While nuclear power plants have been beneficial to countries that use them, they can be problematic and deadly if something were to go wrong, which is what happened at Fukushima and Chernobyl. During these disasters, it was imperative to evacuate everyone as quickly as possible to avoid being harmed by any radioactive material that was released. Although the tools needed to predict where emissions of radiation would disperse, a new report says that researchers have developed a way to now achieve this using artificial intelligence.
Predicting Distribution of Radioactive Fallout Through Artificial Intelligence
According to Science News, researchers have developed a machine-learning-based tool with the ability to predict the direction of radioactive emissions leaked from a nuclear power plant will disperse. This was accomplished once it was taught to use extensive information on prior weather patterns. When the tool was used, it achieved consistently a predictive accuracy of over eighty-five percent while in the winter, the accuracy was up to ninety-five percent when predictable and large weather systems dominate.
This AI tool would be beneficial assisting in immediate evacuation during the aftermath of a nuclear power plant disaster. The study was presented in the journal Scientific Reports and the research team that was responsible is from the University of Tokyo Institute of Industrial Science.
The computer program can determine where the emitted radioactive material will land eventually over thirty hours in advance; this is accomplished through weather forecasts on wind patterns that are expected. This program will assist in implementing health-protective measures and plans for evacuation if another accident involving radioactive material escaping like the Fukushima Daiichi Nuclear Power Plant in 2011 were to happen.
Limitations of Existing Atmospheric Modeling Tools Prompted Study
Apparently, this latest study was motivated by the limits to current atmospheric modeling tools regarding the aftermath of the Fukushima accident; the tools were so unreliable that they decided not to use for evacuation planning immediately after the accident. The team of researchers developed a system that was a formed based on AI known as machine learning, this allows information from prior patterns of weather to predict the direction that emissions of radiation are likely to go.
Lead author Takao Yoshikane said that “our new tool was first trained using years of weather-related data to predict where radioactivity would be distributed if it were released from a particular point. In subsequent testing, it could predict the direction of dispersion with at least 85% accuracy, with this rising to 95% in winter when there are more predictable weather patterns.”
Yoshikane went on to say that “the fact that the accuracy of this approach did not decrease when predicting over 30 hours into the future is extremely important in disaster scenarios. This gives authorities time to arrange evacuation plans in the most badly affected areas, and to issue guidance to people in specific areas about avoiding eating fresh produce and taking potassium iodide, which can limit the absorption of ingested radioactive isotopes by the body.”