[Eoas-seminar] COAPS Short Seminar Series

[eoas-seminar at lists.fsu.edu]

COAPS Short Seminar Series
11:00 Nov. 3rd
Attend F2F (in 255 Research A) or Virtually (via Zoom)
https://fsu.zoom.us/j/92268262553

Meeting ID: 922 6826 2553
Talks are 12 minutes long with an additional 8 minutes for questions.

Neural Synthetic Profiles from Remote Sensing and Observations (NeSPReSO)
By Jose Miranda
Description: Understanding the ocean's subsurface dynamics is crucial for climate science and oceanography, yet it remains one of Earth's least observed domains due to logistical and financial constraints of direct measurements. Our broader research objective is to develop machine learning frameworks that improve data assimilation and initialization conditions for ocean forecast models, enhancing the forecast capabilities in critical regions like the Gulf of Mexico. As a first step toward this goal, we developed Neural Synthetic Profiles from Remote Sensing and Observations (NeSPReSO) to accurately predict subsurface fields from surface data, which can later serve as improved initial conditions for operational ocean models. The model, trained and evaluated in the Gulf of Mexico using Argo profiles and glider data, over-performed other traditional synthetic data generation methods, such as the Gravest Empirical Modes (GEM), Multiple Linear Regression (MLR) and Improved Synthetic Ocean Profile (ISOP). To make NeSPReSO’s output readily usable, we have wrapped the model in a simple web API: a request containing latitude, longitude, and time returns a NetCDF of the predicted temperature and salinity profiles. The service can be for research workflows or data-assimilation in forecast systems. Work now underway extends NeSPReSO by incorporating spatio–temporal context at the surface, a modular architecture with physical constraints, transfer-learning for other basins, and uncertainty quantification. Ultimately, we aim for broadly accessible, accurate ocean state estimates to feed forecast models, and inform data-driven marine-resource management.

Enthalpy Setting Up the HYCOM Global 1/50th degree Bathymetry
By Alan Wallcraft
Description: Ocean models typically get their bathymetry from global high resolution data sets that are based on satellite gravity. GEBCO 2025 is the best source for deep water bathymetry because it is the first to include the SWOT radar altimeter. However, all such data sets have issues in shallow water. I describe the steps taken to improve on GEBCO 2025 in shallow water for our HYCOM global 1/50th degree bathymetry.

NOTE: Please feel free to forward/share this invitation with other groups/disciplines that might be interested in this talk/topic. All are welcome to attend.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fsu.edu/pipermail/eoas-seminar/attachments/20251030/b9b42781/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/calendar
Size: 4642 bytes
Desc: not available
URL: <http://lists.fsu.edu/pipermail/eoas-seminar/attachments/20251030/b9b42781/attachment-0001.ics>