A new computer program could serve as an early warning system for predicting extreme fire weather patterns in northern Alberta.
The self-organizing map (SOM) technology was developed and tested by researchers at the University of Alberta and University of Oklahoma.
“This study takes a look at pressure fields at the surface and another pressure level mid-atmosphere and takes a look at their patterns and says when do we see a pattern for extreme fire weather?” Professor of wildland fire at the U of A and co-author Mike Flannigan tells Mix News.
Flannigan compares the technology to artificial intelligence which mimics the way the human brain works. SOMs can be trained to find patterns in data and model complex relationships helping generate maps that predict where and when extreme fire weather is expected in northern Alberta.
Flannigan says SOMs have been used to predict other weather events, such as monsoons, but this is the first time it’s being used to measure extreme fire weather patterns.
Right now the current model for predicting fire weather and fire danger use precipitation, which Flannigan notes is a very difficult field to predict.
“So the idea here is that it can potentially be an improvement on existing systems, which are fairly good, but what this buys you is hopefully a better forecast for longer periods of time.”
Flannigan says the technology could give officials a four-to-seven day window to detect extreme conditions giving crews enough time to put resources in place if needed.
He adds this won’t make our problems go away, but it’s another tool in the kit for fire management.
“Actually being used in day-to-day operations may take a year or two because fire management is making life and death decisions.”
“Fire management agencies, if they want to give this artificial intelligence a shot, I’m all for that and I’ll help out any way I can.”
The results of the self-organizing map technology study were published in the Canadian Journal of Forest Research on Tuesday.