Drift modelling of marine mammal carcases in coastal waters

  • Michael Bedington

Student thesis: Doctoral ThesisDoctor of Philosophy (awarded by OU/Aberdeen)


A floating object’s drift is governed by its buoyancy, shape, and the wind, waves and currents it experiences. Here, I develop a drift modelling framework for marine mammal carcases in coastal waters. The resulting models were run forwards and backwards in time to provide insights into strategies for environmental monitoring under two scenarios. The first explored the beach search options for carcases resulting from potentially fatal collisions between tidal-stream turbines and marine mammals. The second applied the reverse problem for known-location mass strandings to highlight potential at-sea mortality sites. The drift properties of carcase-like objects were assessed in at-sea experiments. Wave transport was found to be greater than Stokes drift alone and in a complex coastal area could not be represented by a downwind multiplier as many previous models have assumed. A high resolution unstructured grid wave model was set up to complement existing wind and
current models for the West Coast of Scotland, and these components were combined to build a carcase drift model. In the forward case, from tidal turbine locations, the drift model showed a wide spread of potential stranding sites, suggesting monitoring a limited number of beaches is unlikely to be fruitful.
However, selecting beaches in response to immediate wind direction would improve efficiency. Stranding locations alone can only provide evidence of turbine interactions if the number of animals affected is large. In the reverse case, when applied to a mass stranding in Chile, the drift model showed the ability to exclude areas of origin, even though it could not pinpoint an exact mortality site. This work advances understanding of wave transport of surface floating objects, of carcase drift modelling, and of the feasibility of strandings monitoring. The decomposition rate of carcases is a source of uncertainty in the model where further work should be undertaken.
Date of Award3 Sept 2015
Original languageEnglish
Awarding Institution
  • University of Edinburgh
SupervisorAndrew Dale (Supervisor), Ben Wilson (Supervisor) & Mark Inall (Supervisor)

Cite this