{"id":"2ee187b7-f164-4118-96a8-178977cd2492","arxiv_id":"2608.27744","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Existing neutron-star shape functions produce structured radius errors that, when propagated through the oblate Schwarzschild approximation, yield substantially larger local errors in the differential solid angle that underlies X-ray flux calculations.","lead":"This paper tests five approximate formulas for the oblate surface shape of rotating neutron stars, comparing them against exact numerical calculations. It finds that even small radius errors (below 1%) can produce much larger errors (up to tens of percent) in the computed solid angle and therefore in the predicted X-ray flux, especially near the stellar limb.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"The quantitative dOmega error maps (e.g., Figure 8's ∼10% near-limb error) rely entirely on the oblate Schwarzschild approximation; the paper does not verify whether the OS approximation faithfully reproduces the shape-function error propagation compared to a full numerical metric with frame…","rationale":"The reader's weakest_assumption correctly identifies that the OS approximation's accuracy is the main limitation of the paper. My stress-test concurs: the quantitative dOmega error results (e.g., ∼10% local errors in Figure 8) are computed under this approximation, and the paper does not validate that the OS approximation faithfully reproduces shape-function error propagation relative to a full numerical metric. This is a genuine load-bearing concern: if the OS approximation systematically changes the error pattern (e.g., by amplifying or cancelling projection-term contributions), the headline numbers could be misleading. However, the paper is transparent about this limitation and frames its conclusions within the OS context. The core contribution — a careful comparison of shape functions, identification of error morphologies, and derivation of the analytic decomposition of dOmega errors — remains valid and useful. The suggested concrete test would strengthen the paper, but its absence does not warrant a change from ACCEPT, given the explicit caveats and the paper's status as a systematic study under a well-established approximation. The verdict should stand as ACCEPT with the recommendation that the authors consider performing the proposed validation before final publication.","tokens_in":43808,"tokens_out":2326,"duration_ms":23884,"concrete_test":"For two representative models — the NICER-like case (M=1.99 M⊙, Req=12.6 km, ν=348 Hz, ζ=0.234, ϵ=0.036, Figure 8) and a rapid rotator (ν=600 Hz, ζ∼0.155, ϵ∼0.224, Figure 9) — compute the full differential solid angle dOmega using the numerical metric from RNS (which includes frame dragging and quadrupole) for both the RNS reference surface and the AM shape function. Compare the resulting EdOmega maps (percent error relative to RNS reference) with those obtained under the OS approximation. If the maximum absolute difference in EdOmega between the two treatments is less than 2 percentage points for all visible surface points, the OS-based conclusions are robust; if it exceeds this threshold, the quantitative error claims may need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative results on dOmega errors (Section IV, especially Figure 8) are computed under the oblate Schwarzschild (OS) approximation, which neglects frame dragging and the rotational quadrupole of the exterior metric. The paper acknowledges this limitation (Section V B) and defers study of these effects to future work. However, the headline claim that shape-function errors produce ∼10% local dOmega errors for a NICER-like pulsar depends on the assumption that the OS approximation correctly captures the way radius and derivative errors propagate into the differential solid angle. If the exact relativistic treatment systematically amplifies, cancels, or shifts the error patterns, the quantitative conclusions could change substantially. The paper does not provide a cross-check using a full numerical metric (e.g., from RNS or Lorene) to validate that the OS approximation is accurate for error propagation, even for a few representative models. Because the paper's practical recommendations and error budgets rest on these numbers, this missing validation is the most load-bearing concern.