Wildfire evidence study · California, United States · 2017
Atlas illustrates why an operational incident page and a later investigation should be read together. They agree on acreage but differ in the displayed start date. The investigation account also distinguishes two ignition locations, a detail that disappears when the event is reduced to a single map marker.
Documentary review · Sources checked 30 October 2026 · Historical case, not live incident information.
The published event record
| Field | Recorded value |
|---|---|
| Start | 8 October 2017 (investigation account) |
| Later milestone | Archive containment field: 9 February 2018 |
| Reported acres | 51,624 |
The incident page records 51,624 acres in Napa and Solano counties and displays 9 October 2017 as its start date. CAL FIRE’s June 2018 account, reproduced in PG&E’s SEC filing, places the start on the evening of 8 October. It describes tree or limb contact with a PG&E line at two locations. This study uses the investigation account for the start date and flags the archive discrepancy. See primary records below.
Why two origins change the reconstruction question
A single origin coordinate can be useful for a catalog, but it cannot represent two separate ignition locations. Any attempt to calculate spread distance from one point needs to state which origin was used and whether the other front had already influenced the footprint. Otherwise the resulting speed can have no clear physical meaning.
The date mismatch is equally consequential for weather analysis. Selecting the wrong night can pair fire activity with observations from a different wind period. The defensible workflow is to preserve both source values, identify their roles, and choose the investigation chronology for the ignition question rather than average the dates or silently overwrite one.
The unusually late containment field on the archive should also be treated as a recorded field, not proof that the fire advanced continuously until February. Operational closure, containment reporting and active spread are different events. No daily progression series has been reconstructed here.
Atlas can be compared with Tubbs as a regional event only after keeping the incidents separate. Sharing a period and region does not establish a shared ignition mechanism, identical exposure or the same warning sequence. Regional context helps frame the questions; incident-specific evidence answers them.
What this study does and does not establish
The investigation account is accessed through a company filing quoting CAL FIRE. This review does not inspect physical ignition evidence or reconstruct hourly spread. The archive’s date fields are reported without claiming to explain how the discrepancies arose.
This is a documentary evidence review, not fieldwork or an independent forensic investigation. Published numbers retain their source definitions. Differences are identified rather than silently reconciled. The analysis above is Tecnobosque’s interpretation of the cited record.
Primary records and source notes
- Atlas Fire: official incident or annual record — chronology, scope and reported figures discussed above.
- PG&E filing reproducing CAL FIRE’s 8 June 2018 findings — used only for the findings explicitly attributed in this study.
Continue the investigation
Tubbs Fire: cause findings and the evidence timeline · Read our evidence-review method · Browse United States cases · Browse the full fire archive
Cite this record
Tecnobosque. “Atlas Fire 2017: Two Ignition Locations and Conflicting Start Dates.” Documentary evidence review, 30 October 2026. https://tecnobosque.com/atlas-fire-2017/. Cite the underlying primary source when reusing its figures.