EEPS Colloquium: David Sandwell

Mapping the Global Seafloor Using Satellites and Ships: Recent Improvements from SWOT Gravity Using Machine Learning

Variations in Earth’s gravity field create broad bumps and dips in the ocean surface that mirror the mass distribution of seafloor features such as ridges, seamounts, and trenches. This relationship enables scientists to infer seafloor depth from satellite altimetry, though fundamental physics limits the achievable resolution in the deep ocean to about 3–5 km. Over several decades, successive altimeter missions have progressively refined global seafloor maps, and the new SWOT mission now provides the most precise marine gravity measurements to date. In this study, five research groups applied machine learning to merge SWOT gravity data with ship soundings, producing consistent global bathymetry maps that are 23–46% more accurate than traditional approaches. The greatest gains are in resolving shallow seamounts and deep trenches, improving our ability to study ocean circulation, plate tectonics, and geologic hazards, while advancing the goals of the international Seabed 2030 initiative to map the entire ocean floor by 2030.

Host: Roger Michaelides

EEPS Colloquia are made possible by the William C. Ferguson Fund