{"id":"81fdebba-92f5-42d0-bbfc-82d64c0c8a64","arxiv_id":"2506.02348","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"low","formal_verification":"none","parameter_count":7,"one_line_summary":"Extended VLBA monitoring over 1984 days pins the Galactic Center magnetar's proper motion to about 2% precision, consistent with a clockwise-disk origin, and sets acceleration limits consistent with a bound orbit around Sgr A*.","lead":"Astronomers used the Very Long Baseline Array to track the only known pulsar near the Milky Way's central black hole for more than five years, measuring the magnetar PSR J1745-2900's motion to about 2% precision. The measured motion and new acceleration limits support its origin in a nearby disk of young stars and set the stage for a future direct measurement of the black hole's gravitational pull.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The RA acceleration limit is not robust: a 1600-day correlated residual is absorbed by an ad hoc 0.35 mas systematic error, and the fitted a_alpha shifts by more than a factor of two when this term is added.","rationale":"The reader's verdict is CONDITIONAL with high confidence, and my stress-test agrees with that assessment. The headline proper motion is stable across all fit variants and is likely correct; the weak point is the acceleration limit and the treatment of the unexplained RA residual. The paper is transparent in showing the residual and the effect of systematic terms, but the abstract presents the systematics-included acceleration limit as if it were a robust constraint. The most load-bearing issue is not the existence of the residual, which the paper acknowledges, but that the quoted errors assume it is white noise. If the residual arises from Sgr A* centroid wander, the relative astrometry inherits it, and the acceleration limit could be biased by more than its quoted error. This does not overturn the proper motion result, so the CONDITIONAL verdict should stand. The conditions already requested by the reader are appropriate, especially demonstrating that the acceleration upper limit is robust to modeling the 1600-day signal as correlated noise and making the timing analysis that rejects the binary companion available.","tokens_in":17597,"tokens_out":8075,"duration_ms":79431,"concrete_test":"Refit the Table 1 data with a model that explicitly includes a sinusoid at ~1600 days (or a Gaussian-process red-noise term) in addition to proper motion and acceleration, using the exact MJD sampling; also refit with the 0.35 mas systematic error excluded. If the acceleration posterior shifts by more than the quoted error, or if the upper limit moves by more than 2 sigma, the abstract's (0.4, 0.2) limit is not robust. Additionally, compute the projection of a 1 mas, 1600-day sinusoid onto the design matrix for the linear-plus-quadratic model over the observed epochs to quantify the induced bias in mu and a.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The fundamental measurement is the magnetar position relative to Sgr A* (§2); any apparent centroid motion of Sgr A* enters directly. After fitting standard terms, RA residuals show a ~1 mas, ~1600 d quasi-periodic signal (Fig. 4) that none of the considered mechanisms fully explains (§4.1). To make chi^2_alpha ~ 1, the authors add a per-epoch 0.35 mas systematic error in RA (§3 and Table 3). This error is uncorrelated across epochs, while the residual it absorbs is correlated. The acceleration estimate is strongly sensitive to this choice: a_alpha = -0.238 ± 0.089 (bootstrap, no systematic) becomes -0.113 ± 0.074 (with systematic); the abstract's upper limit (0.4, 0.2) is taken from the systematics-included fit. A 1 mas sinusoid at 1600 days projects onto the quadratic term with an apparent acceleration of roughly A(2pi/T)^2, about 0.05 mas yr^{-2}, comparable to the quoted a_alpha and its error. The bootstrap resampling resamples epochs independently, so it does not propagate time correlation. Hence the acceleration limit is not a conservative bound: it depends on treating an unexplained correlated signal as white noise. The proper motion, by contrast, is stable across fit variants, so the 2% accuracy claim is on firmer ground, but the residual could still contribute a common systematic at the ~0.1 mas yr^{-1} level.