The forgotten bright star: Theta Eridani as a millenary stellar transient observed by Hipparchus, Ptolemy and al-Sufi
Idel Waisberg, Boaz Katz
Abstract
Theta Eridani is a V=2.9 star that was nonetheless reported as one of the thirteen brightest stars in the night sky by both Ptolemy in his Almagest (137 AD) and by al-Sufi in his The Book of Fixed Stars (964 AD), in addition to being previously referred by Hipparchus (129 BC) as a particularly bright star. The discrepancy between its historical and modern visual magnitude is the highest among the stars in the Almagest. Theta Eridani is actually a triple star system, and here we combine interferometric data from VLTI/PIONIER and VLTI/GRAVITY, spectroscopic data from ESPaDOns and FEROS, and photometric data from TESS in order to solve for the orbital parameters, masses and radii of the close inner binary Theta Eridani Aa+Ab. We find that it is a tight eccentric binary (, ) of intermediate-mass stars (, ) that are extended to of their Roche lobe radii (), resulting in prominent ellipsoidal oscillations in the lightcurve. We also find that the primary is in a very special phase of its evolution in which it has just finished core hydrogen burning. The remarkable combination of orbital and stellar parameters hints that the historical brightening of Theta Eridani was due to a millenary transient phase powered by orbital energy extraction during a long-lived ``common envelope'' stage triggered by eccentric Roche lobe overflow in a previously more eccentric binary (). This strengthens the case that the apparent brightening was real and not due to an error by three different ancient observers, as has been commonly claimed in the past.
Review reports
xPeerd (AI review)
AI Review (xPeerd — DAReview)
Summary
This manuscript, titled “The Forgotten Bright Star: Theta Eridani as a Millenary Stellar Transient Observed by Hipparchus, Ptolemy and al-Sufi,” argues that the historical brightness discrepancy of Theta Eridani is real and may be explained by a long-lived transient powered by orbital-energy extraction in a close eccentric binary. The paper’s main empirical strength is that it combines interferometric, spectroscopic, and TESS photometric data to constrain the inner binary’s orbit and stellar properties, reporting a tight mildly eccentric system with masses of about 2.33 and 2.19 M⊙ and radii of about 4.3 and 3.95 R⊙ . The observational sections appear technically richer and more concrete than the historical-transient interpretation. The manuscript is interesting, ambitious, and potentially novel because it links historical astronomy to modern binary evolution. However, its central explanatory claim remains substantially more speculative than its observational results. In particular, the manuscript repeatedly advances a physical scenario for a thousand-year brightening episode without direct evidence for the transient itself, while also acknowledging that “a detailed model that explains the transient and its properties is beyond the scope of this paper” . As written, the paper contains a promising observational binary-star study, but the causal bridge from current system properties to an ancient millenary transient is not yet demonstrated with sufficient rigor for the strongest claims made in the abstract and conclusion .
Potential Major Revisions
- The central claim is stronger than the demonstrated evidence
The paper’s strongest language appears in the abstract, where the authors write that “The remarkable combination of orbital and stellar parameters hints that the historical brightening of Theta Eridani was due to a millenary transient phase powered by orbital energy extraction during a long-lived ‘common envelope’ stage triggered by eccentric Roche lobe overflow in a previously more eccentric binary (e ≃ 0.6)” . This is a major interpretive leap. The present-day binary characterization is well supported by multi-technique observations, but the proposed ancient transient is inferred indirectly from historical records plus present system architecture. The manuscript should more clearly separate:
- what is directly measured now,
- what is historically inferred, and
- what is physically hypothesized.
At present, the wording risks overstating causality. The conclusion similarly states that “we propose that the brightening event was triggered when the primary star filled its ‘Roche lobe’ in a previous more eccentric orbit” and that the extracted orbital energy “was reprocessed by a light but optically thick outflowing circumbinary envelope” . These should be framed more explicitly as one possible scenario, not the leading explanation unless stronger quantitative exclusion of alternatives is supplied.
- The manuscript admits model incompleteness but still builds strong conclusions on it
A particularly important internal tension appears in Section 5. The authors explicitly state: “A detailed model that explains the transient and its properties is beyond the scope of this paper. Below we outline a possible mechanism in which the transient is powered by orbital energy extraction” . This admission is appropriate, but it conflicts with the assertive interpretive language elsewhere. If the transient model is only outlined, then the conclusion should not read as though the mechanism has been physically validated.
