Seismic full waveform inversion : nonlocal similarity, sparse dictionary learning, and time-lapse inversion for subsurface flow
Abstract/Contents
- Abstract
- I present three topics in the area of theoretical development and application of Full Waveform Inversion (FWI). The first topic or contribution describes a learning-based adaptive and sparsity promoting regularization method to improve the accuracy of traditional FWI results with the prior knowledge of nonlocal similarity in geological structures. Such a priori is realized by multi-class orthogonal dictionary learning. The second topic extends the learning-based regularization to elastic waves and applies the approach to a field dataset to estimate P-wave and S-wave velocities. This topic highlights the importance of data pre-processing and modification of the FWI algorithm to accommodate field data issues such as radiation pattern estimation, wavelet estimation, and amplitude scaling. The aim of this part is high-resolution reservoir characterization, which is not only an extension of the technique in the first part but also serves as a preparation for the time-lapse inversion in the third part. In the third topic, I describe a framework for using FWI to estimate hidden parameters that are important to geophysical processes such as fluid flow in porous media. This framework extends the power of FWI beyond seismology to other geophysical problems such as reservoir engineering and hydrology by combining seismic observations, rock properties modeling, and flow modeling. This third topic represents a PDE-constrained inverse problem that I solve with an intelligent automatic differentiation method. The method provides three levels of user control with (1) built-in differentiable operators from modern deep-learning infrastructures, and customized operators that can either (2) encapsulate analytic adjoint gradient computation or (3) handle the forward simulation and compute the corresponding gradient for a single time step. This intelligent strategy strikes a good balance between computational efficiency and programming efficiency and would serve as a paradigm for a wide range of PDE-constrained geophysical inverse problems.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2019; ©2019 |
| Publication date | 2019; 2019 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Li, Dongzhuo |
|---|---|
| Degree supervisor | Harris, Jerry M |
| Thesis advisor | Harris, Jerry M |
| Thesis advisor | Biondi, Biondo, 1959- |
| Thesis advisor | Mukerji, Tapan, 1965- |
| Degree committee member | Biondi, Biondo, 1959- |
| Degree committee member | Mukerji, Tapan, 1965- |
| Associated with | Stanford University, Department of Geophysics. |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Dongzhuo Li. |
|---|---|
| Note | Submitted to the Department of Geophysics. |
| Thesis | Thesis Ph.D. Stanford University 2019. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2019 by Dongzhuo Li
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...