Researchers have developed an AI model that can predict with 90 percent accuracy when and where lightning will ignite wildfires.
Last year, lightning strikes ignited close to 7,000 fires in the U.S., which together scorched an area larger than Connecticut. Fires started by lightning can often go unnoticed and grow rapidly, burning through large areas before officials can respond. As warming gives rise to hotter, drier conditions, fires are burning through more land.
To better guard against worsening blazes, Israeli researchers developed a global model that gauges the risk of lightning-induced wildfires. The AI model was trained on detailed satellite data, from 2014 to 2020, on lightning strikes, weather, and other factors. Researchers ran the model with data from 2021, finding it accurately forecast the risk of lightning-induced wildfires more than 90 percent of the time. The research was published in Scientific Reports.
The model, which is not yet making real-time forecasts, is part of a growing push to use artificial intelligence to predict and detect wildfires. The aim is to alert officials to emerging fires before they get out of control.
“With the growing implications of climate change, new modeling tools are required to better understand and predict its impacts,” said lead author Assaf Shmuel, of Bar Ilan University. Artificial intelligence “holds significant potential to enhance these efforts.”
ALSO ON YALE E360
Whiplash: How Big Swings in Precipitation Fueled the L.A. Fires
