New lightning sensors could detect Tahoe wildfires in seconds. Here's how they work

San Francisco Chronicle​

3 days ago

When thunderstorms rolled across the Sierra Nevada in late July, officials warned that lightning could ignite wildfires in forested areas loaded with dry vegetation.
Thousands of lightning bolts struck the Sierra Nevada and sparked dozens of small fires, but officials generally can't determine which strikes ignite wildfires unless people or webcams report smoke. A new network of sensors installed across the Tahoe basin could change that, letting firefighters deploy more rapidly.
The network could allow firefighters to catch blazes before they burn out of control. Many of California's largest wildfires were caused by lightning, most notably in 2020 from a massive thunderstorm outbreak. Eight sensors were installed across the Tahoe basin, including at ski resorts, fire departments and the UC Berkeley Central Sierra Snow Lab, which spearheaded the project.
'It sounds a bit like 'Star Wars,' but they sense a disturbance in the electromagnetic field and are able to map a lightning strike,' said Andrew Schwartz, director of the snow lab.
The high-risk lightning detectors come from Fire Neural Network, a Florida company that has installed detection networks across the West, including San Bernardino County. Unlike other lightning detection technology, the network pinpoints where a lightning strike occurs while simultaneously analyzing meteorological and environmental data to determine which bolts have the highest risk of sparking a wildfire. The company says it can accurately map strikes within 100 feet of where they touch down.
Not all lightning strikes are the same. Some release more energy than others, making them more likely to start wildfires. But these high-risk bolts make up only a small portion of all lightning, said Fire Neural Network co-founder Caroline Comeau.
'Our sensors are able to distinguish those lightning strikes that last extra long,' Comeau said. 'They last up to 1,000 times longer, creating a lot more heating than a regular strike.'
Thunderstorms that rattled the Tahoe region from July 22 to 29 dropped 723 lightning strikes. Fire Neural Network determined that 41 of those bolts were most likely to start fires, according to Comeau.
Wildfire detection technology in California is much more advanced than it was a decade ago. UC San Diego's webcam network utilizes artificial intelligence and has detected more than 1,200 fires across California since 2023.
But Comeau says even more rapid detection is needed. Within 40 seconds, Fire Neural Network's algorithm calculates which strikes have a high probability of starting fires and alerts authorities.
'If a camera system picks up the smoke above the tree line, that's often too late,' Comeau said.
Extra warning time for evacuations could be a game changer for Lake Tahoe. It could take up to 11 hours to flee South Lake Tahoe during a wildfire.
'There's effectively a higher risk around the Tahoe basin because it's a higher populated area and there's limited evacuation routes,' Schwartz said. 'There's a bit of anxiety about fires that start in the Tahoe basin, and these extra detectors could be useful there.'
Still, some scientists are skeptical of Fire Neural Network's claims about pinpoint strike accuracy and efficiency.
'I would like to see validation of those claims,' said Chris Vagasky, a meteorologist and manager of the Wisconsin Environmental Mesonet.
Cal Fire Santa Clara Unit chief Marcus Hernandez has not yet 'seen a correlation between what they determine to be a high-risk lightning strike and an ignition.'
If the network is able to provide firefighters with proper early warning, it could help mitigate the growing number of megafires in a warmer, drier climate.