{"id":"a964de42-d621-495a-9bb3-10a58ac1a87a","arxiv_id":"2505.19718","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Dynamical masses of Kepler-16 AB are remeasured with SOPHIE data and a 2-m class telescope, yielding 1.5% and 0.9% precision and masses 2 to 7% higher than earlier values.","lead":"Using high-resolution cross-correlation spectroscopy on archival spectra from a 1.93-m telescope, the authors detect the faint M-dwarf companion in the circumbinary-planet host Kepler-16 AB and measure dynamical masses for both stars at the percent level. This demonstrates that precise dynamical masses for high-contrast binaries no longer require 8 to 10-m telescopes, broadening access to low-mass star benchmarks.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"SVD detrending's effect on the secondary K2 is extrapolated from ESPRESSO to SOPHIE without re-validation; an injection test is needed to exclude a velocity bias exceeding the claimed precision.","rationale":"The reader's weakest_assumption identifies the same load-bearing concern I find: the SVD removal fraction is assumed from a different instrument and target, not measured for SOPHIE. This is the single point where the central claim (percent-level K2 and masses) is most exposed. I considered other risks—post hoc data exclusions, the small number of time-splitting sets, and the arbitrary 32-component SVD cap—but these affect uncertainty estimation or are secondary to the SVD systematic. The paper's internal checks (two fitting methods, time-splitting, bootstrap, PHOENIX comparison) provide good support for statistical robustness and for the general method, but none of them would catch a constant or velocity-dependent attenuation of the secondary signal by the SVD. The explicit 'we expect' language is a self-acknowledged assumption. Because the reader's verdict is already CONDITIONAL and this concern is the one that motivates conditionality, my read does not change the verdict; it sharpens the specific test needed. I recommend keeping CONDITIONAL pending an injection-based validation of the SVD step on SOPHIE data.","tokens_in":12986,"tokens_out":7667,"duration_ms":79697,"concrete_test":"Inject a synthetic secondary spectrum with known K2 (e.g., exactly 46.88 km/s) and contrast ~6×10^-3 into the 160 SOPHIE spectra at the observed phases, then run the full SVD detrending, CCF, and global time-fit pipeline over many realizations. If the mean recovered K2 differs from the injected value by more than ~0.1-0.2 km/s, the SVD removal is biasing the result. Repeat with different injected K2 values and inspect the phase-resolved recovery to detect velocity-dependent removal.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on the accuracy of K2 = 46.88 ± 0.28 km/s, but the SVD detrending that isolates the secondary is calibrated only via an expectation from Sebastian et al. (2024a), not on these SOPHIE data. Section 3 states: 'According to Table 4 in S24, we thus expect that this SVD detrending has removed less than 5 % of the secondary's spectral features.' This table was produced for ESPRESSO (R≈140 000) data on a different target, while SOPHIE has R=40 000, a different wavelength range, and a different phase sampling. If the removal fraction is larger, or if it is velocity-dependent (e.g., SVD components absorb more secondary flux near phases where the secondary lines approach the primary), the recovered K2 is systematically biased. The mass function scales approximately as K2^3, so a 1% K2 bias implies roughly a 3% mass bias—exceeding the claimed 0.9% mass precision and comparable to the 2.8% K2 offset from Bender et al. (2012). The time-splitting and bootstrap tests in §5 address statistical and correlated-noise uncertainties, not this SVD systematic. The PHOENIX comparison in §7 is qualitative and does not constrain K2.