Unravelling the Earth through software

My current research and outlook, and its place within teaching and research in the School of Environmental Sciences

Leonardo Uieda

University of Liverpool - Jan 9, 2019

Background

Amateur Baker

Guiding questions

1. How can we constrain inverse problems?
2. How can we make more efficient algorithms?
3. What can we learn from machine learning?
4. How can we make science open and reproducible?

Main projects

Tesseroids

Fatiando a Terra

Generic Mapping Tools

Open-source

• Code on Github
• Community of developers
• Documentation
• Best practices

Inverse problems

1. Crustal Thickness

2. 3D Geometry

Crustal thickness
from gravity

3D Geometry from gravity/magnetics

Future work

• Extensions/improvements to methods
• Sedimentary basin inversion
• Melting in Antarctica with GRACE + tesseroids
• Moho of Africa/Antarctica
• Planting inversion on volcanoes

Data processing

GPS interpolation

Future work

• Tests on different data
• Optimize for larger datasets
• Joint interpolation with InSAR
• Apply to gravity and magnetic data

Education

Workshops

• Geophysical inversion
• Programming with Python
• Version control (Git + Github)
• OSS development

Undergraduate

• Exploration geophysics
• Programming & numerical methods
• Introduction to geology
• Geophysics field project

Learning through
guided experimentation

(example: seismic wave propagation)

Future

• Workshops (Software Carpentry)
• EPSRC funding
• Geographic Data Science Lab
• Central Teaching Hub
• Modules: ENVS386, ENVS201, ENVS229,
  ENVS343, ENVS258, ENVS216

Outreach

Conclusion

• Openness and reproducibility
• Method development benefits from collaboration
• Simulation encourages experimentation
• Help implement interactive modules

Acknowledgments

Collaborators

Funding

Contact

• leouieda@gmail.com
• leouieda.com