Nowadays, despite the fact that technology has become an integral part of our lives, there are still natural disasters that are beyond our control. Wildfires are becoming more severe and long lasting due to climate change and other factors that humans can’t control. Unlike hurricanes and tornadoes, which can be predicted days in advance, wildfires can change rapidly and erratically.
That’s why researchers at Aalto University have turned to artificial intelligence to develop a special model that could solve the problem. The AI model was trained on data from peatlands in Indonesia’s Central Kalimantan province recorded between 2002 and 2019. It can make complex predictions about how different interventions could prevent or mitigate future wildfires in the region, taking into account factors such as type of land cover, pre-fire vegetation, and drought indices. The researchers claim that Fire CNN could help cut the number of fires by 50 to 70 percent. The team is currently working to make the model more accurate by testing it in other regions. Their main goal is to quantify how different prevention strategies would work.
All photography courtesy of NASA