I think the manuscript should either:
- substantially strengthen the transient modeling, with clearer energetics, timescale control, mass-loss physics, and stability arguments, or
- narrow its main claim to “the current binary properties are consistent with, but do not establish, a historical transient scenario.”
- The historical-astronomical argument needs a more rigorous uncertainty framework
The paper claims there is “no compelling reason to doubt the accuracy of the ancient records” and that alternative explanations such as confusion or transcription problems “can be excluded,” while atmospheric-extinction over-correction is said to be unsupported by the catalog data . This is one of the most consequential pillars of the paper, yet the standard of exclusion appears too qualitative in the provided text. Since the entire transient hypothesis depends on the historical brightness discrepancy being real, the manuscript should quantify uncertainty in the ancient magnitude interpretation more rigorously.
The authors should provide a clearer error budget for the inferred historical brightnesses and show how sensitive the claimed discrepancy is to:
- extinction assumptions,
- latitude and observing geometry,
- catalog translation/calibration choices,
- stellar identification ambiguity,
- and the historical practice of qualitative brightness ranking.
Without a stronger quantitative treatment of these uncertainties, the astrophysical interpretation remains underdetermined by the historical premise.
- The physical scenario depends on several tuned conditions that need explicit plausibility assessment
The manuscript emphasizes “two rather tantalizing facts”: that the primary is close to filling its Roche lobe and that it is at “the very beginning of the subgiant phase” just after core hydrogen exhaustion . These are indeed suggestive, but suggestiveness is not yet sufficiency. The scenario requires a previous eccentric orbit of about e ≃ 0.58–0.6, pseudo-synchronization arguments, orbital-energy extraction of order 10^47 erg, and a long-lived optically thick circumbinary envelope with very small envelope mass but prolonged luminous output over approximately 1000 years .
These linked assumptions may be physically possible, but the paper should confront how finely tuned the scenario is. For example:
- How robust is the inferred initial eccentricity to uncertainties in rotation period, alignment, and tidal formalism?
- How sensitive is the transient duration to envelope mass, opacity, and energy dissipation efficiency?
- Is the proposed 1000-year timescale naturally produced, or is it effectively chosen to fit the historical interval?
As written, the paper itself acknowledges this unresolved point: “An important remaining question is what sets the timescale of the transient or, equivalently, its luminosity” . That unresolved point is not peripheral; it is central.
- Some reasoning is energetic rather than dynamical, and the distinction should be made explicit
The paper argues that the minimum required radiant energy, “E ∼ 2 × 10^47 erg,” is “comparable (and smaller) than the orbital energy of the close binary” of about 5.4 × 10^47 erg, “suggesting that extraction of its orbital energy is a possible source for the brightening” . This is reasonable as an order-of-magnitude plausibility check, but energetic comparability alone does not establish dynamical realizability. The key question is not merely whether the energy reservoir exists, but whether the binary can dissipate and radiate it on the proposed timescale while preserving the observed present configuration.
The manuscript should more carefully state that the argument presently establishes energetic plausibility, not physical demonstration. Otherwise, readers may overinterpret the result.
- Reproducibility and dependence on adopted assumptions should be more transparent
In Section 4.2, the paper notes that one orbital/distance solution produced more uncertain dynamical values, and then says that “for our official solution (reported in Table 4) we adopted the Gaia DR3 distance and only use the radial velocity semi-amplitudes KAa and KAb to measure the mass ratio, with the total mass following from Kepler’s Third Law” . That may be a justified hybrid approach, but it should be explained more transparently as a modeling choice. Readers need a clearer account of:
- why the hybrid solution is preferred,
- what changes under a fully dynamical solution,
- and whether the main astrophysical interpretation materially depends on that choice.
Similarly, the spectral-abundance section explicitly acknowledges limitations: “a fixed microturbulent velocity of 0.8 km s−1, the assumption of local thermodynamic equilibrium and of a single temperature, when in reality the temperature of the star varies with latitude due to gravity darkening” . These caveats are significant and should be integrated into the confidence level of the abundance interpretation rather than left as a footnote-scale qualification.
- The manuscript shows some anti-fraud resilience, but also some paper-risk features that should be addressed
I do not see obvious paper-mill incoherence. The manuscript is structurally coherent, field-appropriate, and technically anchored in identifiable datasets and established tools such as VLTI, ESPaDOns, FEROS, TESS, PHOENIX, MIST, and PHOEBE . That said, there is a fraud-risk pattern sometimes seen in speculative papers: a very strong headline claim built on a chain of individually plausible but cumulatively weakly constrained inferences. Here, the observational binary-star analysis appears solid, but the millenary-transient explanation may be overextended beyond what the evidence currently secures. I therefore recommend that the authors sharply delimit established results from conjectural reconstruction.
