Lagrangian transport in realistic simulations of ocean submesoscale turbulence

Ocean flows at scales larger than few tens of km are quasi-horizontal due to the pronounced stratification of 
seawater and Earth’s rotation and are characterized by quasi-2D turbulence. Mesoscale (O(100) km) vortices 
contain most of the kinetic energy and are key for ocean dynamics at climatic scales. Submesoscale flows 
(scales below O(10) km), instead, display smaller and faster eddies, and filaments associated with strong 
gradients (e.g. of temperature) and intense vertical transport, which play an important role in both physical 
and biogeochemical budgets. Mesoscale and submesoscale flows also shape the physical and chemical 
environment that conditions the development of marine life. Direct global observations of submesoscale 
surface velocity fields at global scale are still not possible but should be achieved in the near future thanks to 
the satellite SWOT (NASA-CNES, launch in late 2022).

To compute large-scale horizontal transport, surface energy exchanges or global estimates of other 
quantities, it is crucial to assess how well the horizontal velocities provided by the satellite compare to actual 
surface currents and down to what length scale. For this purpose, Lagrangian approaches provide an ideal 
framework, as, unlike standard Eulerian approaches, they integrate in time the signal. As a consequence, they 
may allow a clear separation between fast (ageostrophic) processes, that could contaminate the satellite-
derived velocity, and slower (geostrophic) ones.

In this internship, funded by CNES, we will explore Lagrangian transport in surface ocean turbulence by 
means of state-of-the-art realistic numerical simulations. The analysis will rely on the comparison of different 
statistical indicators of Lagrangian dispersion in the full flow and in some of its subcomponents such as the 
geostrophic one, which should be measured by the satellite. One aim is to determine the effect of high-
frequency, ageostrophic motions on dispersion features. In particular, this study should allow the 
identification of a threshold length scale above which the approximate velocity field is accurate enough, at 
least in a statistical sense, as well as an estimate of the kinetic energy of the missing small scales.

We look for a candidate having good knowledge of fluid mechanics or dynamical systems and an interest for 
numerical methods; education: Fluid Mechanics, Physics, Geophysical Fluid Dynamics, Applied Mathematics. 
Knowing Python will be needed. Good knowledge of oral and written English is required. 

Interested candidates should send their CV, a letter of motivation, transcripts of notes, and possibly contact 
information of one reference.