"UCSC and TU Munich Researchers Introduce RECAST: Deep Learning for Accurate Aftershock Prediction"

Researchers from UCSC and TU Munich have developed a new deep learning-based model called RECAST to forecast aftershocks. The model outperformed the older ETAS model, especially with larger datasets, and showed potential for revolutionizing earthquake forecasting. It can handle different earthquake datasets, combine information from various regions, and predict earthquakes in less-studied areas. The model also allows researchers to access different data types for earthquake prediction, such as continuous ground motion data.