REVIEW 4 major objections 5 minor 18 references
Limb-Darkening Coefficients for the 4-Term and Power-2 Laws for the JWST Space Mission, Adopting PHOENIX Spherical Models at High Resolution
T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read This paper supplies limb-darkening coefficients for the 4-term and power-2 laws in 11 JWST passbands, computed from high-resolution spherical PHOENIX models, and recommends those two laws over the quadratic law.
desk verdict Useful JWST limb-darkening coefficients, but the recommended-law ordering rests solely on in-sample chi2, and the tables themselves are not in the preprint. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing objects are the two parametrizations: the 4-term law, $I(\mu)/I(1) = 1 - \sum_{k=1}^4 a_k(1-\mu^{k/2})$, and the power-2 law, $I(\mu)/I(1) = 1 - g(1-\mu^h)$. The coefficients are obtained by integrating monochromatic PHOENIX specific intensities over each passband response function, then applying the Levenberg–Marquardt least-squares fit with the critical cosine $\mu_{\rm crit}$ chosen as the point where the derivative of intensity with respect to radius $r = \sqrt{1-\mu^2}$ is maximal. This choice forces the fit to respect the sharp spherical drop-off at the limb, where plane-parallel models differ most.
What would settle it
Compare the coefficients' predicted center-to-limb intensity profiles against interferometric measurements of a bright star in one of the 11 passbands, or fit an independent spherical model grid with different opacities and check whether the coefficients shift by more than the internal fit scatter; if either test disagrees, the tables' accuracy claim fails.
Extended reading notes
Core claim
The central claim is that the limb-darkening coefficients computed from 306 spherically symmetric PHOENIX NewEra models, with about $1.08\times10^6$ wavelength points and solar composition at microturbulent velocity $1.0$ km/s, match the passband-integrated specific intensities of the models with the highest fidelity when expressed through the 4-term law $I(\mu)/I(1) = 1 - \sum_{k=1}^4 a_k(1-\mu^{k/2})$ and the power-2 law $I(\mu)/I(1) = 1 - g(1-\mu^h)$. Using the adopted critical-point prescription to locate $\mu_{\rm crit}$ at the maximum of the derivative near the limb, the authors obtain $\chi^2$ values for the 4-term law that are often smaller than those of the power-2 law, sometimes by a factor of 100, and both are dramatically better than the quadratic law. The paper therefore supplies Tables 2 and 3 with coefficients for the 11 passbands and recommends the 4-term or power-2 laws, in that order.
Load-bearing premise
The entire tabulation assumes the PHOENIX NewEra spherical models faithfully represent real stellar atmospheres across 0.1–6.0 $\mu$m; if their opacities, abundances, or spherical treatment are wrong, every fitted coefficient inherits that error.
Editorial extensions
If this is right
- JWST transit and eclipse analyses should adopt the 4-term or power-2 coefficients rather than the quadratic law, reducing systematic light-curve biases.
- The tables cover all three major near-IR instruments and both the G235 and G395 NIRSpec gratings, so users of NIRCam, NIRISS, and NIRSpec have a consistent set of coefficients.
- Because the models span $T_{\rm eff}$ 2400–7800 K and $\log g$ 3.0–5.5, the coefficients apply to the cool dwarfs and subgiants that dominate exoplanet host stars.
- The $\chi^2$ comparisons provide quantitative guidance on which law to use for a given passband and stellar parameter.
Reading between the lines
- If the quadratic law is as biased as the fitted $\chi^2$ ratios suggest, some published JWST transmission spectra that used quadratic limb darkening may need reanalysis, and retrieved abundances could shift by more than the reported precision.
- For grazing transits that sample the limb region directly, spherical-model coefficients of this type should be especially valuable, since that is exactly where the intensity drops sharply to zero and plane-parallel fits fail.
