Home /Software Release 2016 (January 18, 2016)

Software Release 2016 (January 18, 2016)

Software Releases — January 2016

We continue to push hard to produce software to reproduce our results. As you can see from the list below, we are starting to port our codes to 3D, include time-stepping modelling and inversion, anisotropy, and constraints. The latter is developed for “black box” applications allowing for direct integration with existing code bases.

Master branch

  1. Fast imaging with surface-related multiples by sparse inversion. (update in master branch) This package includes 1) the algorithm 2) synthetic examples to produce the main results shown in paper “N. Tu and F. J. Herrmann, Fast imaging with surface-related multiples by sparse inversion, Geophysical Journal international, 2015, 201, 304-317”. For questions contact Ning Tu. [Read more] [GitHub]
  2. Fast least-squares imaging with source estimation using multiples. (update in master branch) This package includes 1) the algorithm 2) synthetic examples to produce the main results shown in the manuscript “N. Tu et al., Source estimation with surface-related multiples—fast ambiguity-resolved seismic imaging, submitted to Geophysical Journal international, 2015”. For questions contact Ning Tu. [Read more] [GitHub]
  3. 2D/3D Time-domain modeling kernel for inversion. (update in master branch) This package contains a MATLAB implementation of a 2D time-domain acoustic modeling operator and its Jacobian. You can use up to 8th-order finite difference in space and 2nd-order in time. We also use PML boundary layer. This code is also implemented with domain-decomposition. Multiple matlab works are highly recommended for the simulations of large model. Hard or mirrored free-surface boundary condition is also included (optional). For questions contact Mathias Loubouting. [GitHub]

Beta in master branch

  1. Sparsity-promoting denoising of seismic data. (beta in master branch) This package contains a MATLAB implementation of sparsity-promoting denoising of seismic data in the curvelet domain using L1 minimization. For questions contact Haneet Wason. [Read more] [GitHub]
  2. Time-domain 2D/3D modelling and linearized modelling. (beta in master branch) This package contains basic examples of time-domain modelling. The 2D version allows acoustic and TTI acoustic anisotropic modelling and linearized modelling. The 3D modelling and linearized modelling has domain decomposition (parallel matlab necessary) and is acoustic only. For questions contact Mathias Louboutin. [Read more] [GitHub]
  3. Time-domain 2D FWI with TTI anisotropy. (beta in master branch) This package contains basic examples of time-domain FWI. The data is modelled and inverted with an TTI anisotropic kernel. We only invert for velocity. This simple example shows the result after 5 gradient step and also compare it with a purely acoustic kernel inversion. For questions contact Mathias Loubouting. [Read more] [GitHub]
  4. Constrained FWI. (beta in master branch) This package contains a set of routines that imposes multiple constraints onto “black-box” implementations of nonlinear inverse problems—such as full-waveform inversion—that aim to minimize an objective using local derivative (gradient) information. This package can work with any black box code that provides an objective function value and a gradient. Constraints include box constraints, TV-norm constraint, transform-domain sparsity constraints, various nonconvex constraints, several implementations of minimum smoothness constraints and several others. When provided a function value and gradient, this package sets up the constraints and solves the constrained optimization problem. For questions contact Bas Peters. [Read more] [GitHub]