Reading method · evidence
Start with the sentence each layer can actually prove
Two overlaid colors can appear to tell the same story while coming from completely different times, sensors and methods.
- Published
- Revised
The evidence ladder
Potential, observation, geometry, inference and official decision answer different questions. Moving up the ladder requires a new source, not more assertive wording.
- 01
- 02
- 03
- 04
- 05
Short answer
To read a wildfire map, first identify the layer’s nature — potential, observation, geometry or model — then its source, time, coverage and limits. Compare several forms of evidence without merging them. Finally, return to official alerts and authorities before acting: an analysis map never replaces an instruction.
Key points
Name the verb
“Favors,” “detected,” “mapped” and “projects” are four different claims.
Read time before color
A dramatic but old layer may be less relevant than a modest, recent observation.
Preserve contradictions
A model and observation that diverge create a question to examine, not a reason to erase one.
1. Classify the layer before interpreting it
A danger index describes conditions favorable to ignition or spread; it locates no fire. A thermal detection observes an anomaly in a footprint and at a time. A perimeter draws surveyed or interpreted geometry. A scenario computes change under assumptions.
This classification turns the legend into a testable sentence. If the sentence becomes stronger than the method — for example “the fire is here” for a thermal pixel — interpretation has exceeded the data.
2. Check source, acquisition and publication
Look for the organization, product, acquisition time, update time and time zone. An image may have been published now while representing an earlier pass.
The page date is not necessarily the event date. Fire Maps places times near the layer so a cache or screenshot cannot silently turn an old observation into the present.
3. Ask what empty space means
Empty space may mean measured absence, no detection, no coverage, delayed response or withdrawn data. Those states are not interchangeable.
Check resolution, cloud, terrain, receiver density and served extent. An honest map says “outside coverage” when it cannot say “nothing observed.”
4. Cross-check without manufacturing super-evidence
A recent detection near an old perimeter may indicate activity to examine. It does not automatically update that perimeter. Converging aircraft tracks may suggest a mission; they do not prove the fire edge.
Preserve each layer’s provenance and state the conclusion at the strength of the weakest link. Several independent clues can raise confidence without becoming a measurement no sensor made.
5. Separate understanding from safety decisions
The map can help understand a situation, prepare questions and find sources. Evacuations, closures, access and instructions belong to the responsible authorities and may change faster than a public layer.
When danger exists, follow official alerts and emergency services. Never approach a fire, drop zone or aviation operation to verify what the screen shows.
Common questions
- Which layer is most reliable?
- No layer is universally superior. Reliability depends on the question: an official alert is for action, a detection for finding heat, a perimeter for a dated boundary and a model for exploring a scenario.
- Does “updated” mean “observed now”?
- No. Update time may be publication or retrieval time. Look for the acquisition or validity time of the phenomenon itself.
- Can a screenshot be shared without context?
- Sharing the URL while preserving legend, source and time is better. An isolated capture quickly loses context and can continue circulating after it expires.
Primary sources consulted
Revised · July 29, 2026