Automatic picking of multi-mode dispersion curves using CNN-based machine learning

Picking dispersion curves is required for many approaches for shallow-subsurface characterization using surface waves but is typically a labor-intensive process. Reliable, automatic picking would make these methods more efficient and practicable. We present a convolutional neural network (CNN) based machine learning (ML) approach to automatically pick the curves for the fundamental and higher modes along the two azimuths of any 2D seismic profile. The approach can be extended to 3D by applying 2D processing along multiple azimuths. Various attributes such as amplitudes, coherency, local phase velocity as well as frequency and wavenumber of dispersion curves are derived; different sub-sets of these are tested in the training process to assess the best combinations. We use the U-net architecture that is modified to convert the conventional 2D image segmentation problem in (f, k) domain into a direct multi-mode curve fitting and a subsequent picking process. The effectiveness of the automatic picking process is demonstrated in this study through applications to a field OBN dataset where different modes of Scholte waves were recorded.