Earthquake Nowcasting with Deep Learning
This work addresses earthquake forecasting for disaster preparedness, but it appears incremental as it builds on existing deep learning methods with new models and data.
The paper tackles earthquake prediction by applying deep learning models, including recurrent neural networks and transformers, to forecast seismic activity in Southern California from 1950 to 2020, achieving results measured by Nash Sutcliffe Efficiency.
We review previous approaches to nowcasting earthquakes and introduce new approaches based on deep learning using three distinct models based on recurrent neural networks and transformers. We discuss different choices for observables and measures presenting promising initial results for a region of Southern California from 1950-2020. Earthquake activity is predicted as a function of 0.1-degree spatial bins for time periods varying from two weeks to four years. The overall quality is measured by the Nash Sutcliffe Efficiency comparing the deviation of nowcast and observation with the variance over time in each spatial region. The software is available as open-source together with the preprocessed data from the USGS.