- The underlying high-resolution model grid spans 0.1–6.0 $\mu$m, so a natural extension would be to generate matching coefficients for MIRI passbands or for upcoming infrared missions such as Ariel, using the same pipeline.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. This manuscript computes limb-darkening coefficients for 11 JWST passbands (NIRCam, NIRISS, NIRSpec) by fitting the 4-term and power-2 laws to specific intensities from 306 spherical PHOENIX (NewEra) atmosphere models with Teff = 2400–7800 K, log g = 3.0–5.5, solar abundance, and microturbulent velocity 1.0 km/s. The fitting uses Levenberg–Marquardt minimization with the Wittkowski et al. μcrit prescription, passband intensities are integrated via Eq. (9), and the quality of each fit is assessed with Eq. (10). The paper compares the quadratic, power-2, and 4-term laws, recommends the 4-term and power-2 laws in that order, and makes the coefficient tables available only through CDS.
Significance. If the tables are accurate, this is a useful reference resource for JWST exoplanet and close-binary analyses, updating spherical-model limb-darkening coefficients to a new high-resolution PHOENIX grid and covering passbands in widespread current use. Strengths include a transparently described fitting procedure, high spectral resolution models, explicit passband integration, and a well-defined treatment of the spherical limb drop-off. However, the central deliverable is not inspectable in the submitted version, coefficient uncertainties are absent, and the recommendation rests on in-sample χ2 ratios; the abstract's claim that the coefficients match the model intensities 'with the highest fidelity' is therefore not yet substantiated. The paper's value is primarily as a tabulation rather than a methodological advance, and that tabulation still needs to be made fully verifiable.
major comments (4)
- [§3, Eq. (10)] The merit function is an unweighted sum of squared residuals, and the manuscript gives no uncertainties, covariances, or reduced χ2 for the fitted coefficients. As a result, the quoted χ2 values are unit- and N-dependent, and the 4-term versus power-2 comparison in Figs. 3–5 cannot be evaluated as a statistical statement. For a product intended for high-S/N JWST light-curve analyses, the absence of coefficient uncertainties is a substantive gap: users cannot propagate LDC errors into fitted system parameters. Please report reduced χ2 and parameter covariances, or justify their omission.
- [§4, Figs. 3–5] The recommendation of the 4-term and power-2 laws 'in that order of preference' is based entirely on χ2 ratios computed on the same 306 models used to derive the coefficients. This is an in-sample comparison; because the 4-term law contains four free parameters, a smaller residual sum is expected by construction regardless of predictive performance. No cross-validation, hold-out test, or synthetic light-curve retrieval is presented. Consequently, the abstract's claim that these coefficients match the intensity distribution 'with the highest fidelity' exceeds what the evidence supports, and the ordering of the recommendation is not robustly established. A simple flux-conservation or light-curve retrieval test would address this directly.
- [§3, μcrit and §4] The fits are restricted to μ ≥ μcrit, and the spherical PHOENIX intensities drop to zero below that point. The manuscript does not check whether the fitted laws preserve the disk-integrated flux, ∫ μ I(μ) dμ, over each passband. Since transit and eclipse light curves depend on this integral, a fit that is excellent in unweighted χ2 over the fitted domain can still bias the light curve if the small-μ behavior is misrepresented. The paper states that flux is not conserved for the quadratic law (Section 4) but provides no corresponding flux-conservation test for the recommended 4-term and power-2 coefficients; such a test should be added before the 'highest fidelity' wording is retained.
- [§3, Tables 2 and 3] The central product of the paper—the coefficient tables—is not included in the manuscript or in any submitted supplement; only CDS placeholders appear. It is therefore impossible for a reader or referee to verify the sign conventions, coefficient ranges, number of significant digits, or interpolation behavior claimed in Section 3. For a paper whose entire purpose is to deliver these tables, I request that the full tables be made available for review, or at least a representative subset with a complete description of the columns.
minor comments (5)
- [Fig. 2] The horizontal axis label reads 'co. γ' rather than 'cos γ', the caption begins 'Mode):' rather than 'Model:', and the inset label contains 'uni03BC crit', which appears to be a Unicode encoding artifact.
