Reproducible material for Stochastic Multi-dimensional Deconvolution - Ravasi M., Selvan, T., Luiken N. - ArXiv Paper.
Project structure
This repository is organized as follows:
📂 stochmdd: python library containing routines for stochastic mdd📂 notebooks: set of jupyter notebooks reproducing the experiments in the paper (see the README file inside this folder for more details)
📂 data: folder where input data must be placed
NOTE: due to their large size, the various datasets cannot be shared directly in this repository. You can download the VolveSynthetic dataset from this Zenodo link. If interested in any of the other datasets, contact the authors directly!
Data and Experiements
Four experiments are considered:
- Hyperbolic: set of synthetically generated hyperbolic events for both model and kernel operator.
- Dipping_OBC: layered model with dipping seabed and receivers at the seabed in OBC style acquisition. Data are created via FD modelling followed by up/down separation.
- Salt: salt model from Vargas et al. (2021). Data are created via Scattering-Rayleigh-Marchenko redatuming.
- Synthetic_Volve: synthetic model that resembles the Volve field subsurface model. See VolveSynthetic for more details regarding the generation of the model and seismic data. Data are created via FD modelling followed by up/down separation.
- Field_Volve: field Volve OBC dataset, data are pre-processed by up/down separation.
Getting started 👾 🤖
To ensure reproducibility of the results, we suggest using the environment.yml file when creating an environment.
When working within KAUST computational environment, first install Miniconda using https://github.com/kaust-rccl/ibex-miniconda-install.
Once you have made sure that Anaconda (or Miniconda) is available in your system, to create a new environment simply run:
conda env create -f environment.yml
to create an environment called mdd-stochastic.