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"This paper compares five approximate shape functions for rotating neutron stars against numerical reference surfaces computed with the RNS code for seven equations of state. It characterizes errors in radius and its derivative, finds structured patterns that depend on compactness and spin, demonstrates that simple polynomial corrections can reduce radius errors to within ~0.5%, and propagates these errors into the differential solid angle dΩ under the oblate Schwarzschild (OS) approximation. The analysis shows that derivative errors dominate dΩ errors near the limb, leading to local errors of ~10% for a NICER-like pulsar even when radius errors are sub-percent. The paper provides guidelines for constructing improved shape functions, emphasizing accurate polar-radius modelling and smooth, small derivative errors.","tokens_in":44097,"tokens_out":2604,"duration_ms":26850,"significance":"If the results hold, this paper makes a practically important contribution for neutron star radius inference from X-ray flux modelling. It is the first systematic, EOS-spanning comparison of all existing shape functions, revealing not just error magnitudes but the distinct morphologies and the critical role of the derivative. The demonstration that simple corrections can reduce errors to ~0.5% supports the feasibility of improved empirical shape functions. The physics insight that derivative errors dominate near the limb, and that integrated solid angle can hide cancellation, is valuable for the community developing pulse-profile modelling codes. The work is thorough, the numerical methodology (RNS at high resolution, NS-SWORD) is sound, and the code is publicly available.","major_comments":[{"comment":"The quantitative dΩ error maps (e.g., the ~10% near-limb error for the NICER-like model) rely entirely on the oblate Schwarzschild (OS) approximation, which neglects frame dragging and the rotational quadrupole of the exterior metric. The paper acknowledges this limitation (Section V B) and defers a full-metric calculation to future work. However, because these numbers are the headline quantitative result and the basis for the statement that shape-function errors can produce substantial local flux errors, the authors should provide at least a brief estimate of how much the OS approximation itself might alter the error propagation. For a few representative models (e.g., the NICER-like case and the 600 Hz case), comparing dΩ computed with OS against dΩ computed with a full numerical metric (from RNS or Lorene) would bound the uncertainty. If such a comparison is impractical, a detailed, quantitative discussion of the expected systematic bias from the OS approximation should be added to Section IV or V, including an argument for why the general error pattern (area vs. projection, derivative dominance) is robust even if the exact percentages change.","section":"Section IV and Figure 8"},{"comment":"The analysis of the PVLS shape function correctly identifies that the polynomial fit for the polar radius limits the overall accuracy, raising the maximum error from 0.25% to 2.77%. This is a key finding. However, the paper does not show whether the remaining errors after applying the correction fit (Section III E) are dominated by the polar radius error or by latitudinal shape errors. The authors should quantify, for PVLS, how much of the residual error after correction (panels b–e in Figure 6) stems from the polar radius fit versus the ANN surface reconstruction. This would sharpen the recommendation that “modeling the polar radius … may be one of the key problems.”","section":"Section III B, Figure 3"}],"minor_comments":[{"comment":"The abstract states “errors in dΩ may be understood as coming from an area contribution … and a projection contribution.” It would be helpful to explicitly mention that the projection contribution grows near the limb, as this is a central result.","section":"Abstract"},{"comment":"The sentence “We will not (in this paper) provide a new, improved shape function” is clear, but the paper does provide correction surfaces (Section III E). The term “correction surface” could be misleading if a reader thinks it is a full shape function. Consider a brief clarification that the correction surface is a proof-of-principle demonstration, not a final product.","section":"Section I, page 3"},{"comment":"The notation uses the same symbol for the impact parameter b and the metric function B in equation (14). While the context is different, this could cause confusion. Consider a footnote or different symbol for the metric potential.","section":"Section II A, equations (7)–(8)"},{"comment":"The sentence “The same extremal radius error is shown for three of the shape functions in panel (a) of Figure 6” references panel (a), but Figure 6 has five panels (a)–(e). In the online version, panel (a) is identified, but ensure the caption and text are explicit for each panel.","section":"Section III D, end of page 12"},{"comment":"The description of Figure 7 for the i=0° panel states “the error in the differential solid angle is approximately twice the error in the radius at this latitude.” This is correct from equation (43). However, the reference to the “lower panels” is ambiguous; there are two rows (AM and