Karen Miniature, a research ecologist with the Forest Service who participated in the study and maintains a historical database of national wildfire reports, says understanding what causes wildfires is crucial to preventing them and educating the public. Strategic prevention appears to be working: According to the National Fire Protection Association, the number of home fires in the U.S. has dropped by approximately almost half since the 1980s.
In 2024, Miniature expanded its wildfire archive to include more information useful to investigators, such as weather, altitude, population density and when the fire started. “We need to include these elements in the data so we can track them over time. We are still following the events of the 20th century,” she said.
According to Miniature, wildfire trends in the western United States have changed with human activity. In recent decades, ignitions from power lines, fireworks and firearms have become increasingly common, in contrast to fires caused by railroads and sawmills, which were once more common.
The study found that vehicles and equipment are likely the main culprits, potentially causing 21 percent of fires without a known cause since 1992 last fall Airport fire there was just such an event in California that burned over 23,000 acres. An increasing number of fires are the result of arson or accidental ignition – whether from smoking, gunshots or bonfires – accounting for another 18 percent. In 2017, an Arizona couple chose a blue, smoking firework for their baby’s gender reveal party. Sawmill firesetting fire to nearly 47,000 acres.
But these results are not final. Machine learning models, such as those used in the study, are trained to predict the likelihood of a given cause of fire occurring, rather than proving that a specific ignition occurred. Although the model developed in the study showed 90 percent accuracy in selecting lightning or human activity as the ignition source when tested on fires with known causes, it was more tough to determine exactly which of 11 possible human behaviors was to blame, but only succeeded half the time.
Yavar Pourmohamad, a data scientist at Boise State University who led the study, said knowing the likely causes of a fire could support authorities warn people in high-risk areas before the fire actually starts. “It might give people a clue as to what to watch out for the most,” he said. “Perhaps in the future, artificial intelligence can become a trusted tool for real-world operations.”
Synolakis, the USC professor, says Pourmohamad and Miniature’s research is significant for understanding changes in risk. He advocates proactive measures, such as burying power lines underground where they can’t be blown away by the wind.
AND 2018 study found that fires started by downed power lines – such as the Camp Fire in Paradise, California, that same year – were getting worse. Although the authors note that while power lines do not cause many fires, they are associated with larger swaths of burned land.
“We need to really make sure that our communities are more resilient to climate change,” Synolakis said. “As we see in the extreme conditions in Los Angeles, simply suppressing a fire is not enough.”