AI predicts aftershocks in seconds: how this will change the response to earthquake consequences.

According to Korrespondent.net: Artificial intelligence can now predict the risk of aftershocks just a few seconds after the main earthquake, according to ScienceX. Results from a new study show that machine learning models can quickly determine where and how many aftershocks to expect in the coming hours.

Speed of prediction

Traditionally, methods for detecting aftershocks can take from several hours to several days. Such delays complicate the ability to make timely decisions regarding safety and resource distribution in areas affected by earthquakes. Fast predictions provided by artificial intelligence could be critically important for emergency response teams.

Research and results

These tools were developed by scientists from the University of Edinburgh, the British Geological Survey, and the University of Padua. The models were trained on earthquake data from California, New Zealand, Italy, Japan, and Greece – regions with frequent seismic activity. The team analyzed how accurately the models could predict the number of aftershocks within 24 hours of earthquakes with a magnitude of 4 or greater.

The results were compared with the most widely used forecasting system – the ETAS model, which is used in Italy, New Zealand, and the USA. Although the accuracy of both approaches was similar, ETAS requires significantly more computation time – sometimes taking hours or days, as it is based on a large number of simulations.

Global application of models

Scientists note that AI models trained on data from various tectonic regions can be used for predictions in most seismically active parts of the world.

Recall that it was previously reported that Microsoft AI predicts weather and natural disasters more accurately than humans.

A new religion has emerged in the world: people pray to artificial intelligence

News from Korrespondent.net on Telegram and WhatsApp. Subscribe to our channels.

The development of artificial intelligence technologies opens new possibilities in the field of seismology. Fast and accurate forecasting models can significantly improve safety measures in areas where the risk of earthquakes is high. This is especially important for regions that suffer from frequent natural disasters since every second counts during emergencies.