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports 25 epochs (41 measurements) of VLBA astrometry of PSR J1745–2900 over 1984 days, extending the 2014–2015 data with 15 new epochs. The authors fit a model including proper motion, acceleration, parallax, and core shift relative to Sgr A*, and find μ_α ≈ 2.01–2.04 mas yr^{-1}, μ_δ ≈ 6.07–6.10 mas yr^{-1}; quote an acceleration limit of ≲ (0.4, 0.2) mas yr^{-2}; report no secular change in scattering size; find a mean core shift consistent with zero; and interpret the proper motion as consistent with an origin in the clockwise stellar disk. They also discuss a ~1600-day quasi-periodic residual in RA, test and reject a binary companion using pulse-timing data, and constrain a possible second scattering screen.","tokens_in":17881,"tokens_out":10125,"duration_ms":91631,"significance":"The proper motion measurement is a substantial improvement over Bower et al. (2015) and, if the relative-astrometry calibration is sound, provides the strongest astrometric constraint on a Galactic Center pulsar. The paper is honest about systematics: it explicitly states in §4.1 that no tested mechanism fully explains the residuals, and it uses multiple fitting schemes and bootstrap resampling. The CW-disk origin claim is tested against external stellar kinematics and is robust to the acceleration/core-shift additions. The main scientific payoff—future detection of gravitational acceleration by Sgr A* if the magnetar re-brightens—is clearly articulated. However, the acceleration limit is more fragile than the abstract suggests because it depends on treating a correlated residual as white noise.","major_comments":[{"comment":"The acceleration upper limit is not robust to the treatment of correlated residuals. The RA residuals after standard fits show a ~1 mas, ~1600-day quasi-periodic signal (Fig. 4; §4.1). The paper makes χ²_α ≈ 1 by adding an uncorrelated per-epoch systematic error of 0.35 mas in RA (§3), but this changes the fitted a_α from –0.252 ± 0.052 mas yr^{-2} (LSQ, no systematic) to –0.111 ± 0.057 mas yr^{-2} (with systematic; Table 3), and the bootstrap resamples epochs independently, so it does not propagate the time correlation. A ~1 mas, ~1600-day quasi-periodic residual can project substantially onto the quadratic term, and the factor-of-two shift in a_α demonstrates the sensitivity. The abstract's limit of ≲ (0.4, 0.2) mas yr^{-2} and the statement that it is consistent with the ~0.03 mas yr^{-2} expectation therefore rest on treating an unexplained correlated signal as white noise. This should either be modeled explicitly (e.g., as a sinusoid or Gaussian process) or the acceleration claim should be reported as a provisional constraint with a clear caveat. The proper-motion and CW-disk-origin conclusions are not affected by this issue.","section":"Section 3, Table 3, Abstract"},{"comment":"The quoted '≲2% proper-motion accuracy' is a precision relative to Sgr A* under the assumption that Sgr A*'s apparent centroid is fixed in time. The manuscript itself finds that none of the candidate mechanisms fully explains the RA residuals, and if part of the ~1 mas, ~1600-day residual is apparent centroid motion of the reference source (refractive wander or Sgr A* structure changes), it would contribute a common systematic of order 0.1 mas yr^{-1}—larger than the reported 0.04 mas yr^{-1} random error in μ_α. I ask the authors to state explicitly that the ≲2% figure is statistical precision relative to Sgr A*, and to provide a bound on the reference-source systematic or cite independent evidence that Sgr A*'s centroid is stable at this level over the full multi-year span at these frequencies.","section":"Section 2 and §4.1"}],"minor_comments":[{"comment":"The text gives π < 0.4 mas while the Figure 3 caption says 'upper limit at 95% confidence of π < 0.6 mas'; please reconcile the value and state the confidence level.","section":"Section 4.1 vs. Figure 3"},{"comment":"Typo: 'indistuinghisable' should be 'indistinguishable'.","section":"Figure 1 caption"},{"comment":"The bootstrap description says '10 4 astrometric series'; this should be '10^4', and the resampling