ExTraSS: a Domain Decomposed 3D NLTE Radiative Transfer spectral synthesis code for nebular phase transients

Anders Jerkstrand; Bart F.A. van Baal

arxiv: 2511.07539 · v2 · submitted 2025-11-10 · 🌌 astro-ph.HE · astro-ph.IM

ExTraSS: a Domain Decomposed 3D NLTE Radiative Transfer spectral synthesis code for nebular phase transients

Bart F.A. van Baal , Anders Jerkstrand This is my paper

Pith reviewed 2026-05-17 23:16 UTC · model grok-4.3

classification 🌌 astro-ph.HE astro-ph.IM

keywords radiative transferNLTEsupernovaenebular phasespectral synthesisdomain decomposition3D modelingphotoexcitation rates

0 comments

The pith

A domain decomposition algorithm manages millions of photoexcitation rates to enable full 3D NLTE radiative transfer across large nebular structures.

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

The paper presents an upgraded version of the ExTraSS code that performs 3D non-local thermodynamic equilibrium radiative transfer calculations for the nebular phase of explosive transients such as supernovae. In this phase the ejecta are optically thin, radioactive decay powers the emission, and the entire asymmetric structure contributes to the observed spectrum, requiring both NLTE level populations and radiative transfer. The central technical step is a new domain decomposition method that splits the generation and storage of photoexcitation rates over more than 100,000 cells into manageable sub-domains and recombines the results. This removes the previous memory and computational barrier that had prevented routine 3D NLTE modeling of realistic supernova geometries.

Core claim

ExTraSS now solves the coupled 3D NLTE radiative transfer problem by applying a domain decomposition algorithm that distributes the calculation of photoexcitation rates across spatial sub-domains, thereby reducing the storage requirement from millions of rates per cell while preserving the global solution.

What carries the argument

Domain Decomposition algorithm that partitions the computational domain, computes photoexcitation rates locally within each sub-domain, and recombines the results to obtain the full 3D NLTE solution.

If this is right

Synthetic spectra can now be generated for fully three-dimensional, asymmetric supernova ejecta in the nebular phase without the previous restriction to one-dimensional or LTE approximations.
The code can treat the entire optically thin nebula as a single emitting volume powered by distributed radioactive decay.
Convergence tests reported in the paper show that the decomposed solution reproduces the accuracy of the monolithic solver for the models examined.

Where Pith is reading between the lines

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

The same decomposition strategy could be applied to other large-scale NLTE problems such as planetary nebulae or H II regions where cell counts exceed 10^5.
If the recombination step scales well, the method opens the door to higher-resolution 3D grids or inclusion of additional atomic species without proportional memory growth.
Time-dependent extensions would allow tracking the spectral evolution as the nebula continues to expand and thin.

Load-bearing premise

That recombining the photoexcitation rates from separate domains yields the same level populations and emergent spectra as solving the entire coupled system at once.

What would settle it

Run both the full 3D NLTE solver and the domain-decomposed version on the same modest-sized model (where the full solver still fits in memory) and compare the resulting line profiles and luminosities for agreement within numerical tolerance.

Figures

Figures reproduced from arXiv: 2511.07539 by Anders Jerkstrand, Bart F.A. van Baal.

**Figure 1.** Figure 1: A pseudo-code schematic for the NLTE_solver and the EXCION_solver subroutines. The EXCION_solver is responsible for determining the level populations of all elements present inside a cell, while the NLTE_solver combines this with finding the temperature 𝑇cell through the T_solver. maximum d𝑇max=200 K). After 75 iterations of the NLTE_solver, this is reduced to 0.2, and after 110 iterations it drops to 0.02… view at source ↗

**Figure 2.** Figure 2: A pseudo-code schematic for the full program flow of ExTraSS, and how the new RayTraceGrid interacts with the previous NLTE_solver. If there is no previous solution present, the NLTE_solver has to run first, to be able to generate the rays which are transported by the RayTraceGrid module. If the ray hits a line (i.e. 𝜆ray = 𝜆line) 12, the Sobolev optical depth 𝜏line is used to attenuate 𝑁 phot by photoexci… view at source ↗

