Hierarchical Physics-Embedded Learning for Prediction and Discovery in Spatiotemporal Dynamical Systems

TL;DR

This paper introduces a hierarchical physics-embedded learning framework that enhances prediction and discovery of complex spatiotemporal dynamics by integrating prior physical knowledge and symbolic regression within a neural operator architecture.

Contribution

It presents a novel two-level hierarchical architecture that embeds physical laws and enables interpretable discovery of governing equations from sparse, noisy data.

Findings

Improves data efficiency and physical consistency in modeling complex systems

Enables symbolic regression for discovering unknown physical laws

Effectively captures non-local dependencies with Fourier Neural Operators

Abstract

Modeling complex spatiotemporal dynamics, particularly in far-from-equilibrium systems, remains a grand challenge in science. The governing partial differential equations (PDEs) for these systems are often intractable to derive from first principles, due to their inherent complexity, characterized by high-order derivatives and strong nonlinearities, coupled with incomplete physical knowledge. This has spurred the development of data-driven methods, yet these approaches face limitations: Purely data-driven models are often physically inconsistent and data-intensive, while existing physics-informed methods lack the structural capacity to represent complex operators or systematically integrate partial physical knowledge. Here, we propose a hierarchical physics-embedded learning framework that fundamentally advances both the forward spatiotemporal prediction and inverse discovery of physical laws from sparse and noisy data. The key innovation is a two-level architecture that mirrors the process of scientific discovery: the first level learns fundamental symbolic components of a PDE, while the second learns their governing combinations. This hierarchical decomposition not only reduces learning complexity but, more importantly, enables a structural integration of prior knowledge. Known physical laws are directly embedded into the models computational graph, guaranteeing physical consistency and improving data efficiency. By building the framework upon adaptive Fourier Neural Operators, we can effectively capture the non-local dependencies and high-order operators characteristic of dynamical systems. Additionally, by structurally decoupling known and unknown terms, the framework further enables interpretable discovery of underlying governing equations through symbolic regression, without presupposing functional forms.