pith. sign in

arxiv: 1907.01027 · v1 · pith:V6JHTLIZnew · submitted 2019-07-01 · 🧮 math.NA · cs.NA

Low-memory, discrete ordinates, discontinuous Galerkin methods for radiative transport

Zheng Sun , Cory D. Hauck This is my paper

Pith reviewed 2026-05-25 11:20 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords radiative transport equationdiscrete ordinatesdiscontinuous Galerkinlow-memory methodasymptotic diffusion limitmesh sweepingupwind reconstruction
0
0 comments X

The pith

Coupling spatial unknowns across angles produces a low-memory S_N-DG method that preserves the diffusion limit and sweeping structure for radiative transport.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a low-memory variant of the upwind discrete ordinates discontinuous Galerkin method for the radiative transport equation. By coupling spatial unknowns across collocation angles it builds a smaller finite element space with fewer degrees of freedom than the standard approach. This reduced space keeps the asymptotic diffusion limit and the directional structure required for mesh sweeping algorithms. The method is first-order accurate in general and second-order only in scattering-dominated diffusive regimes; upwind reconstruction is added to restore second-order accuracy everywhere. Upwind-sweep procedures are also given to shrink the linear system size inside the Krylov solver.

Core claim

The low-memory S_N-DG method is obtained by coupling spatial unknowns across collocation angles to form a reduced finite element space. This space retains the upwind structure and asymptotic diffusion limit of the original method. The resulting scheme is first-order accurate except in the diffusive regime where second-order convergence appears; upwind reconstruction restores second-order accuracy in all regimes. Sweep-based numerical procedures further reduce the dimension of the system passed to the Krylov solver.

What carries the argument

The reduced finite element space formed by coupling spatial unknowns across collocation angles, which keeps the upwind structure of the original S_N-DG discretization.

If this is right

  • Memory use drops because the total number of degrees of freedom is smaller than in standard S_N-DG.
  • The asymptotic diffusion limit remains intact, so the method behaves correctly in optically thick regimes.
  • Mesh sweeping algorithms continue to apply without change because the characteristic structure is unchanged.
  • Upwind reconstruction raises the accuracy to second order in all regimes.
  • The dimension of the linear system inside the Krylov solver is reduced by the sweep-based procedures.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same coupling idea might be tried on other angular discretizations that use discrete ordinates.
  • Lower memory could allow finer spatial meshes in three-dimensional problems where storage is the bottleneck.
  • Because the method stays first-order outside the diffusive regime, hybrid schemes that switch reconstruction based on local optical depth may be useful.
  • The reduced system size could improve the performance of iterative solvers on distributed-memory architectures.

Load-bearing premise

Coupling spatial unknowns across collocation angles produces a well-defined smaller finite element space that retains the upwind structure and diffusion-limit preservation without new instabilities or order reductions beyond those stated.

What would settle it

A test in which the low-memory method loses the diffusion limit or becomes unstable under mesh sweeping on a simple slab geometry with known analytic solution would falsify the preservation claims.

Figures

Figures reproduced from arXiv: 1907.01027 by Cory D. Hauck, Zheng Sun.

