This movie shows the buoyancy (or temperature) field from a large eddy simulation (LES) of mixed layer instability. Here, the simulation is initialized in thermal wind balance with a domain size of 1km x 1km x 120m and the model resolution is approximatly 2 meters. Blue and red colors denote cold and warm water. As the instability develops, a strong front forms with intense vertical velocities and small-scale turbulence. The flow eventually organizes into a coherent submesoscale eddy with fine-scale stratified turbulence still evident.
This movie shows the same simulation as above, but now color indicates the concentration of phytoplankton, with blue and green indicating high concentrations. The phytoplankton model includes light limitation. A reduction in mixing induced by the increase in stratification triggers a phytoplankton bloom in otherwise light limited conditions.
As part of ongoing work, we are examining the impact of turbulence on the ability of motile bacteria to swim up chemical gradients, a process known as chemotaxis. In this direct numerical simulation, a patch of dissolved nutrients was injected into fully-developed homogeneous, isotropic turbulence. The red iso-surface indicates where the nutrient concentration is 10% of its initial maximum value. Bacteria start out uniformly distributed throughout the computational volume, but quickly begin to cluster around the nutrient filaments. The color volume rendering shows where the bacteria concentration is larger than the initial value.

Phytoplankton cells can grow if they are exposed to sufficient nutrients and light for photosynthesis. Traditionally deep mixed layers have been thought to be unable to support phytoplankton growth because the cells spend too much time at depth away from light. This simulation from Taylor and Ferrari (2011) shows that when turbulence is weak, phytoplankton growth is still possible in a deep mixed layer. Here, a deep mixed layer is seeded with a uniform phytoplankton concentration (green). Turbulence is driven by cooling the surface with a constant weak surface heat flux (1 W/m2). Although the mixed layer is deeper than the 'critical depth' phytoplankton still grow near the surface (yellow) and are organized by the convective cells.
This movie shows a simulation from a study in progress in collaboration with Shafer Smith and Raffaele Ferrari. Color shows the vertical vorticity, and the white line is a constant density surface. The flow is initialized with four jets in thermal wind balance with the density field and a stable stratification intended to mimic the ocean interior. An initial perturbation is added which is very visible in the vorticity field. Soon the flow becomes unstable to baroclinic instability and eddies form. This simulation solved the full Navier-Stokes code, while a second simulation solved the Quasi-Geostrophic equations under the same initial conditions. The large scale structures are very similar in both simulations, but the Navier-Stokes simulation develops a forward energy cascade at small scales and a much shallower energy spectrum.
Symmetric instability can develop in the ocean and atmosphere when the potential vorticity is less than zero. In this large-eddy simulation, a density front in the upper ocean was forced by cooling the surface of the ocean with a constant heat flux, which also acts to decrease the potential vorticity. As symmetric instability develops, the flow (vectors) becomes nearly aligned with the isopycnal surfaces (color). Eventually, the vertical shear associated with symmetric instability becomes unstable to a secondary Kelvin-Helmholtz instability which can be clearly seen when the green isopycnal sheet rolls up into multiple billows.
In this direct numerical simulation, a turbulent Ekman layer is generated as a uniform flow in geostrophic balance encounters a flat, no-slip bottom boundary. The external flow has a stable temperature stratification which inhibits the development of the turbulent boundary layer. Temperature is shown in color, instantaneous stream-ribbons are shown in white, and an iso-surface indicates low speed regions near the wall.
Symmetric instability occurs when the Potential Vorticity (PV) of a front takes the opposite sign of the Coriolis parameter. Here, a two-dimensional simulation from Taylor and Ferrari (2009) starts with a two-layer fluid with the upper layer unstable to symmetric instability. As the instability develops, velocity bands appear oriented along isopycnals. The vertical shear associated with the symmetric instability soon becomes unstable to a secondary Kelvin-Helmholtz instability which breaks down into turbulence. The secondary shear instability plays an important role in equilibrating the primary symmetric instability by modifying the potential vorticity.