We are pleased to announce an upcoming 4D-BGC webinar on Wednesday, July 8th, at 13:00 UTC / 15:00 CEST. Said Ouala from IMT Atlantique will present Neural-BGC: Neural Network emulation of BGC from T/S data for interpolation and coupled Physics-BGC simulations. More information can be found below:
Wednesday, 8 July, 2026 at 3 pm CEST / 1 pm UTC
Registration Link: https://washington.zoom.us/meeting/register/MbV5ougcSqWOJBd0NF-oHw
Meeting ID: 976 9119 4859 ● Passcode: 634738
Find your local number: https://us06web.zoom.us/u/kdy6pLJAeC
SCOR Working Group 168 Webinar #10
Title: Neural-BGC: Neural Network emulation of BGC from T/S data for interpolation and coupled Physics-BGC simulations
Presenter:
Said Ouala
IMT Atlantique
Ocean biogeochemistry (BGC) regulates key variables such as oxygen and nutrients that are essential for marine ecosystems and the global carbon cycle. However, BGC observations remain sparse, and BGC models rely on poorly constrained parameters and processes. To address these limitations, we developed Neural-BGC, an AI model trained on millions of ocean observations collected since 1965. Neural-BGC emulates BGC variables from temperature and salinity fields. It is used both for interpolation and coupled physics-BGC simulations. For interpolation, the model estimates BGC variables from available temperature and salinity observations, and these estimates are combined with direct oxygen measurements through optimal interpolation to produce gridded products. When coupled to a physical ocean model, Neural-BGC reproduces biogeochemical features, including the oxygen minimum zones of the Indian Ocean and nutrient-rich blooms in the Canary Current Upwelling System.
You can view and subscribe to the 4D-BGC webinar calendar here:
https://calendar.google.com/calendar/u/2?cid=c2Nvci53Zy4xNjhAZ21haWwuY29t
4D-BGC webinars are recorded, and previous presentations can be found at https://www.youtube.com/@4D-BGC