LAUSR

Retrieving subsurface velocity information from recorded seismograms generally involves solving an inverse problem using various optimization methods. Deep learning (DL) has recently become an emerging alternative technique to provide solutions for such velocity inversion tasks. However, due to the lack of labeled data in field applications, supervised DL methods proposed for velocity model estimations often rely on training with synthetic modeled data. Applying these DL models to field data can be severely limited by differences between the configurations used in the synthetic training data and those for field surveys. We investigate the impacts of acquisition geometries on DL velocity inversion and propose an acquisition-robust DL framework to improve the model’s applicability by mitigating the impacts of different acquisition geometries. We train the proposed DL model with synthetic training data transformed to an alternative domain, e.g., wavenumber–time domain in this study. We demonstrate that DL models trained in this way are more robust and allow generalization over different acquisition configurations.