Chinese researchers have released the world's first near real-time intelligent prediction model for earthquake-induced landslides.

From Chengdu University of Technology, it was announced that Professor Fan Xuanmei's team has established the world's largest database of earthquake-induced landslides, based on nearly 400,000 landslides triggered by 38 strong earthquakes in the past 50 years. Combining deep learning algorithms, they have developed the world's first near-real-time intelligent prediction model for earthquake-induced landslides. This model can predict the spatial probability of any earthquake-induced landslide globally within one minute, with an average accuracy of 82%. This research achievement, titled "Deep Learning Enables Global Earthquake-Induced Landslide Prediction," was recently published in the international top journal National Science Review.