This library contains all the required building blocks to peform 2D turning ray tomography and associated examples codes.

Project structure

This repository is organized as follows:

• 📂 refrtomo: python library containing routines for 2D turning ray tomography;
• 📂 data: folder containing data
• 📂 notebooks: set of jupyter notebooks reproducing the experiments in the paper (see below for more details);

Notebooks

The following notebooks are provided:

• 📙 Rayshooting_straight.ipynb: notebook testing raytracer and tomographic matrix engine with straight rays;
• 📙 Rayshooting.ipynb: notebook testing raytracer and tomographic matrix engine with turning rays;
• 📙 Rayshooting_Stryde.ipynb: notebook testing raytracer and tomographic matrix engine with small refraction dataset acquired with Stryde sensors in KAUST;
• 📙 RefrTomo.ipynb: notebook performing refraction tomographic of a portion of synthetic Marmousi model;
• 📙 RefrTomo_minimal.ipynb: notebook performing refraction tomographic of a portion of synthetic Marmousi model (minimal number of steps);
• 📙 FirstArrivalPicking_Stryde.ipynb: notebook performing automated first arrival picking of a small refraction dataset acquired with Stryde sensors in KAUST;
• 📙 RefrTomo_Stryde.ipynb: notebook performing refraction tomographic of a small refraction dataset acquired with Stryde sensors in KAUST.
• 📙 RefrTomo_Strydeminimal.ipynb: notebook performing refraction tomographic of a small refraction dataset acquired with Stryde sensors in KAUST (minimal number of steps).

Getting started 👾 🤖

To ensure reproducibility of the results, we suggest using the environment.yml file when creating an environment.

Simply run:

./install_env.sh

It will take some time, if at the end you see the word Done! on your terminal you are ready to go. After that you can simply install your package:

pip install .

or in developer mode:

pip install -e .

Remember to always activate the environment by typing:

conda activate refrtomo

Improvements ☝️

• Create routine that takes two rays reaching the surface and recursively interpolates a ray in the middle until a given distance from a surface point of interest (to be used to take the two closest rays on either side of a receiver, which are however above the allowed threshold, and find one in the middle below this threshold);

• Finalize inversion routines using scipy.optimize.least_squares (currently the Levenberg-Marquardt cannot be used with non-dense Jacobians and the trust-region methods do not seem to converge as nicely as our Gauss-Newton);

• Add directional smoothing constraints.