Meteorologists tracking storm using new AI tools to increase accuracy

When major storms bear down on Saskatchewan, meteorologists have their work cut out for them.

Colorado lows, like the one that hit recently, tend to be hit-or-miss systems, so tracking where they'll end up is a top priority.

New tools help to make those predictions more accurate, with one company working to integrate AI into their forecasts.

President and Chief Scientist of Weatherlogics Scott Kehler says that technology has taken off in the last few years.

"There's a bunch of AI models available now for us to use. Companies like Google have developed these models and a number of other agencies as well. What I noticed is that a lot of the models based on artificial intelligence have the weather system tracking a little bit farther to the West compared to the traditional weather models, and so that gives us a bit of a hint that maybe the impacts of the storm will differ from the data that we traditionally use."

"As the storm has now become a little bit closer in the forecast, we're actually starting to see these traditional weather models start to come into better agreement with the AI models. So it looks like these AI models may have given us a little bit of extra lead time on the track of this winter storm." 

Kehler says that AI technology is well suited to track those more narrow hit-or-miss systems further ahead than traditional methods.

"The record over all of the AI models shows that they tend to outperform the traditional weather models in the long range. So that helps to give us a little bit of a longer lead time in terms of predicting when these storms will hit and then as you get into the shorter range, let's say one or two days from when the system is expected to hit, then the difference between the model starts to become smaller and less noticeable."

You can find more info about upcoming weather on the WestCentralOnline weather page HERE.