Figure 4.1
Figure 4.1. Figure 4.1: Numerical efficiency in Example 4.6. The first row is for isotropic test ψ = cos(x) and the second row is for anisotropic test ψ = cos(x + µ). Example 4.1. We first examine convergence rates of the methods using fabricated solutions. Let ε = 1, σs = 1, σa = 1 and D = [0, 1]. Assuming the exact solution ψ, we compute the source term q and the inflow boundary conditions ψl and ψr accordingly. With this app… view at source ↗
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p016_4.png] view at source ↗
Figure 4.2
Figure 4.2. Figure 4.2: Profiles of numerical scalar fluxes in Example 4.2. scheme, both LMDG and RLMDG schemes provide correct solution profiles. Since the problem is diffusive, the LMDG scheme gives accurate approximations that are almost indistinguishable with the P 1 -DG solutions. The reconstructed scheme has difficulty resolving the kink at x = 0.5, likely because the reconstruction is no longer accurate at this point. Th… view at source ↗
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p017_4.png] view at source ↗
Figure 4.3
Figure 4.3. Figure 4.3: Profiles of numerical scalar fluxes in Example 4.3, h = 0.1. 0 0.2 0.4 0.6 0.8 1 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 (a) σs,2 = 100. 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 1.2 1.4 (b) σs,2 = 1000. 0 0.2 0.4 0.6 0.8 1 0 1 2 3 4 5 6 7 8 9 10 (c) σs,2 = 10000 [PITH_FULL_IMAGE:figures/full_fig_p017_4_3.png] view at source ↗
Figure 4.4
Figure 4.4. Figure 4.4: Profiles of numerical scalar fluxes in Example 4.3, h = 0.02. Example 4.4. In this numerical test, we solve a test problem from [21] with dis￾continuous cross-sections. We take q = 0 with the left inflow ψl = 1 at xa = 0 and [PITH_FULL_IMAGE:figures/full_fig_p017_4_4.png] view at source ↗
Figure 4
Figure 4. Figure 4: b [PITH_FULL_IMAGE:figures/full_fig_p018_4.png] view at source ↗
Figure 4.5
Figure 4.5. Figure 4.5: Profiles of numerical scalar fluxes in Example 4.4. Example 4.5. This test is also from [21], with D = [0, 20] and ψl = ψr = 0. The cross-sections are σs ε = ( 90, 0 < x < 10 100, 10 < x < 20 and εσa = ( 10, 0 < x < 10 0, 10 < x < 20 . We solve the problem using the 16-point Gauss quadrature rule and the spatial mesh is uniform with h = 1. For this numerical test, the system has smaller changes among dif… view at source ↗
Figure 4.6
Figure 4.6. Figure 4.6: Profiles of numerical scalar fluxes in Example 4.5. ψ = sin(x + y) P 0 -DG P 1 -DG Q1 -DG LMDG RLMDG h/√ 2 error order error order error order error order error order 1/20 2.04e-2 - 1.45e-4 - 1.40e-4 - 1.24e-4 - 4.59e-4 - 1/40 1.10e-2 0.89 3.42e-5 2.08 3.53e-5 1.98 3.12e-5 1.99 1.18e-4 1.96 1/80 5.77e-3 0.94 8.28e-6 2.04 8.88e-6 1.99 7.82e-6 2.00 2.98e-5 1.98 1/160 2.96e-3 0.96 2.04e-6 2.02 2.26e-6 2.00 … view at source ↗
Figure 4
Figure 4. Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p019_4.png] view at source ↗
Figure 4.7
Figure 4.7. Figure 4.7: Profiles of numerical scalar fluxes in Example 4.7. Acknowledgment. ZS would like to thank Oak Ridge National Laboratory for hosting his NSF internship and to thank the staff, post-docs, interns and other visitors at ORNL for their warm hospitality. Appendix A. Assembly of the matrices. From the variational form (2.12), we can derive a matrix system LΨ = SΨ + Q. The matrices are defined as L = [L (l,p,r)… view at source ↗
read the original abstract

The discrete ordinates discontinuous Galerkin ($S_N$-DG) method is a well-established and practical approach for solving the radiative transport equation. In this paper, we study a low-memory variation of the upwind $S_N$-DG method. The proposed method uses a smaller finite element space that is constructed by coupling spatial unknowns across collocation angles, thereby yielding an approximation with fewer degrees of freedom than the standard method. Like the original $S_N$-DG method, the low memory variation still preserves the asymptotic diffusion limit and maintains the characteristic structure needed for mesh sweeping algorithms. While we observe second-order convergence in scattering dominated, diffusive regime, the low-memory method is in general only first-order accurate. To address this issue, we use upwind reconstruction to recover second-order accuracy. For both methods, numerical procedures based on upwind sweeps are proposed to reduce the system dimension in the underlying Krylov solver strategy.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a low-memory variant of the upwind S_N-DG method for the radiative transport equation. By constructing a smaller finite element space that couples spatial unknowns across collocation angles, the method reduces degrees of freedom while claiming to preserve the asymptotic diffusion limit and the characteristic structure required for mesh sweeping. The low-memory scheme is generally first-order accurate, but upwind reconstruction is introduced to recover second-order accuracy in scattering-dominated diffusive regimes; upwind sweep procedures are also proposed to reduce the dimension of the system solved by Krylov methods.

