How early can a wildfire
be detected?
The Signal Lab is the research program behind Fire Nearby. Around the clock, our fusion engine watches independent early indicators of fire, and every candidate it raises is tracked to a verified real-world outcome. This page is where we publish what that work is teaching us, updated as the research progresses.
What we’re building
Official wildfire maps wait for an incident to be declared, which can lag hours behind a brand-new fire. We are working on the opposite end of that gap: identifying likely new fires from open signals alone, as early as the data physically allows, and proving how often we get it right. The work runs in three loops:
Fuse
Independent signals (satellite heat, satellite fire-detection AI, live dispatch traffic, lightning, air-quality anomalies) are merged into scored candidates when they agree in the same place and window of time.
Verify
Every candidate is matched against official incident records.[4] Hits record how many minutes we were ahead of the official discovery; candidates that never become fires are logged as misses.
Learn
The labeled outcomes become training data. A statistical model learns which combinations of evidence actually precede real fires, and each version has to beat the last on held-out data before it ships.
A five-year labeled dataset, built from scratch
A live system only accumulates evidence at real-time speed, so in August 2026 we rebuilt the past instead. We swept five years of archived satellite fire detections[1] through the exact clustering, suppression, and scoring rules our live engine uses, then matched every resulting cluster against official incident records[4] to label its outcome. Each cluster was also stamped with the conditions at the moment it formed: archived lightning activity,[2] drought status,[3] and the location’s own fire history. The result is a research dataset that does not exist anywhere else.
Only 1.3% of these clusters ever earned an official fire name. That imbalance is the core of the problem. Almost everything hot a satellite sees (controlled agricultural burns, industrial heat, small fires that are out before anyone logs them) never becomes an incident, and a useful early-warning system has to tell the 1.3% apart from the rest.
What five years of outcomes taught us
With every cluster labeled, we can measure exactly how much each piece of evidence shifts the odds that a satellite heat cluster becomes a real, named wildfire. The figure below shows the strongest observed condition for each factor, expressed as a multiple of the 1.46% baseline conversion rate for US clusters (95% confidence intervals computed on every bin):
History dominates
The strongest single predictor is the location’s own fire record. A heat cluster in a cell that historically sees 40+ fires a year is 6.2x more likely to become a named fire; a cluster somewhere with no fire history converts at a quarter of the base rate.
Intensity was underused
Satellites report the radiative power of each detection,[1] and outcomes track it monotonically: detections above 150 megawatts convert at 3.1x base rate, the faintest at 0.7x. That finding is now incorporated into Ember.
Lightning surprised us
The literature on lightning-initiated fires[5] led us to expect recent strikes to raise the odds. In our data they lower them: clusters with strikes in the prior 24 hours converted at 1.0% vs 1.8% with none in two weeks, largely because thunderstorms bring rain and are concentrated in regions where heat detections rarely become named incidents. Findings like this are why we measure instead of assume.
Meet Ember, our early-detection model
The five-year dataset trains Ember, a statistical model that estimates the probability a brand-new heat cluster will become an officially named wildfire. The name fits the job: an ember is the small, persistent heat that exists before anyone calls it a fire, which is exactly the moment this model lives in. Each Ember release is versioned, so you can always see which iteration produced a number and what changed between them.
| Version | Ember v0 |
| Status | Shadow evaluation (predictions logged, not shown publicly) |
| Population | US satellite heat clusters |
| Training window | August 2021 to April 2026 |
| Validation | Untouched May to August 2026 holdout |
| Discrimination (AUC) | 0.878 on the holdout |
| Top 1% of clusters | 18.6% became named fires, a 13.3x lift over base rate |
| Median warning lead | 23.9 hours before official discovery, on matched fires |
| Early by 12h or more | 62% of matched fires |
Why you don’t see Ember’s numbers on the map yet
Ember v0 currently runs in shadow mode: it scores every eligible live cluster and its predictions are logged, but nothing it says affects what the public site shows. Strong results on historical data are a promising start, not proof; the model earns its way onto the map only by beating our existing system on live fires it has never seen. That comparison is running now.
Accountability, in public
A prediction system you can’t audit is just marketing. Everything on Fire Nearby is timestamped and sourced, and Ember will be held to the same standard: when it goes live, this page will carry a running public scoreboard of every call it makes, correct and incorrect, with lead times on the hits and an honest count of the false alarms. Until then, here is the live research pipeline at work:
Methods, briefly
For readers who want the rigor without the full paper:
- Temporal holdout. Ember trains only on data through April 2026 and is evaluated on the untouched May to August 2026 fire season, so its headline numbers describe prediction, not memorization.
- No peeking at the future. The fire history feature for any cluster is built strictly from records that predate the evaluation window, so the model never gets credit for information it could not have had at the time.
- Uncertainty is reported. Every conversion rate in the audit carries a Wilson 95% confidence interval, and thin bins are treated as inconclusive rather than as findings.
- Known limitations. The historical dataset is satellite-driven; the live engine also sees dispatch traffic and other fast human signals that archives cannot reconstruct. Canadian clusters are currently excluded from training because official Canadian labels lag by years. We publish limitations here for the same reason we timestamp everything else.
Research log
New entries are added as the work progresses.
Five-year dataset completed; Ember v0 trained and moved to shadow evaluation
Rebuilt five fire seasons of satellite history through the live engine’s rules (887,732 labeled clusters), audited every scoring assumption against outcomes, corrected the ones the data disagreed with (heat intensity now weighted in production), and trained Ember v0 (holdout AUC 0.878). The model now scores live clusters silently while we compare it against the current system.
Ember v1
Adds the environmental layers already being attached to the historical dataset (fire weather, fuel dryness, terrain, vegetation condition), plus the live public scoreboard once shadow evaluation clears our precision bar.
Sources
- [1]NASA FIRMS: VIIRS and MODIS active fire detections (LANCE/ESDIS). firms.modaps.eosdis.nasa.gov/
- [2]NOAA GOES-R Series Geostationary Lightning Mapper (GLM). www.goes-r.gov/spacesegment/glm.html
- [3]US Drought Monitor, National Drought Mitigation Center, University of Nebraska-Lincoln / USDA / NOAA. droughtmonitor.unl.edu/
- [4]NIFC Wildland Fire Interagency Geospatial Services (WFIGS): official incident records and discovery times. data-nifc.opendata.arcgis.com/
- [5]Schultz, C. J. et al. (2019). Spatial, Temporal and Electrical Characteristics of Lightning in Reported Lightning-Initiated Wildfire Events. Fire 2(2), 18. doi.org/10.3390/fire2020018
Dataset statistics and model metrics on this page are from the Fire Nearby historical replay and evaluation pipeline, August 2026. Questions about the methodology are welcome: firenearbysupport@gmail.com.
See the signals yourself
The early-signal layers this research is built on (satellite heat, 5-minute geostationary scans, live dispatch reports) are already on the live map, and the About page documents every source and its freshness in real time.