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
×6\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})
P(md)P(dm)P(m)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

https://benmather.info