Long-range distillation is introduced, whereby a long-timestep probabilistic ``student''model is trained to forecast directly at long-range using a huge synthetic dataset generated by an autoregressive ``teacher''model, representing the first demonstration that real-world skill can be improved by training on AI-generat...
Scott A. Martin, Noah D. Brenowitz, Dale R. Durran et al.· arXiv.org· 5 citations
Recent progress in applying deep learning to ocean dynamics is reviewed, including spatiotemporal interpolation of satellite and in situ data, estimation of unobserved ocean variables, neural data assimilation, ocean forecasting, and the development of eddy parameterizations for hybrid models.
G. Manucharyan, Scott A. Martin, D. Balwada et al.· Annual Review of Marine Scie...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.