Two Dimensional Acoustic Full Waveform Inversion using Discrete Cosine Transform
Abstract
Full waveform inversion (FWI) has been carried out to reconstruct high-resolution velocity model in the subsurface. However, as geophysical data acquisition techniques are developed for obtaining high dimensional velocity model, computational cost increases tremendously. Because we set unknowns on each modeling grid point, and these unknowns are updated in FWI. So FWI always suffers from computational burden even though it has many advantages. To reduce computational burden, we incorporate Discrete Cosine Transform (DCT) that is one of transformations used widely in image processing. In this paper, DCT is applied to FWI for reducing the number of unknowns. The velocity model is transformed by DCT, and a small part of DCT coefficients are used in FWI to delineate velocity model features without much of loss in reconstruction accuracy. In selecting DCT coefficients, we sort the absolute value of DCT coefficients in descending order and choose DCT coefficients with compression ratio for using more dominant DCT coefficients. To investigate the applicability of our DCT-based FWI method, we compare reconstructed velocity model with true velocity model. We find that velocity model is reconstructed with satisfactory accuracy despite of using a small number of DCT coefficients. After that, we apply our FWI method to numerical examples. It is shown that our method can produce satisfactory results along with alleviating the computational burdens significantly. From numerical results, we expect that our method can have advantage to improve computational efficiency for 3D FWI.
ACKNOWLEDGEMENTS This work was supported by Korea Institute of Energy Technology Evaluation and Planning (KETEP) and the Ministry of Trade, Industry & Energy (MOTIE) of the Republic of Korea (No.20182510102470). Jonghyun Lee was supported by Hawai'i Experimental Program to Stimulate Competitive Research (EPSCoR) provided by the National Science Foundation Research Infrastructure Improvement (RII) Track-1: 'Ike Wai: Securing Hawai'i's Water Future Award OIA 1557349 and Faculty Research Participation Program at the U.S. Engineer Research and Development Center, Coastal and Hydraulics Laboratory administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U.S. department of Energy and ERDC.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.S53D0485K
- Keywords:
-
- 0555 Neural networks;
- fuzzy logic;
- machine learning;
- COMPUTATIONAL GEOPHYSICS;
- 1942 Machine learning;
- INFORMATICS;
- 3260 Inverse theory;
- MATHEMATICAL GEOPHYSICS;
- 3275 Uncertainty quantification;
- MATHEMATICAL GEOPHYSICS