Ben Mather PRO
Computational Geophysicist at the Sydney Informatics Hub, University of Sydney
Dr. Ben Mather
EarthByte Group & Sydney Informatics Hub
University of Sydney
GPlates
- Configuration of plate boundaries through time
- Downloaded by 102,259 individuals
- Tutorials, manuals, free data and models
- Industry standard for model-data integration
- 926 Google Scholar mentions of GPlates software
GPlates reconstruction
Position of Australia back through geological history
pyGPlates
- Jupyter notebooks hosted inside Docker images
- Seamless integration with data processing in Python
import pygplates
time = 40 # Ma
# point on the sphere
pt = pygplates.PointOnSphere(lat, lon)
# create a point feature
point = pygplates.Feature()
point.set_geometry(pt)
point.set_reconstruction_plate_id(plate_ID)
pygplates.reconstruct(point,
reconstruction_file,
output_file,
time)GPlates Portal
Gateway to virtual globes and cloud-based computing services.
Potential application of spatio-temporal data mining for education, resource exploration.
pyGPlates examples run in the cloud
Underworld
- Framework to solve mantle convection problems
- 92,479 "events" in the last year
- UW2 has a Python wrapper with tutorials, manuals, and workflows
- Simplified geodynamics framework - UWGeo
- UW3 will solve mantle convection on the sphere
UWGeodynamics
- High level functions within the UW ecosystem
- Rapid prototyping of geodynamic models
- Assumes little knowledge in coding
import UWGeodynamics as GEO
Model = GEO.Model(elementRes=(64, 64),
minCoord=(0. * u.kilometre, 0. * u.kilometre),
maxCoord=(64. * u.kilometre, 64. * u.kilometre))Coupling surface processes with UWGeo
UW3 on the sphere!
Tackle whole earth problems using a familiar Python interface
\times 6
Cubed
Sphere
Whole earth model
Badlands
- Framework to solve landscape evolution
- pyBadlands is a Python wrapper to interface with data-driven frameworks
- Docker images for easy deployment
Influence of mantle flow on the drainage of E Australia
Salles, Flament, Muller 2017
Integrated Earth Models
Future directions...
ARC Basin Genesis Hub
- 5-year joint ARC/Industry funded centre
- 80 members from academia and industry
- All BGH software development is open-source
GPlates
Underworld
Badlands
Probability
Understanding solid earth evolution:
- Model-data assimilation
- Uncertainty quantification
- Visualisation
Bayes Theorem
P(\mathbf{m}|\mathbf{d}) \propto P(\mathbf{d}|\mathbf{m}) P(\mathbf{m})
Infer parameters from the available data and what we already know about the model
Model-data integration
- Assimilate high precision and high resolution geochemical data from the surface
- Develop inverse/adjoint modelling capabilities
- Build statistical approaches to evaluate uncertainty in complex models
Thank you
Dr. Ben Mather
Madsen Building, School of Geosciences,
The University of Sydney, NSW 2006
NCRIS
By Ben Mather
