A new AI-based program, developed by researchers from the University of Alberta, utilizes historical meteorological data to predict precise location of forest fires
Increasing temperature, global warming, and climate changes are propelling the risk of natural disasters such as wildfires in many countries around the world. New software developed by researchers uses artificial neural networks to sift through the data and predict climatic conditions where extreme weather events are likely occur. The system is trained on historical data, before then, making probabilistic predictions, which are in turn approved by expert researchers. If the predictions are not approved, the system modifies its approach and tries again.
The technology is able to analyze a richer dataset more efficiently as compared to conventional forecasting systems. Whereas existing systems usually rely on precipitation, temperature, wind, and humidity data, the new system also incorporates pressure fields. It’s an approach the team believes gives the system a more realistic perspective on potential fire-risk conditions. To identify high risk conditions, drones might be required to track ensuing fire. Researchers at the University of Nebraska have developed a drone to help safely manage the growing number of wildfires unfolding around the world.
“Unmanned aerial devices have the potential to carry out key resource management strategies and could help us deal with something as big as the international increase in severe wildfires,” the team explained. Furthermore, team believes that the new drones could eventually take the place of manned aircraft and hotshot firefighting teams that are currently used in wildfire fighting scenarios. Such new technologies are expected to help in overcoming challenge that has huge human and financial implications.