A new study of scientists published in the Monthly Weather Review showed that machine learning technology or AI can possibly predict extreme weather.
The AI will use facial recognition to predict dangerous weather conditions.
We all know that hailstorms have a damaging impact on our agriculture, property, and even wildlife. In fact, last week there’s a severe hailstorm in Eastern Montana that killed 13,000 shorebirds and waterfowl.
Aside from that, hailstorms can also cost us a lot of money because of damages in property and to people. To be exact, these storms cost us $22 billion every year according to CBS News.
Therefore, the prediction of these storms can prevent us from all those damages if what the scientists found in their study is true.
The artificial intelligence of the National Center for Atmospheric Research (NCAR) will start testing it.
In fact, they have created or rather trained a deep learning model called “convolution neural network” which will pinpoint specific storm features that will determine whether hail will be formed or not.
If there’s a possibility of a hailstorm, the AI will also identify the largeness of the hailstones. “The structure of a storm affects the possibility to produce hail,” said NCAR scientist David John Gagne in a statement.
Moreover, determining the structure of the storm is important because it will allow us to know the size, condition (if severe), and the path of it.
Experiments in Predicting Extreme Weather
Recently, NCAR scientists presented the AI software with images of computer-generated storms paired with figures about temperature, pressure, and wind speed and direction.
The AI now started to predict the structure of storms and whether it will hail or not.
Surprisingly, the AI was able to confirm the storm features based on the computer-generated data.
However, the scientists don’t want to celebrate yet because the data they used is just simulated storms and it’s different than the actual ones.
Nonetheless, the team of scientists says that their study could sooner or later be operational and potentially predict the extreme weather that our world is experiencing.