This repo contains materials for the course ErSE328 Advanced Seismic Inversion course taught by Professor Tariq Alkhalifah in King Abdullah University of Science. An inversion example is shown below:
Throughout the computational part of the course, we will mainly be utilizing the Deepwave Python library which you can access from their repository. To get yourself started, you can directly learn the Deepwave fundamentals from their documentations.
To install the environment, run the following command:
sh install_env.sh
It will take some time, but if, in the end, you see the word Done! on your terminal, you are ready to go.
Remember to always activate the environment by typing:
conda activate erse328asi
First connect to Ibex using your KAUST credential.
Then, clone this repository and install the erse328asi conda environment. If it is installed please go ahead with the following command. If not, please refer to this documentation to install conda. Inside the ErSE328-AdvancedSeismicInversion folder, run the following command to submit the slurm jupyter notebook request.
sbatch erse328asi_notebook.slurm
You can connect from your workstation to access the notebook with the instructions from the output slurm job. The file is in the format of slurm-JOBID.out.
ssh -L 6789:GPUID:6789 [email protected]
where the JOBID and GPUID are the unique identifiers from the slurm output job request. Then, access the http://localhost:6789/ link from your workstation and fill in the credential from the slurm-JOBID.out.
| Assignment # | Due date | Objectives |
|---|---|---|
| Assignment 0 | 04/03/2024 | Ensure the Deepwave package is installed properly. |
| Assignment 1 | 07/03/2024 | Perform FWI to the Marmousi and suggest ways to improve the results. |
| Assignment 2 | 24/03/2024 | Perform multiscale FWI to improve the results. |
| Assignment 3 | 05/05/2024 | Remedy the ill-posedness of FWI using regularization. |
| Assignment 4 | 05/05/2024 | Moving beyond acoustic media assumption. |