Significance. If the diffusion-limit preservation and order-recovery claims hold under the coupled-space construction, the approach would provide a memory-efficient alternative to standard S_N-DG discretizations for radiative transport, with direct relevance to large-scale simulations in optically thick media where sweeping algorithms are essential.

major comments (2)
  1. The central claim that the reduced finite-element space (with cross-angle coupling of spatial unknowns) preserves the asymptotic diffusion limit is load-bearing, yet the abstract provides no derivation or error analysis. Standard proofs for S_N-DG rely on independent per-ordinate discretization so that discrete moments close correctly and the upwind flux yields the proper boundary-layer behavior in the σ_s → ∞ limit; the coupling introduces linear dependencies between ordinates whose effect on the O(1) cross terms in the diffusion limit is not addressed.
  2. The statement that the low-memory method 'still preserves the asymptotic diffusion limit' (abstract) is asserted without reference to a specific theorem, lemma, or asymptotic expansion that would confirm the moment closure remains consistent after coupling. This omission makes it impossible to verify whether the construction satisfies the angle-independent consistency conditions required by the original method.
minor comments (2)
  1. The abstract notes second-order convergence 'in scattering dominated, diffusive regime' but does not indicate the specific test problems, mesh families, or reference solutions used to observe this; a table or figure reference should be added.
  2. The phrase 'numerical procedures based on upwind sweeps are proposed to reduce the system dimension in the underlying Krylov solver strategy' is stated without naming the particular Krylov method or quantifying the dimension reduction; this should be clarified with an equation or algorithm reference.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful review and the focus on the diffusion-limit claim, which is indeed central to the method's utility. Both major comments address the same issue: the absence of an explicit derivation or theorem reference supporting preservation of the asymptotic diffusion limit under the cross-angle coupling. We respond point-by-point below and will revise the manuscript to strengthen this part of the presentation.

read point-by-point responses

  1. Referee: The central claim that the reduced finite-element space (with cross-angle coupling of spatial unknowns) preserves the asymptotic diffusion limit is load-bearing, yet the abstract provides no derivation or error analysis. Standard proofs for S_N-DG rely on independent per-ordinate discretization so that discrete moments close correctly and the upwind flux yields the proper boundary-layer behavior in the σ_s → ∞ limit; the coupling introduces linear dependencies between ordinates whose effect on the O(1) cross terms in the diffusion limit is not addressed.

    Authors: We agree that the abstract itself contains no derivation, as is typical due to length limits. The manuscript constructs the reduced space so that the coupling is identical for all ordinates, thereby preserving the angle-independent moment closure and the upwind flux structure required for the diffusion limit. However, the referee correctly notes that the effect of the introduced linear dependencies on the O(1) cross terms is not analyzed in detail. We will add a short lemma in the revised version that extends the standard S_N-DG diffusion-limit argument to the coupled space and verifies consistency of the relevant terms. revision: yes

  2. Referee: The statement that the low-memory method 'still preserves the asymptotic diffusion limit' (abstract) is asserted without reference to a specific theorem, lemma, or asymptotic expansion that would confirm the moment closure remains consistent after coupling. This omission makes it impossible to verify whether the construction satisfies the angle-independent consistency conditions required by the original method.

    Authors: The claim follows from the fact that the reduced space is obtained by a uniform (angle-independent) coupling that does not alter the discrete ordinate moments or the upwind numerical flux. We acknowledge that no explicit theorem or expansion is referenced in the current text. In revision we will insert a pointer to the relevant consistency conditions (or a brief asymptotic argument) immediately after the method definition so that readers can verify the angle-independent closure directly. revision: yes

Circularity Check

0 steps flagged

No circularity: construction and preservation claims are independent of inputs

full rationale

The paper describes a direct modification of the standard upwind S_N-DG method by coupling spatial unknowns across collocation angles to reduce degrees of freedom. It asserts that the resulting space retains the upwind structure and asymptotic diffusion limit preservation of the original method. No equations, definitions, or self-citations in the abstract or described construction reduce any claimed property (e.g., diffusion-limit preservation or sweeping compatibility) to a fitted parameter, renamed input, or load-bearing self-reference. The derivation chain is self-contained as an explicit finite-element space alteration whose properties are analyzed separately from the original method's proofs. This is the expected non-finding for a methods paper whose central modification is presented without circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review performed on abstract only; no explicit free parameters, invented entities, or non-standard axioms are stated. The work rests on the standard mathematical framework of discontinuous Galerkin and discrete ordinates methods.

