You are here

New way to predict lightning strikes: study

Category: 

CALGARY—Apparently lightning can strike the same place twice after all.

A study by researchers at the University of Calgary's Schulich School of Engineering suggests it's possible to predict where lightning will strike and how often.

They say satellite data and artificial intelligence can help foresee where lightning poses a greater risk to spark wildfires.

“That will give us a more precise description about the patterns happening in the lightning and the wildfire hazard,” said Dr. Xin Wang, one of three researchers involved in the study.

"It also can be used for the future predictions about those hazards.

“Those events don't just randomly happen," Wang stressed. "They also have spatial and temporal patterns.”

Lightning-caused wildfires are a growing problem in Western Canada and have been stretching fire resources to their limits.

The study used historical data from 2010-16 and examines how lightning is affected by a number of elements, including land use, soil type, elevation, vegetation cover, and surface temperature.

The study, published in “Sensors” open access journal, was written by Wang, Jeong Woo Kim, and DongHwan Cha.

Kim said the group used advanced technology, including artificial intelligence.

“We analyzed a number of different types of data over a number of years so we can pinpoint the location that has a high chance of wildfire,” he noted.

“Statistically, I would say it is more than 90 percent accurate.”

But a professor of wildland fire at the University of Alberta questioned whether the method would be helpful to firefighters on the ground.

“It doesn't tell them anything useful for day-to-day operations,” said Mike Flannigan.

“Maybe strategically there might be something, but they've already mapped where fires have occurred in the past and any firefighters worth their salt know where those areas are,” he noted.

Free story: 
Twitter icon
Facebook icon
Google icon
LinkedIn icon
Pinterest icon
Reddit icon
e-mail icon