Pith. sign in

REVIEW 3 major objections 4 minor 2 cited by

Differentiable Stellar Atmospheres with Physics-Informed Neural Networks

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read Kurucz-a1 claims that a physics-constrained neural network can replace the atmospheric-structure step in stellar spectroscopy, matching ATLAS-12 accuracy while enforcing hydrostatic equilibrium at least as tightly, and often more tightly…

desk verdict Solid differentiable ATLAS-12 emulator with genuine interpolation value; the superiority-over-ATLAS-12 claim is circular and should be revised. read the letter →

arxiv 2507.06357 v1 pith:Y7N4YYQ4 submitted 2025-07-08 astro-ph.SR astro-ph.EPastro-ph.GAastro-ph.IM

classification astro-ph.SRastro-ph.EPastro-ph.GAastro-ph.IM
keywords physics-informedneuralnetworksstellaratmosphereshydrostaticequilibriumdifferentiablespectroscopyATLAS-12spectralsynthesisradiativetransferemulator
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

Kurucz-a1 is a physics-informed neural network trained on 104,269 ATLAS-12 model atmospheres that returns the six atmospheric structure fields at 80 optical-depth points from four stellar parameters ($T_{\rm eff}$, $\log g$, [Fe/H], [$\alpha$/Fe]). Its training loss adds a hydrostatic-equilibrium term, $dP/d\tau = g/\kappa$, computed by automatic differentiation, and the paper reports median errors below 0.12% in temperature, 1.1% in pressure and density, and 1.5% in opacity. The aim is to provide a differentiable replacement for the legacy atmosphere-structure solvers that are the remaining bottleneck in end-to-end stellar spectroscopy. The paper further claims that the emulator satisfies hydrostatic equilibrium at least as well as ATLAS-12, on average better, and that synthetic solar spectra built from its structures match observed spectra more closely than spectra built from ATLAS-12 structures.

What carries the argument

The central object is the dual-encoder PINN: one multi-layer perceptron embeds global stellar parameters into 512-dimensional vectors, a second embeds the 80 Rosseland optical-depth coordinates, and the concatenated 1024-dimensional embeddings pass through a three-layer MLP that predicts six atmospheric parameters at every depth point. The load-bearing mechanism is the physics loss in Eq. (2), which penalizes $(dP/d\tau - g/\kappa)^2$ summed over depth points, with the pressure gradient obtained by automatic differentiation and the term weighted by $\alpha = 0.03$ against the data-loss MSE. This constraint is what makes the emulator physically consistent rather than a purely data-driven interpolator, and it is the reason the trained network is smoother in the gradient quantities that matter for radiative transfer.

What would settle it

Take the hottest quarter of the validation grid ($T_{\rm eff}\gtrsim 20{,}000$ K), compute the radiative acceleration from the network's own ACCRAD output, and evaluate the full momentum balance $dP/d\tau = g/\kappa + a_{\rm rad}$ on both Kurucz-a1 and ATLAS-12 structures; if ATLAS-12 is closer to the full balance, or if Kurucz-a1's predicted structures depart from ATLAS-12 by more than the reported median errors in this regime, the claim of superior physical consistency fails. A simpler version is to generate synthetic spectra for hot stars with both structures and compare line profiles, since the regime where radiation pressure shifts the structure is where the 'better than ATLAS-12' result should disappear.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a neural network with a physics loss can stand in for a classical stellar-atmosphere solver: given a point in the four-parameter space of effective temperature, surface gravity, metallicity, and alpha enhancement, Kurucz-a1 reconstructs column density, temperature, gas pressure, electron number density, Rosseland opacity, and radiative acceleration as functions of Rosseland optical depth, with accuracy comparable to, and in hydrostatic equilibrium tighter than, the ATLAS-12 code used to generate its training grid. Because the network is built in a differentiable framework, the whole mapping from stellar parameters to atmospheric structure is differentiable, and chaining it with a differentiable radiative-transfer code makes the complete spectrum a differentiable function of stellar and atomic parameters. The claim is that this closes the last gap in gradient-based, data-driven optimization of universal physical parameters across stellar populations.

