This Twitterbot keeps you up-to-date on fires burning near you

Pulling from a list of burning fires, @WildfireSignal uses satellite imagery to let you know if smoke or flames are headed your direction.


As fire seasons in the U.S. gets hotter and drier, a new Twitterbot will show you if a wildfire is burning near your house, where the fire is headed, and if a plume of smoke is traveling in your direction by posting an updated time-lapse video and infrared images every six hours.


The tool, called @WildfireSignal, went live on Twitter on July 18. Scientists and programmers at Descartes Labs, a startup that processes images from satellites, designed the tool to pull a list of active fires from a government database, then clean up near-real-time images from the GOES-16 satellite at each fire’s location. Using the massive amount of data generated by the satellite, it automatically builds a time-lapse video of each fire and embeds it in a tweet with a hashtag of the fire’s name.

Burro, CO. [Image: @WildfireSignal]
“There just isn’t a tool like this right now that’s actively monitoring fires and generating that sort of visual,” says Caitlin Kontgis, applied scientist lead at Descartes Labs. “You can very clearly see the fire come to life in the video.” On the left-hand side of the image, the video shows billowing smoke; on the right, the image shows spots of heat that can help track a fire at night when smoke isn’t visible.

Someone evacuated from their home can use the tool to track a fire’s progress, information that can otherwise be difficult to find. Firefighters could also potentially use the service, though the startup’s next iteration–which will automatically detect new wildfires–will be more helpful.

Martin, Oregon [Image: @WildfireSignal]
The team is testing an early version of this fire-spotting technology with staff at Santa Fe National Forest, which was closed for several weeks recently because of a very high risk of fire. Rather than waiting for the Twitterbot’s six-hour updates, it can update continuously so officials can get an early warning of a fire. Descartes Labs is now building in machine learning to automatically trigger notifications when a heat signature develops in a way that indicates a fire. The earlier a fire is identified, the better chance firefighters have of controlling it.

“I think the real power in this is the fact that it [will operate] at night,” says Kontgis. During the day, at least in populated areas, it’s likely that someone might spot a plume of smoke as a fire starts and call 911. In the dark, when smoke isn’t visible and people may be asleep, the new service could be the first to notify firefighters that a fire exists. “What we don’t want to have happen is have a fire reported because it’s bearing down on somebody’s yard and they have to escape in the middle of the night without much warning.”

About the author

Adele Peters is a staff writer at Fast Company who focuses on solutions to some of the world's largest problems, from climate change to homelessness. Previously, she worked with GOOD, BioLite, and the Sustainable Products and Solutions program at UC Berkeley, and contributed to the second edition of the bestselling book "Worldchanging: A User's Guide for the 21st Century."