Potential Minor Revisions
- Typographic and grammatical issues
There are multiple small language issues that should be corrected for polish and clarity.
- Page 2, Section 1/transition into Section 2: “used the demark the end of the Eridanus constellation” should be corrected to “used to demarcate” or similar .
- Page 2, Introduction: “Whether this discrepancy is real or the result of misjudgement” should likely read “misjudgment” for consistency in modern English, unless the journal style prefers the current spelling .
- Page 2, organization sentence: “we discourse on the history of Theta Eridani” is stylistically awkward in scientific prose; “we discuss” would be more standard .
- Page 7, Table 3/Section 3.2 vicinity: “The spec- tra” is split by lineation in extraction, which likely reflects typesetting, but similar hyphenation artifacts should be checked throughout the manuscript .
- Page 13, Section 4.5.1: “We found (v sin i)Aa ≃ 70 km s−1 and (v sin i)Ab ≃ 60 km s−1 . If the stellar spin axes are aligned...” is clear, but later “The radial velocities of both components were them measured” should be “were then measured” .
- Page 13, Section 4.5.1: “we dot not find the flagship marks” should be “we do not find” .
- Page 13, Section 4.5.1: “YII 4900.1Å” should likely be standardized to “Y II 4900.1 Å” to match neighboring notation .
- Page 14, Section 5 discussion lead: “caused by its non negligible rotation” should be “non-negligible rotation” .
- Page 15, Section 6: “not far form the main-sequence” should be “not far from the main sequence” .
- Page 15, Section 6: “a rather uncharacteristic error made by three independent ancient observers” is understandable, but “an uncharacteristic error” would be grammatically preferable if that exact phrase appears in the manuscript context visible around the conclusion/discussion passages .
- Formatting and presentation issues
- The manuscript contains visible arXiv/preprint-style formatting artifacts in the extracted text, including line wraps and split equations/tables. Before submission, the authors should carefully inspect all equations, table headings, figure captions, and footnotes for readability and consistency .
- Some section transitions would benefit from firmer signposting, especially the move from observational results to speculative transient modeling. A subsection explicitly titled “Hypothesis status and limitations” would improve reader orientation.
- Table-heavy sections should ensure all symbols are defined in captions or notes. For instance, parameters such as q, γ, and specific flux-ratio notation should be explicitly defined near the relevant tables for ease of reading .
- The smartphone photo in Figure 1 is unusual in tone for a technical paper. It may be acceptable for context, but the authors should ensure that its scientific necessity is clear and that it does not weaken the manuscript’s formal presentation .
- AI content analysis for this post-2021 work
Estimated percentage of AI-generated content: Low, approximately 5–15%.
Rationale:
- The manuscript has a coherent expert voice, domain-specific technical detail, and a layered argument structure that is more characteristic of human specialist authorship than generic AI drafting.
- It includes concrete observational specifics, instrumental details, and field-appropriate caveats that do not read like formulaic machine-generated padding .
- However, a few phrases have a slightly rhetorical or headline-like quality, especially where the argument becomes speculative, for example “The remarkable combination of orbital and stellar parameters hints...” in the abstract . That rhetoric alone is not evidence of AI generation; it is more likely authorial style.
Highlighted AI-detected sections:
- No section stands out as obviously AI-generated.
- The most machine-like passages are not generic filler, but rather strong, compressive claims in the abstract and conclusion that summarize speculative links with unusual confidence .
- I do not see obvious markers of synthetic hallucination such as internally inconsistent terminology, citation-pattern collapse, or abrupt conceptual drift.
Assessed epistemic impact:
- Even if a small amount of AI-assisted editing was used, its likely impact on the manuscript’s epistemic contribution appears limited.
- The real epistemic concern is not AI generation, but claim calibration: the observational results are considerably firmer than the historical-transient interpretation.
Recommendations
-
Reframe the paper around its strongest contribution: the modern characterization of Theta Eridani Aa+Ab as a close, mildly eccentric, evolved binary with noteworthy Roche-lobe proximity and evolutionary state. Those results appear publishable on their own merits .
-
Downgrade the transient explanation from near-conclusion to structured hypothesis unless the authors can add more rigorous quantitative modeling. In particular, revise the abstract, discussion, and conclusion so the distinction between evidence and conjecture is unmistakable .
