Supervisor
- Hakase Hayashida ( hakase.hayashida@utas.edu.au )
Ice algae colonise the base of Antarctic sea ice in early spring, serving as a food source for grazers. Quantifying ice algal biomass is therefore a key to understand polar marine ecosystems as a whole. Despite its importance, field measurements of ice algal biomass are limited spatially as well as temporally due to technical methodological challenges in these remote and harsh environments.
Mathematical optimisation such as machine learning can be a powerful tool to fill in this observational gap, help interpret the observed variability, and therefore, improve our understanding of the role of ice algae in polar marine ecosystems.
In this project, the selected student will develop a numerical algorithm to generate a spatial map of Antarctic ice algal biomass using Machine Learning. The student will use the existing database of discrete ice algal biomass measurements to train the algorithm. Depending on the interest and the progress of the student, this project can be further expanded to:
- Evaluate the skill and the sensitivity of the algorithm.
- Apply the algorithm for the Arctic domain.
- Generate spatial maps under various future climate scenarios.
This project may be suitable for students who are interested in polar marine science and programming. A strong programming skill in Python is essential, and previous experience in Machine Learning is desirable.