• Geophysics and Geophysical Exploration
• /
• v.23 no.1
• /
• pp.38-49
• /
• 2020

Full-waveform inversion (FWI) is an optimization process of fitting observed and modeled data to reconstruct high-resolution subsurface physical models. In acoustic FWI (AFWI), pressure data acquired using a marine streamer has mainly been used to reconstruct the subsurface P-wave velocity models. With recent advances in marine seismic-acquisition techniques, acquiring multi-component data in marine environments have become increasingly common. Thus, AFWI strategies must be developed to effectively use marine multi-component data. Herein, we proposed an AFWI strategy using horizontal and vertical particle-acceleration data. By analyzing the modeled acoustic data and conducting sensitivity kernel analysis, we first investigated the characteristics of each data component using AFWI. Common-shot gathers show that direct, diving, and reflection waves appearing in the pressure data are separated in each component of the particle-acceleration data. Sensitivity kernel analyses show that the horizontal particle-acceleration wavefields typically contribute to the recovery of the long-wavelength structures in the shallow part of the model, and the vertical particle-acceleration wavefields are generally required to reconstruct long- and short-wavelength structures in the deep parts and over the whole area of a given model. Finally, we present a sequential-inversion strategy for using the particle-acceleration wavefields. We believe that this approach can be used to reconstruct a reasonable P-wave velocity model, even when the pressure data is not available.

• Geophysics and Geophysical Exploration
• /
• v.22 no.4
• /
• pp.202-209
• /
• 2019

In this study, an acoustic full-waveform inversion using Adam optimizer was proposed. The steepest descent method, which is commonly used for the optimization of seismic waveform inversion, is fast and easy to apply, but the inverse problem does not converge correctly. Various optimization methods suggested as alternative solutions require large calculation time though they were much more accurate than the steepest descent method. The Adam optimizer is widely used in deep learning for the optimization of learning model. It is considered as one of the most effective optimization method for diverse models. Thus, we proposed seismic full-waveform inversion algorithm using the Adam optimizer for fast and accurate convergence. To prove the performance of the suggested inversion algorithm, we compared the updated P-wave velocity model obtained using the Adam optimizer with the inversion results from the steepest descent method. As a result, we confirmed that the proposed algorithm can provide fast error convergence and precise inversion results.