PhD Studentship: Using generation ocean observations and machine learning to transform mixed layer depth

The Mixed Layer Depth (MLD) governs ocean-atmosphere exchanges, vertical mixing and ocean dynamics. Ocean models can simulate high-frequency MLD dynamics—such as realistic storm mixing—they often fail to resolve them temporally, which means they cannot be used directly to resolve or predict day-to-day changes. This project addresses a key science question:
What physical processes and high-resolution inputs are required to temporally resolve hourly MLD dynamics?

Answering this is increasingly critical. Precise, temporally resolved MLD forecasts provide essential acoustic and operational advantages for maritime activities, including Navy operations and sustainable management of marine infrastructure deployment. Furthermore, offshore wind farms physically alter the local MLD through wake turbulence, impacting regional stratification and the ecosystem it supports.

Methods:
This project will address this using the MaLCOM machine learning framework. The researcher will develop neural networks to transition MLD prediction from a purely statistical baseline to a temporally resolved forecast, utilizing high-resolution, in-situ and Earth observational datasets for rigorous testing and validation.

Researcher Ownership:
We emphasize project flexibility. The Doctoral Researcher will actively steer the research direction by selecting specific geographic test sites, defining target temporal scales, and choosing which oceanographic and meteorological variables to integrate into their ML architecture.

On-the-job Training:
The supervisory team will provide comprehensive skill development. Exeter will mentor the student in advanced Python programming, AI development, and MLD modelling. PML will train the student in handling complex in-situ MLD observations, while the Met Office will provide instruction on operational forecast requirements and output evaluation.

Collaborative Strength:
This partnership tightly integrates Exeter’s ML expertise, PML’s observational science, the Met Office’s forecasting capabilities, and autonomous vessel data collected by Oshen. This multi-institutional synergy ensures the researcher gains cross-disciplinary mentorship and access to unique datasets, significantly enhancing their PhD experience.

For eligible students the studentship will cover home tuition fees plus an annual tax-free stipend