{"id":"3fcfb7f3-ae71-4ad3-9bcb-a2d709eb1ce7","arxiv_id":"2505.08042","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":5,"one_line_summary":"GJ 105 AC now has a fully measured 76-year orbit and dynamical masses M1 = 0.78 ± 0.02 M_sun and M2 = 0.098 ± 0.002 M_sun, precise to about 2%.","lead":"This paper measures the orbit of the nearby binary GJ 105 AC using 569 new radial velocity measurements that catch the fast swing near closest approach, plus astrometry of the pair. The result is a model-independent mass for each star to about 2% precision, making the pair useful benchmarks for testing stellar models.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the astrometric error model is the weakest link, but the paper's own published-error comparison bounds its impact and the benchmark mass claim survives.","rationale":"The reader's weakest_assumption identified the astrometric error model, and I agree that this is the least externally anchored part of the analysis. I examined whether it actually threatens the central claim: the paper's dynamical masses are robust to the choice of error weighting (centroids shift about 2%, within the quoted uncertainties), the degraded precision is still about 3%, and the benchmark definition in the paper is 3%. Independent SED masses and the previous Feng et al. (2021) dynamical mass are consistent. The RV data set (569 new MINERVA points plus archival Lick, HARPS, HIRES, APF) tightly constrains P, e, omega_1, and K1, so the masses are not predominantly carried by the small astrometric sample. I therefore find no significant objection that changes the ACCEPT verdict. The suggested removal test is a worthwhile verification step because the NIRC2 epochs are the only astrometry inside the periastron passage, but the available evidence indicates the conclusion is stable.","tokens_in":24902,"tokens_out":14319,"duration_ms":154124,"concrete_test":"Remove the three NIRC2 epochs from the preferred joint fit (reweighted errors) and recompute M1 and M2 with all other data unchanged; if either mass shifts by more than its quoted 68% interval, the 2% precision claim is over-reliant on the least-characterized astrometry. As a second check, re-reduce the 2021 NIRC2 epoch using the Service et al. (2016) distortion solution and compare the resulting separation and position angle against the published adopted values.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most plausible load-bearing concern is the self-referential astrometric error treatment. Section 4.6 derives per-instrument reweighted errors from leave-one-out residuals of the joint model itself, then refits the same model; an additional global error scale S_ast is fitted (ln S_ast = 0.37 for the reweighted fit). If this procedure underestimates systematic errors, the quoted uncertainties M1 = 0.782 +/- 0.019 and M2 = 0.098 +/- 0.002 would be too small. The three NIRC2 epochs are the only astrometry near the periastron passage, and they carry assumed 3 mas / 0.5 deg errors plus the Yelda et al. (2010) distortion model; the LOO reweighting inflates their errors to 32 mas, which is a large but unverified correction. However, the paper directly tests the alternative published-error weighting: the mass centroids shift by only 2.1% (M1) and 2.0% (M2), and the text concedes that including this source brings precision to about 3%. Because the stated benchmark threshold is +/-3%, the central claim of a ~2-3% model-independent mass remains intact even under the less favorable error model. The masses also agree with the independent EXOFASTv2 SED masses (1.4 sigma) and with Feng et al. (2021), and the RV data dominate the period and eccentricity constraints. The concern therefore does not land as load-bearing for the paper's core assertion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The authors present a joint Keplerian fit to new MINERVA radial velocities and 13 relative astrometry points (including new Keck/NIRC2 measurements) for the K3+M7 binary GJ 105 AC. They derive P = 76.0 ± 1.3 yr, M1 = 0.782 ± 0.019 M_sun, and M2 = 0.098 ± 0.002 M_sun, with 2.5% and 2.3% precision, respectively. They compare these dynamical masses with independent EXOFASTv2 SED-based masses, discuss the system's status as a benchmark 'effectively single' star, and assess the prospects for asteroseismic follow-up.","tokens_in":25245,"tokens_out":11783,"duration_ms":115807,"significance":"If the results hold, this paper adds two stars to the small population of 'effectively single' benchmark objects with model-independent masses precise to ~2-3%, and GJ 105 C becomes one of the lowest-mass stars with a dynamical mass. The new MINERVA RV data spanning the periastron passage are a valuable contribution, and the joint fit is carefully tested against the choice of astrometric