skip to content

Cambridge NERC Doctoral Training Partnerships

Graduate Research Opportunities
Brief summary: 
Detect and track great baleen whales and study their migration patterns using new seismic and hydrophone data from the North Atlantic seafloor.
Importance of the area of research concerned: 
The great baleen whales are one of the most spectacular parts of the marine ecosystem. Their populations also have an increasingly appreciated role in the Earth’s carbon cycle and, thus, an influence on climate change. Populations of some species have been recovering, after decades of depletion previously, but other species, such as the North Atlantic right whale, remain on the brink of extinction. Recordings of the vocalisations—the songs—of the whales offer valuable information on their populations. They have been made previously by hydrophones, but these are often short-term or discontinuous. In 2018-2020, the project SEA-SEIS ( collected 19 months of continuous, hydrophone and 3-component seismic data on the North Atlantic seafloor. The frequency range of the data comprises the entire ranges of the Blue and Fin Whale vocalisations and the low-frequency parts of the Humpback and North Atlantic Right Whale ones. The data show clear whale song recordings and can be used to locate and track the animals. The SEA-SEIS network stretches across over half the width of the NE Atlantic and offers unique new evidence on the abundance and migration patterns of the whales.
Project summary : 
Broadband data from ocean-bottom seismometers and hydrophones can be used to locate and track baleen whales. The project supervisor has recently collected unique new data in the North Atlantic (, which will be used here. State of the art methods for the detection of whale vocalisations in continuous waveform data use artificial intelligence (AI) and machine learning (ML) and are available from the co-supervisors and collaborators. Building on this, new methods for the detection, location and tracking of the animals will also be developed and applied to the very large dataset of 19 months of continuous data. The analysis of the dataset will provide important new evidence on the whales’ behaviour, abundance, and migration patterns. The project will also present excellent opportunities for outreach and public engagement (e.g.,
What will the student do?: 
The dataset available for this project is a treasure trove of new data, but detecting and identifying whale vocalisations in the very large data volume is a challenge. Vocalisation examples for different species are available from previous studies and will be used as templates. AI and ML tools will be used to detect vocalisations in the continuous data and match signals to species. The student will first use ML to detect fin whale vocalizations and build a catalogue of them. The recently developed Perceiver neural network architecture, proven effective across a broad range of signals, will be tested and applied. Second, other whale species will be identified systematically. Large datasets will be assembled and used for the quantitative analysis of whale abundance and migration patterns. The work will be conducted in collaboration with marine mammal acoustic experts at the Scottish Association for Marine Sciences and, possibly, the Alan Turing Institute. The student will, first, obtain important new information on the populations of species known to inhabit the North Atlantic and then, perhaps, make some unexpected discoveries, given the unique nature of the data.
References - references should provide further reading about the project: 
Lebedev, S., R. Bonadio, M. Tsekhmistrenko, J.I. de Laat, C.J. Bean, 2021. Seafloor seismometers look for clues to North Atlantic volcanism, EOS, 102,
Vickers, W., Milner, B., Risch, D. and Lee, R., 2021. Robust North Atlantic right whale detection using deep learning models for denoising. The Journal of the Acoustical Society of America, 149(6), pp.3797-3812.
Dréo, R., Bouffaut, L., Leroy, E., Barruol, G. and Samaran, F., 2019. Baleen whale distribution and seasonal occurrence revealed by an ocean bottom seismometer network in the Western Indian Ocean. Deep Sea Research Part II: Topical Studies in Oceanography, 161, pp.132-144.
You can find out about applying for this project on the Department of Earth Sciences page.
Prof Sergei Lebedev