-
Strengthen the historical analysis by adding a formal uncertainty framework for the magnitude discrepancy. This is essential because the paper’s astrophysical narrative depends critically on the reliability of ancient brightness reports .
-
Expand the physical modeling or provide a dedicated limitations subsection addressing the envelope structure, dissipation mechanism, transient duration, and sensitivity to assumed prior orbital parameters. The current admission that the detailed model is outside scope is honest but leaves the central mechanism underdeveloped .
-
Clarify dependence on adopted distances, rotation assumptions, synchronization arguments, and abundance-model simplifications. These do not necessarily invalidate the study, but they must be better integrated into the uncertainty narrative .
-
Thoroughly proofread the manuscript for grammar, notation spacing, and stylistic consistency before submission.
Overall, I judge this to be an intriguing and potentially valuable manuscript with a solid observational core, but one whose most consequential claim currently outruns the evidentiary and modeling support. The paper would improve substantially if it becomes more conservative, more quantitative about historical uncertainty, and more explicit about the speculative status of the proposed millenary transient mechanism.
Reviewer3 (AI review)
AI Review (reviewer3.com)
Overview
Orbital and Stellar Parameters of Theta Eridani Aa+Ab
Contribution
The authors combine interferometric, spectroscopic, and photometric data to determine the orbital parameters, dynamical masses, and radii of the inner binary Theta Eridani Aa+Ab, establishing that the primary star has recently finished core hydrogen burning and expanded to approximately 80% of its Roche lobe radius.
Mechanism for Millenary Historical Transient
Contribution
This study proposes a physical model explaining that the ~10-fold historical brightening of Theta Eridani recorded by ancient astronomers was real and powered by orbital energy extraction via tidal dissipation during an eccentric Roche lobe overflow phase.
Comments (14)
1. Contradictory claim on whether extinction over-correction explains discrepancy — Editorial
derestimated rather than over-corrected. Arguably the strongest argument that atmospheric extinc- tion over-correction does explain the θ Eri discrep- ancy is that, as can be seen in Figure 2, two other southern mP=1 stars, Canopus and Alpha Cen- tauri, have lower culmination altitudes than θ Eri and yet their brightnesses were not overestimated in the ancient catalogs ( Canopus was rather un- 6 It should be noted that it is still somewhat of a mystery how
In Section 2, under item '3. Over-correction of atmospheric extinction', the text first states: 'This explanation is nonetheless not supported by the data.' However, a few sentences later in the same paragraph, a missing 'not' reverses the author's intended meaning: 'Arguably the strongest argument that atmospheric extinction over-correction does explain the \theta Eri discrepancy is that, as can be seen in Figure 2, two other southern mP=1 stars, Canopus and Alpha Centauri, have lower culmination altitudes than \theta Eri and yet their brightnesses were not overestimated in the ancient catalogs...'
2. Inconsistent sign for Equation 4 versus prose definition — Editorial
extinction and level of background light. The Almagest magnitude of θ Eri converts to a Johnson magnitude of 0.22 (0.11 without the correction terms), compared to its modern V magnitude of 2.88. The discrepancy ∆ V ≃ 2 . 7 is not only the highest among the magnitude 1 stars in the Almagest but also among all 992 stars. Figure 2 shows the magnitude discrepancy between the
In Section 2 (page 5) and Figures 2 and 3, the visual magnitude discrepancy \Delta V is defined and plotted as ancient minus modern ('\Delta V (Almagest-modern)' and '\Delta V (al-Sufi-modern)'), yielding negative values for Acamar (around -2.7 and -2.4 in the plots) because its ancient numerical magnitude was smaller (brighter) than its modern value. However, throughout the prose (Abstract, Section 2 on page 3, and Section 6 on page 15), the discrepancy is reported as a positive quantity: '\Delta V \simeq 2.7' and '\Delta V \simeq 2.4'. The sign convention between the text and the figure axes is inverted without explanation.
3. Symbol typo R_{Aa} instead of R_{Ab} for secondary radius — Editorial
Figure 10 shows the photometric data (blue) together
In Section 4.4 ('Photometric temperatures') and again in the caption of Figure 10, the radius of the secondary star (Ab) is denoted with the primary's subscript R_{Aa} instead of R_{Ab}. Earlier in Section 4.4, the radii were properly defined as 'The radii were fixed to R_{Aa} = 4.3 R_\odot and R_{Ab} = 4.0 R_\odot', but the text later says 'and for the secondary (T_{Ab} = 7800 K, \log g = 3.5, R_{Aa} = 4.0 R_\odot)' and the Figure 10 caption repeats 'and for the secondary (T_{Ab} = 7800 K, \log g = 3.5, R_{Aa} = 4.0 R_\odot)'.
