Deep Learning Methods and Applications to Geophysics
Deep learning (DL) methods have been widely applied in various fields. This course focuses on employing neural network (NN)-based methods to solve forward and inverse problems of geophysics, especially seismic rock physics. We first introduce some methods and approaches summarized by the lecturer based on published literature, and then introduce their applications in geophysics, including background knowledge, methods and numerical examples. There will be interactive time in each class, and all attendees are welcome to ask questions and communicate with the lecturer.
Lecturer
Date
14th October ~ 30th December, 2025
Location
| Weekday | Time | Venue | Online | ID | Password |
|---|---|---|---|---|---|
| Tuesday | 13:30 - 16:55 | A3-1-301 | ZOOM 06 | 537 192 5549 | BIMSA |
Prerequisite
Basic knowledge on deep neural networks, geophysics, partial differential equations, and the Python language.
Syllabus
1. Introduction to neural network (NN)-based approaches, some important knowledge points
NN-based Surrogate Modeling Methods:
2. Time stepping operator based on different network architectures
3. Different neural ODE benchmark models and the involved constraints
Several NN-based Methods:
4. Physics-informed neural network (PINN)
5. Deep neural operator (DeepONet)
6. Reduced order modeling (ROM)
7. Other new methods
The Application in Geophysics Problems:
8. Learning explicit/implicit models from data
9. Seismic rock physics
10. Joint inversion of seismic and electromagnetic methods
11. Full waveform inversion
12. Some review literatures, communication, and interaction
NN-based Surrogate Modeling Methods:
2. Time stepping operator based on different network architectures
3. Different neural ODE benchmark models and the involved constraints
Several NN-based Methods:
4. Physics-informed neural network (PINN)
5. Deep neural operator (DeepONet)
6. Reduced order modeling (ROM)
7. Other new methods
The Application in Geophysics Problems:
8. Learning explicit/implicit models from data
9. Seismic rock physics
10. Joint inversion of seismic and electromagnetic methods
11. Full waveform inversion
12. Some review literatures, communication, and interaction
Reference
1. Some methods and approaches summarized from published literature.
2. Published literature related to deep learning for solving differential equations, and the application in geophysics.
2. Published literature related to deep learning for solving differential equations, and the application in geophysics.
Audience
Graduate
, Postdoc
Video Public
No
Notes Public
No
Language
Chinese
Lecturer Intro
Fansheng Xiong (熊繁升) is currently an Assistant Professor of BIMSA. Before that, he received his doctoral degree in 2020 from Tsinghua University, and he was a visiting research assistant at Yale University during 2018-2019. His research interest mainly focuses on solving forward/inverse problems of PDEs based on machine learning method, and the application of PINN, DeepONet, Reduced Order Model in exploration geophysics, especially seismic rock physics. He is PI for grant from National Natural Science Foundation of China, and he has published papers in journals like Journal of Geophysical Research-Solid Earth, Geophysical Journal International, and Geophysics.