Stable spectral neural operator for learning stiff PDE systems from limited data

TL;DR

This paper introduces SSNO, a spectral neural operator framework that effectively models stiff PDE systems from limited data, achieving high accuracy and data efficiency without requiring explicit PDE knowledge.

Contribution

The paper proposes a novel spectral neural operator architecture with inductive biases for physics learning, capable of modeling stiff PDEs from minimal data without explicit equation knowledge.

Findings

Achieves 10-100x lower prediction errors than existing models.

Requires only 2-5 training trajectories for robust generalization.

Demonstrates effectiveness across 2D and 3D benchmarks in various geometries.

Abstract

Accurate modeling of spatiotemporal dynamics is crucial to understanding complex phenomena across science and engineering. However, this task faces a fundamental challenge when the governing equations are unknown and observational data are sparse. System stiffness, the coupling of multiple time-scales, further exacerbates this problem and hinders long-term prediction. Existing methods fall short: purely data-driven methods demand massive datasets, whereas physics-aware approaches are constrained by their reliance on known equations and fine-grained time steps. To overcome these limitations, we introduce an equation-free learning framework, namely, the Stable Spectral Neural Operator (SSNO), for modeling stiff partial differential equation (PDE) systems based on limited data. Instead of encoding specific equation terms, SSNO embeds spectrally inspired structures in its architecture, yielding strong inductive biases for learning the underlying physics. It automatically learns local and global spatial interactions in the frequency domain, while handling system stiffness with a robust integrating factor time-stepping scheme. Demonstrated across multiple 2D and 3D benchmarks in Cartesian and spherical geometries, SSNO achieves prediction errors one to two orders of magnitude lower than leading models. Crucially, it shows remarkable data efficiency, requiring only very few (2--5) training trajectories for robust generalization to out-of-distribution conditions. This work offers a robust and generalizable approach to learning stiff spatiotemporal dynamics from limited data without explicit \textit{a priori} knowledge of PDE terms.