Load-bearing premise

The load-bearing premise is that the simplified hydrostatic equation $dP/d\tau = g/\kappa$ is the right benchmark for physical consistency, so when ATLAS-12 deviates from it, that deviation counts as solver error rather than missing physics; if radiation pressure or turbulent pressure is important, notably at the hottest grid temperatures, Kurucz-a1's tighter agreement with this equation is a bias, not a gain.

Editorial extensions

If this is right

  • Atmospheric structures can be generated on demand for arbitrary stellar parameters in about 0.37 ms per model, so sparse-grid interpolation errors in abundance analysis can be eliminated without significant computational cost.
  • Linking Kurucz-a1 with differentiable radiative-transfer codes yields an end-to-end differentiable spectrum, enabling gradient-based fitting of atomic and model parameters against large spectroscopic surveys.
  • Because the emulator covers $T_{\rm eff}$ from 2500 to 50,000 K, $\log g$ from $-1$ to 5.5, [Fe/H] from $-4$ to +1.46, and [$\alpha$/Fe] from $-0.2$ to +0.62, a single network replaces the need to store and interpolate a large multi-dimensional grid.
  • Solar spectral synthesis with Kurucz-a1 structures matches observations at least as well as synthesis with ATLAS-12 structures, and the paper reports better agreement in some line wings due to tighter hydrostatic equilibrium.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: the most decisive validation would be hot-star spectra near the top of the grid ($T_{\rm eff}\sim 50{,}000$ K), because the training constraint omits radiation pressure while the network outputs ACCRAD; if full momentum balance is the benchmark, the claimed superiority over ATLAS-12 may reverse.
  • Editorial extension: since the emulator inherits ATLAS-12's opacity and equation-of-state choices, it makes those choices differentiable but does not correct them; the same framework could in principle be retrained on any other grid or on a full solver that includes radiation pressure.
  • Editorial extension: the natural next step, which the paper motivates but does not perform, is a survey-scale joint optimization in which universal line-formation parameters are inferred simultaneously across many stars while stellar parameters are marginalized.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper presents Kurucz-a1, a physics-informed neural network emulator for 1D LTE stellar atmospheres. It is trained on 104,269 ATLAS-12 models to predict six atmospheric parameters at 80 optical depth points from four stellar parameters (Teff, log g, [Fe/H], [α/Fe]), using a dual-encoder architecture and a composite loss that combines data fidelity with a hydrostatic-equilibrium penalty (Eqs. 1–3). The authors report held-out median relative errors below 0.12% for temperature, 1.1% for pressure and density, and 1.5% for opacity; show that the physics loss reduces the hydrostatic-equilibrium residual relative to an unconstrained MLP baseline; and compare Kurucz-a1 and ATLAS-12 synthetic solar spectra against observed solar data. The central claims are that Kurucz-a1 achieves hydrostatic equilibrium comparable to or better than ATLAS-12 and that it is more consistent with observed solar spectra, making it a viable differentiable replacement for atmospheric structure solvers.

Significance. If the central claims hold, Kurucz-a1 would fill a concrete bottleneck in end-to-end differentiable stellar spectroscopy: it provides a fast, differentiable surrogate for ATLAS-12 atmospheric structures, enabling joint optimization of stellar parameters and universal atomic physics. The paper's strengths include a large training grid, held-out validation on 10,427 models, a direct comparison against an unconstrained MLP baseline that supports the value of the physics loss, and open-source code. The interpolation-accuracy results are valuable and likely robust. However, the physical-consistency and solar-spectrum claims need additional support before the results can be accepted at face value; the hydrostatic-equilibrium comparison is tied to the same simplified equation used in training, and the solar comparison is a single spectrum without quantitative diagnostics.

