- Computational oceanography has come of age. We are approaching the time when numerical “experiments will…duplicate the circulation to any desired degree of accuracy” (E. Lorenz).
- Simulation output must be “democratized” so that barriers to dissemination are minimized. The data should be freely available, including to non-professional users. We must sample the simulations the way we sample the real ocean (and in other ways).
- Exascale computational science is on the horizon, and ocean dynamics is a vanguard use case. We envision scalable cyberinfrastructure resources with generic application in science and engineering.