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper applies the High-Resolution Cross-Correlation Spectroscopy (HRCCS) method to 160 SOPHIE spectra of the circumbinary binary Kepler-16 AB, using SVD detrending to remove the primary and isolate the M-dwarf secondary. Two fitting methods (K-focusing with the Saltire model, and a new global time-domain fit) yield consistent values of the secondary semi-amplitude, K2 = 46.88 ± 0.28 km/s, with a detection significance of 9.5 sigma. Combining this with K1 from Triaud et al. (2022) and the photometric inclination from Doyle et al. (2011), the authors derive dynamical masses M1 = 0.704 ± 0.011 Msun and M2 = 0.2054 ± 0.0019 Msun, claiming 1.5% and 0.9% precision, and conclude that 2-m-class telescopes can deliver such measurements. The paper also presents time-splitting and bootstrap analyses to assess systematic uncertainties, and a phase/wavelength-resolved extraction of the secondary's signal.","tokens_in":13270,"tokens_out":3377,"duration_ms":38014,"significance":"If the result holds, this is a valuable methodological demonstration: it extends HRCCS dynamical mass measurements from an 8-m class instrument to a 2-m class instrument, which would significantly broaden the sample of EBLM systems with model-independent masses. The paper is thorough in its uncertainty treatment: two independent fitting methods agree, time-splitting and bootstrap give consistent systematic estimates, and the final masses are compatible with the photometric-dynamical model at about 1.5 sigma. The authors are also explicit about the discrepancy with Bender et al. (2012) and about the assumptions involved in the SVD detrending. The central load-bearing measurement is K2, and the main weakness is that its accuracy depends on an untested extrapolation of the SVD removal fraction from ESPRESSO to SOPHIE data.","major_comments":[{"comment":"The claim that SVD detrending removes less than 5% of the secondary's spectral features is based on 'Table 4 in S24' (Sebastian et al. 2024a), which was derived for ESPRESSO data (R≈140,000) on a different binary. The present SOPHIE data have R=40,000, a different wavelength range, and different phase sampling. This is a load-bearing extrapolation because a velocity-dependent or larger removal fraction would bias K2 directly. Since the mass function scales approximately as K2^3, a 1% K2 bias corresponds to roughly a 3% mass bias, comparable to the claimed 0.9% precision and to the 2.8% K2 offset from Bender et al. (2012). The time-splitting and bootstrap tests in Section 5 address statistical and correlated-noise uncertainties, but not this SVD systematic. I recommend adding an injection-recovery test: inject a synthetic secondary spectrum with known K2 into the SOPHIE data, run the full detrending and fitting pipeline, and verify that the recovered K2 is unbiased at the claimed level.","section":"Section 3, SVD detrending"},{"comment":"The paper attributes the ~3-sigma discrepancy with Bender et al. (2012) to their use of only six spectra, but this is an assertion rather than a demonstration. A systematic offset in either analysis could equally explain the discrepancy. Since the central claim is that the new masses are accurate, the authors should either quantify the expected scatter from a six-spectrum analysis (e.g., by re-fitting subsets of their own data or by simulation) or present a more hedged discussion of the possible common-mode systematics. As written, the explanation is plausible but not load-bearing evidence for the accuracy of the SOPHIE-based K2.","section":"Section 6, Table 3"},{"comment":"The global time fit adopts tight uniform priors on P, T0, e, and omega from Triaud et al. (2022), and the K-focusing fit keeps them fixed. Since the T22 orbital parameters were derived partly from the same SOPHIE dataset, the K2 measurement is not fully independent of those parameters. The authors do re-fit the Keplerian parameters in the global fit and find values consistent with T22, which mitigates the concern, but the text does not explicitly address the degree of overlap. I would like a sentence clarifying how the overlapping data affect the claimed independence of the dynamical masses.","section":"Section 4 and Table 1"}],"minor_comments":[{"comment":"There is a typo: 'SOHPIE spectra' should be 'SOPHIE spectra'.","section":"Appendix A"},{"comment":"The affiliation line contains a typo: 'Birmimgham' should be 'Birmingham'.","section":"Author affiliations"},{"comment":"The lower panel of Figure 2 shows the CCF slice and Saltire model, but the caption does not define the green/blue colors in the shaded uncertainty region; please clarify.","section":"Section 