Research map / Probabilistic and scientific modeling / Differential equations and PDEs
SSTODE: Ocean-Atmosphere Physics-Informed Neural ODEs for Sea Surface Temperature Prediction
Zheng Jiang, Wei Wang, Gaowei Zhang, Yi Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add learnable global diffusion term to advection ODE (ablation) | SSTODE · OceanVP global (q=5) | MSE: 0.0596 -> 0.0527 | Improved |
| use learnable global scalar kappa instead of fixed kappa (ablation) | SSTODE · OceanVP global (q=5) | MSE: 0.0562 -> 0.0527 | Improved |
| use spatially varying 2D kappa map instead of global scalar (ablation) | SSTODE · OceanVP global (q=5) | MSE: 0.0527 -> 0.0578 | Hurt |
| add EEI source term with all four heat fluxes (ablation) | SSTODE · OceanVP global (q=5) | MSE: 0.0595 -> 0.0527 | Improved |
| add only shortwave radiation flux to the source term (ablation) | SSTODE · OceanVP global (q=5) | MSE: 0.0595 -> 0.0546 | Improved |
| add only longwave radiation flux to the source term (ablation) | SSTODE · OceanVP global (q=5) | MSE: 0.0595 -> 0.0560 | Improved |
| add only latent heat flux to the source term (ablation) | SSTODE · OceanVP global (q=5) | MSE: 0.0595 -> 0.0559 | Improved |
| add only sensible heat flux to the source term (ablation) | SSTODE · OceanVP global (q=5) | MSE: 0.0595 -> 0.0554 | Improved |
| use SSTODE (advection-diffusion plus EEI) instead of advection-only ClimODE | Neural ODE SST forecaster · OceanVP global (q=5) | MSE: 0.0622 -> 0.0527 | Improved |
| use SSTODE instead of ClimODE on ERA5 | Neural ODE SST forecaster · ERA5 global (q=12) | MSE: 0.0408 -> 0.0349 | Improved |
| use SSTODE instead of SimVPv2 in Equatorial Pacific at q=5 | regional SST forecasting · OceanVP Equatorial Pacific (q=5) | MSE (x1000): 0.5160 -> 0.5157 | No clear effect |
| use SSTODE instead of ClimODE | parameter count · 2.8125 degree resolution | parameters: 3.67M -> 4.62M parameters, about 10% better accuracy | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ClimODE — ClimODE: Climate and Weather Forecasting with Physics-informed Neural ODEs (ICLR 2024) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- OceanCBM: A Concept Bottleneck Model for Mechanistic Interpretability in Ocean Forecasting (NeurIPS 2026 · similarity 0.88)
- ClimODE: Climate and Weather Forecasting with Physics-informed Neural ODEs (ICLR 2024 · similarity 0.86)
- FuXi-Ocean: A Global Ocean Forecasting System with Sub-Daily Resolution (NeurIPS 2025 · similarity 0.85)
- ChaosBench: A Multi-Channel, Physics-Based Benchmark for Subseasonal-to-Seasonal Climate Prediction (NeurIPS 2024 · similarity 0.85)
- Geometric and Physical Constraints Synergistically Enhance Neural PDE Surrogates (ICML 2025 · similarity 0.84)
- OxyGenerator: Reconstructing Global Ocean Deoxygenation Over a Century with Deep Learning (ICML 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