**Figure 4.** Figure 4 [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗

**Figure 5.** Figure 5: A pseudo-code schematic of how the non-blocking MPI communication is implemented. This shows the communication steps taken by the manager towards the workers on the same node (lines 2-5), as well as cross-node to other managers (lines 6-14). The same logic is used for the “downstream” side. The data received from another node is copied to a local buffer, such that the non-blocking IRECV request to the othe… view at source ↗

**Figure 6.** Figure 6: A look at the FPE and activity threads during one full iteration of RayTraceGrid (highlighted region). The green region (80 %) corresponds to main thread computations, while the blue corresponds to MPI (20 %). The MPI starts picking up when the first cores complete the ray tracing of all rays that they generate themselves, at which point they start checking for incoming data onto their domain through MPI c… view at source ↗

**Figure 7.** Figure 7: Comparison of the 3D photoionization rates from ExTraSS for the s9.0 model, in solid lines, against the rates in the 1D model from Jerkstrand et al. (2018), dashed lines, for oxygen (blue), iron (orange) and calcium (green). The decrease in the photoionization rate for Fe I around 800 km s−1 does lead to a corresponding rise in the neutral Fe fraction. 3.2.3 Photoexcitation In [PITH_FULL_IMAGE:figures/ful… view at source ↗

**Figure 8.** Figure 8: Comparison of the median 3D photoexcitation rates from ExTraSS (solid lines), against the rates from SUMO, using the 1D model from Jerkstrand et al. (2018) (dashed lines), for a series of elements. H𝛼 is shown in blue, Ca II 𝜆 8498 in orange, Fe I 𝜆 5060 in green and [O I] 𝜆 6300 in red. Also for this Fe I line, at the highest velocities the optical depth is too low to be considered, but now also in 3D thi… view at source ↗

**Figure 9.** Figure 9: Comparison of the median Fe I level populations from ExTraSS for the s9.0 model (exploded by Stockinger et al. 2020, see van Baal et al., submitted for a detailed nebular phase anaylsis), against the rates in the 1D model from Jerkstrand et al. (2018) in the top panel, and the relative fractions of the levels in the bottom panel. Lighter colours correspond to radii with higher velocities. For both ExTraSS … view at source ↗

**Figure 10.** Figure 10: The same as in [PITH_FULL_IMAGE:figures/full_fig_p012_10.png] view at source ↗

read the original abstract

In the nebular phase, supernovae are powered by radioactive decay and continuously fade, while their densities have decreased enough such that the expanding nebula becomes (largely) optically thin and the entire structure contributes to the emission. Models for the nebular phase need to take Non-Local Thermodynamic Equilibrium (NLTE) effects into account, while at the same time radiative transfer effects often cannot be ignored. To account for the asymmetric morphologies of SNe, 3D input ejecta models must be used. In this work, we present the $\texttt{ExTraSS}$ (EXplosive TRAnsient Spectral Simulator) code, which has been upgraded to be fully capable of 3D NLTE radiative transfer calculations in order to generate synthetic spectra for explosive transients in the nebular phase, with a focus on supernovae. We solve a long-standing difficulty of 3D NLTE radiative transfer -- to manage generation and storage of millions of photoexcitation rates over $\gtrsim10^{5}$ of cells -- by developing a new Domain Decomposition algorithm. We describe this new methodology and general code operations in detail, and analyse convergence and accuracy for $\texttt{ExTraSS}$.

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.

