Using network theory to understand changes in the Atlantic Ocean transport patterns

The ocean currents transport large amounts of tracers such as heat, salt, nutrients and debris around the globe. Local currents can produce global transport features such as large-scale accumulation, coherent transport structures or transport barriers, which can be relevant for many applications in ocean and climate sciences. However, the relationship between local (Eulerian) currents and large-scale transport properties is not straightforward because of the high-dimensionality and complexity of the ocean circulation, making it difficult to identify global transport patterns without costly numerical experiments. The analysis of large-scale ocean transport phenomena therefore requires a low-dimensional representation that preserves important features of the real ocean, and that can be analyzed with a reasonable computational cost.

We use Lagrangian particle tracking in combination with ocean circulation model output to construct low dimensional representations of tracer transport in the ocean as networks that represent the flow of particles within and between different geographic regions. Using tools from network science such as clustering algorithms, we identify flow features such as coherent transport structures of the global ocean transport. We then analyze how transport patterns in the Atlantic Ocean change under a changing climate, using a forecast ocean model simulation for the next 100 years.