OceanCBM: A Concept Bottleneck Model for Mechanistic Interpretability in Ocean Forecasting
This work addresses the need for interpretable and physically faithful ML models in oceanography, enabling mechanistic diagnosis of extreme events like marine heatwaves.
OceanCBM introduces the first concept bottleneck model for spatiotemporal ocean forecasting, predicting mixed layer heat content while routing through prescribed physical concepts and a free concept. It achieves interpretable, physically grounded representations without sacrificing predictive skill, outperforming baselines in consistency of mechanistic representations.
Extreme ocean phenomena are challenging not only to predict but to diagnose, as accurate forecasts alone do not reveal the underlying physical drivers. While recent machine learning approaches achieve strong predictive skill, they remain largely opaque and provide limited guarantees of fidelity to ground-truth physics. We introduce OceanCBM, the first concept bottleneck model (CBM) for spatiotemporal prediction and mechanistic interrogation of ocean dynamics. OceanCBM uses mixed supervision to predict mixed layer heat content, a key precursor of marine heatwaves, while routing information through an intermediate layer of prescribed concepts derived from geophysical fluid dynamics and a 'free' concept. This design imposes soft physical structure without over-constraining the model, and the free concept both regularizes concept predictions and captures residual physical processes. Across ensemble initializations, we show that mixed supervision yields consistent mechanistic representations, whereas prediction-only and prescription-only baselines learn highly variable latent structures despite similar predictive performance. OceanCBM achieves interpretable, physically grounded representations without sacrificing skill, explicitly characterizing the interpretability-performance trade-off.