- [Eq. (7)] The denominator is written as I(1) while all other laws use I(μ = 1); this is cosmetic but should be made uniform across the equations.
- [§3] The manuscript states that the original 127 μ points are interpolated to 60,000 regularly spaced points but does not specify the interpolation method; a sentence describing the algorithm would allow reproducibility.
- [§3, Eq. (9)] The passband response functions are referenced only through web URLs; please cite versioned throughput files (e.g., JWST pipeline reference data) so users can reproduce the passband integrals.
- [Abstract and §3] The abstract says 'approximately 10^6 wavelengths' while Section 3 gives 1.08 × 10^6; the two statements should be made consistent.
Circularity Check
No significant circularity: the tabulated coefficients are the fitted product, and the 4-term/power-2 recommendation is an in-sample fit-quality comparison rather than a disguised prediction or a self-citation chain.
full rationale
This paper's product is a set of limb-darkening coefficients obtained by Levenberg–Marquardt least-squares fitting of two parametrized laws (Eqs. 6 and 7) to passband-integrated PHOENIX spherical-model intensities (Eq. 9). There is no external prediction step: the coefficients are the fitted parameters, and the abstract's 'highest fidelity' language refers to the quality of that fit to the model intensities, not to a quantity outside the fitted data. The recommendation of the 4-term over the power-2 over the quadratic law rests on chi-square ratios computed with Eq. 10 on the same 306 models used for fitting; this is in-sample model comparison. It is a validity limitation that no reduced chi-square, flux-conservation check, or light-curve retrieval test is given, but it is not circular, because the comparison does not redefine an input as an output. Self-citations are present (Hauschildt et al. 2025 for the PHOENIX NewEra grid; Claret 2000 for the 4-term law; Claret & Hauschildt 2003 for quasi-spherical models; Claret & Southworth 2022 for the power-2 law), but none is used as an external uniqueness theorem or to forbid alternatives, and the paper recomputes the fits and displays the chi-square ratios. The paper also states the limitation that none of the three laws achieves a perfect fit. No equation in the paper reduces to its own input by construction, and no fitted parameter is renamed as a prediction.
Assumptions & free parameters
free parameters (2)
- 4-term law coefficients a1-a4 =
Tables 2 (CDS)
- power-2 law coefficients g and h =
Tables 3 (CDS)
assumptions (4)
- domain assumption PHOENIX NewEra spherical models accurately represent stellar disc intensities for Teff 2400 to 7800 K, log g 3.0 to 5.5, solar composition, and xi=1.0 km/s.
- domain assumption JWST passband response functions used in Eq. 9 are correct.
- domain assumption The derivative-based mu_crit of Wittkowski et al. (2004) correctly defines the fitting cutoff at the stellar limb.
- standard math Unweighted least-squares fits using the merit function of Eq. 10 are a sufficient statistical treatment.
Cite this review
Pith. "Pith review of Limb-Darkening Coefficients for the 4-Term and Power-2 Laws for the JWST Space Mission, Adopting PHOENIX Spherical Models at High Resolution." pith.science (2026). https://pith.science/paper/KGGD45AT
@misc{pith2026250607265,
author = {Pith},
title = {Pith review of: Limb-Darkening Coefficients for the 4-Term and Power-2 Laws for the JWST Space Mission, Adopting PHOENIX Spherical Models at High Resolution},
year = {2026},
howpublished = {\url{https://pith.science/paper/KGGD45AT}},
note = {Machine review of arXiv:2506.07265}
}
abstract
Modeling observations of transiting exoplanets or close binary systems by comparing the observations with theoretical light curves requires precise knowledge of the distribution of specific intensities across the stellar disk. We aim to facilitate this type of research by providing extensive tabulations of limb-darkening coefficients for 11 frequently used near- and mid-infrared passbands on the NIRCam, NIRISS, and NIRSpec instruments installed on board the James Webb Space Telescope. The calculation of the limb-darkening coefficients was based on spherically symmetric atmosphere models from the PHOENIX series, with high spectral resolution (approximately $10^{6}$ wavelengths), and covering the wavelength range $0.1-6.0~\mu$m. The models were computed for solar composition, and a microturbulent velocity of 1.0 km s$^{-1}$. We adopted two of the more accurate parametrizations for the coefficients: the 4-term law, and the power-2 law. We applied the Levenberg-Marquardt least-squares minimization method, with a strategy to determine the critical value $\mu_{\rm crit}$ of the cosine of the viewing angle near the limb that is designed to improve numerical accuracy. The limb-darkening coefficients were derived based on a total of 306 atmosphere models covering an effective temperature range of $2400-7800$ K, and a $\log g$ interval between 3.0 and 5.5. We discuss the quality of the fits to the specific intensities provided by the power-2 and 4-term laws, as well as by the often used quadratic law. Based on a comparison, we recommend the use of the 4-term or power-2 laws, in that order of preference.
