|dc.description.abstract||The problem of plastic in our oceans has received much attention in science and policy in recent years. In this context, Lagrangian particle modelling has been used to better understand the causes behind the plastic accumulation in the subtropical gyres. However, most studies do not take into account the currents which are not resolved by Oceanic General Circulation Models (OGCM) beyond simple linear interpolation.
In this thesis we study to what level of accuracy these sub-grid scale flows can be approximated by three different diffusion models: simple Brownian motion, wind-dependent Brownian motion and the Kinematic Lagrangian Model of Lacorata. We simulate an ensemble of 70,000 particles with the Lagrangian particle tracking program Parcels forced by data from the GlobCurrent OGCM for three and thirteen years, and compare the obtained distributions to the drifter-based results of Maximenko et al.
The results showed that Brownian motion provides a better fit with observations than simulations with just the large-scale advection components. It was also shown that the optimal amount of diffusivity needed to be higher for the best fit with the Maximenko result than an experimentally determined diffusivity relationship would prescribe. The `Kinetic Lagrangian Model' was proven not to provide better results than the base-line test and was, therefore, not seen as conducive to providing a better plastic distribution.
This thesis was concluded by suggesting that further research needed to be done into the oceanic-coastal interface.||