major comments (3)
  1. [Section 3, Eq. (3)] The physics loss enforces dP/dτ = g/κ using gas pressure only, while the network also predicts radiative acceleration (ACCRAD) and the grid extends to Teff = 50,000 K, where radiation pressure is non-negligible. ATLAS-12 solves a more general momentum balance, so its residual against Eq. (3) may reflect omitted radiative acceleration rather than finite-difference error. The assertion in Section 4 that ATLAS-12's deviations are numerical ('~0.1% level', 'finite-difference discretization') is not supported by any independent test. I recommend either including ACCRAD in the physics loss or reporting the residual of the full momentum balance, dP_gas/dτ = g/κ − ACCRAD (plus turbulent pressure where relevant), for both ATLAS-12 and Kurucz-a1, split by effective temperature.
  2. [Section 4, Figures 3–4] The claim that 'Kurucz-a1 ... achieves even better hydrostatic equilibrium than ATLAS-12' is evaluated with the same loss function (Eq. 2) that is minimized during training. Because the network is explicitly optimized against this metric, smaller residuals on validation models are expected even if the equation is incomplete; the comparison therefore does not by itself establish physical consistency. Please provide an independent diagnostic, for example comparison of pressure scale heights or residuals of the full momentum equation, and report the data-loss and physics-loss terms separately at the hot end of the grid.
  3. [Section 4, Figure 5] The abstract's claim of better agreement with observed solar spectra rests on a single spectrum with no quantitative goodness-of-fit statistic or uncertainty estimate. The bottom panel's residuals are small, and the interpretation ('likely due to better hydrostatic equilibrium adherence') is speculative because synthetic spectra also depend on line lists, abundances, microturbulence, and atomic parameters, which are not varied in the comparison. Please report residual statistics (e.g., RMS or χ² over the plotted window), specify all PySME settings and solar parameters, and ideally validate on at least a small sample of stars.
minor comments (4)
  1. [Abstract] The phrase 'more consistent with the solar observed spectra' is grammatically awkward; please rephrase as 'more consistent with observed solar spectra'.
  2. [Section 3, Eq. (1)] Please state how each output quantity is normalized in the data MSE and whether the loss is computed in linear or logarithmic space; this is essential for interpreting the relative errors in Figure 2 and for reproducing the training setup.
  3. [Section 4, Figure 4] The discussion refers to '~0.1% level' residuals for ATLAS-12, but no numerical values are given for the loss distributions; please report the median and interquartile ranges of the hydrostatic equilibrium loss for ATLAS-12, Kurucz-a1, and the MLP baseline.
  4. [Appendix A] The Figure 2 caption says the validation set 'spans Galactic stellar populations', but the appendix describes a general grid; please clarify whether the validation set is stratified by parameter region and how the evolutionary tracks in Figure 6 are used to define the sampling.

Circularity Check

1 steps flagged · score 5.0 of 10

Headline claim of superior hydrostatic equilibrium is measured by the same simplified residual used as the training loss; held-out ATLAS-12 accuracy and solar-spectrum comparisons remain independent.

  1. self definitional [Section 3, Eq. (2); Section 4, Figures 3-4 and accompanying text]
    "The physics loss enforces hydrostatic equilibrium: Lphysics = 1 Nb ·N d X i,j ( dP dτ − g κ ) 2 i,j ... Notably, Kurucz-a1, on average, achieves even better hydrostatic equilibrium than ATLAS-12, likely due to ATLAS-12's discrete numerical scheme and optimization methods that predate modern gradient-based techniques. The small deviations from precise equilibrium in ATLAS-12 (~0.1% level) reflect limitations of finite-difference discretization."