unit (individual measurements vs. epochs) should be stated explicitly.","section":"Section 3"},{"comment":"The acceleration upper limit should state the confidence level (apparently 3σ) and clarify that it is an upper limit on the absolute value in each coordinate, rather than leaving the confidence implicit.","section":"Abstract and Table 3"},{"comment":"The ellipsis rows for multi-band epochs make the table hard to parse; use explicit repeated MJD values or a footnote to indicate multiple bands at the same epoch.","section":"Table 1"},{"comment":"The reference list gives Bower et al. 2006a and 2006b with identical journal, volume, and page; please verify that the 2006b entry is correct and distinct.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid observational contribution, and the central proper-motion result is robust across fit variants. My recommendation is driven by the acceleration claim in the abstract, which needs either a proper correlated-noise treatment or a more qualified statement. Once that is addressed, I expect the paper to be acceptable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I've read this paper and the stress-test memo. The stress-test lands: the proper motion is the real result, and it's solid. The acceleration upper limit is softer than the abstract implies.\n\nWhat's new: 15 new VLBA epochs, extending the baseline from 477 to 1984 days and the epoch count from 10 to 25. That yields a five-fold improved proper motion of (2.04, 6.07) ± (0.04, 0.02) mas/yr, stable across every fit variant. That's a genuine, careful measurement, and the CW-disk origin story holds. The null results on angular-broadening variability and on Sgr A* core shift are useful, and the size-stability limit is sensible. The paper is transparent about the residual scatter and the post hoc systematic errors, which is good practice.\n\nWhere it's soft: the RA residuals show a ~1 mas, ~1600-day quasi-periodic signal that no proposed mechanism explains. The authors add a per-epoch 0.35 mas systematic to bring chi^2 to ~1. That systematic is uncorrelated, while the residual is correlated, so it's absorbing a red signal as white noise. The acceleration estimate shifts: a_alpha goes from -0.238 ± 0.089 without the systematic to -0.113 ± 0.074 with it, and the abstract quotes the systematics-included fit. A correlated signal of this amplitude projects onto the quadratic term at roughly the same level as the quoted a_alpha, so the acceleration limit is not conservative. The bootstrap resamples epochs independently, so it doesn't propagate the correlation. Also, the binary-companion rejection leans on Eatough et al. (2025, in prep); that dependency should be flagged.\n\nThe abstract's (0.4, 0.2) limit should be reconciled with Table 3, where the a_alpha error is ~0.07 and the fitted value is -0.11. The proper motion, by contrast, is robust to all these choices, so the 2% precision claim is on firm ground.\n\nWho this is for: people working on GC pulsars, scattering screens, and Sgr A* astrometry. It's a careful data paper, worth citing for the proper motion and the stability limits.\n\nRecommendation: this deserves peer review. The central measurement is trustworthy, but the acceleration limit should be reframed as tentative, with the systematic-error issue addressed or de-emphasized. I'd ask for a revision, not a rejection.","headline":"Solid, transparent astrometry extension; the proper motion is robust, but the headline acceleration limit leans on an ad hoc systematic that absorbs a correlated residual.","tokens_in":18574,"tokens_out":2262,"would_cite":true,"duration_ms":17639,"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":"Five years of VLBA monitoring of the Galactic Center magnetar PSR J1745–2900 constrain its proper motion to about 2% and set acceleration limits consistent with a bound orbit around Sgr A*.","keywords":["PSR J1745-2900","Galactic Center magnetar","VLBA astrometry","proper motion","acceleration","Sgr A*","interstellar scattering","core shift"],"falsifier":"A simultaneous VLBI campaign that ties Sgr A* directly to distant extragalactic sources over the same 1984-day span would reveal any reference-frame wander; if the fitted magnetar proper motion