error weighting. The masses agree with independent SED modeling at 1.4 sigma and with the Feng et al. (2021) mass at 0.5 sigma, which strengthens confidence in the result. The explicit 'effectively single' criteria in Section 6.1 are a useful framework for placing the system in the context of the broader benchmark-star population, even though the thresholds are approximate.","major_comments":[],"minor_comments":[{"comment":"The text 'located in the thin disk ( ?)' contains an unresolved placeholder citation that should be replaced with the appropriate reference.","section":"Section 2"},{"comment":"The HARPS data span is given as 'between 2003 October 27 and 200 September 6'; the year is presumably 2009 and should be corrected.","section":"Section 3.1"},{"comment":"The abstract contains a duplicated article in 'as well as the the second-widest true separation', which should be corrected to 'the second-widest'.","section":"Abstract"},{"comment":"The astrometry-only fit with published errors gives Tperi, omega1, and Omega that differ from the preferred joint fit by 5.6-6.6 sigma, yet the text only notes that the periods are consistent within 1.3 sigma; the authors should add a brief discussion of whether this tension reflects the need for the LOO reweighting, a known degeneracy in astrometry-only solutions, or an unresolved systematic in the astrometry.","section":"Section 4.6 and Table 7"},{"comment":"The LOO reweighting procedure is self-referential in that the same joint model is used both to calibrate the astrometric errors and to perform the final fit; the published-error comparison bounds the impact on the masses, but a sentence in Section 5 or the abstract clarifying that the ~2% uncertainties are conditional on this reweighting and that the alternative error model yields ~3% precision would make the claim easier to parse.","section":"Section 4.6"},{"comment":"The sentence 'all of the low-mass stars in our sample with a/R_star < 100 are isolated' appears to contradict the stated criterion a/R_star > 100; please check whether the inequality should be reversed.","section":"Section 6.1.1"},{"comment":"The text 'a class of Markov Markov Chain Monte Carlo' contains a duplicated word and should read 'Markov Chain Monte Carlo'.","section":"Section 4.5"}],"recommendation":"minor_revision","confidential_remarks":"The manuscript is well within the scope of the journal and the dynamical mass measurement is a solid contribution. My one substantive request is that the authors address the large differences in Tperi, omega1, and Omega between the published-error astrometry-only fit and the preferred joint fit in Table 7; this is discussable in revision rather than a reason to reject. The LOO error reweighting is the weakest methodological link, but the paper's own robustness check against the published-error model keeps the central claim intact."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a good paper and the central result holds up. The authors add 569 new MINERVA RVs covering the periastron passage and RV minimum for the first time, pull three previously unpublished NIRC2 astrometric epochs from the Keck archive, and jointly fit RVs plus relative astrometry to get M1 = 0.78 ± 0.02 Msun and M2 = 0.098 ± 0.002 Msun. That is a real improvement over Feng et al. (2021), which had the secondary at 9% uncertainty. The masses agree with independent EXOFASTv2 SED masses and with the previous dynamical mass, so nothing looks suspicious. The orbital conventions are carefully rederived, the data are in machine-readable tables, and the Keplerian fitting is standard and reproducible.\n\nThe weakest part is the astrometric error treatment, and the paper is upfront about it. Section 4.6 reweights the 13 relative astrometry points using leave-one-out residuals from the joint model itself, then refits the same model with an additional global S_ast scaling. That is self-referential, and the three NIRC2 points get inflated to 32 mas. Also, the likelihood ignores RA/Dec covariance from the PA/separation conversion, and the authors exclude some data (MINERVA T4, a few photometric points) for stated reasons. But here is the thing: the paper directly tests the alternative published-error weighting. The masses shift by only about 2%, and the precision degrades to roughly 3%—still at the benchmark threshold. So the concern does not land as load-bearing. A stricter referee might ask for a covariance-aware or hierarchical error model, but the central claim would survive.\n\nThe benchmark-star framing is reasonable. The \"effectively single\" criteria in Section 6 involve approximate thresholds, but those do not feed back into the masses; they are classification, not derivation. The TESS asteroseismology non-detection is a minor side note, appropriately hedged.