    Kurucz-a1 is explicitly trained to minimize the residual (dP/dτ − g/κ)^2 via Lphysics (Eq. 2). The 'superior hydrostatic equilibrium' claim is then evaluated with exactly the same residual in Figures 3 and 4. The comparison is therefore by construction: the network is optimized against this specific simplified equation, while ATLAS-12 is not. The paper further assumes ATLAS-12's residual is purely finite-difference error, but Eq. (3) omits radiative acceleration (ACCRAD, which the network also outputs) and turbulent pressure, so part of ATLAS-12's residual may reflect physics excluded from the training objective. Thus the headline improvement over ATLAS-12 partly reports the fitted objective itself rather than an independent physical validation.

full rationale

The paper's central interpolation claims are independent: Kurucz-a1 is benchmarked against a held-out 10% of ATLAS-12 models (10,427 models) in Figure 2, with median errors below 0.12% for temperature, 1.1% for pressure/density, and 1.5% for opacity. That is genuine held-out-data evidence and does not reduce to the training objective. The solar-spectrum comparison with PySME against the Melchior database is an external, non-fitted validation and therefore also not circular. The circularity is concentrated in the abstract and Section 4 claim that Kurucz-a1 achieves 'superior hydrostatic equilibrium' and is 'more consistent' with the solar spectrum 'due to superior hydrostatic equilibrium adherence.' The hydrostatic-equilibrium metric is the same Lphysics (Eq. 2) minimized during training, so the 'superiority' over ATLAS-12 is at least partly a statement about optimizing the chosen simplified benchmark, not about agreement with the full physical momentum balance. Since the radiative acceleration output (ACCRAD) is omitted from Eq. (3), the simplified benchmark may be biased, especially for hot models near 50,000 K. This does not invalidate the emulator's practical accuracy, but it means one of the three headline claims reduces by construction. Self-citations (e.g., Ting et al. 2016) are motivational and not load-bearing for the main derivation, so they do not raise the score further. Overall score 5: partial circularity in one headline sub-claim, while the central differentiable-emulator content retains independent support.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central claim leans on a supervised fit to an external grid (ATLAS-12), so the key ledger items are the assumed form of the physics constraint, the adequacy of the training data, and the ML hyperparameters. No new physical entity is introduced.

free parameters (3)
  • physics loss weight alpha = 0.03
    Set by grid search (Section 3, Eq. 1) to balance data MSE and hydrostatic-equilibrium loss; tuned on the validation set, not derived from physics. The central 'superior to ATLAS-12' result depends on this weighting.
  • network architecture and depth encoding = hidden dims [1024, 512, 256], embedding dim 512, GeLU; depth encoder unspecified
    Chosen by design; no ablation or theoretical justification is given. Model capacity likely affects reported accuracy, but the paper does not show sensitivity to these choices.
  • training hyperparameters = not stated
    Batch size, learning rate, epochs, normalization, and seed are absent from the paper; these are free choices that affect the reported errors and are not auditable from the text.
assumptions (5)
  • domain assumption Hydrostatic equilibrium in the simplified form dP/dtau = g/kappa is the correct physical constraint for all models in the range 2500-50000 K.
    Invoked in Eq. (3) and enforced in Eq. (2). It omits radiation pressure (ACCRAD is output but not used) and turbulent pressure, which are non-negligible for hot stars and cool stars respectively.
  • domain assumption ATLAS-12 grid outputs are adequate ground truth for training.
    The entire dataset is ATLAS-12 models (Appendix A). If ATLAS-12's systematic errors are significant, the emulator inherits them; the paper's claim to outperform ATLAS-12 assumes these outputs are not the target of correctness.
  • domain assumption LTE, 1D plane-parallel atmospheric modeling is sufficient for the intended spectral analysis.
    Inherited from ATLAS-12 and stated in the abstract; non-LTE and 3D effects are outside scope but matter for many lines.
  • domain assumption The four parameters Teff, logg, [Fe/H], [alpha/Fe] span the relevant stellar population for the claimed universality.
    Used as the input space (Section 3); individual abundance variations are deferred to future work (Section 5), so the 'universal physical parameters' claim is limited.
  • standard math Neural networks with automatic differentiation can represent and differentiate the atmosphere structure mapping.
    Standard universal approximation and autodiff assumptions used throughout the method.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Differentiable Stellar Atmospheres with Physics-Informed Neural Networks." pith.science (2026). https://pith.science/paper/Y7N4YYQ4