3, Figure 2"},{"comment":"The bootstrap description uses 'K_p' interchangeably with 'K2' in the text; please use a consistent notation throughout.","section":"Section 5.2"},{"comment":"The comparison to PHOENIX models in the right lower panel of Figure 4 is described as reproducing 'dominant features', but the model's average CCF contrast is smaller than the data; the text should state whether this offset is expected from the assumed Teff, metallicity, or line mask, or whether it indicates a systematic in the extracted contrast.","section":"Section 7"}],"recommendation":"major_revision","confidential_remarks":"The paper is well-executed and the two fitting methods agree, but the accuracy of the central mass measurement hinges on an untested assumption about SVD detrending on SOPHIE data. An injection-recovery test is the natural and probably sufficient remedy. If the authors can provide one, I would expect the paper to become acceptable; without it, the 0.9% precision claim is not fully supported. The discrepancy with Bender et al. also deserves a more quantitative treatment, but that is secondary. No concerns about novelty or scope: the paper fits MNRAS well."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe short version: this is the first HRCCS dynamical-mass measurement on a 2-m class telescope and the first for Kepler-16 AB. It delivers M1 and M2 at 1.5% and 0.9% precision, consistent with the photometric-dynamical model within 1.5 sigma, and introduces a global time-domain CCF fit that is a clear improvement over K-focusing for low-SNR trails. It deserves a serious referee.\n\nWhat is genuinely new: adapting the S24 method from ESPRESSO/VLT to SOPHIE/1.93-m; the global time fit that lets the four orbital parameters float instead of keeping them fixed; and a systematic-uncertainty analysis (time-splitting plus bootstrap) that is honest about correlated noise. The paper also argues that bootstrap can substitute for time-splitting when the signal is too weak to split, which is directly useful for future faint-companion work.\n\nThe soft spot is the SVD detrending. The authors quote S24's table to expect less than 5% removal of secondary features, but that table came from ESPRESSO at R=140,000 on a different target, while SOPHIE has R=40,000 and a different wavelength range. If the removal fraction is larger, or velocity-dependent near phases where the secondary approaches the primary, K2 could be biased. I don't think this is fatal: because the secondary moves in the primary rest frame, its lines are mostly orthogonal to the common SVD modes, and two independent fitting methods agree. But an injection test on the SOPHIE data would settle it in one paragraph and should be requested in revision. The lack of released code is a smaller weakness; the Saltire model should be available.\n\nThe 3-sigma offset from Bender et al. is plausibly their six spectra; the photometric-dynamical model agrees within 1.5 sigma, so the mass shift is not alarming. The paper is appropriately cautious about the planet mass.\n\nWho this is for: people measuring dynamical masses of high-contrast binaries, especially with 2-m facilities, and anyone using HRCCS for exoplanet atmospheres since the uncertainty analysis transfers. My recommendation: send it to review. The SVD question is a fixable revision, not a fatal flaw.","headline":"First 2-m HRCCS masses for Kepler-16 AB; sound analysis, with one transferability caveat on SVD that a revision can fix.","tokens_in":13869,"tokens_out":3368,"would_cite":true,"duration_ms":34248,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A 1.93-meter telescope equipped with the SOPHIE spectrograph can resolve the faint companion of Kepler-16 AB well enough to weigh both stars to about one percent.","keywords":["binaries: spectroscopic","stars: fundamental parameters","stars: low-mass","binaries: eclipsing","techniques: spectroscopic","circumbinary planets","dynamical masses","M-dwarf companions"],"falsifier":"An injection-recovery test on the actual SOPHIE spectra would settle the matter: add a synthetic M-dwarf spectrum with a known injected orbital semi-amplitude into the 160 detrended spectra, run the full pipeline, and check that the recovered $K_2$ is unbiased at the $0.28\\ \\mathrm{km\\,s^{-1}}$ level. A second check is an independent measurement of $K_2$ for Kepler-16 B with an 8-meter-class spectrograph, which would test whether the 2–7% mass offset from the older six-spectrum measurement is a bias in the old data or a systematic in the new method.","tokens_in":12823,"feed_emoji":"🔭","tokens_out":17867,"duration_ms":132673,"temperature":0.7,"pith_summary":"The paper aims to show that high-resolution cross-correlation spectroscopy (HRCCS), a technique developed to read exoplanet atmospheres, can extract the orbital motion of a faint stellar companion from spectra taken with a 2-meter-class telescope, not just with 8–10 meter telescopes. Applying it to 160 archived SOPHIE spectra of the circumbinary-planet host Kepler-16 AB, the authors recover the M-dwarf secondary at $9.5\\sigma$ significance and measure its velocity semi-amplitude $K_2 = 46.88 \\pm 0.28\\ \\mathrm{km\\,s^{-1}}$. Combining this with the primary's well-measured orbit and the known binary inclination yields dynamical masses $M_1 = 0.704 \\pm 0.011\\,M_\\odot$ (1.5%) and $M_2 = 0.2054 \\pm 0.0019\\,M_\\odot$ (0.9%). If correct, this opens the large existing sample of single-lined eclipsing binaries to precise, model-independent mass measurements on modest telescopes, and the slightly higher masses of Kepler-16 AB leave the known circumbinary planet's mass unchanged within uncertainties.","feed_headline":"Weigh Kepler-16's M-dwarf to 0.9% on a 2-m telescope","feed_subtitle":"A 2-meter-class telescope now weighs the faint star in the first circumbinary-planet system to percent accuracy.","key_machinery":"The load-bearing object is the cross-correlation trail of the secondary, which is built by aligning all spectra to the primary's rest frame, removing the primary's absorption with a truncated singular-value decomposition (the number of removed components chosen by the 'effective rank' criterion), and then cross-correlating the residuals with an M-dwarf line mask. The secondary's weak signal is co-added either in the $K_2$–$V_{\\rm rest}$ plane ('K-focusing', fitted by the Saltire model) or fitted simultaneously across the whole orbit with a double-Gaussian Keplerian model. The parameter carrying the mass measurement is $K_2$, the secondary's semi-amplitude: with the primary's semi-amplitude $K_1$ from earlier radial-velocity work and the inclination from Kepler light-curve modelling, the standard two-body mass equations give both masses directly.","core_discovery":"The central discovery claimed is that HRCCS works on 2-meter-class data: after a singular-value decomposition removes the dominant star's absorption spectrum from the 160 SOPHIE spectra, the M-dwarf secondary's cross-correlation signal is detected at $9.5\\sigma$, and its Keplerian trail is visible in the time domain. Two independent fitting approaches — the K-focusing map fit with the Saltire model and a new global time-domain fit of the full cross-correlation trail — return consistent values of $K_2$, and bootstrapping plus time-splitting tests show the quoted $0.28\\ \\mathrm{km\\,s^{-1}}$ uncertainty is robust. The resulting masses are 2–7% higher than previous measurements, within $1.5\\sigma$ of the photometric-dynamical model but only $3\\sigma$ of the earlier six-epoch dynamical measurement, and they leave no measurable change in the circumbinary planet's mass.","pith_inferences":["If the method scales to fainter companions, the same 2-meter pipeline could turn the radius-inflation tension at the bottom of the main sequence into a direct mass benchmark for evolutionary models of fully convective stars.","The assumed survival fraction of the companion's lines during SVD detrending is the single most important thing to calibrate next; an injection-recovery test on SOPHIE data would convert the under-5% removal assumption into a measured bias correction across orbital phase.","If the 2–7% offset against the older six-epoch measurement is a real feature, it suggests HRCCS masses, not the photometric-dynamical model, become the reference values for Kepler-16 AB and its planet."],"forward_implications":["A large existing sample of single-lined eclipsing binaries can be re-observed or re-analysed with this pipeline to turn them into