. 6, 2026.

  1. New model puts the cold front at the fire at 3:28 p.m.Working model v5, Rubicon Ridge: a 2,160-run ensemble fitted to the burn outline, with the front-arrival time read from 38 weather stations.
  2. CFB Wainwright fire: exclusive 70 cm satellite imagesMeasurements from a 70 cm Satellogic image of Oct. 20: the burn outline, unburned islands, and distances from the ranges to the burn edge.
  3. CFB Wainwright fire timeline: DND vs. the weather stationNational Defence's stated timeline set against the hourly observations recorded at the base.
  4. The flank became the frontHow the wind shift turned the side of the fire into its leading edge.
  5. What most likely happened at CFB Wainwright, and whyEvery warning light was on: National Defence's account set against the government's own records.
  6. The storm line, then the fire flaredThe passage of the storm line read from satellite and station data, and what the fire did next.
  7. A week of satellite data on the CFB Wainwright fireBurning by 2:49 p.m., likely since morning: a week of detections, including Oct. 14 and Oct. 18.
  8. DND said the fire “ignited” at 4 p.m. The military now says it was known for hoursThe first change in National Defence's own account of when the fire began.
  9. Satellite data fits an exercise-fire scenarioWhat the detection record is and is not consistent with.
  10. Satellites saw fire on the range before DND's ignition timeThe piece that opened the investigation: detections on the range hours before the time National Defence gave.

Fact checks

Claims we tested against the public record, including our own numbers.

  1. How we found the 82 km/h gust at CFB WainwrightOur method, step by step, so the number can be checked.
  2. Was there an explosion at CFB Wainwright? We checked the seismographsThe seismic record does not show one.
  3. Gas wells played no part in the CFB Wainwright fire, records showWell and pipeline records against the fire's location and timing.

Records and accountability

What we asked for, what was released, and what was never recorded.

  1. No GPS trail at CFB Wainwright: fire probe rides on range logsWhat the board of inquiry has to work from, and what was never recorded.
  2. Watching you watching me: Ottawa reads, DND won't answerGovernment traffic to our Wainwright reporting, against the replies we did not get.
  3. CFB Wainwright deaths: Alberta asserts base jurisdictionThe provincial claim to jurisdiction over deaths on the base.
  4. CFB Wainwright fire: ATIP request trackerEvery access-to-information request we have filed, with its status.

The running timeline

Published the evening of the fire and updated through Oct. 21.

  1. LIVE: CFB Wainwright fire — what the data and records showOur live coverage from the first local reports, through the ministerial statements, to the ATIP tracker going up.

The open archives

Two public repositories. Raw files are kept unaltered, hashed with SHA-256 and timestamped on the Bitcoin blockchain through OpenTimestamps. Alberta was on Mountain Daylight Time (UTC−6) on Oct. 17, 2026; satellite data is in UTC.

Part 1 · frozen at working model v4

github.com/Prime-Rogue-Inc-Corporate/cfb-wainwright-fire-archive

NASA FIRMS detections (raw responses, merged table and map, and every detection around the site since 2012), NOAA GOES-18/19 loops and the 10-minute fire timeline, NOAA Hazard Mapping System points and smoke outlines, Environment Canada forecasts, warnings and station observations, NRCan fire-weather files, Copernicus Sentinel-2 burn-scar products, ADS-B positions from Oct. 17–18, the location analysis and competing-hypotheses work, the 3D terrain reconstruction, SHA-256 manifests with OpenTimestamps proofs, and records released under access-to-information law. Frozen as of Oct. 27, 2026 at tag v4-published.

Open part 1 →

Part 2 · Rubicon Ridge, working model v5

github.com/Prime-Rogue-Inc-Corporate/cfb-wainwright-fire-archive-2

Raw pulls for Oct. 16–18 (FIRMS, PurpleAir, GOES-18/19 fire extracts, RAP, ECCC HRDPS and radar, 38 weather stations, Alberta air quality and ACIS, Sentinel-3 SLSTR), the hourly maximum-gust map, the front-arrival analysis, the GOES visible imagery showing no identifiable smoke plume, the polar-orbiter pass review, the 2,160-run model v5 ensemble and its refit to the 70 cm burn outline, the air-response timeline built from ADS-B, and the measurements taken from the Satellogic image of Oct. 20. Commercial imagery is never stored: only our own measurements from it, as vectors and tables.

Open part 2 →

Code is MIT. Prime Rogue's own data and analysis are CC BY-NC 4.0, credited to Prime Rogue Inc.; commercial use, including by commercial media, needs written permission. Government, NASA, NOAA, Copernicus, USGS, Environment Canada and Alberta data keep their own licences, credited in each repository.

Everything we have published on the Wainwright fire is now in one index, with the open archives beside it

Housekeeping

  • This page gathers the eighteen pieces we have published on the fire since the evening of Oct. 17, sorted into reconstruction and analysis, fact checks, records and accountability, and the running timeline.
  • It also indexes the two public archives that the reporting rests on, so a reader can go from a claim to the raw file behind it.
  • The reporting itself still sits on primerogueinc.com while that site is rebuilt. When each piece moves, it is redirected, and the links on this page follow it.

Working model v5 puts the cold front at the fire at 3:28 p.m.

Analysis: our model

  • Rubicon Ridge, working model v5, fits a 2,160-run ensemble to the burn outline measured from a 70 cm Satellogic image, and reads the front's arrival from 38 weather stations. Full piece
  • The inputs, the code and the run outputs are in part 2 of the archive, which also notes which results are working results and not yet published.

If you have something

Please do not open a GitHub issue or pull request with documents or anything that identifies you: that history is permanent and public. Use Signal (Rubicon49.50) or Proton email instead. Contributors are credited or protected as they prefer, under the Canadian Association of Journalists' ethics guidelines. Corrections to our code or methods are welcome as issues; corrections to the reporting go to our corrections page.