@misc{pith2026250706357,
  author       = {Pith},
  title        = {Pith review of: Differentiable Stellar Atmospheres with Physics-Informed Neural Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y7N4YYQ4}},
  note         = {Machine review of arXiv:2507.06357}
}
read the original abstract

We present Kurucz-a1, a physics-informed neural network (PINN) that emulates 1D stellar atmosphere models under Local Thermodynamic Equilibrium (LTE), addressing a critical bottleneck in differentiable stellar spectroscopy. By incorporating hydrostatic equilibrium as a physical constraint during training, Kurucz-a1 creates a differentiable atmospheric structure solver that maintains physical consistency while achieving computational efficiency. Kurucz-a1 can achieve superior hydrostatic equilibrium and more consistent with the solar observed spectra compared to ATLAS-12 itself, demonstrating the advantages of modern optimization techniques. Combined with modern differentiable radiative transfer codes, this approach enables data-driven optimization of universal physical parameters across diverse stellar populations-a capability essential for next-generation stellar astrophysics.

Figures

Figures reproduced from arXiv: 2507.06357 by the authors.

Figure 1
Figure 1. Kurucz-a1 neural network architecture with dual￾encoder design. Stellar parameters (Teff, log g, [Fe/H], [α/Fe]) and optical depth points are encoded separately, then concatenated and processed through fully connected layers (FC) to predict atmo￾spheric parameters (ρx, T, P, κ, XNE, ACCRAD) at each depth point. separately. Modern frameworks such as Korg (Wheeler et al., 2024) have successfully modernized radiative t… view at source ↗
Figure 2
Figure 2. Relative error distributions for atmospheric parameters predicted by Kurucz-a1compared to ATLAS-12 models. The four panels show column mass density (RHOX), temperature (T), pressure (P), and Rosseland opacity (ABROSS) versus optical depth. The validation dataset spans Galactic stellar populations (see appendix). 10 1 10 2 10 3 10 4 10 5 10 6 10 7 A b s olute P r e s s u r e G r a die nt (|d P / d |) ATLAS-12 (Ground… view at source ↗
Figure 4
Figure 4. Distribution of hydrostatic equilibrium loss values across the validation dataset. Kurucz-a1(purple) achieves a tight distri￾bution comparable to ATLAS-12 (gray), while the MLP baseline (red) shows higher and more scattered values. Values near zero in￾dicate perfect adherence to the hydrostatic equilibrium constraint. physics enforcement: Ltotal = (1 − α) · Ldata + α · Lphysics (1) with α = 0.03 by grid search, wher… view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Solar spectrum synthesis validation. Top: Observed solar spectrum (green dashed) compared to synthetic spectra from Kurucz-a1 (blue) and ATLAS-12 (orange). Middle: Direct comparison between Kurucz-a1 and ATLAS-12 synthetic spectra, showing minimal differences. Bottom: …
Figure 6
Figure 6. Figure 6: Distribution of stellar atmospheric models in the training dataset. Left panel: Kiel diagram showing coverage in effective temperature (log Teff) versus surface gravity (log g) parameter space. The colored evolutionary tracks from PARSEC models (Bressan et al., 2012) r…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. The Payne Zero Project I: Stellar Spectra from Physical Models in Seconds

    astro-ph.SR 2026-07 conditional novelty 7.0 of 10

    GPU-native Kurucz synthesis and multicore atmospheres deliver second-scale physical spectra, enabling direct multi-element APOGEE fitting and joint 10^5-parameter line calibration without flux emulators.