axioms (1)
  • standard math The standard upwind S_N-DG discretization preserves the asymptotic diffusion limit and admits mesh sweeping.
    The low-memory variant is defined as a modification of this established property.

pith-pipeline@v0.9.0 · 5685 in / 1236 out tokens · 20730 ms · 2026-05-25T11:20:38.902346+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

32 extracted references · 32 canonical work pages · 1 internal anchor

  1. [1]

    M. L. Adams , Discontinuous finite element transport solutions in thick diffusive problems , Nucl. Sci. Engrg., 137 (2001), pp. 298–333

    work page 2001

  2. [2]

    Agoshkov, Boundary Value Problems for Transport Equations, Springer Science & Business Media, 2012

    V. Agoshkov, Boundary Value Problems for Transport Equations, Springer Science & Business Media, 2012

    work page 2012

  3. [3]

    D. N. Arnold, F. Brezzi, B. Cockburn, and L. D. Marini, Unified analysis of discontinuous Galerkin methods for elliptic problems , SIAM J. Numer. Anal., 39 (2002), pp. 1749–1779

    work page 2002

  4. [4]

    Babuˇska and M

    I. Babuˇska and M. Suri , On locking and robustness in the finite element method , SIAM J. Numer. Anal., 29 (1992), pp. 1261–1293

    work page 1992

  5. [5]

    Bardos, R

    C. Bardos, R. Santos, and R. Sentis , Diffusion approximation and computation of the critical size, Trans. Amer. Math. Soc., 284 (1984), pp. 617–649

    work page 1984

  6. [6]

    Bensoussan, J

    A. Bensoussan, J. L. Lions, and G. C. Papanicolaou, Boundary layers and homogenization of transport processes, Publ. Res. Inst. Math. Sci., 15 (1979), pp. 53–157

    work page 1979

  7. [7]

    Cockburn and C.-W

    B. Cockburn and C.-W. Shu , Runge–Kutta discontinuous Galerkin methods for convection- dominated problems, J. Sci. Comput., 16 (2001), pp. 173–261

    work page 2001

  8. [8]

    Dautray and J

    R. Dautray and J. L. Lions, Mathematical Analysis and Numerical Methods for Science and Technology, Volume 6: Evolution Problems II , Spinger-Verlag, Berlin, 2000

    work page 2000

  9. [9]

    Davison, Neutron Transport Theory, Oxford University Press, London, 1973

    B. Davison, Neutron Transport Theory, Oxford University Press, London, 1973

    work page 1973

  10. [10]

    Egger and M

    H. Egger and M. Schlottbom , A mixed variational framework for the radiative transfer equation, Mathematical Models and Methods in Applied Sciences, 22 (2012), p. 1150014

    work page 2012

  11. [11]

    G. H. Golub and C. F. Van Loan , Matrix Computations , Johns Hopkins University Press, 3 ed., 2012

    work page 2012

  12. [12]

    Graziani, Computational Methods in Transport, vol

    F. Graziani, Computational Methods in Transport, vol. 48, Springer, 2006

    work page 2006

  13. [13]

    Guermond and G

    J.-L. Guermond and G. Kanschat , Asymptotic analysis of upwind discontinuous Galerkin approximation of the radiative transport equation in the diffusive limit , SIAM J. Numer. Anal., 48 (2010), pp. 53–78

    work page 2010

  14. [14]

    Guermond, G

    J.-L. Guermond, G. Kanschat, and J. C. Ragusa , Discontinuous Galerkin for the radia- tive transport equation, in Recent Developments in Discontinuous Galerkin Finite Element Methods for Partial Differential Equations, Springer, 2014, pp. 181–193

    work page 2014

  15. [15]

