{"id":"666eada4-e9f2-4fed-aa16-3428486ed8ee","arxiv_id":"2509.06215","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"First proper-motion measurements of Andromeda V and VI from a 20-year Hubble baseline show And VI co-orbits M31's disk, while And V remains poorly constrained.","lead":"Using Hubble images taken 20 years apart, the authors measured the sky motions of two dwarf galaxies orbiting Andromeda. One of them, Andromeda VI, moves in the same direction as Andromeda's disk, while Andromeda V's motion is too uncertain to confirm the opposite direction.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The And VI zero-point shows a 2.2σ RA discrepancy between the two ACS-chip solutions; unquantified galaxy-centroid systematics could rotate the orbit and weaken the co-rotation claim.","rationale":"Good-faith read: this is a careful, data-rich measurement paper. It uses public HST data, describes the astrometric reductions in detail, and the orbit conclusions are appropriately hedged, especially for And V. The reader's weakest assumption, that background galaxies provide an unbiased absolute-proper-motion zero point, is indeed the load-bearing one. I do not claim the zero point is wrong; the concern is that the paper's own Table 2 contains a 2.2σ RA discrepancy between the two And VI ACS-chip galaxy solutions, and no systematic error is added to the adopted weighted mean. The weighted mean is dominated by chip 1 because its formal errors are smaller, but the paper does not test whether the co-rotation conclusion survives the alternative chip-2 zero point. This is a concrete, checkable vulnerability rather than a demonstrated error. If the alternative zero point leaves the orbit conclusion intact, the ACCEPT verdict is fully justified. Pending that check, I would move to CONDITIONAL rather than reject, because the measurement itself and the transparent uncertainty budget are valuable and the requested test uses only numbers already in Table 2. Agreement with the reader is partial: we identify the same underlying assumption, but I add the internal chip discrepancy as a specific place where that assumption may fail.","tokens_in":14178,"tokens_out":9634,"duration_ms":95266,"concrete_test":"Recompute And VI's orbital-pole distribution and orbit Monte Carlo (Section 5.1, Figure 5) using only the ACS chip-2 absolute PM from Table 2, (mu_alpha, mu_delta) = (-38±26, -63±22) microas/yr, instead of the weighted average (-1.6±12.3, -52.6±11.2); also run the chip-1-only case (9±14, -49±13). If the median disk-alignment angle and the fraction of co-orbiting realizations remain consistent, within about 1σ, with the printed conclusion in both cases, the zero-point discrepancy does not invalidate the central claim.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Absolute proper motions and all orbit conclusions depend on the Section 4 background-galaxy zero point. The zero point assumes that galaxy centers measured with the WFPC2 DL model (which for F450W uses the F555W model and PSF library, Sec. 2.1) and the ACS hst1pass stellar PSF are unbiased between epochs and filters. A filter- or morphology-dependent centroid shift would appear directly as a spurious proper motion, and the paper adds no systematic error for this. Table 2 provides a check that is already in hand: for And VI, the two independent ACS-chip zero points are (mu_cor_alpha, mu_cor_delta)_1 = (-27±12, 72±11) and _2 = (34±25, 69±21) microas/yr. The RA values disagree by 61±28 microas/yr (~2.2σ), while the adopted mean RA proper motion is only -1.6±12.3 microas/yr. Because And VI's disk alignment depends on the direction of its ~50 microas/yr PM vector, a 40 microas/yr RA shift, well within the chip-2 solution's uncertainty, rotates the orbital pole and could weaken the central co-planar, co-rotating claim. And V is less affected because its orbit is already weakly constrained and the text hedges accordingly.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper measures absolute proper motions of the M31 dwarf satellites Andromeda V and Andromeda VI using HST WFPC2 (epoch ~2000) and ACS/WFC (epoch ~2020) images spanning a 20-year baseline. The WFPC2 data are processed with a deep-learning centering model and updated astrometric calibrations; the absolute zero-point is set by roughly 100 background galaxies per field. The resulting proper motions are combined with published distances, line-of-sight velocities, and two M31 proper-motion determinations to integrate orbits under two M31 mass models. The authors report that And VI is consistent, within uncertainties, with a co-planar, co-rotating orbit around M31's disk, while And V is consistent (with large errors) with a counter-orbiting alignment. They also find both satellites are well bound to