  2. Foundation Models for Astrophysics

    astro-ph.IM 2026-08 conditional novelty 3.0 of 10

    Astronomical 'foundation models' largely reuse transformers and self-supervised pretraining, but evidence of transfer to new instruments, populations, or tasks remains rare; the paper argues such evidence, not archite...

Reference graph

Works this paper leans on

38 extracted references · 15 canonical work pages · cited by 2 Pith papers

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    The PHOENIX Model Atmosphere Grid for Stars

    Allard , F. The PHOENIX Model Atmosphere Grid for Stars . In Reyl \'e , C., Richard , J., Cambr \'e sy , L., Deleuil , M., P \'e contal , E., Tresse , L., and Vauglin , I. (eds.), SF2A-2016: Proceedings of the Annual meeting of the French Society of Astronomy and Astrophysics, pp.\ 223--227, December 2016

  3. [3]

    Uncertainties in stellar abundance analyses

    Asplund, M. Uncertainties in stellar abundance analyses. Highlights of Astronomy, 13: 0 542–547, 2005. doi:10.1017/S1539299600016555

  4. [4]

    Asplund , M., Nordlund , A ., Trampedach , R., Allende Prieto , C., and Stein , R. F. Line formation in solar granulation. I. Fe line shapes, shifts and asymmetries . , 359: 0 729--742, July 2000 a . doi:10.48550/arXiv.astro-ph/0005320

  5. [5]

    Asplund , M., Nordlund , A ., Trampedach , R., and Stein , R. F. Line formation in solar granulation. II. The photospheric Fe abundance . , 359: 0 743--754, July 2000 b . doi:10.48550/arXiv.astro-ph/0005321

  6. [6]

    Asplund , M., Grevesse , N., and Sauval , A. J. The Solar Chemical Composition . In Barnes , III, T. G. and Bash , F. N. (eds.), Cosmic Abundances as Records of Stellar Evolution and Nucleosynthesis, volume 336 of Astronomical Society of the Pacific Conference Series, pp.\ 25, September 2005

  7. [7]

    PARSEC: stellar tracks and isochrones with the PAdova and TRieste Stellar Evolution Code

    Bressan , A., Marigo , P., Girardi , L., Salasnich , B., Dal Cero , C., Rubele , S., and Nanni , A. PARSEC: stellar tracks and isochrones with the PAdova and TRieste Stellar Evolution Code . , 427 0 (1): 0 127--145, November 2012. doi:10.1111/j.1365-2966.2012.21948.x

  8. [8]

    and Kurucz , R

    Castelli , F. and Kurucz , R. L. New Grids of ATLAS9 Model Atmospheres . In Piskunov , N., Weiss , W. W., and Gray , D. F. (eds.), Modelling of Stellar Atmospheres, volume 210 of IAU Symposium, pp.\ A20, January 2003. doi:10.48550/arXiv.astro-ph/0405087

Show all 38 references
  1. [9]

    P., Magee , N

    Colgan , J., Kilcrease , D. P., Magee , N. H., Sherrill , M. E., Abdallah , Jr., J., Hakel , P., Fontes , C. J., Guzik , J. A., and Mussack , K. A. A New Generation of Los Alamos Opacity Tables . , 817 0 (2): 0 116, February 2016. doi:10.3847/0004-637X/817/2/116

  2. [10]

    and Yu , B

    E , W. and Yu , B. The Deep Ritz method: A deep learning-based numerical algorithm for solving variational problems . arXiv e-prints, art. arXiv:1710.00211, September 2017. doi:10.48550/arXiv.1710.00211

  3. [11]

    M., Magg , E., Plez , B., Bergemann , M., Heiter , U., Olander , T., and Hoppe , R