changes by more than the quoted errors when the reference frame is re-anchored, the claimed acceleration limits are biased. Alternatively, a future epoch after re-brightening that measures a tangential acceleration significantly different from zero—or a proper-motion change exceeding the quoted errors at a ~10-year separation—would either confirm the bound orbit or falsify the current limits.","tokens_in":17313,"feed_emoji":"🔭","tokens_out":12977,"duration_ms":99462,"temperature":0.7,"pith_summary":"This paper reports 25 epochs of VLBA astrometry of PSR J1745–2900, the only known pulsar near the Galactic Center black hole Sgr A*, spanning 1984 days. The measurements tighten the magnetar's proper motion to roughly 2% uncertainty and set upper limits on its acceleration of about 0.4 and 0.2 milliarcseconds per year squared in right ascension and declination, consistent with the maximum acceleration expected if the magnetar orbits Sgr A* at its projected separation. The authors argue that if the magnetar re-brightens, an additional epoch of monitoring about a decade later could directly detect that acceleration, confirming the magnetar is bound to the black hole and measuring their separation. They also investigate a ~1600-day oscillation in the astrometric residuals and find no fully satisfactory explanation: a binary companion model fits the positions but is contradicted by pulse-timing measurements.","feed_headline":"Five-year watch pins down the Galactic Center magnetar's motion to 2%","feed_subtitle":"Motion is consistent with a bound orbit around Sgr A*; a re-brightening could reveal the black hole's pull.","key_machinery":"The fundamental observable is the position of the magnetar relative to Sgr A*, obtained by calibrating the VLBA data on Sgr A* and transferring the solutions to the magnetar; the Sgr A* reference frame is fixed by an updated ICRF position and proper motion. The astrometric model fits a reference offset, a parallax term relative to Sgr A*, a linear proper motion, a quadratic acceleration term, and a wavelength-dependent core shift. Bootstrap resampling over $10^4$ synthetic series is used to confirm the error estimates, and Lomb-Scargle periodograms of the residuals are used to search for a binary signal. The angular broadening is tracked by fitting Gaussians to the magnetar's image at 2 cm and deconvolving the synthesized beam.","core_discovery":"The central claim is that the extended astrometric baseline, 41 independent measurements over 1984 days, constrains the proper motion of PSR J1745–2900 to $\\lesssim$2% uncertainty ($\\mu_\\alpha = 2.01 \\pm 0.04$, $\\mu_\\delta = 6.09 \\pm 0.02$ mas yr$^{-1}$) and places upper limits on the absolute tangential acceleration of $\\lesssim (0.4, 0.2)$ mas yr$^{-2}$ in the two coordinates. These limits are consistent with the maximum acceleration of $\\sim 0.03$ mas yr$^{-2}$ expected for a bound orbit around Sgr A* at the projected separation of about 0.1 pc. The paper interprets the proper-motion direction as consistent with the magnetar originating in the clockwise stellar disk with a modest kick. It further reports that the astrometric residuals contain an apparent sinusoidal ~1600-day variation in right ascension that cannot be fully explained: a stellar companion fits the astrometry but is ruled out by pulse-timing measurements, and no other candidate mechanism (refractive wander, changes in Sgr A* structure) is fully satisfactory. No secular change in the magnetar's angular broadening is detected, and the mean core shift of Sgr A* is consistent with zero and with expectations for a compact jet or symmetric accretion flow.","pith_inferences":["If the unexplained ~1600-day right-ascension residual is real wandering of Sgr A*'s centroid rather than motion of the magnetar, the quoted proper motion and acceleration limits absorb that drift, and the true systematic may exceed the per-epoch errors the paper adopts (0.35 mas RA, 0.15 mas Dec).","A future campaign that ties Sgr A* to extragalactic reference sources, rather than only measuring the magnetar relative to Sgr A*, could separate reference-frame wander from the magnetar's acceleration and sharpen the test.","If the magnetar stays too faint for another