PVLS). Specify which row.","section":"Section IV B, page 18"},{"comment":"In the paragraph starting “The median locations and angular widths …”, the green circle and lines in Figure 8 are described. The color scale for the error maps in Figure 8 uses a different range than previous figures. The caption should include the percentile-based color scaling, as done for Figures 7 and 9.","section":"Section IV B, page 20"},{"comment":"The table includes spin frequencies for several stars but no corresponding compactness or spin parameter. Since these parameters are central to the paper, consider adding columns with typical ζ and ϵ values (or referencing Table II) to help the reader connect observed stars to the parameter space studied.","section":"Table I"},{"comment":"The recommendation to model the polar radius accurately is well supported, but the text “improving the modeling of Rpole is a method to reduce the relative error to as low as 0.25%” implicitly assumes the polar radius is the only source of error. Qualify this statement to note that the ANN itself contributes ~0.25% error in the best case.","section":"Section V C, item 3"}],"recommendation":"minor_revision","confidential_remarks":"The paper is a solid, well-organized contribution. The main potential weakness is the absence of a cross-check of the OS approximation for dΩ error propagation against a full numerical metric. However, the paper is upfront about this limitation, and the core findings about shape-function errors (radius and derivative) are independent of OS. The recommendations for future shape functions are well justified. I recommend minor revision primarily to address the quantitative dependence on OS (major comment 1) and to clarify the source of PVLS residuals (major comment 2). The paper is appropriate for this journal and likely to be of immediate use to the neutron star astrophysics community."},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"The paper does something genuinely useful: it compares five shape functions against a common reference (RNS) over a broad parameter range, quantifies both radius and derivative errors, and propagates them into the differential solid angle dOmega under the OS approximation. The finding that derivative errors dominate near the limb and that polar-radius modeling is the main bottleneck for the PVLS shape function is actionable. The cross-validation test for universality is well done. Credit where it’s due: the methodology is sound, the public codes are used properly, and the limitations are stated honestly.\n\nThe soft spot is the one the stress-test flags. The headline numbers—Figure 8’s ~10% local dOmega error for a NICER-like pulsar—rest entirely on the oblate Schwarzschild approximation. The paper acknowledges that OS neglects frame dragging and the rotational quadrupole, and defers a check to future work. That’s a real hole: if the exact metric amplifies or shifts the error patterns, the quantitative conclusions could change. The authors don’t verify that OS faithfully propagates shape-function errors for even one representative model. That said, the paper’s main recommendations (model the derivative, improve the polar radius, stay within training range) are robust regardless of the metric choice, because they follow from the radius and derivative errors alone. The dOmega numbers are suggestive, not definitive.\n\nThe paper is worth engaging. It’s the first joint error analysis of these shape functions, and it gives practical guidance for building better ones. I’d accept it for peer review, with a request that the authors add a cross-check using a full numerical metric (even for a few cases) or at least strengthen the caveat that the dOmega error maps are OS-specific.","headline":"Solid error analysis of shape functions, with a clear caveat about the OS approximation for dOmega.","tokens_in":44606,"tokens_out":971,"would_cite":true,"duration_ms":10905,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek/deepseek-v4-flash","headline":"Five shape functions for rotating neutron stars produce systematic radius errors that propagate into large local errors in the observable solid angle, with corrections reducing radius errors to ~0.5%.","keywords":["rotating neutron stars","shape functions","oblate Schwarzschild approximation","differential solid angle","pulse profile modeling","NICER","equation of state","neutron star radii"],"falsifier":"Compute the full pulsed flux (including atmospheric beaming, Lorentz/Doppler factors, and redshift) for a NICER-like pulsar using both a shape function (e.g., AM) and the full RNS reference surface, then perform a Bayesian inference of the radius; if the bias in the recovered radius is smaller than the ~10% local dΩ error suggests, the flux-weighted error cancellation is stronger than the geometric analysis