My research pairs methods-driven model development with scientific question-driven applications.
This work couples ice-sheet, sea-level and climate models to disentangle the individual sea-level contributions of the four paleo ice sheets over the last 21,000 years. We use large ensembles of simulations, representing uncertainty in forcing and model dynamics, to investigate how distinct ice-sheet configurations may be consistent with the sea-level record.
Large model runs are only useful if others can check and build on them. I develop automated, documented, version-controlled pipelines for running and post-processing simulations on shared HPC clusters, with an emphasis on workflows that a new student or collaborator can pick up without months of onboarding.
My PhD work developed emulators of a popular subglacial hydrology model that are are up to 100,000 times faster than solving the numerical model. The emulators were used in a Bayesian framework to calibrate model parameters and reduce uncertainty in model predictions.
Simulations that couple the emulators with an ice-flow model add to existing evidence that modellling choices, such as the representation of basal hydrology and friction, strongly influence model predictions for the coming 100 and 300 years.
I completed my PhD at Simon Fraser University (Earth Sciences, 2026), working on