    Gerber , J. M., Magg , E., Plez , B., Bergemann , M., Heiter , U., Olander , T., and Hoppe , R. Non-LTE radiative transfer with Turbospectrum . , 669: 0 A43, January 2023. doi:10.1051/0004-6361/202243673

  4. [12]

    Gray , D. F. The Observation and Analysis of Stellar Photospheres . 2008

  5. [13]

    G., Nordlund , A ., and Plez , B

    Gustafsson , B., Edvardsson , B., Eriksson , K., J rgensen , U. G., Nordlund , A ., and Plez , B. A grid of MARCS model atmospheres for late-type stars. I. Methods and general properties . , 486 0 (3): 0 951--970, August 2008. doi:10.1051/0004-6361:200809724

  6. [14]

    V., Hansteen , V

    Hayek , W., Asplund , M., Carlsson , M., Trampedach , R., Collet , R., Gudiksen , B. V., Hansteen , V. H., and Leenaarts , J. Radiative transfer with scattering for domain-decomposed 3D MHD simulations of cool stellar atmospheres. Numerical methods and application to the quiet...

  7. [15]

    J., Soubiran , C., and Th \'e venin , F

    Heiter , U., Jofr \'e , P., Gustafsson , B., Korn , A. J., Soubiran , C., and Th \'e venin , F. Gaia FGK benchmark stars: Effective temperatures and surface gravities . , 582: 0 A49, October 2015. doi:10.1051/0004-6361/201526319

  8. [16]

    and Gimpel , K

    Hendrycks , D. and Gimpel , K. Gaussian Error Linear Units (GELUs) . arXiv e-prints, art. arXiv:1606.08415, June 2016. doi:10.48550/arXiv.1606.08415

  9. [17]

    and Mihalas , D

    Hubeny , I. and Mihalas , D. Theory of Stellar Atmospheres. An Introduction to Astrophysical Non-equilibrium Quantitative Spectroscopic Analysis . 2015

  10. [18]

    Kurucz , R. L. Atlas: a Computer Program for Calculating Model Stellar Atmospheres . SAO Special Report, 309, February 1970

  11. [19]

    Kurucz , R. L. Status of the ATLAS 12 Opacity Sampling Program and of New Programs for Rosseland and for Distribution Function Opacity . In Adelman , S. J., Kupka , F., and Weiss , W. W. (eds.), M.A.S.S., Model Atmospheres and Spectrum Synthesis, volume 108 of Astronomical Soc...

  12. [20]

    Kurucz , R. L. ATLAS12: Opacity sampling model atmosphere program . Astrophysics Source Code Library, record ascl:1303.024, March 2013

  13. [21]

    Kurucz , R. L. Model Atmosphere Codes: ATLAS12 and ATLAS9 . In Niemczura , E., Smalley , B., and Pych , W. (eds.), Determination of Atmospheric Parameters of B, pp.\ 39--51. 2014. doi:10.1007/978-3-319-06956-2_4

  14. [22]

    Kurucz , R. L. and Avrett , E. H. Solar Spectrum Synthesis. I. A Sample Atlas from 224 to 300 nm . SAO Special Report, 391, May 1981

  15. [23]

    The Belgian repository of fundamental atomic data and stellar spectra (BRASS)

    Laverick , M., Lobel , A., Merle , T., Royer , P., Martayan , C., David , M., Hensberge , H., and Thienpont , E. The Belgian repository of fundamental atomic data and stellar spectra (BRASS). I. Cross-matching atomic databases of astrophysical interest . , 612: 0 A60, April 20...