Figures
Figures from the paper (1 more)
Reference graph
Works this paper leans on
-
[1]
Alderson, L., Batalha, N. E., W akeford, H. R. et al., 2024, AJ , 167, 216
work page 2024
-
[2]
Chandrasekhar, S., 1944, ApJ, 99, 180
work page 1944
-
[3]
Claret, A., 2000, A&A, 363, 1081
work page 2000
- [4]
- [5]
-
[6]
Claret, A., Southworth 2022, A&A, 664, A128
work page 2022
-
[7]
2024, AJ, 168, 227 Díaz-Cordovés, J., Giménez, A., 1992, A&A, 259, 227
Coulombe, L.-P., Roy, P.-A., & Benneke, B. 2024, AJ, 168, 227 Díaz-Cordovés, J., Giménez, A., 1992, A&A, 259, 227
work page 2024
-
[8]
Feinstein, A. D., Radica, M., W elbanks, L. et al., 2023, Natu re, 614, 670
work page 2023
Show all 18 references
-
[9]
H., Barman; T., Baron, E
Hauschildt, P. H., Barman; T., Baron, E. et al., 2025, A&A, in press (arXiv:2504.17597) Hestroffer, D., 1997, A&A, 327, 199
2025 arXiv
-
[10]
A., Sobieski, S., 1970, AJ, 75, 175
Klinglesmith, D. A., Sobieski, S., 1970, AJ, 75, 175
1970
-
[11]
Kopal, Z., 1950, Harvard College Observatory Circular, 454 , 1
1950
-
[12]
et al., 2025, AJ, 169, 79
Liu, Q., Zhu, W., Zhou, Y. et al., 2025, AJ, 169, 79
2025
-
[13]
M., 2023, Nature Astronom y, 7, 1317
Lustig-Yaeger, J., Fu, G., May, E. M., 2023, Nature Astronom y, 7, 1317
2023
-
[14]
A., 1921, MNRAS, 81, 361
Milne, E. A., 1921, MNRAS, 81, 361
1921
-
[15]
1947, Phys
Placzek, G. 1947, Phys. Rev. 72, 556
1947
-
[16]
K., Mukherjee, S
Rustamkulov, Z., Sing, D. K., Mukherjee, S. et al., 2023, Nat ure, 614, 659
2023
-
[17]
43 Van’t Veer, F., 1960, Rech
Schwarzschild, K., 1906, Nachrichten von der Gesellschaft der Wis- senschaften zu Göttingen, Mathematisch-Physikalische Kl asse, p. 43 Van’t Veer, F., 1960, Rech. Astr. Obs. Utrecht 14, No. 3 W allack, N. L., Batalha, N., Alderson, L. et al., 2024, AJ, 16 8, 77
1906
-
[18]
P., Kervella, P., 2004, A&A, 413, 711 Acknowledgements
Wittkowski, M., Aufdenberg, J. P., Kervella, P., 2004, A&A, 413, 711 Acknowledgements. W e thank the anonymous referee for helpful comments. The Spanish MINC/AEI (PID2022-137241NB-C43 and PID2019-107061GB-C64) are gratefully acknowledged for th eir sup- port during the develop...
2004 doi
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.