    Positive asymptotic preserving approximation of the radiation transport equation

    J.-L. Guermond, B. Popov, and J. Ragusa , Positive asymptotic preserving approximation of the radiation transport equation , arXiv preprint arXiv:1905.03390, (2019)

    work page internal anchor Pith review Pith/arXiv arXiv 1905

  16. [16]

    G. J. Habetler and B. J. Matkowsky , Uniform asymptotic expansions in transport theory with small mean free paths, and the diffusion approximation , J. Math. Phys., 16 (1975), pp. 846–854

    work page 1975

  17. [17]

    W. Han, J. Huang, and J. Eichholz , Discrete-ordinate discontinuous Galerkin methods for solving the radiative transfer equation , SIAM J. Sci. Comput., 32 (2010), pp. 477–497

    work page 2010

  18. [18]

    Heningburg and C

    V. Heningburg and C. Hauck, Hybrid solver for the radiative transport equation using finite volume and discontinuous Galerkin , submitted, (2019)

    work page 2019

  19. [19]

    Jin , Efficient asymptotic-preserving (AP) schemes for some multiscale kinetic equations , SIAM J

    S. Jin , Efficient asymptotic-preserving (AP) schemes for some multiscale kinetic equations , SIAM J. Sci. Comput., 21 (1999), pp. 441–454

    work page 1999

  20. [20]

    Jin and C

    S. Jin and C. D. Levermore , Numerical schemes for hyperbolic conservation laws with stiff relaxation terms, J. Comput. Phys., 126 (1996), pp. 449–467

    work page 1996

  21. [21]

    Larsen and J

    E. Larsen and J. Morel , Asymptotic solutions of numerical transport problems in optically thick, diffusive regimes II , J. Comput. Phys., 83 (1989), pp. 212–236

    work page 1989

  22. [22]

    E. W. Larsen, J. Morel, and W. F. Miller , Asymptotic solutions of numerical transport problems in optically thick, diffusive regimes , J. Comput. Phys., 69 (1987), pp. 283–324

    work page 1987

  23. [23]

    E. W. Larsen and J. E. Morel , Advances in discrete-ordinates methodology , in Nuclear Computational Science, Springer, 2010, pp. 1–84. 24 Z. SUN AND C. D. HAUCK

    work page 2010

  24. [24]

    Lesaint and P

    P. Lesaint and P. A. Raviart, On a finite element method for solving the neutron transport equation, in Mathematical Aspects of Finite Elements in Partial Differential Equations, Proceedings of a Symposium Conducted by the Mathematics Research Center, the Univer- sity of Wisconsin-Madison, Madison, WI, USA, 1974, pp. 1–3

    work page 1974

  25. [25]

    E. E. Lewis and J. W. F. Miller, Computational Methods in Neutron Transport, John Wiley and Sons, New York, 1984

    work page 1984

  26. [26]

    Mihalis and B

    D. Mihalis and B. Weibel-Mihalis , Foundations of Radiation Hydrodynamics, Dover, Mine- ola, New York, 1999

    work page 1999

  27. [27]

    G. C. Pomraning, Radiation Hydrodynamics, Pergamon Press, New York, 1973

    work page 1973

  28. [28]

    J. C. Ragusa, J.-L. Guermond, and G. Kanschat , A robust SN -DG-approximation for radiation transport in optically thick and diffusive regimes , J. Comput. Phys., 231 (2012), pp. 1947–1962

    work page 2012

  29. [29]

    W. H. Reed and T. Hill , Triangular mesh methods for the neutron transport equation , tech. report, Los Alamos Scientific Lab., N. Mex.(USA), 1973

    work page 1973

  30. [30]

    Wu and Y

    L. Wu and Y. Guo , Geometric correction for diffusive expansion of steady neutron transport equation, Commun. Math. Phys., 336 (2015), pp. 1473–1553

    work page 2015

  31. [31]

    Xu and C.-W

    Y. Xu and C.-W. Shu , Local discontinuous Galerkin methods for high-order time-dependent partial differential equations, Commun. Comput. Phys., 7 (2010), pp. 1–46

    work page 2010

  32. [32]

    Yan and C.-W

    J. Yan and C.-W. Shu, Local discontinuous Galerkin methods for partial differential equations with higher order derivatives , J. Sci. Comput., 17 (2002), pp. 27–47

    work page 2002