IWMO 2025 abstract submission

Abstract submission form

Deadline: June 10, 2025

The abstract submission form will require the following information:

• Title
• Registration confirmation number: Received via email after registration. Example: W7NZ3HWJFNR
• Presentation type: "Oral" or "Poster" or "Oral and Poster".
• Author names, titles, and affiliations. Please be sure to indicate your title so we can determine your eligibility for the Outstanding Young Sciensist Award (OYSA).
• Funding acknowledgments (if relevant). An example abstract is shown below. Please limit the abstract text to roughly half a page.

Example abstract submission

Title: Toward direct numerical simulation (DNS) of the entire ocean

Registration confirmation number: W7NZ3HWJFNR

Presentation type: Oral

Authors:

Oliver B. Fringer, Professor, Dept. of Civil and Environmental Engineering, Stanford University, Stanford, CA, USA

Sekond A. Artheur, Research Scientist, Computational Oceanography Group, Institute of the Ocean, Acitysomewhere, Countryonearth

Therdar Thor, Postdoc, Dept. of Oceans, Anytown University, Notacity, USA

Abstract:

We conduct a direct numerical simulation (DNS) of the global ocean with a uniform horizontal and vertical grid resolution of 1 mm to resolve the finest Kolmogorov scales. This requires 1.3e27 grid points (1000 trillion trillion) in three dimensions. The numerical method employs a second-order in space discretization scheme that is stepped forward in time with an explicit, second-order accurate time advancement scheme. The free surface is computed with a marker-in-cell (MAC) scheme which simulates surface wave breaking. The grid extends into the global surf zone and resolves wave breaking at the big-wave surfing sites around the world including Nazare, Portugal, Mavericks, California, and Teahupo'o, French Polynesia. Explicit time advancement requires a time-step size of 1.5 microseconds to ensure stability of long waves propagating at a speed of 328 m/s over the Mariana Trench, where the ocean depth reaches 11000 m. Numerical simulation of the last 1000 years to study climate evolution requires 2.1e16 time steps. Using 11 million cores on the El Capital supercomputer at Los Alamos National Laboratory delivering 1742 petaFlops, the simulation is still running and will take 5000 petaYears (5 billion billion years) to complete. To properly compute turbulent statistics, data will be output at every time step, thus requiring 2.2e29 petabytes of storage that will be available on a solar-powered storage facility covering the surface of Mercury. Owing to a bandwidth of 1 Pb/s to transfer data to the server, the first time step will take 338 billion years to upload. Thus, we do not have results to present at this time. Nevertheless, we suspect the simulation will not agree with observations because of an inability to properly initialize the model with an accurate, mm-scale global bathymetry data set. Insufficient atmospheric model resolution to properly force the free surface also presents limitations to accurately represent ocean-atmosphere interactions. Research is ongoing.

Acknowledgments: This work is supported by a grant from the National Computational Fluid Dynamics Foundation Grant #22344