4. Incorrect cross-reference for period measurement in Section 4.2 — Editorial
In practice, we fix the period to the extremely precise value obtained from the TESS lightcurve (4.4). We then
In Section 4.2 ('Orbital parameters'), the text states: 'In practice, we fix the period to the extremely precise value obtained from the TESS lightcurve (4.4).' However, the TESS lightcurve analysis and the period measurement (P = 4.107704 \pm 0.000008 day) are presented in Section 4.3 ('Radii from ellipsoidal modulation'), whereas Section 4.4 is 'Photometric temperatures'.
5. Qualitative visual inspection without quantitative abundance fitting — Major
Through visual inspection we find that, within the limitations of the atmosphere model 9 , the vast majority of elements have surface abundances quite close to solar. The only exception are s-process elements, whose surface abundance appears enhanced (see e.g. Sr II 4077.7˚ A, Sr II 10327.3˚ A, Y II 4883.7˚ A, YII 4900.1˚ A, Ba II 4934.1˚ A, Ba II 5853.7˚ A, BaII 6141.7˚ A). While such
The claim that s-process elements are enhanced while Ca, Sc, and iron-peak elements are solar across Eri Aa, Ab, and B relies solely on visual inspection of spectra against a single solar-metallicity model with fixed microturbulence (). Intermediate-mass A stars and subgiants typically exhibit higher microturbulence (–), which disproportionately strengthens saturated strong resonance lines such as Sr II and Ba II ; without fitting or deriving quantitative abundances , microturbulent line broadening can mimic an s-process overabundance.
6. Rayleigh-Jeans approximation used at near-infrared wavelengths — Major
where f_{NIR} \simeq 0.88 is the near-infrared (H and K bands) flux ratio from the interferometry, \frac{T_{Aa}}{T_{Ab}} \simeq \frac{7600}{7800} (4.4) and the NIR is at the Rayleigh-Jeans tail of the stars' spectral energy density.
Equation (6) estimates the radius ratio by assuming the H () and K () bands lie well within the Rayleigh-Jeans tail where . At –, is in H band and in K band, so the Planck function scales closer to – rather than , slightly altering the derived secondary radius.
7. Circular use of Gaia DR3 Rp photometry — Major
). It was also important to set the “third light” contamination fraction due to compo- nent θ Eri B to 33% based on ∆ R p = 1 . 19 between B and A from Gaia DR3 (the R p filter is very similar to the TESS filter). Figure 9 shows the phase-averaged TESS lightcurve
In Section 4.4 you exclude Gaia DR3 and photometry from temperature fitting because Gaia DR3 photometry is not optimized for very bright stars and yields large residuals, yet in Section 4.3 you rely on the Gaia DR3 measurement between components A and B to fix the TESS third-light contamination at 33%. Because the derived stellar radii (, ) scale directly with the assumed third-light dilution, any bright-star systematic error in Gaia for the primary would bias the inferred radii and should be propagated into their uncertainties.
8. Coevality inconsistency between Aa and Ab radii — Minor
The second intriguing fact is that the primary appears to be at a very special stage in its evolution, namely at the very beginning of the subgiant phase i.e. when the star has just finished core hydrogen burning.
If Eri Aa () and Ab () are coeval at and Aa has just reached the terminal-age main sequence () where radius expands steeply with age (Figure 12), a less massive coeval companion (, whose main-sequence lifetime is – longer) should still be well on the main sequence with a significantly smaller radius (–), rather than the measured . This radius ratio discrepancy should be reconciled with the proposed evolutionary and mass-transfer history.
9. Overstated independence of ancient magnitude estimates — Major
Although it is ultimately impossible to completely exclude the possi- bility of errors due to an over-correction of atmospheric extinction, we find that this hypothesis is not supported by the overall catalog data and would have to be a rather uncharacteristic error made by three independent ancient observers.
You argue that Hipparchus, Ptolemy, and al-Sufi act as three independent observers confirming the brightness of Eri. However, because Ptolemy's and al-Sufi's catalogs directly inherited their baseline star lists and magnitudes from their predecessors, observing a star's position or rising/setting time does not guarantee an independent re-estimation of its magnitude—especially since Hipparchus left no numerical magnitude and Eri was the brightest star at the southern end of the River even at .
