Researchers have developed a novel neural operator framework, TI-DeepONet, for predicting seismic velocity fields in layered elastic media. This system combines a time-integrator-embedded Deep Operator Network with a Fourier Neural Operator to achieve temporal extrapolation beyond the training interval. The framework was tested on four cases involving SH waves in soil columns, demonstrating its ability to reproduce key velocity arrivals and depth-dependent response patterns with varying accuracy. AI
IMPACT Introduces a novel neural operator framework for enhanced temporal extrapolation in seismic wave prediction, potentially improving geophysical modeling accuracy.
RANK_REASON Academic paper detailing a new method for seismic wave propagation prediction using neural operators. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →