Computational Geology and Geophysics

Machine Learning of Seismic and Image Data

Some volcanoes display a persistent lava lake and frequent eruptions, called Strombolian. We use Convolutional Neural Networks to analyze over 9 million infrared images collected at the Erebus Volcano, Antarctica, from 2013 to 2016. (Dr. Morra)

Numerical Modeling of Flow through Porous Media

While Enhanced Oil Recovery (EOR) such as water injection is one of the main techniques used today to extract oil from reservoirs, it is only partially known what are the optimal viscosity and wettability conditions of the injecting fluids to (1) reach breakthrough and (2) increase saturation and therefore maximize extraction. The project consists in using Colour-Gradient Lattice Boltzmann Method (LBM) to calculate these properties for various porous mediums. (Dr. Morra)

Formation and Evolution of Planets, and Implications for Habitability

Planets go through several stages during their evolution, where the most critical moment is the formation of their iron core. This phase ultimately dictates the initial distributions of key life-essential elements and fundamental to the formation of habitable worlds. The project consists in executing fluid dynamic models for metal-silicate interactions in magma oceans with variable initial and boundary conditions that couple changing chemical-physical properties. (Dr. Morra)