Desk Editor's Note private letter to a colleague

The paper adds a domain decomposition method to ExTraSS that tackles the photoexcitation rate storage problem for 3D NLTE on large grids, but the recombination step needs direct validation against full-domain results.

read the letter

The main takeaway is that van Baal and Jerkstrand have implemented a domain decomposition algorithm inside ExTraSS to make 3D NLTE radiative transfer practical for nebular-phase supernovae on grids with more than 10^5 cells. They target the specific bottleneck of generating and storing millions of photoexcitation rates across the volume, which has been a known limit for this kind of modeling. The paper describes how they split the calculations, recombine them, and then test convergence and accuracy. This is a concrete engineering response to a real scaling issue when moving from 1D or 2D to full 3D asymmetric ejecta models. For specialists who already work with NLTE codes, the detailed methodology section gives a usable blueprint for similar problems. The code focus on nebular transients also aligns with current observational needs, since many supernovae show clear asymmetry in late-time spectra. The approach is presented as newly developed for this purpose rather than a rehash of existing techniques. The soft spot is the recombination step and whether it fully restores the global radiation field coupling. In the optically thin nebular regime, photoexcitation rates depend on the mean intensity integrated over the entire structure. Any approximation at domain boundaries could shift level populations and emergent spectra away from what a monolithic solver would produce. The abstract states that convergence and accuracy were analyzed, yet the letter would be stronger with explicit side-by-side comparisons on a benchmark geometry that has an independent reference solution. Minor implementation details such as overlap handling or boundary conditions would also benefit from more quantitative error metrics. This paper is aimed at supernova modelers who need 3D NLTE spectra and at code developers working on radiative transfer scalability. A reader already running or extending similar codes would extract the most value from the algorithmic description and the reported tests. It is not a foundational theoretical result, but it is a practical advance in a narrow but active area. I would send it to peer review rather than desk reject because the core claim is well-defined and the application is timely enough to warrant expert scrutiny on the validation.

Referee Report

1 major / 2 minor

Summary. The manuscript presents the ExTraSS code for 3D NLTE radiative transfer spectral synthesis of nebular-phase transients, with emphasis on supernovae. The central contribution is a Domain Decomposition algorithm that addresses the generation and storage of millions of photoexcitation rates across ≳10^5 cells; the paper describes the algorithm, general code operations, and reports an analysis of convergence and accuracy.

Significance. If the domain decomposition is shown to recover the same global NLTE solution as a monolithic calculation, the work would provide a practical route to modeling asymmetric 3D ejecta structures under NLTE conditions in the optically thin nebular regime, where the entire volume contributes to the emergent spectrum.

major comments (1)

[Abstract and Domain Decomposition section] Abstract and the section describing the Domain Decomposition algorithm: the claim that the new algorithm 'delivers a correct 3D NLTE solution' rests on the recombination step restoring the full inter-cell radiative coupling. In the optically thin limit, photoexcitation rates depend on the volume-integrated radiation field; any truncation or approximation at domain boundaries would alter level populations. The manuscript states that convergence and accuracy are analysed, but does not indicate whether these tests include direct, quantitative comparisons (e.g., level populations or emergent spectra) between decomposed and full-domain runs on a benchmark geometry with a known reference solution.

minor comments (2)

The abstract would benefit from a concise statement of the specific benchmark problems and quantitative metrics (e.g., maximum fractional difference in level populations) used in the convergence analysis.
Notation for the photoexcitation rate matrix and the recombination operator should be introduced with explicit equations to make the domain-splitting procedure reproducible.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their careful reading of our manuscript and for providing constructive feedback. We address the major comment below and will incorporate revisions as indicated.

read point-by-point responses

Referee: [Abstract and Domain Decomposition section] Abstract and the section describing the Domain Decomposition algorithm: the claim that the new algorithm 'delivers a correct 3D NLTE solution' rests on the recombination step restoring the full inter-cell radiative coupling. In the optically thin limit, photoexcitation rates depend on the volume-integrated radiation field; any truncation or approximation at domain boundaries would alter level populations. The manuscript states that convergence and accuracy are analysed, but does not indicate whether these tests include direct, quantitative comparisons (e.g., level populations or emergent spectra) between decomposed and full-domain runs on a benchmark geometry with a known reference solution.

