Abstract: The ocean is a turbulent fluid in contact with another turbulent system (the atmosphere) and influenced by other so-far unpredictable external forces such as solar variations, ice sheet melt, runoff, Etc. Over the past approximately 20 years, the oceanographic and allied communities have developed observing systems that are quasi-global in scope, albeit still limited in coverage, subject to difficult calibration problems, and sometimes outright contradiction. Making best-use of these observations requires combining them with understanding of the presumed governing equations for fluid and thermo-dynamics on a grand scale. Mathematically, the combination problem is best-formulated as one of control theory in a stochastic environment, but of dimension vastly exceeding conventional systems (e.g., a control vector of 2x10^9 elements). Examples of what has been accomplished in practice raise numerous mathematical and physical problems, many of them related to the issues of uncertainty quantification.