A Hyperbolic Neural Closure for M1 Radiation Transfer

2026-07-11

AI Summary

This work addresses the numerical instability in the M1 radiative transfer method, which stems from the lack of high-order moment closure models and the tendency of conventional or unconstrained machine learning closures to produce non-real characteristic speeds. To overcome this, the authors propose the first neural closure framework that explicitly embeds hyperbolicity constraints. Their approach employs two neural networks to parameterize the system Jacobian: a symmetric network and a strictly convex entropy network. By leveraging the entropy Hessian to construct a positive-definite symmetrizer, the method guarantees real eigenvalues and thus hyperbolicity. The closure relation is further refined through a numerical integration path. Implemented within a discontinuous Galerkin scheme, this approach significantly enhances both closure accuracy and solution fidelity while ensuring long-term numerical stability.

Abstract

In radiation transfer simulations, an M1 method achieves substantial computational savings by replacing the full angular transport equation with a low-order moment system. Because this reduced system is not closed, a closure model is required to represent the unknown higher-order moments using lower-order moments. While machine learning (ML)-based closures can improve accuracy beyond classical analytic closures, unconstrained learned closures may produce non-real characteristic speeds and consequently cause numerical solver breakdown. To guarantee real eigenvalues of the Jacobian associated with ML closures, we propose a hyperbolic neural closure for the M1 radiative transfer system. Rather than directly predicting closure terms, we parameterize the Jacobian through two neural networks: (i) a symmetric matrix network and (ii) a strictly convex entropy network whose Hessian defines a positive definite symmetrizer. These components are combined to yield a Jacobian that is similar to a symmetric matrix, thereby ensuring real eigenvalues. The closure is then reconstructed by numerical integration of the learned Jacobian field along a prescribed integration path. Numerical experiments show that the proposed closure not only achieves higher closure accuracy than classical analytic closures, but also improves solution accuracy and remains stable in discontinuous Galerkin simulations for radiative transfer problems.

Problem

• radiation transfer
• M1 closure
• hyperbolicity
• machine learning
• numerical stability

Innovation

• hyperbolic neural closure
• M1 radiative transfer
• machine learning closure
• real eigenvalues
• convex entropy