The Spatiotemporal Distribution of Dissolved Inorganic Carbon in the Global Ocean Interior: Reconstructed through Machine Learning
A full-depth reconstruction of the complete inventory of DIC, from the 1990s to 2019, using "ResNet-DIC", a deep neural network trained on GLODAPv2, shows good predictive power when validated 10 using independent data sets and model subsampling experiments.
Tobias Ehmen, N. Mackay, A. Watson
· 0 citations