Seismic event P phase picking project
Main framework: Obspy, Seisan, Tensorflow with Keras
Using U-net to generate pick probability
This version is unstable. Do not use it now.
The code is still in the development state, API will change frequently.
Please star us for upcoming updates!
Prerequisite:
-
S-File catalog from SEISAN
-
SeisComP Data Structure (SDS) database. The directory and file layout of SDS is defined as:
SDSROOT/YEAR/NET/STA/CHAN.TYPE/NET.STA.LOC.CHAN.TYPE.YEAR.DAY
Installation:
-
Follow the instructions in the Docker folder to create a Docker container.
-
SSH into the Docker container you create.
ssh username@localhost -p49154
-
Copy
/SeisNN/jupyter.sh
to your workspace and execute to start jupyter lab servercp /SeisNN/jupyter.sh ~/. chmod 777 jupyter.sh ./jupyter.sh
-
Copy
/SeisNN/notebook
to your workspacecp -r /SeisNN/notebook ~/.
-
Paste the URL with generate token into your local browser
http://127.0.0.1:8888/?token=36b31a373a9d18cc9b30a50883ad5a3638b19bed47be8074
-
Go through notebook/00_initialize.ipynb to generate config.yaml
In the notebook folder:
Reference:
Zhu, W., & Beroza, G. C. (2018). PhaseNet: A Deep-Neural-Network-Based Seismic Arrival Time Picking Method. arXiv preprint arXiv:1803.03211.
Ronneberger, O., Fischer, P., & Brox, T. (2015, October). U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention (pp. 234-241). Springer, Cham.
Zhou, Z., Siddiquee, M. M. R., Tajbakhsh, N., & Liang, J. (2018). Unet++: A nested u-net architecture for medical image segmentation. In Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support (pp. 3-11). Springer, Cham.
Personal Blog (Traditional Chinese only):