Why discretize?#
Inverse problems are common across the geosciences: imaging in geophysics, history matching, parameter estimation, and many of these require constrained optimization using partial differential equations (PDEs) where the derivative of mesh variables are sought. Finite difference, finite element and finite volume techniques allow subdivision of continuous differential equations into discrete domains. The knowledge and appropriate application of these methods is fundamental to simulating physical processes. Many inverse problems in the geosciences are solved using stochastic techniques or external finite difference based tools (e.g. PEST); these are robust to local minima and the programmatic implementation, respectively, however these methods do not scale to millions of parameters to be estimated. This sort of scale is necessary for solving many of the inverse problems in geophysics and increasingly hydrogeology (e.g. electromagnetics, gravity, and fluid flow problems).
In the context of the inverse problem, when the physical properties, the domain, and the boundary conditions are not necessarily known, the simplicity and efficiency in mesh generation are important criteria. Complex mesh geometries, such as body fitted grids, commonly used when the domain is explicitly given, are less appropriate. Additionally, when considering the inverse problem, it is important that operators and their derivatives are accessible to interrogation and extension. The goal of this work is to provide a high level background to finite volume techniques abstracted across four mesh types:
tensor product mesh
discretize.TensorMesh
cylindrically symmetric mesh
discretize.CylMesh
curvilinear mesh
discretize.CurvilinearMesh
octree and quadtree meshes
discretize.TreeMesh
discretize
contributes an overview of finite volume techniques in the
context of geoscience inverse problems, which are treated in a consistent way
across various mesh types, highlighting similarities and differences.
Citing discretize#
There is a paper about discretize as it is used in SimPEG, if you use this code, please help our scientific visibility by citing our work!
Cockett, R., Kang, S., Heagy, L. J., Pidlisecky, A., & Oldenburg, D. W. (2015). SimPEG: An open source framework for simulation and gradient based parameter estimation in geophysical applications. Computers & Geosciences.
BibTex:
@article{Cockett2015,
title={SimPEG: An open source framework for simulation and gradient based parameter estimation in geophysical applications},
author={Cockett, Rowan and Kang, Seogi and Heagy, Lindsey J and Pidlisecky, Adam and Oldenburg, Douglas W},
journal={Computers \& Geosciences},
year={2015},
publisher={Elsevier}
}
License#
The MIT License (MIT)
Copyright (c) 2013-2023 SimPEG Developers
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.