The United States has experienced devastating wildfires in recent years. In January 2025, the country watched as Southern California burned, with the Palisades and Malibu fires eventually claiming 23,448 acres, over 6,800 buildings, and twelve lives. The Southern California fires were driven by powerful winds called the Santa Anna winds, which picked up embers and caused the fires to spread rapidly, making them difficult to contain.
Wind Data: A Powerful Force for Fighting Fires
Wind information – its location, direction, and speed – is vital to fighting fires. A company called SkyTL proposed to address this need by creating a system called WindTL to provide updated wind models and real-time heat maps that show how winds change during fires. This system, SkyTL said, would give firefighters and first responders up-to-the-moment information about how and where a fire is likely to go.
NOAA recognized the opportunity that SkyTL offered and used the Small Business Innovation Research (SBIR) program to support SkyTL’s creation of the WindTL system. The SBIR program provides funding to small businesses to develop innovative technologies and move them from concepts to real, marketable products that help the American public.
WindTL is exactly the kind of product that the SBIR program is designed to support, using cutting-edge technology to combat a threat to people’s lives, homes, and property. The WindTL system takes in a vast amount of information from satellites, weather stations and sensors, unmanned aerial systems (UAS, also known as drones), user input, and historical data. It then uses AI and simulators to interpret and process the data, running it through Physics-Informed Neural Networks (PINNs), research-grade open-source simulators, proprietary physics and AI models, and LLM-powered data extractors to create models that predict how fires will spread and heat maps that show how active fires react to changing winds.
An SBIR Success Saves Homes
NOAA’s SBIR investment in SkyTL was successful, and WindTL has already borne fruit for fire safety. In 2024, firefighters used both traditional models and WindTL to determine where to send their teams when fighting the Thompson fire in Oroville, California. The traditional models predicted that the fire would be contained by a river acting as a firebreak and natural boundary.
The WindTL model, however, used hyperlocal wind patterns to predict that strong winds would send embers across the river, putting the area on the far bank at risk. The WindTL model proved correct, and because the firefighters had advanced warning they were redeployed and ready when embers did cross the river and could prevent the fire from reaching people’s homes.
NOAA’s SBIR program and funding played a key role in the development of the WindTL system. According to SkyTL CEO Rocio Vitalle, “The SBIR NOAA grant allowed us to get the initial funding to de-risk our approach for wildfire modeling and make the necessary modifications to ensure our approach met the technical challenges and stakeholder requirements. With the NOAA SBIR, we were able to have the time and resources to develop an approach that significantly enhanced the current state of the art, and work with stakeholders via no-cost pilots to continuously refine and improve the technology.”