10. Unexplained tidal synchronization and circularization timescales — Critical
An intriguing possibility is that the stellar spins were pseudo-synchronized with the previous more eccentric orbit and maintained their stellar angular momenta.
The proposed mechanism requires dynamical tidal dissipation to circularize the orbit from to within $\sim 1000\text{ yrs}$ while leaving the stellar spins frozen at their pre-outburst pseudo-synchronous period ($P_{\text{rot}} \simeq 2\text{ days}$). Because the moment of inertia of the stars is much smaller than that of the orbit (), tidal synchronization typically proceeds orders of magnitude faster than orbital circularization, and radiative damping of dynamical tides in A-type stars normally operates on timescales much longer than $10^3\text{ yrs}$; without demonstrating how orbital energy is extracted so rapidly without spinning down the stars to the modern pseudo-synchronous period ($\sim 3.8\text{ days}$), the physical viability of the scenario remains unverified.
11. Hydrostatic equilibrium required for dynamical tide energy extraction — Major
1000 years. The pseudo-Roche lobe size of the primary at periastron for such an orbit r L ≃ 0 . 384 a i (1 − e i ) = 4 . 3 R ⊙ is equal to its current radius, implying that it would indeed be subject to eccentric Roche lobe overflow.
The model posits that the primary star fills its periastron Roche lobe () at to supply the envelope via eccentric Roche-lobe overflow, while simultaneously extracting per orbit via dynamical tides over orbits. However, once a star loses mass over periastron passages and is subject to strong nonlinear tidal heating in its outer layers, its radius and internal structure will evolve on a thermal/dynamical timescale rather than remaining statically poised at as expands by from to .
12. Unjustified residency time for circumbinary envelope mass — Major
In order for the process to last over n ∼ 1000 yrs 4 days ∼ 10 5 orbits, the mass in the circumbinary envelope must be much smaller than that of the system. In this case, it is likely that a non-negligible fraction of the circumbi- nary mass is ejected every orbit due to the instability of a test particle in this equal mass eccentric binary. The total mass ejection is limited by the available en- ergy to be less than about M ej ∼ 0 . 05 M ⊙ . This implies that the circumbinary mass at any given time is of order M env ∼ M ej n ∼ 5 × 10 − 7 M ⊙ . The optical depth of such an envelope is large
You estimate the instantaneous circumbinary envelope mass as (and thus optical depth ) by assuming that the ejected mass resides in the region for roughly one orbital period ($P \simeq 4\text{ days}$). However, the dynamical/escape timescale at around a binary () is only $R_{\text{env}}/v_{\text{esc}} \sim 13\text{ hours}$ ( orbits), which would reduce the steady-state mass and optical depth at unless the outflow velocity is fine-tuned to be far below the local escape speed.
13. Single-star evolution applied to post-mass-transfer binary — Minor
Figure 12 shows the evolution of the stellar radius (left) and of the core mass fractions of hydrogen, helium and carbon (right) for a 2.30M_\odot star from a MESA Isochrones and Stellar Tracks (MIST; Choi et al. 2016; Dotter 2016; Paxton et al. 2011, 2013, 2015) Equivalent Evolutionary Points (EEP) file.
You infer the evolutionary state and age () of Eri Aa by comparing its current radius () to a single-star MIST track for , while simultaneously proposing that Aa recently lost and absorbed or dissipated $\sim 10^{47}\text{ erg}$ of tidal energy within the last millennium. Because rapid mass loss and tidal heating in the radiative envelope drive a star out of thermal equilibrium, its present radius may remain inflated relative to a single-star equilibrium model on a Kelvin-Helmholtz timescale (–$10^6\text{ yrs}$), complicating a direct readout of its core hydrogen exhaustion state from a standard track.
14. Orbital angular momentum loss from mass ejection neglected — Critical
Given that only a small mass is lost from the system, the angular momentum (which is dominated by the or- bital angular momentum; the spin angular momentum is only about 7% of the orbital angular momentum in the current binary) is approximately conserved.
The reconstruction of the pre-outburst orbit (, , $\Delta E_{\text{orb}} \approx 1.8 \times 10^{47}\text{ erg}$) assumes strict conservation of orbital angular momentum, . However, if is ejected from the circumbinary envelope (at ) by gravitational torques from the inner binary, it carries away a specific angular momentum several times the binary's specific orbital angular momentum, removing a substantial fraction (–) of and significantly altering the inferred initial semi-major axis, periastron distance, and extracted energy budget.
