Models, datasets, observing systems, and software tools developed by OOMG and made available to the research community.
A principled Fourier neural operator–based digital twin for regional ocean forecasting. OceanNet enables skillful long-term prediction of Gulf Stream meander position, Loop Current eddy evolution, and other mesoscale ocean dynamics at a fraction of the computational cost of traditional numerical models.
An ensemble machine learning framework for forecasting significant wave heights and periods along the U.S. East Coast, trained on multi-decadal hindcast data and validated against buoy observations.
A coupled atmosphere–ocean–wave simulation framework used to study hurricane–ocean interactions, validated during Florence (2018) and applied to coastal flooding and compound event studies.
A coordinated network of autonomous underwater gliders providing sustained subsurface ocean observations along the U.S. Southeast continental shelf, operated in collaboration with SECOORA and NOAA (2016–2020).
Modeling and observing design support for the NSF Ocean Observatories Initiative Pioneer Array, including rapid assessment of relocation options in the Southern Mid-Atlantic Bight (2021–2022).
Processing pipelines and analysis tools for satellite-derived sea surface temperature, altimetry, and ocean color — used across OOMG's modeling and AI projects for data assimilation and validation.
High-resolution hindcast and forecast products for Gulf of Mexico ocean circulation, supporting fisheries stock assessments, offshore industry safety planning, and natural hazard response (2022–2027).
A multi-decadal hindcast of coastal circulation, temperature, salinity, and biogeochemical tracers for the U.S. Southeast shelf, generated using a data-assimilative ROMS configuration.
A large-scale nutrient distribution climatology for the South China Sea derived from a novel algorithm applied to historical in situ profiles, published alongside analysis code (Du et al. 2021).