Elastic near-surface model estimation from full waveforms by deep learning

Vladimir Kazei, Oleg Ovcharenko, Pavel Plotnitskii, Daniel Peter, Tariq Ali Alkhalifah, Ilya Silvestrov, Andrey Bakulin, Paul Zwartjes

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Scopus citations


Strong near-surface heterogeneity poses a major challenge for seismic imaging of deep targets in arid environments. Inspired by the initial success of deep learning applications to inverse problems, we investigate the possibility of building nearsurface models directly from raw elastic data including surface and body waves in arid conditions. Namely, we train a convolutional neural network to map the data into the model directly in a supervised way on a part SEAM Arid synthetic dataset and evaluate its performance on a different part of the same dataset. The main feature of our approach is that we estimate the model as a set of 1D vertical velocity profiles, utilizing relevant subsets of input data from neighboring locations. This effectively reduces the data and label spaces for a more practical neural network application.
Original languageEnglish (US)
Title of host publicationSEG Technical Program Expanded Abstracts 2020
PublisherSociety of Exploration Geophysicists
StatePublished - Oct 1 2020

Bibliographical note

KAUST Repository Item: Exported on 2020-10-08
Acknowledgements: We thank Jan Walda from the University of Hamburg, mem-bers of the Seismic Modeling and Inversion group (SMI) andthe Seismic Wave Analysis Group (SWAG) at KAUST for con-structive discussions. The research reported in this publica-tion was supported by funding from King Abdullah Universityof Science and Technology (KAUST), Thuwal, 23955-6900,Saudi Arabia and Saudi Aramco.


Dive into the research topics of 'Elastic near-surface model estimation from full waveforms by deep learning'. Together they form a unique fingerprint.

Cite this