Authors: We appreciate the referee pointing out the need for explicit validation of the domain decomposition approach. While the recombination step is intended to restore the full inter-cell radiative coupling, we agree that direct comparisons are necessary to confirm this. In the revised version of the manuscript, we will add a subsection detailing quantitative comparisons of level populations and emergent spectra between the domain-decomposed calculations and equivalent full-domain runs on a benchmark geometry. These tests will use a simple asymmetric structure with a known reference solution to demonstrate that the algorithm delivers the correct 3D NLTE solution without boundary-induced alterations in the optically thin limit. This addresses the concern directly and strengthens the evidence for the method's accuracy. revision: yes

Circularity Check

0 steps flagged

New domain decomposition algorithm for 3D NLTE photoexcitation rates is an independent algorithmic contribution with no circular derivation.

full rationale

The paper introduces ExTraSS as an upgraded code for 3D NLTE radiative transfer in nebular-phase transients, focusing on a new Domain Decomposition algorithm to handle generation and storage of millions of photoexcitation rates across ≳10^5 cells. It describes the methodology in detail and analyzes convergence and accuracy. No load-bearing steps in the provided text reduce by construction to fitted inputs, self-citations, or renamings; the central claim is a practical algorithmic solution whose validity is assessed through direct testing rather than derived from prior results by the same authors. The derivation chain is self-contained as code development against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work rests on standard assumptions of radiative transfer and NLTE atomic physics in expanding media; no new free parameters, invented entities, or ad-hoc axioms are introduced beyond the algorithmic innovation itself.

axioms (1)

domain assumption Nebular-phase supernovae are powered by radioactive decay and become largely optically thin, requiring full NLTE treatment with radiative transfer.
Stated directly in the abstract as the physical regime the code targets.

pith-pipeline@v0.9.0 · 5526 in / 1200 out tokens · 24162 ms · 2026-05-17T23:16:10.966245+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

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

IndisputableMonolith/Foundation/AlexanderDuality.lean alexander_duality_circle_linking unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

The domain decomposition scheme implemented in ExTraSS... is inspired by Brunner & Brantley (2009). ... phi-based decomposition, i.e. each node takes an equal part of the number of azimuthal slices

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.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Mapping 3-D Explosive Nucleosynthesis with Type II Supernova Infrared Emission Lines
astro-ph.HE 2026-05 conditional novelty 6.0

Late-time IR spectroscopy of SN 2024ggi shows varied line morphologies implying chemical inhomogeneity and aspherical ionization, with modeling favoring 12-15 solar mass progenitors but only high-mass energetic 3D sim...
The Double-Peaked Calcium-Strong SN 2025coe: Progenitor Constraints from Early Interaction and Ejecta Asymmetries
astro-ph.HE 2026-01 unverdicted novelty 5.0

SN 2025coe's double-peaked light curve and nebular spectra are consistent with either an asymmetric core-collapse explosion of a low-mass He-core progenitor or a thermonuclear hybrid white dwarf merger.

Reference graph

Works this paper leans on

1 extracted references · 1 canonical work pages · cited by 2 Pith papers

[1]

J., Rodrigue G

Alme H. J., Rodrigue G. H., Zimmerman G. B., 2001, The Journal of Super- computing, 18, 5 AlpD.,LarssonJ.,FranssonC.,GablerM.,WongwathanaratA.,JankaH.-T., 2018, ApJ, 864, 175 Axelrod T. S., 1980, PhD thesis, University of California, Santa Cruz BarmentlooS.,JerkstrandA.,IwamotoK.,HachisuI.,NomotoK.,Sollerman J., Woosley S., 2024, MNRAS, 533, 1251 Baron E....

work page doi:10.1063/1.32213 2001

[1] [1]

J., Rodrigue G

Alme H. J., Rodrigue G. H., Zimmerman G. B., 2001, The Journal of Super- computing, 18, 5 AlpD.,LarssonJ.,FranssonC.,GablerM.,WongwathanaratA.,JankaH.-T., 2018, ApJ, 864, 175 Axelrod T. S., 1980, PhD thesis, University of California, Santa Cruz BarmentlooS.,JerkstrandA.,IwamotoK.,HachisuI.,NomotoK.,Sollerman J., Woosley S., 2024, MNRAS, 533, 1251 Baron E....

work page doi:10.1063/1.32213 2001