MITgcm Movie Mania

Website Mailing List
September 30, 2010 by Helen Hill


MITgcm Goes to the Movies

story by Helen Hill


020209-1838-xugglered54What is it about MITgcm that it spawns so many great movies? In a break with science stories, this month MITgcm shares 3 recent movies showcasing projects using MITgcm.


You may remember our 2009 story about the efforts of Jean-Michel Campin and co-workers to reproduce a convective chimney using a multi-scale technique involving embedding high resolution 2d plumes in a substantially lower resolution 3d model to capture the essence of plume-scale convection for a fraction of the computational cost incurred for a full high resolution 3d model “Multi Scale Superparameterization in Ocean Modeling”. The movie below provides a beautiful illustration of this technique in action. Each panel corresponds to a cut-away section through the temperature field of a chimney evolving over a period of 3 days. Top left shows the temperature field from the coarse resolution run employing a simple adjustment algorithm to reproduce the gross effects of convection. On the right, the same view but for the full high-resolution, non-hydrostatic simulation. Bottom right, we see the corresponding evolution of the multi-scale simulation demonstrating the inter-play between the embedded 2-d plume models and their continually changing orientation in response to the large scale circulation obtained from the coarse scale simulation. To find out more about this work contact Jean-Michel.

Temperature Evolution in a Baroclinically Unstable Channel

The next movie, created by Ryan Abernathey a grad student working with John Marshall, depicts the 3D temperature field evolving over a period of one year in a zonally-periodic channel meant to represent the ACC, forced with wind stress and heating at the surface. The forcing leads to a baroclinically-unstable mean state, which produces the vigorous mesoscale eddies seen in the movie. To find out more about this work or to learn how to create similar movies of your own from MITgcm output contact Ryan.


Our third selection is a preliminary movie created by Thomas Zweigle working with Martin Losch at Germany’s AWI – Martin was previously featured in a story about sea-ice modeling back in May 2009 and is a regular mitgcm-devel contributor. Zwiegle has been using MITgcm to perform direct numerical simulations (DNS) of  double-diffusive systems; situations when two quantities (e.g. temperature and salt) have different molecular diffusivities. The simulation below represents a domain on the interface between two steps in a thermohaline staircase. At the start of the simulation salinity and temperature are linearly stratified with saltier water overlying less salty water, warmer water overlying colder. The domain is 150 x 10 x 50 grid points, with uniform horizontal and vertical resolution of 0.00165m ie the volume displayed corresponds to a box only around 8cm wide by 24cm deep. As the simulation proceeds temperature and salinity are observed to be transported through the interface by the salt-fingers. While fingers lose heat by molecular diffusion, salinity remains elevated because of lower molecular diffusivity with the result that denser, more saline fingers sink down, while neighboring fluid gains heat causing it to rise upwards. The time step in this DNS is about 0.001 sec. Random noise in the layer interface is used to initialize convection. To find out more about this work contact Thomas.

East Greenland Spill Jet

Our final movie, from Marcello Magaldi working with Tom Haines takes as its focus the character and variability of the East Greenland Spill Jet. Using the same configuration as that described in July’s  story Modeling Nordic Seas, this simulation corresponds to a case study of the 2003 circulation. Shown are isopynic surfaces south of the Denmark Strait overflow. Surfaces are stacked on top of each other with transparency decreasing towards greater values. The movie runs from July 1st, 2003 through September 1st, at a rate or 4 frames per day (1 every 6 hours). Greenland (to the left) gives way to the Kangerdlugssuaq Trough and beyond it the Denmark Strait. Iceland is to the right. The domain is 10 degrees East-West by 2 degrees North-South, with a maximum depth of just over 3000m. Read more about this work in Magaldi M.G., Haine T. W. N., Pickart R. S., 2010. On the nature and variability of the East Greenland Spill Jet: a case study in summer 2003. J. Phys. Oceanogr., Submitted.

Got MITgcm movies of your own you would like to share? Contact Helen Hill