XSci (AI review)
Summary
Theta Eridani, now V = 2.9, was listed among the brightest stars in the sky by Ptolemy (137 AD) and al-Sufi (964 AD), and singled out for its brightness by Hipparchus (129 BC). On the calibrated Almagest scale of Protte & Hoffmann (2020) the implied fading is ~2.7 mag, which the authors argue is real rather than an observing error. They then solve the inner pair Aa+Ab from archival VLTI/PIONIER and GRAVITY astrometry, ESPaDOnS spectra and TESS photometry, obtaining a tight, mildly eccentric orbit (a = 0.083 au, e = 0.105, P = 4.107704 d) of two ~2.2-2.3 M_sun stars filling ~80% and 74% of their periastron Roche lobes, the primary just past core hydrogen exhaustion. They propose the brightening was a millennium-long transient powered by orbital energy extracted by dynamical tides during eccentric Roche-lobe overflow in a formerly more eccentric orbit.
Contribution and Significance
The durable contribution is Section 4: the first dynamical masses, radii and three-dimensional orbit for theta Eri Aa+Ab, from archival data not taken for the purpose — the PIONIER frames were calibrator observations, re-reduced as science. Astrometric orbit, SB2 radial velocities and ellipsoidal modulation are made to agree, and the dynamical parallax, 53.7 (+3.2/-2.6) pc, reproduces the Gaia DR3 value for theta Eri B.
What makes the paper more than that is the coincidence it uncovers: the star with the largest historical-to-modern magnitude discrepancy in the Almagest is a near-Roche-lobe-filling binary whose primary sits at the terminal main sequence. Against the closest prior work — the century-old exchange (Anderson 1893; Gore 1900; Webb 1916), the calibrations of Schaefer (2013) and Protte & Hoffmann (2020), and the authors' own GRAVITY note (Waisberg et al. 2025) — this connection is new; none offers a physical mechanism. The mechanism matters less than the abstract implies: eccentric Roche-lobe overflow leading to a common envelope is established already (Regös et al. 2005; Lajoie & Sills 2011), and the new element — a long-lived low-mass envelope powered by dynamical tides rather than drag — remains a sketch (concern 1).
Strengths
- Section 4.2: the joint astrometric and spectroscopic solution is cleanly built, the degeneracies are reported honestly (Figure 14), and adopting the Gaia distance of the single companion is well justified.
- Section 4.3: using the ellipsoidal amplitude with a, i and q fixed by the orbit is the right way to obtain radii here, and Figure 9 shows it genuinely discriminates between 4.1, 4.3 and 4.5 R_sun.
- Section 2 and Table 1 treat the historical material quantitatively — culmination altitude, airmass and extinction per observer — and the alpha Cen error in the Protte & Hoffmann tables is caught and corrected (footnote 5).
- The extinction over-correction hypothesis is tested, not asserted: Canopus and Alpha Centauri culminated lower at Alexandria yet were not over-bright (Figure 2).
Recommendation
Major Revision for Astronomy & Astrophysics. The observational characterisation is sound and the central observation is striking, but the quantitative support for the transient mechanism — the claim of the title and abstract — does not survive checking against the manuscript's own numbers, and the radius underpinning two further claims carries an unstated systematic error. Both are repairable within the paper's approach, but the repair is substantial.
Major Revision for the Astrophysical Journal: the same concerns apply unchanged.
Major Revision for Monthly Notices of the Royal Astronomical Society: likewise, though its referees would press harder for Section 5 to be quantified or labelled a conjecture.
Major Concerns
-
The transient energy budget does not close, and the inferred initial orbit has no uncertainty (Section 5, Equations 7-14). The manuscript needs E ~ 2 x 10^47 erg to sustain ~1500 L_sun for 1000 years and obtains Delta E_orb ~ 1.8 x 10^47 erg, which it calls "just what is needed". But 1.8 is less than 2, and its own timeline is longer than 1000 years: the star was bright to Hipparchus in 129 BC, still magnitude 1 to al-Sufi in 964 AD, and faded only by 1598 — between ~1100 and ~1700 years. At 1700 years the requirement is ~3 x 10^47 erg, exceeding the extractable energy and more than half the current orbital energy (5.4 x 10^47 erg); the 0.05 M_sun of ejecta must be unbound out of the same reservoir.
