Full text loading...
We developed a machine learning-based approach for P- and S-wave separation in DAS data. We used the simulation method of separated P- and S-waves for single component data to generate DAS training data. We then trained a machine learning model to simultaneously output DAS-P and DAS-S data, effectively separating P- and S-waves from input DAS recordings. The model was tested on the vertical DAS configuration, demonstrating its ability to effectively separate and image even weaker wave components.