P and S Waves Detection with Deep Learning

Installation

pip install --extra-index-url https://epos-apps.grid.cyfronet.pl/api/packages/epos-ai/pypi/simple epos_ai_picking_tools

Generalized Phase Detection (GPD) Model

Documentation

Usage: gpd_tool [OPTIONS] COMMAND [ARGS]...

Options:
  --help  Show this message and exit.

Commands:
  citation         Prints citation of the model
  list-pretrained  Show pretrained model names
  pick             Detect phases in streams
  version          Prints version

Pick

Usage: gpd_tool pick [OPTIONS] [STREAM_FILE_NAMES]...

  Detect phases in streams

Options:
  -w, --weights TEXT        for possible options see output of 'list-pretrained'
                            [default: original]
  -o, --output PATH         directory to store results  [default: .]
  -s, --stride INTEGER      stride in samples for point prediction models
                            [default: 10]
  -tp, --threshold-p FLOAT  detection threshold for the P phase  [default: 0.75]
  -ts, --threshold-s FLOAT  detection threshold for the S phase  [default: 0.75]
  --help                    Show this message and exit.

Citation

Ross, Z. E., Meier, M.-A., Hauksson, E., & Heaton, T. H. (2018). Generalized Seismic Phase Detection with Deep Learning. ArXiv:1805.01075 [Physics]. https://arxiv.org/abs/1805.01075

PhaseNet

Documentation

Usage: phasenet_tool [OPTIONS] COMMAND [ARGS]...

Options:
  --help  Show this message and exit.

Commands:
  citation         Prints citation of the model
  list-pretrained  Show pretrained model names
  pick             Detect phases in streams
  version          Prints version

Pick

Usage: phasenet_tool pick [OPTIONS] [STREAM_FILE_NAMES]...

  Detect phases in streams

Options:
  -w, --weights TEXT        for possible options see output of 'list-pretrained'
                            [default: original]
  -o, --output PATH         directory to store results  [default: .]
  -tp, --threshold-p FLOAT  detection threshold for the P phase  [default: 0.3]
  -ts, --threshold-s FLOAT  detection threshold for the S phase  [default: 0.3]
  -b, --blinding TUPLE      number of prediction samples to discard on each side
                            of each window prediction  [default: 0, 0]
  --help                    Show this message and exit.

Citation

Zhu, W., & Beroza, G. C. (2019). PhaseNet: a deep-neural-network-based seismic arrival-time picking method. Geophysical Journal International, 216(1), 261-273. https://doi.org/10.1093/gji/ggy423