implies.","tokens_in":45397,"feed_emoji":"🌟","tokens_out":2869,"duration_ms":28122,"temperature":0.7,"pith_summary":"The paper examines five shape functions used to model the oblate surface of rotating neutron stars in the oblate Schwarzschild (OS) approximation. It finds that while radius errors are typically below 1% for moderate spins, errors in the derivative of the shape function cause substantially larger local errors in the differential solid angle (up to ~10% for a NICER-like pulsar) because the projection contribution dominates near the limb. Simple polynomial corrections reduce the radius errors of several shape functions to within approximately 0.5%. The paper provides guidelines for constructing more accurate shape functions, emphasizing that both the shape and its derivative must be modeled accurately and that the polar radius must be predicted precisely.","feed_headline":"Shape-function errors can bias neutron star radii estimates by 10%","feed_subtitle":"Five popular oblate-surface models produce large local solid-angle errors near the limb, but simple corrections shrink radius errors to…","key_machinery":"The central objects are the shape function R(θ) and its derivative R'(θ) for the oblate surface of a rotating neutron star. The error propagation into the differential solid angle dΩ (a key factor in the observed X-ray flux) is analyzed within the oblate Schwarzschild (OS) approximation. An analytic approximation derived from the Beloborodov formula separates the dΩ error into an area contribution (≈2E_R) and a projection contribution dominated by E_{R'}, explaining why small radius errors can produce large dΩ errors near the limb where cosσ → 0.","core_discovery":"The central claim is that the five existing shape functions—MLCB, BBPO, AM, SPYY, and PVLS—all introduce systematic errors in the surface radius and its derivative, and these errors propagate into the differential solid angle dΩ, a geometric factor in the observed flux. Although radius errors are often <1% at moderate spins, the derivative error generates a projection term that amplifies dΩ errors near the limb. For a NICER-like pulsar (PSR J0740+6620), local dΩ errors reach ~10% over about 20% of the rotational phase. Simple polynomial corrections in colatitude, compactness, and spin reduce the bulk radius errors to within ±0.5%, but residual EOS-dependent offsets hint at a possible limit to quasi-universality. The analysis also reveals that the PVLS shape function's accuracy is dominated by errors in its analytical fit for the polar radius rather than the neural network's surface interpolation.","pith_inferences":["The ~10% dΩ error near the limb may be partially offset by the beaming of the Hydrogen atmosphere, which suppresses limb emission; the net effect on inferred radius could be smaller than the geometric error alone.","The EOS-dependent offsets remaining after correction hint at a fundamental floor to quasi-universality; a truly universal shape function may be impossible, and separate calibrations for different EOS classes (e.g., hadronic vs. bare quark stars) may be necessary.","The error propagation framework could be extended beyond the OS approximation to full general-relativistic metrics (e.g., Kerr-like) to assess whether frame-dragging and quadrupole terms alter the shape-function error budget.","The PVLS polar radius fit error (0.25% to 2.77%) suggests that improving the analytic R_pole relation is a more impactful target than refining the neural network itself."],"forward_implications":["Current shape functions, particularly AM used in NICER analyses, may introduce systematic biases in inferred radii for faster-rotating pulsars; the paper recommends using SPYY or PVLS for rapidly rotating accreting millisecond X-ray pulsars.","Future shape functions must be calibrated over the compactness and spin range of the targeted observations and must model the polar radius accurately, as errors there dominate the PVLS surface error.","The error decomposition shows that shape functions should minimize errors in both R(θ) and R'(θ); a function with small R error but large derivative error will still produce large dΩ errors.","Gravitational lensing and atmospheric beaming may partly mitigate the limb errors, but the paper argues that integrated solid angle is a poor metric because cancellations mask local inaccuracies.","The universality tests suggest that a better shape function could be constructed by fitting a polynomial correction to an existing one like SPYY, using a larger and more representative EOS set."],"supporting_citations":[{"why":"Introduces the first shape function (MLCB) and the oblate Schwarzschild approximation; the paper tests this shape function.","marker":"[32]"},{"why":"Presents the BBPO shape function based on the Hartle-Thorne slow-rotation approximation; the