This is the final report of a three-year, Laboratory-Directed Research and Development (LDRD) project at the h s Alamos National Laboratory &ANI,). The ocean is a very complex nonlinear system that exhibits turbulence on essentially all scales, multiple equilibria, and significant intrinsic variability. Modeling the Ocean's dynamics at mesoscales is of fundamental importance for long-time-scale climate predictions. A major goal of this project has been to coordinate, strengthen, and focus the efforts of applied mathematicians, computer scientists, computational physicists and engineers (at LANL and a consortium of Universities) in a joint effort addressing the issues in mesoscale Ocean dynamics. The project combines expertise in the core competencies of high performance computing and theory of complex systems in a new way that has great potential for improving ocean models now running on the Connection Machines CM-200 and CM-5 and on the Cray T3D.