Overview

Florian Wagner, Carsten Rücker, Thomas Günther, Andrea Balza

Tutorial Info:

Many tasks in applied geosciences cannot be solved by a single measurement method and require the integration of geophysical, geological, geotechnical and hydrological methods. Such data integration tasks need flexible open-source software. PyGIMLi is a Python library for modeling and inversion, which provides finite element and finite volume solvers on structured and unstructured 2D/3D meshes, various ready-to-use geophysical forward operators and method managers, as well as a generalized deterministic Gauss-Newton inversion algorithm with flexible means of incorporating prior knowledge (e.g., through region-specific regularization). In this tutorial, we will provide an overwiew of pyGIMLi‘s functionality and applications, solve stationary heat conduction on a 2D mesh in a Jupyter Notebook with pyGIMLi‘s modeling level and create and invert synthetic geophysical data on a user-specified subsurface geometry. The tutorial will conclude with further development plans, integration with other software packages and guidelines to contribute.

Content

  • 00:00:00 - Introduction
  • 00:01:09 - Main features, conda installer, API doc
  • 00:09:30 - 2D meshtools demonstration
  • 00:31:10 - Equation level: 2D heat equation
  • 00:48:19 - Crosshole traveltime forward modeling
  • 01:04:42 - Method Manager: Traveltime inversion
  • 01:24:47 - Inverting electrical resistivity field data
  • 01:41:18 - Inversion with own forward operator
  • 01:49:54 - Homepage with examples, papers, contribution guide