Abstract
Marine plastic pollution is a serious problem, affecting life both in
and out of the oceans. To plan and implement successful mitigation
efforts, it is essential to have a complete understanding of the envir
onmental impacts of this pollution. This includes knowledge about
the distribution of plastic in the marine environment, such as cir
culation patterns, accumulation hotspots, and transport pathways
from source to sink. Transport modelling can be a very valuable
tool in this context.
This thesis consists of three scientific publications that all address
ways that Lagrangian modelling of marine plastic pollution can be
improved.
The first paper discusses how using modelled hydrodynamic data
introduces discontinuities in the velocity field of the transport equa
tion, and how these discontinuities can have a negative effect on the
expected accuracy of numerical integration. We introduce a scheme
that can be applied to handle the discontinuities in the velocity field
during integration, and demonstrate that this can improve accuracy
by several orders of magnitude. This is of particular importance in
backtracking applications, which are commonly used in attempts to
identify sources of observed marine pollution.
The final two papers in this thesis address a component of the dy
namics of the transport of plastic litter in the marine environment
that is often overlooked in simulations. It is well-known by now that
beaches are not always permanent sinks for marine litter, but in
models, they are still often treated as such. Based on an extensive
review of literature from field studies on beached plastic litter, as
well as a set of simple laboratory experiments, we have developed a
new model for beaching and resuspension of floating objects. The
model is based on the assumption that waves are instrumental in
facilitating transport of objects between land and water, and has a
probability parameter that can be used to fit the model to observa
tions.
In the first of these two papers, our new wave-based model is presen
ted as a stand-alone one-dimensional model, and we present and
discuss its properties and implementation in detail. In the second
paper, we integrate the model with a large-scale Lagrangian trans
port model for the first time. We perform a case-study where we
compare simulations using this model to simulations using a simpler
model for resuspension that is based on expected residence time of
litter on the beach. We find that the wave-based model can capture
variations in beaching and resuspension dynamics on shorter times
cales, which can influence the overall transport. More studies are
needed to determine whether the new wave-based model is better,
but regardless, we believe that it is a good starting point for further
model development.