Ocean Modelling

Our group is a member of the Climate Modeling Alliance (CliMA) to advance climate modeling and prediction leveraging advances in machine learning and artificial intelligence to train the model with available observations. The work is done in partnership with scientists at Caltech, the Naval Postgraduate School, and the Jet Propulsion Laboratory. You can read about the overall project at CliMA website.

At MIT we develop the ocean component of the Climate Machine, the sane of the new climate model, leveraging the MIT expertise in the development of ocean models and in the representation of sub-gridscale mixing processes in the ocean. To take advantage of new computer architectures, languages, and machine learning techniques, we have partnered with Alan Edelman‘s group in CSAIL at MIT, and they are helping us write CLiMa in the Julia computing language. This has already enabled us to run seamlessly the model on GPUs and CPUs. The strategy of the CliMA group is, to the extent that is possible, to develop common hydro-dynamical cores, parameterizations, and machine-learning techniques for both atmosphere and ocean, so that development in one fluid can inform the other.

Previously working on this theme: Brandon Allen, Jean-Michel Campin, Valentin Churavy, Alan Edelman, Chris Hill, John Marshall, Ali Ramadhan, James Schloss, Andre Souza, Gregory Wagner.