\n\nWho is this for? People who care about low-mass stellar evolution benchmarks, model-independent masses, and orbit fitting with long-period binaries. It is not a method paper, but it is a clean, useful data paper with a genuinely new result. I would bring it to reading group and would cite it. The authors have done the work, disclosed the caveats, and the result is stable under the main alternative assumption. Send it to peer review; I would accept after minor comments, mostly asking for a slightly more honest discussion of the LOO reweighting limitations and any future astrometric epochs that could break the degeneracy.","headline":"A solid benchmark-star paper: new MINERVA RVs and archival NIRC2 astrometry sharpen the dynamical masses of GJ 105 AC to ~2%, and the result survives the main error-model worry.","tokens_in":25904,"tokens_out":1112,"would_cite":true,"duration_ms":13685,"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":"This paper derives model-independent dynamical masses for both components of the nearby binary GJ 105 AC, at about 2% precision, and argues the stars are effectively single benchmark objects.","keywords":["dynamical masses","binary stars","orbital solution","radial velocities","relative astrometry","benchmark stars","GJ 105 AC","low-mass stars"],"falsifier":"Measure the binary's orbit independently with future astrometry that does not use the same error model, for example a Gaia astrometric orbit or additional high-resolution imaging over the next decade, and compare the total mass $M_1+M_2$ and individual masses with the paper's values; a discrepancy larger than the combined uncertainties at the level of a few percent would show the quoted precision overstates what the data support.","tokens_in":24652,"feed_emoji":"⭐","tokens_out":7706,"duration_ms":72011,"temperature":0.7,"pith_summary":"This paper reports a complete three-dimensional orbit for the nearby binary GJ 105 AC and uses it to weigh both stars without relying on stellar models. The joint analysis of new radial velocities and relative astrometry yields $M_1 = 0.78 \\pm 0.02\\,M_\\odot$ for the K3 primary and $M_2 = 0.098 \\pm 0.002\\,M_\\odot$ for the M7 companion, on a $76.0 \\pm 1.3$ year orbit. That precision puts both stars in a rare class: stars with model-independent dynamical masses good to roughly 2–3% that are far enough apart to be treated as effectively single and to give clean, unblended spectra. Because GJ 105 A is a bright nearby K dwarf, it becomes a practical benchmark for testing stellar evolution models and for future asteroseismic age work.","feed_headline":"Two nearby stars now have masses precise to 2 percent","feed_subtitle":"A new orbital fit makes GJ 105 A and C benchmark stars with clean spectra, the widest such pair after alpha Centauri.","key_machinery":"The central object is the three-dimensional Keplerian orbit of the secondary about the primary, parametrized by the period, time of periapsis, eccentricity, argument of periapsis, longitude of the ascending node, inclination, and the two masses, with the radial velocity and astrometry likelihoods linked through Kepler's third law and the velocity semi-amplitude $K_1$. The argument is carried by the joint fit: radial velocities fix the period, eccentricity, and periapsis timing, while relative astrometry fixes the orientation and physical angular scale; together they break the degeneracies that plague either data set alone. A secondary mechanism is the leave-one-out reweighting of the 13 astrometric errors, which the authors use to avoid overfitting the small astrometry sample, and they check that the preferred masses change by only $\\sim$2% when the published errors are used instead.","core_discovery":"The authors establish that GJ 105 AC is a benchmark binary: combining 569 new MINERVA radial velocities that capture the full periapsis passage and the RV minimum with 13 relative astrometry points spanning 27 years and seven instruments, they fit a single Keplerian model and read off the masses directly from the orbit. The resulting dynamical masses are $M_1 = 0.78 \\pm 0.02\\,M_\\odot$ and $M_2 = 0.098 \\pm 0.002\\,M_\\odot$, with both components agreeing with independent SED and isochrone model masses at the 1.4$\\sigma$ level. The paper argues this makes GJ 105 A and C the newest members of a small population of stars with $\\sim$2–3% model-independent masses that are effectively single, and notes the system has the widest on-sky separation of any such pair after $\\alpha$ Centauri AB, so both stars can be observed without spectral blending.","pith_inferences":["One extension the authors do not pursue is using the dynamical mass of GJ 105 A to calibrate asteroseismic scaling relations directly; the predicted extreme-precision RV oscillation amplitude of about 1.9 m/s