Antarctic ecosystems are complex, and data is limited since it is expensive to collect. Species interact in food webs which can be modelled as mathematical networks. The relationships between species are not always known, or we might know they interact but not how strongly. Noisy (or imperfect) data can be used to model these species interactions to give more certainty about how the ecosystem works as a whole – although the worse the data is, the less information it contributes. This data might not be perfect, but it can still be very useful for planning management of the ecosystems. In this project, we build upon a study from Dr Matthew Adams and Dr Kate Helmstedt to understand how this noisy data can be useful.
Reference: Adams et al. (2020) Informing management decisions for ecological networks, using dynamic models calibrated to noisy time‐series data. Ecology Letters 23: 607-619.
- Adapt existing code to an Antarctic case study
- explore mathematical approaches for a Value-of-Information analysis
- write a report detailing findings for a technical audience.
We aim to determine what kinds of noisy data might be useful for different kinds of ecosystem management.
Skills and experience
Programming skills (preferably in R or Matlab), mathematical or statistical undergraduate experience, good written communication skills, and an interest in using mathematics to address applied ecology problems to inform environmental management.
You may be eligible to apply for a research scholarship.
Contact the supervisor for more information.