Wildfire Black Box · Colorado · Historical evidence record
Black Forest is an evidence problem as well as a wildfire disaster. A broad ignition category can be established while the specific mechanism remains unresolved. Published loss totals can differ. An evacuation-area population is a different quantity from the number of people whose departure was observed.
What happened in the Black Forest Fire?
The El Paso County Sheriff’s investigation summary, prepared 17 November 2014, describes a smoke report near Shoup Road and Highway 83 at approximately 1:45 p.m. on 11 June 2013. It reports 14,280 acres, 489 homes destroyed, two resident deaths and full containment on 20 June 2013. These figures belong to that document. [1]
The summary concludes that the fire was human caused, but does not identify a definitive ignition mechanism. Analysis of metallic particles was inconclusive. Intentional ignition was not supported by the evidence or circumstances, although it was not completely excluded. [1]
Tecnobosque interpretation: preserve the level at which a finding is supported. An origin category, a particular heat source, an intentional act and responsibility are separate claims. A broad category does not fill the remaining fields automatically.
Nor does finding a candidate material establish that it caused ignition. The report’s stated uncertainty should remain attached to that possibility. This record names no suspected individual and makes no finding about present-day criminal proceedings.
How many homes were destroyed? Preserve the source-defined totals
Three public sources use different home-loss figures. The current fire district’s educational page gives 486; the sheriff’s 2014 summary gives 489; the NWS Pueblo retrospective gives 511. The latter also displays a containment date in 2016 that conflicts with the dated investigation’s 2013 chronology. [1] [2] [3]
| Evidence object | Homes destroyed | Date and scope |
|---|---|---|
| Sheriff’s investigation summary | 489 | Document prepared 17 November 2014 |
| Black Forest Fire Rescue educational page | 486 | Webpage reviewed 13 October 2026; not an audited parcel inventory |
| NWS Pueblo historical retrospective | 511 | Webpage reviewed 13 October 2026; chronology discrepancy retained |
The answer in this version is therefore source-specific. The existence of three values does not establish why they differ. This review has not matched addresses, loss definitions, assessment dates or subsequent corrections across inventories.
Do not average the numbers into a new “official” total. Nor should their minimum and maximum become a statistical uncertainty interval: these are different published statements, not independent samples drawn under one measurement protocol.
A future reconciliation needs the actual underlying records. Retain a unique property identifier, assessment date, damage classification, building-versus-housing-unit definition and revision history. Count changes can then be explained by documented additions, removals or recategorization instead of speculation.
Why a communications scenario is not an incident timeline
NISTIR 8181, June 2017, compiles historical and generated incident scenarios for public-safety communications research. Its Mountain-West WUI scenario cites the Black Forest after-action report. The scenario includes an evacuation area of 94,000 acres, 13,000 homes and 38,000 people, alongside a 14,280-acre burned area. These are figures reproduced in an adapted scenario, not new field measurements by NIST. [4]
Derived arithmetic: 94,000 ÷ 14,280 ≈ 6.58. The scenario’s evacuation-area figure is about 6.58 times its burned-area figure. This comparison does not show the spatial overlap, the date of each boundary or how many residents actually left. Those require other evidence.
It is tempting to treat a large evacuation footprint as a measure of over-warning or successful protection. Neither conclusion follows from two areas alone. The relevant analysis would need the warning’s purpose, conditions when it was issued, threatened population, later exposure and observed movement.
NIST explains that scenario attributes were standardized for research use. Such a document can frame technology-testing questions without replacing an original timestamped dispatch log. This record does not use its vignette’s time or weekday to overwrite the investigation’s dated chronology. [4]
Worked example: people contacted are not people observed leaving
Invented dataset—not Black Forest observations. Suppose a follow-up survey targets 500 unique households. At a defined cutoff, 200 respond: 160 report having departed and 40 report not having departed. The other 300 provide no response. Assume the responses are accurate and refer to the same cutoff.
| Quantity | Calculation | Meaning |
|---|---|---|
| Survey response | 200 ÷ 500 = 40% | Share of targeted households that answered |
| Departure among respondents | 160 ÷ 200 = 80% | Share of respondents reporting departure |
| Confirmed departure relative to survey target | 160 ÷ 500 = 32% | Reported departures divided by all targeted households |
The 80% figure cannot be presented as the departure rate of all 500 households. Respondents may differ from nonrespondents. Equally, the 32% figure does not mean the other 68% stayed: most of them supplied no movement information.
Under the stated assumptions, actual departures could range from 160 to 460 households. The lower bound counts the reported departures; the upper adds all 300 unknown households. That is 32–92% of the target. This is a logical bound for the invented dataset, not a confidence interval or an estimate for Black Forest.
An evacuation zone, a notification target, a delivered message, a survey response and an observed departure should occupy separate fields. Counting duplicate notifications also requires a stable recipient identifier; repeated messages to one address do not create additional households.
Known At and evolution of knowledge
| Checkpoint | Evidence | Boundary |
|---|---|---|
| June 2013 | Incident later described in the investigation | A retrospective source does not prove what every resident knew then |
| 17 November 2014 | Sheriff’s prepared investigation summary | Broad cause finding with an unresolved specific mechanism |
| June 2017 | NIST communications-scenario collection | Adaptation for research context, not a new incident investigation |
| 13 October 2026 | This review of agency educational webpages | Access date is not their original publication date |
This sequence tracks evidence objects. It cannot establish receipt of a warning by a specific household. That would need a dated message, identifiable recipient and an appropriate record of delivery or acknowledgment. Availability of information and receipt of information are different claims.
Fire DNA: the questions this case supports
- Origin: a broad finding coexists with an unresolved mechanism.
- Loss accounting: public home-loss totals remain source-specific.
- Evacuation: area and population descriptors do not establish observed departures.
- Data reuse: a research scenario retains a different purpose from an investigation.
- Remaining gaps: no reconciled property inventory or household-level movement dataset was audited here.
No numerical severity ranking or TB-ECI score is assigned. The record does not independently resolve every building ignition or evaluate a particular mitigation measure’s effect.
Sources, limits and citation
- El Paso County Sheriff’s Office: Black Forest Fire Investigation Executive Summary 13-7655. Prepared 17 November 2014. Primary agency document preserved by The Gazette; the host is not the issuing authority.
- Black Forest Fire Rescue: Lessons Learned. Current agency educational page; used for its explicitly attributed loss figure.
- NWS Pueblo: 2013 Wildfires retrospective. Used to preserve a differing public loss figure and flag its displayed chronology discrepancy.
- Choong and colleagues: NISTIR 8181. June 2017, Incident Scenarios Collection for Public Safety Communications Research. Introduction and scope, Mountain-West WUI scenario and bibliography.
Method: document review, source comparison and separately labeled arithmetic. The complete Black Forest after-action report could not be retrieved in this review, so its indexed appendix figures are not used as audited incident data. NIST’s adaptation is identified as such. No parcel reanalysis, fieldwork, criminal-case update or present-day evacuation advice is claimed.
Cite: Tecnobosque, Black Forest Fire 2013, record TB-WBB-US-2013-BLACK-FOREST, version 1.1, evidence review 13 October 2026.