HiBasin

The Hierarchical BAyesian Source INversion (HiBASIN) is a Python package to perform seismic moment tensor (MT), single force (SF), or joint MT and SF inversion within a hierarchical Bayesian framework incoporating uncertainty estimates for data noise and theory error. Hibsasin is based on MTUQ and emcee.

Installation

1. Requirements:

   • MTUQ (https://github.com/mtuqorg/mtuq)
   • emcee (https://github.com/dfm/emcee)
   • pyrocko
   • basemap
   • corner

2. Install HiBasin:

git clone git@github.com:mtuqorg/HiBasin.git
cd HiBasin
pip install -e .

Documentation

Read the MTUQ documentation for Acquiring seismic data, Acquiring Green's functions, and Data processing. Note that, at least one-hour long pre-event ambient noise should be included in the downloaded seismic data. A cutting noise window will be used to estimate the noise.

Examples

Note: Before running any correlated noise or model error case, execute the following in your terminal:

ulimit -n unlimited

This is required because the implementation uses multiprocessing.shared_memory to reduce memory shared by hundreds of chains, which demands a large number of open file descriptors.

1. Full moment tensor inversion using HiBasin for a tectonic earthquake.

   uncorrelated noise treatment	correlated noise treatment
   Script, Figure	Script, Figure

   • Check here for inverted time shifts and noise parameters.
   • Check here for estimated covariance matrix for correlated noise.

2. Full moment tensor inversion using HiBasin for six DPRK explosions in 2006–2017 by considering data noise.

   2006	2009	2013	2016a	2016b	2017
   Script, Figure	Script, Figure	Script, Figure	Script, Figure	Script, Figure	Script, Figure

   • Check here for inverted time shift and noise parameters.

3. Tutorial for 1D model uncertainty treatment via covariance matrix.

   • Perturb the reference 1D model
   • Compute the ensemble of Green's functions
   • Estimate the covariance matrix by giving a reference moment tensor solution

   Check the script for details and the script and figure for an example.

   Check a synthetic experiment here for using the estimated covariance matrix.

Citation:

1. Hu, J., T.-S., Phạm, & H., Tkalčić, (2023). Seismic moment tensor inversion with theory errors from 2-D Earth structure: implications for the 2009–2017 DPRK nuclear blasts. Geophysical Journal International, 235(3), 2035–2054.
2. Mustać, M. & H., Tkalčić. (2016). Point source moment tensor inversion through a Bayesian hierarchical model. Geophysical Journal International, 204 (1), 311-323.
3. Phạm, T.-S. and H., Tkalčić. (2021). Toward improving point‐source moment‐tensor inference by incorporating 1d earth model's uncertainty: Implications for the long valley caldera earthquakes. Journal of Geophysical Research: Solid Earth 126 (11), e2021JB022477.