The agreement is also fragile. Equation (13) gives e_i ~ 0.58 only because P_p-syn is set to exactly 2 days; the v sin i values (70 and 60 km s^-1, Table 4) carry no uncertainties, and the two components in fact imply 2.1 and 2.3 days. P_rot = 2.3 d gives e_i ~ 0.48 and Delta E_orb ~ 1.2 x 10^47 erg — a third less, and clearly insufficient. Please propagate errors on v sin i, R and spin-axis alignment through to e_i and Delta E_orb.
-
The primary radius, and the evolutionary phase inferred from it, carry unquantified systematics (Sections 4.3 and 5, Figure 12). The third light from theta Eri B is "set ... to 33% based on Delta R_p = 1.19". That magnitude difference gives f_B/f_A = 0.334, but a fraction of total light of only 25.0%. If 0.33 was entered as PHOEBE's fractional third light, the ellipsoidal amplitude is over-corrected by ~13%; since Delta F/F scales as R^3, R_Aa is biased by ~4%, or ~0.17 R_sun, more than the quoted +/- 0.1 R_sun. Please state which convention was used, and refit if needed: R_Aa propagates into the filling factor, log g and the evolutionary phase.
That the primary has "just finished core hydrogen burning" rests on matching R_Aa to one MIST track at exactly 2.30 M_sun and solar metallicity, with no propagation of the +/- 0.13 M_sun mass uncertainty and no variation of metallicity. Figure 12 begins at 700 Myr with R already falling from 4.0 R_sun, so R = 4.3 R_sun is also reached on the pre-hook branch below the plotted range — a second solution hidden by the axis limits. The measured T_eff = 7600 K and log g = 3.54 are never used to break this degeneracy; a Kiel diagram against a grid of tracks would settle it.
-
The independence of the three historical records is overstated (abstract; Sections 2 and 6). Section 6 speaks of "three independent ancient observers", the abstract of "three different ancient observers". Section 2 argues the opposite about the records: the Almagest is at least partly based on Hipparchus' lost catalogue (the 1 deg longitude offset; Duke 2003), and al-Sufi's is "heavily based on the Almagest", positions copied exactly plus 12 deg 42 arcmin. Three independent sightings are not three independent magnitude assignments: for al-Sufi the evidence is non-revision of an inherited value, and for Hipparchus a descriptive superlative rather than a catalogue magnitude. The argument is stronger stated precisely: al-Sufi revised magnitudes elsewhere (chi Car, beta Sgr) but not this one. Gysembergh, Williams & Zingg (2022), Journal for the History of Astronomy 53, 383, should also be discussed.
Minor Comments
- In Section 2, item 3, "the strongest argument that atmospheric extinction over-correction does explain the theta Eri discrepancy" inverts the evidence that follows and should read "does not explain".
- The first column of Table 1 is headed "Hipparcos" where Hipparchus is meant, confusing since the Hipparcos catalogue is used in the same table.
- The abstract and Section 6 claim the "thirteen brightest" for al-Sufi too, but that count is derived only for Ptolemy, and Figure 3 gives him fourteen magnitude-1 stars.
- Figure 2 scatters over roughly +/- 2 mag, so quantify the outlier claim with the rms of Delta V for the other fourteen magnitude-1 stars.
- In Section 5, R_L,Aa/a = 0.384 corresponds to M_Aa/M_Ab = 1.06, not to q = 0.94 as defined in Equation (3), which gives 0.374.
- The asynchronous Roche-lobe correction made for the current orbit (6.1 to 5.6 R_sun) is not made for the initial orbit, where the spin is sub-synchronous at periastron and the correction acts in the opposite sense, weakening the coincidence r_L = R_Aa (see Sepinsky, Willems & Kalogera 2007, ApJ 660, 1624).
- State the gyration constant behind "the spin angular momentum is only about 7% of the orbital angular momentum" (I obtain 2-3%), and confront the ~3.9 d pseudo-synchronisation period Equation (10) gives for the current orbit against the observed ~2 d.
- Give the units of the limit Pdot <~ 10^-6 (Section 4.3); it lies an order of magnitude below the mean rate implied by a shrinkage from 8 d to 4.1 d in 1000 yr.
- Table 4 gives no uncertainties for T_Aa, T_Ab or the v sin i values, and R_Ab is 3.95 R_sun there but 4.0 R_sun in Section 4.3 and the abstract.
- Since the ellipsoidal amplitude is the only radius measurement and the star saturates TESS, report that amplitude per sector rather than calling the lightcurves "virtually identical".
- The s-process enhancement rests on visual inspection at v sin i ~ 70 km s^-1 (Section 4.5.1) yet carries the primordial-enrichment conclusion, so equivalent widths with uncertainties are needed.