Analysis

SWbatch User Guide


SWbatch is a user-friendly, web-based application for performing batch-style surface wave inversions using the Dinver module of the open-source software GeopsySWbatch allows the user to rapidly and conveniently invert experimental dispersion data considering multiple inversion parameterizations to address the problem’s non-uniqueness and multiple trials per parameterization to address the problem’s nonlinearity as detailed in the SWinvert workflow (Vantassel and Cox, 2020). To facilitate the potentially large amounts of pre- and post-processing required when performing batch surface-wave inversions a Python package, SWprepost, (Vantassel, 2020) has been released open-source. More information about SWprepost can be found on its GitHub page.

If you use SWbatch in your research we ask that you please cite the following:

Vantassel, J.P., Gurram, H., and Cox, B.R., (2020). jpvantassel/swbatch: latest (Concept). Zenodo. https://doi.org/10.5281/zenodo.3840546

Vantassel, J.P., Cox, B.R., (2020). SWinvert: A workflow for performing rigorous 1D surface wave inversions. Geophysical Journal International https://doi.org/10.1093/gji/ggaa426

Note: For software, version specific citations should be preferred to general concept citations, such as that listed above. To generate a version specific citation for SWbatch, please use the citation tool for that specific version on the SWbatch archive.

Getting Started

There are two ways of using SWbatch:

  1. As part of a developed Jupyter workflow called SWinvert. (Recommended)
  2. Or directly through the DesignSafe-CI Research Workbench.

Instructions for using the Jupyter Workflow

  1. Visit the SWprepost GitHub and follow the Getting Started instructions. The advanced example walks you through using the SWinvert surface wave inversion Jupyter workflow. (30 minutes)
  2. Login to DesignSafe. Transfer the advanced example and follow the instructions provided therein to repeat the tutorial. This time be sure to use the computational power of SWbatch to perform the inversion rather than viewing the results provided. Be sure to remove the previous inputs and results before running your inversion with SWbatch. Note more detailed instructions for completing this step are provided in the Jupyter notebook. (20 minutes, excludes inversion runtime)
  3. Upload your own experimental dispersion data and repeat the workflow. Be sure to remove the previous inputs and results before running your inversion with SWbatch. (20 minutes, excludes inversion runtime)
  4. Enjoy!

Instructions for using the DesignSafe-CI Research Workbench

  1. Visit the SWprepost GitHub and follow the Getting Started instructions. This will introduce you to SWprepost and the SWinvert workflow, which is required before proceeding to step 2 in these instructions. (30 minutes)
  2. Login to DesignSafe. Create a directory for your inversion, inside of which mimic the directory structure of the advanced example you completed as part of the previous step. Place your .target and .param files in the appropriate directories. (45 minutes)
  3. Launch SWbatch, by going to Research Workbench>Workspace>Simulation>SWbatch on DesignSafe and following the instructions provided there. (30 minutes, excludes inversion runtime)
  4. To see the status of your simulation refer to the Job Status bar. When your job is complete use the View button to view your inversion results. (5 minutes)
  5. Enjoy!