Sources and further reading
This guide links scientific explanations to the NeuralPOM manuscript and implementation. The references below support the physical framework and related differentiable-model context. The full manuscript bibliography remains with the paper.
NeuralPOM
Shu, R., Liu, G., Gou, R., Gao, T., Wen, H., Li, H., and Huang, X. (2026). New Wine in an Old Bottle — NeuralPOM: A Hybrid Ocean Forecast Model with Differentiable Dynamics and Staged Learning. arXiv:2611.02460 [physics.ao-ph]. arXiv, PDF, supplement, and source data.
Ocean dynamics and mixing
Blumberg, A. F. & Mellor, G. L. (1987). A description of a three-dimensional coastal ocean circulation model. In Three-Dimensional Coastal Ocean Models, pp. 1–16. American Geophysical Union. doi:10.1029/CO004p0001. The POM physical and numerical foundation.
Mellor, G. L. & Yamada, T. (1982). Development of a turbulence closure model for geophysical fluid problems. Reviews of Geophysics, 20, 851–875. doi:10.1029/RG020i004p00851. The vertical turbulence closure.
Mellor, G. L. (1991). An equation of state for numerical models of oceans and estuaries. Journal of Atmospheric and Oceanic Technology, 8, 609–611. Publisher record. The POM equation of state used to connect tracers and density.
Differentiable physical models
Kochkov, D. et al. (2024). Neural general circulation models for weather and climate. Nature, 632, 1060–1066. doi:10.1038/s41586-024-07744-y. A related atmospheric hybrid model; see the NeuralPOM manuscript for the ocean-specific method and comparisons.
Implementation sources
| Topic | Primary source |
|---|---|
| Configuration and dependencies | Portable examples, pyproject.toml |
| Physical cores | Core source tree |
| Training and preparation | v1 source tree |
| Forecast evaluation | Evaluation source tree |
| Regression checks | Tests |