double-lined systems, yielding model-independent masses of the M-dwarf companions.","The measured $K_2 = 46.88 \\pm 0.28\\ \\mathrm{km\\,s^{-1}}$ implies the binary is 2–7% more massive than the photometric-dynamical value, which within uncertainties does not change the mass of the circumbinary planet.","The global time-domain CCF fit and the bootstrap/partial-data uncertainty protocol give a recipe for deriving trustworthy $K_2$ uncertainties even when the companion's CCF is barely above the noise.","For SOPHIE, the method should be applicable to companions of roughly $0.35\\,M_\\odot$ and above, roughly doubling the reach of precision dynamical masses from 2-meter telescopes."],"supporting_citations":[{"why":"Supplies the HRCCS/K-focusing method and the simulation-based estimate that SVD detrending removes less than 5% of the secondary's spectral features.","marker":"S24"},{"why":"Provides the primary's semi-amplitude $K_1$, the orbital elements used to align spectra and set priors, and the planet's radial-velocity solution.","marker":"T22"},{"why":"The previous six-epoch dynamical mass measurement from a 9.2-meter telescope that the new masses are compared against (2–7% offset, about 3-sigma).","marker":"Bender et al. (2012)"},{"why":"Provides the binary inclination from Kepler light curves and the photometric-dynamical masses used in the mass derivation.","marker":"Doyle et al. (2011)"},{"why":"Defines the effective-rank criterion used to choose how many SVD components to remove in each spectral chunk.","marker":"Roy & Vetterli (2007)"},{"why":"Supplies the emcee MCMC sampler used for both the Saltire map fit and the global time-domain fit.","marker":"Foreman-Mackey et al. (2013)"},{"why":"Documents the SOPHIE spectrograph on the 1.93-meter telescope, the instrument whose data enable the 2-meter-class claim.","marker":"Perruchot et al. (2008)"}],"fun_headline_variants":["2-m scope weighs Kepler-16's M-dwarf to 0.9%","HRCCS on a 2-m telescope measures Kepler-16's M-dwarf to 0.9%","Kepler-16's faint star mass measured to 0.9% with 2-m scope","M-dwarf in Kepler-16 detected at 9.5σ and weighed to 0.9%","2-m class HRCCS weighs the secondary in Kepler-16 to 0.9%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The chain of inference assumes that the data-cleaning step (a singular-value decomposition that removes the primary star's spectral lines) leaves more than 95% of the companion's spectral features intact, a fraction adopted from earlier simulations rather than re-derived for the SOPHIE spectra; if the removal is larger or velocity-dependent, the measured semi-amplitude $K_2$ and both dynamical masses would be systematically biased.","fun_headline_variants_meta":{"raw":{"variants":["2-m scope weighs Kepler-16's M-dwarf to 0.9%","HRCCS on a 2-m telescope measures Kepler-16's M-dwarf to 0.9%","Kepler-16's faint star mass measured to 0.9% with 2-m scope","M-dwarf in Kepler-16 detected at 9.5σ and weighed to 0.9%","2-m class HRCCS weighs the secondary in Kepler-16 to 0.9%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00187,"raw_usage":{"total_tokens":7411,"prompt_tokens":1089,"completion_tokens":6322,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":705,"completion_tokens_details":{"reasoning_tokens":6195}},"tokens_in":705,"tokens_out":6322,"duration_ms":34643,"temperature":1.0,"reasoning_tokens":6195,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:07:06.549970+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"An injection-recovery test on the actual SOPHIE spectra would settle the matter: add a synthetic M-dwarf spectrum with a known injected orbital semi-amplitude into the 160 detrended spectra, run the full pipeline, and check that the recovered $K_2$ is unbiased at the $0.28\\ \\mathrm{km\\,s^{-1}}$ level. A second check is an independent measurement of $K_2$ for Kepler-16 B with an 8-meter-class spectrograph, which would test whether the 2–7% mass offset from the older six-spectrum measurement is a bias in the old data or a systematic in the new method.","supporting_citations":[{"cited_title":"pp 606--610","cited_arxiv_id":null,"evidence_quote":"Defines the effective-rank criterion used to choose how many SVD components to remove in each spectral chunk."}],"review_version":1}