decade, the acceleration detection may need to come from stacking relative astrometry of nearby S-stars (already measured with NIR astrometry) rather than this pulsar alone."],"forward_implications":["If PSR J1745–2900 re-brightens, a second astrometric measurement separated by ~10 years from the mean epoch would detect the acceleration of Sgr A* at 5–10$\\sigma$, directly confirming a bound orbit and measuring the separation.","The proper motion direction matches the clockwise stellar disk, supporting the magnetar's origin in that disk with a modest natal kick.","Over four years, no change in the magnetar's apparent size is seen, ruling out a significant time-variable scattering screen near the Galactic Center and implying any obscuring ionized gas is patchy on scales larger than ~200 AU.","The mean core shift of Sgr A* is consistent with zero and with the magnitude expected from GRMHD jet models, so the data do not discriminate between jet and symmetric-accretion-flow interpretations."],"supporting_citations":[{"why":"Supplies the first 10 epochs and the relative-astrometry fitting procedures that this paper extends.","marker":"Bower et al. (2015)"},{"why":"Defines the Sgr A* proper motion used to anchor the reference frame.","marker":"Reid & Brunthaler (2020)"},{"why":"Supplies the updated ICRF position of Sgr A* adopted as the reference origin.","marker":"Gordon et al. (2023)"},{"why":"Provides the Galactic Center distance used to compute the expected maximum acceleration of a bound orbit.","marker":"Do et al. (2019)"},{"why":"Provides the pulse-timing data whose periodogram is inconsistent with the binary-companion interpretation.","marker":"Eatough et al. (2025)"},{"why":"Gives the jet-model core-shift predictions that the null core-shift measurement is compared against.","marker":"Moscibrodzka et al. (2014)"},{"why":"Supplies the angular-broadening/temporal-scattering relation used to locate the scattering screen and constrain a second screen.","marker":"Cordes & Lazio (1997)"},{"why":"Sets the angular-broadening model of Sgr A* used in the calibration.","marker":"Bower et al. (2006a)"}],"fun_headline_variants":["Galactic Center magnetar's motion measured to 2%","5-year watch locks magnetar's path to 2% accuracy","Magnetar's acceleration limit aligns with Sgr A* orbit","Future watch could detect Sgr A*'s pull on magnetar"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The measurement is relative to Sgr A*, and the calibration forces Sgr A* to have a fixed apparent position and motion, so any real drift of Sgr A*'s centroid—which the paper's own ~1600-day residual hints at—would be absorbed into the magnetar's fitted proper motion and acceleration.","fun_headline_variants_meta":{"raw":{"variants":["Galactic Center magnetar's motion measured to 2%","5-year watch locks magnetar's path to 2% accuracy","Magnetar's acceleration limit aligns with Sgr A* orbit","Future watch could detect Sgr A*'s pull on magnetar"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.002469,"raw_usage":{"total_tokens":9580,"prompt_tokens":1150,"completion_tokens":8430,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":766,"completion_tokens_details":{"reasoning_tokens":8355}},"tokens_in":766,"tokens_out":8430,"duration_ms":68415,"temperature":1.0,"reasoning_tokens":8355,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:26:45.938464+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A simultaneous VLBI campaign that ties Sgr A* directly to distant extragalactic sources over the same 1984-day span would reveal any reference-frame wander; if the fitted magnetar proper motion changes by more than the quoted errors when the reference frame is re-anchored, the claimed acceleration limits are biased. Alternatively, a future epoch after re-brightening that measures a tangential acceleration significantly different from zero—or a proper-motion change exceeding the quoted errors at a ~10-year separation—would either confirm the bound orbit or falsify the current limits.","supporting_citations":[{"cited_title":"2025, in prep","cited_arxiv_id":null,"evidence_quote":"Provides the pulse-timing data whose periodogram is inconsistent with the binary-companion interpretation."}],"review_version":1}