  16. [24]

    LeBlanc , F., Monin , D., Hui-Bon-Hoa , A., and Hauschildt , P. H. Stellar model atmospheres with abundance stratification . , 495 0 (3): 0 937--944, March 2009. doi:10.1051/0004-6361:200810848

  17. [25]

    F., and Nordlund , A

    Magic , Z., Collet , R., Asplund , M., Trampedach , R., Hayek , W., Chiavassa , A., Stein , R. F., and Nordlund , A . The Stagger-grid: A grid of 3D stellar atmosphere models. I. Methods and general properties . , 557: 0 A26, September 2013. doi:10.1051/0004-6361/201321274

  18. [26]

    E., Gustafsson , B., Majewski , S

    M \'e sz \'a ros , S., Allende Prieto , C., Edvardsson , B., Castelli , F., Garc \' a P \'e rez , A. E., Gustafsson , B., Majewski , S. R., Plez , B., Schiavon , R., Shetrone , M., and de Vicente , A. New ATLAS9 and MARCS Model Atmosphere Grids for the Apache Point Observatory...

  19. [27]

    Neilson , H. R. and Lester , J. B. Spherically symmetric model stellar atmospheres and limb darkening. II. Limb-darkening laws, gravity-darkening coefficients and angular diameter corrections for FGK dwarf stars . , 556: 0 A86, August 2013. doi:10.1051/0004-6361/201321888

  20. [28]

    Pasetto , S., Chiosi , C., Cropper , M., and Grebel , E. K. Theory of stellar convection: removing the mixing-length parameter . , 445 0 (4): 0 3592--3609, December 2014. doi:10.1093/mnras/stu1933

  21. [29]

    and Valenti , J

    Piskunov , N. and Valenti , J. A. Spectroscopy Made Easy: Evolution . , 597: 0 A16, January 2017. doi:10.1051/0004-6361/201629124

  22. [30]

    Raissi , M., Yazdani , A., and Karniadakis , G. E. Hidden Fluid Mechanics: A Navier-Stokes Informed Deep Learning Framework for Assimilating Flow Visualization Data . arXiv e-prints, art. arXiv:1808.04327, August 2018. doi:10.48550/arXiv.1808.04327

  23. [31]

    Raissi , M., Perdikaris , P., and Karniadakis , G. E. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations . Journal of Computational Physics, 378: 0 686--707, February 2019. doi...

  24. [32]

    Royer , P., e. a. MELCHIORS. The Mercator Library of High Resolution Stellar Spectroscopy . , 681: 0 A107, January 2024. doi:10.1051/0004-6361/202346847

  25. [33]

    MOOG: LTE line analysis and spectrum synthesis

    Sneden , C., Bean , J., Ivans , I., Lucatello , S., and Sobeck , J. MOOG: LTE line analysis and spectrum synthesis . Astrophysics Source Code Library, record ascl:1202.009, February 2012

  26. [34]

    Accelerated Fitting of Stellar Spectra

    Ting , Y.-S., Conroy , C., and Rix , H.-W. Accelerated Fitting of Stellar Spectra . , 826 0 (1): 0 83, July 2016. doi:10.3847/0004-637X/826/1/83

  27. [35]

    Wehrhahn , A., Piskunov , N., and Ryabchikova , T. PySME. Spectroscopy Made Easier . , 671: 0 A171, March 2023. doi:10.1051/0004-6361/202244482

  28. [36]

    J., Casey , A

    Wheeler , A. J., Casey , A. R., and Abruzzo , M. W. Korg: Fitting, Model Atmosphere Interpolation, and Brackett Lines . , 167 0 (2): 0 83, February 2024. doi:10.3847/1538-3881/ad19cc

  29. [37]

    York , D. G., e. a. The Sloan Digital Sky Survey: Technical Summary . , 120: 0 1579--1587, September 2000. doi:10.1086/301513. URL http://adsabs.harvard.edu/abs/2000AJ....120.1579Y

  30. [38]

    LAMOST spectral survey An overview

    Zhao , G., Zhao , Y.-H., Chu , Y.-Q., Jing , Y.-P., and Deng , L.-C. LAMOST spectral survey An overview . Research in Astronomy and Astrophysics, 12 0 (7): 0 723--734, July 2012. doi:10.1088/1674-4527/12/7/002

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.