Sensitivity of advective transfer times across the North Atlantic Ocean to the temporal and spatial resolution of model velocity data: Implication for European eel larval transport

European eel (Anguilla anguilla) larvae achieve one of the longest larval migrations of the marine realm, i.e., more than 6000 km from their spawning grounds in the Sargasso Sea to European continental shelves. The duration of this migration remains debated, between 7 months and 3 years. This information is, however, crucial since it determines the period over which larvae are affected by environmental conditions and hence the subsequent recruitment success.We investigate the pathways and duration of trans-Atlantic connections using 3 years of high-resolution (daily, 1/12°) velocity fields available from a Mercator-Océan model configuration without data assimilation. We study specifically the effect of spatial and temporal resolutions on our estimates by applying various filters in time (from daily to 12-day averages) and space (from 1/12° to 1° gridcell aggregation) to the nominal model outputs. Numerical particles are released in the presumed European eel spawning area and considered as passive tracers at three specific depths (around 0, 50, and 200 m).We diagnose particularly the intensity of the water transfer between suitable control sections that encompass the eel larva distribution. Transit ages are also investigated, with a particular focus on the pathways that minimize the connection times between the western and eastern North Atlantic. We show that small-scale structures (eddies and filaments) contribute to faster connections though they also correspond to additional complexity in trajectories. The shortest pathways mostly follow the Gulf Stream and the North Atlantic Drift, whereas interior connections require longer transfers that prove less compatible with biological observations.

► We assess trans-Atlantic transfers with trajectories computed in model velocity data.
► The intensity of the connection depends on the scales kept in the velocity dataset.
► The shortest connections follow the Gulf Stream and other pathways take months longer.
► Filtered velocity fields lead to increased travel times and less complex trajectories.
► Equivalent parameters inferred from surface drifters do not refute model results.