paper evaluates its errors.","marker":"[39]"},{"why":"Presents the AM shape function used in all NICER analyses; the paper focuses heavily on its errors at moderate spins.","marker":"[40]"},{"why":"Presents the SPYY shape function with slow and fast coefficient sets; the paper tests and corrects it.","marker":"[41]"},{"why":"Presents the PVLS neural-network shape function; the paper identifies the polar-radius fit as its main limitation.","marker":"[42]"},{"why":"Provides the NICER code-verification framework for the OS approximation; the paper uses its notation and justifies the need for OS modeling.","marker":"[15]"},{"why":"Provides the Beloborodov approximation used to derive the analytic error decomposition for dΩ.","marker":"[111]"}],"fun_headline_variants":["Neutron star shape models skew flux near the limb by 10%","Shape function errors amplify neutron star flux errors at the limb","Polynomial fixes shrink neutron star shape radius errors to 0.5%","Five neutron star shape functions produce structured radius and flux errors","Oblate shape errors trace polar radius fit, not neural surface model"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The analysis assumes that the oblate Schwarzschild (OS) approximation is accurate enough to isolate the impact of shape-function errors on the flux, specifically that the error in dΩ computed within OS meaningfully represents the error in the full relativistic flux calculation; if the OS systematically amplifies or cancels errors differently from the exact rotating metric, the quantitative conclusions about the magnitude and pattern of dΩ errors would change.","fun_headline_variants_meta":{"raw":{"variants":["Neutron star shape models skew flux near the limb by 10%","Shape function errors amplify neutron star flux errors at the limb","Polynomial fixes shrink neutron star shape radius errors to 0.5%","Five neutron star shape functions produce structured radius and flux errors","Oblate shape errors trace polar radius fit, not neural surface model"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":0.000305,"raw_usage":{"cost":0.00030478,"is_byok":false,"cost_details":{"upstream_inference_cost":0.00030478,"upstream_inference_prompt_cost":0.0001715,"upstream_inference_completions_cost":0.00013328},"total_tokens":1701,"prompt_tokens":1225,"completion_tokens":476,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":0,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":398}},"tokens_in":1225,"tokens_out":476,"cost_source":"openrouter_usage","duration_ms":5540,"temperature":1.0,"reasoning_tokens":398,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-09-09T08:19:54.539126+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"falsifier":"Compute the full pulsed flux (including atmospheric beaming, Lorentz/Doppler factors, and redshift) for a NICER-like pulsar using both a shape function (e.g., AM) and the full RNS reference surface, then perform a Bayesian inference of the radius; if the bias in the recovered radius is smaller than the ~10% local dΩ error suggests, the flux-weighted error cancellation is stronger than the geometric analysis implies.","supporting_citations":[{"cited_title":"A NICER View of the Nearest and Brightest Millisecond Pulsar: PSR J0437$\\unicode{x2013}$4715","cited_arxiv_id":"2407.06789","evidence_quote":"Introduces the first shape function (MLCB) and the oblate Schwarzschild approximation; the paper tests this shape function."},{"cited_title":"The Oblate Schwarzschild Approximation for Light Curves of Rapidly Rotating Neutron Stars","cited_arxiv_id":"astro-ph/0703123","evidence_quote":"Presents the BBPO shape function based on the Hartle-Thorne slow-rotation approximation; the paper evaluates its errors."},{"cited_title":"Bounds on the Compactness of Neutron Stars from Brightness Oscillations","cited_arxiv_id":"astro-ph/9711325","evidence_quote":"Presents the AM shape function used in all NICER analyses; the paper focuses heavily on its errors at moderate spins."},{"cited_title":"On the Nature of the X-ray Emission from the Accreting Millisecond Pulsar SAX J1808.4-3658","cited_arxiv_id":"astro-ph/0303084","evidence_quote":"Presents the SPYY shape function with slow and fast coefficient sets; the paper tests and corrects it."},{"cited_title":"General purpose ray-tracing and polarized radiative transfer in General Relativity","cited_arxiv_id":"1804.04670","evidence_quote":"Presents the PVLS neural-network shape function; the paper identifies the polar-radius fit as its main limitation."},{"cited_title":"Constraining the Mass and Radius of Neutron Stars in Globular Clusters","cited_arxiv_id":"1709.05013","evidence_quote":"Provides the NICER code-verification framework for the OS approximation; the paper uses its notation and justifies the need for OS modeling."}],"review_version":1}