Pushing stuck particles used in Lagrangian ocean analysis, from land back to the ocean using PARCELS

Abstract

One of the larger global problems is plastic in the ocean. Every year millions of tonnes of plastic waste enters the ocean and it it not clear what happens to the plastics and where it all ends up. A few models have been built to use Lagrangian ocean analysis to predict pathways of particles, like plastic, in the ocean. One example is PARCELS, which is a open source framework in Python. It uses numerical integration to calculate trajectories of particles in hydrodynamic models. But at some locations, with a large timestep or when using random motion, it is possible that the particles end up on land unintentionally, outside of the domain, and get stuck. In this thesis, we develop a method to prevent particles getting stuck on the land by pushing the particles back into the ocean. It is found that it is possible to get less or even no particles getting stuck. Using the hydrodynamic fields from GlobCurrent, a push with a velocity of 10^(−2) meters per second halves the number of particles that get on land and a velocity in the order of 1 meters per second reduces that number to zero.