Modeling turbulent and self-gravitating fluids with Fourier neural operators

Abstract

Neural Operators (NOs) are a leading method for surrogate modeling of partial differential equations. Unlike traditional neural networks, which approximate individual functions, NOs learn the mappings between function spaces. While NOs have been predominantly tested on simplified 1D and 2D problems, such as those explored in prior works, these studies fail to address the complexities of more realistic, high-dimensional, and high-dynamic range systems. Moreover, many real-world applications involve incomplete or noisy data, which has not been adequately explored in current NO literature. In this work, we present a novel application of NOs to astrophysical data, which involves high-dynamic range projections into an observational space. We train Fourier NO (FNO) models to predict the evolution of incomplete observational proxies with density variations spanning four orders of magnitude. We demonstrate that FNOs can predict the effects of unobserved dynamical variables. Our work lays the groundwork for future studies that forecast direct astronomical observables.

References in corpus (22)

• Theory of Star Formation
• The Athena++ Adaptive Mesh Refinement Framework: Design and Magnetohydrodynamic Solvers
• FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators
• PDEBENCH: An Extensive Benchmark for Scientific Machine Learning
• Spherical Fourier Neural Operators: Learning Stable Dynamics on the Sphere
• FORGE'd in FIRE II: The Formation of Magnetically-Dominated Quasar Accretion Disks from Cosmological Initial Conditions
• Plasma Surrogate Modelling using Fourier Neural Operators
• Towards Multi-spatiotemporal-scale Generalized PDE Modeling
• Application of Convolutional Neural Networks to Identify Stellar Feedback Bubbles in CO Emission
• Clifford Neural Layers for PDE Modeling
• Magnetohydrodynamics with Physics Informed Neural Operators
• Denoising Diffusion Probabilistic Models to Predict the Density of Molecular Clouds
• Multi-Grid Tensorized Fourier Neural Operator for High-Resolution PDEs
• Learning Dissipative Dynamics in Chaotic Systems
• Fourier Neural Operator for Plasma Modelling
• Neural Operators with Localized Integral and Differential Kernels
• Application of Convolutional Neural Networks to Predict Magnetic Fields Directions in Turbulent Clouds
• The Cost-Accuracy Trade-Off In Operator Learning With Neural Networks
• Guaranteed Approximation Bounds for Mixed-Precision Neural Operators
• Magnetized compressible turbulence with a fluctuation dynamo and Reynolds numbers over a million
• Coarse Graining with Neural Operators for Simulating Chaotic Systems
• Generative AI for fast and accurate statistical computation of fluids