makes this comparison feasible.","At roughly $0.098\\,M_\\odot$, GJ 105 C sits near the bottom of the main sequence, so a future measurement of its radius and temperature would make it a pointed test of the mass–luminosity relation in a regime where dynamical masses are rare.","The effectively single criteria introduced here could be applied to the growing sample of astrometric binaries to estimate how many clean-spectrum benchmark stars exist; the paper's own census of 805 precise-mass stars finds only 122 qualifiers, suggesting the usable population is small and worth cataloging."],"forward_implications":["GJ 105 A and C join the small set of stars with model-independent masses at roughly 2–3% precision that are effectively single, meaning their properties can be compared directly with single-star evolution models.","The new MINERVA radial velocities, which cover the full periapsis passage and the RV minimum for the first time, remove the earlier period ambiguity and pin the period to $76.0 \\pm 1.3$ years.","With an average on-sky separation of $2.67''$ and a maximum of $3.51''$, the system is, after $\\alpha$ Centauri AB, the widest such precise-mass pair, so both components can be observed with clean, essentially unblended spectra.","GJ 105 A is a viable target for extreme-precision RV asteroseismology, with a predicted oscillation amplitude of about 1.9 m/s, which could deliver an independent age for the system; TESS photometry alone does not detect the oscillations."],"supporting_citations":[{"why":"Supplied the previous full orbital solution and the APF radial velocities; the paper's mass for GJ 105 C agrees with it.","marker":"Feng et al. (2021)"},{"why":"Provided discovery astrometry and HST/WFPC2 relative positions that anchor the long baseline of the orbit.","marker":"Golimowski et al. (2000)"},{"why":"Supplied the distortion solution applied to the new NIRC2 images to convert pixel positions to astrometry.","marker":"Yelda et al. (2010)"},{"why":"Gaia DR3 parallax is used as the distance prior that turns angular separations into physical masses.","marker":"Gaia Collaboration et al. (2023)"},{"why":"Contributed the Lick/Hamilton radial velocities used in the RV and joint fits.","marker":"Fischer et al. (2014)"},{"why":"Contributed the HARPS radial velocities used in the fits.","marker":"Butler et al. (2017)"},{"why":"Contributed the Keck/HIRES radial velocities used in the fits.","marker":"Santos et al. (2010)"},{"why":"Provided the long-period alternative orbit from the same astrometry, which the new periapsis coverage rules out.","marker":"Roberts & Mason (2018)"}],"fun_headline_variants":["GJ 105 AC joins benchmark stars with 2% masses","Full periapsis passage yields 2% masses for GJ 105 AC","GJ 105 AC: widest benchmark pair with 2% masses","GJ 105 AC: precise masses for effectively single stars"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The ~2 percent mass errors assume the adopted uncertainties on the 13 astrometric measurements, including the assumed 0.5-degree and 3-milliarcsecond errors and the camera distortion model for the new Keck epochs, contain no hidden systematic error; the paper itself notes that with the published errors the precision drops to about 3 percent.","fun_headline_variants_meta":{"raw":{"variants":["GJ 105 AC joins benchmark stars with 2% masses","Full periapsis passage yields 2% masses for GJ 105 AC","GJ 105 AC: widest benchmark pair with 2% masses","GJ 105 AC: precise masses for effectively single stars"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001502,"raw_usage":{"total_tokens":6091,"prompt_tokens":1077,"completion_tokens":5014,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":693,"completion_tokens_details":{"reasoning_tokens":4936}},"tokens_in":693,"tokens_out":5014,"duration_ms":35464,"temperature":1.0,"reasoning_tokens":4936,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:05:54.202770+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the binary's orbit independently with future astrometry that does not use the same error model, for example a Gaia astrometric orbit or additional high-resolution imaging over the next decade, and compare the total mass $M_1+M_2$ and individual masses with the paper's values; a discrepancy larger than the combined uncertainties at the level of a few percent would show the quoted precision overstates what the data support.","supporting_citations":[{"cited_title":"A., Henry, T","cited_arxiv_id":null,"evidence_quote":"Provided discovery astrometry and HST/WFPC2 relative positions that anchor the long baseline of the orbit."},{"cited_title":"C., & Mason, B","cited_arxiv_id":null,"evidence_quote":"Provided the long-period alternative orbit from the same astrometry, which the new periapsis coverage rules out."}],"review_version":1}