M31 and that And VI remains beyond ~90 kpc, implying weak tidal effects.","tokens_in":14489,"tokens_out":4355,"duration_ms":40168,"significance":"If the measurements are correct, this is a valuable observational contribution: it increases the sample of M31 satellites with measured proper motions to six, adds the most distant satellite (And VI, ~280 kpc) to that sample, and provides the first orbit determinations for two satellites far from the Great Plane of Andromeda. The paper is careful with random-error propagation (Monte Carlo orbit realizations), uses an external reference frame (background galaxies) for the absolute zero-point, and makes the data sets available via MAST DOIs. The orbit analysis explores multiple M31 mass models and proper-motion choices, and the conclusions are appropriately hedged for And V. The central results, if robust, bear on the kinematics of M31's satellite system, the lopsidedness of the satellite distribution, and the tidal history of And VI.","major_comments":[{"comment":"The absolute zero-point rests on the assumption that background-galaxy centroids measured with a stellar PSF model (hst1pass for ACS, the deep-learning model for WFPC2) are unbiased between the 2000 WFPC2 and 2020 ACS epochs and across filters. The manuscript presents no test of this assumption and assigns no systematic error to it. The data in Table 2 already contain a check: for And VI, the two ACS-chip zero-points are (µ_cor_alpha, µ_cor_delta) = (−27±12, 72±11) and (34±25, 69±21) µas/yr, which disagree in RA by 61±28 µas/yr (≈2.2σ), while the adopted absolute RA proper motion is only −1.6±12.3 µas/yr. Because And VI's disk-alignment conclusion depends on the direction of its ~50 µas/yr proper-motion vector, a shift of this size—well within the chip-2 solution's uncertainty—can rotate the orbital pole and weaken the co-planar, co-rotating claim. I request either a quantitative systematic-error term for the galaxy-centroid zero-point or a validation (e.g., splitting galaxies by morphology or color, or using the available Gaia EDR3 stars as a cross-check), and a discussion of the chip discrepancy.","section":"§4, Table 2"},{"comment":"The F450W WFPC2 data are processed with the F555W deep-learning centering model and PSF library (Section 2.1), while the ACS data use filter-specific models. For unresolved stars this may be acceptable, but for background galaxies—which are the sole absolute reference—a filter- or morphology-dependent centroid offset between epochs would appear directly as a spurious proper motion. The paper notes only that random positional errors for F450W are a few percent larger; it does not address the possibility of a systematic bias. This is a load-bearing point for the absolute proper motions and should be investigated or explicitly argued to be negligible.","section":"§2.1, §4"}],"minor_comments":[{"comment":"The caption reads 'Same an in Fig. 6' and should be corrected to 'Same as in Fig. 6'.","section":"Figure 7 caption"},{"comment":"The word 'immediatly' in the summary is a typo and should be 'immediately'.","section":"§6"},{"comment":"The sentence 'The final zero-point correction is a weighted average of all galaxies with total proper-motion values less than 2 mas yr−1' should specify whether the cut is on the quadrature sum of the two proper-motion components or on each component separately.","section":"§4"},{"comment":"The statement that the choice of M31 mass and proper motion 'does not strongly change the expected orbital characteristics' is followed by pericenter-time ranges of 1.5–2.0 Gyr for And VI; please quantify what 'not strongly' means in the text.","section":"§5.2"},{"comment":"The keyword list appears incomplete; consider adding 'dwarf galaxies' and 'galaxy kinematics' to improve discoverability.","section":"Keywords"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid observational contribution with good data availability and a careful treatment of random errors. The main concern is the unquantified systematic in the absolute zero-point, which is load-bearing for the orbit conclusions. The chip discrepancy in Table 2 should be addressed explicitly. I recommend major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First things first: this is a solid, incremental but genuinely useful astrometry paper. It adds the first proper-motion measurements for And V and And VI, bringing the M31 sample with 3D velocities to six, and And VI is the most distant M31 satellite with an orbit. The measurements use a 20-year HST baseline, the astrometric calibration is detailed and mostly inherited from the group's earlier work, and the paper ships the data via MAST DOIs. The orbit analysis uses two M31 masses and two M31 proper motions, propagates uncertainties with Monte Carlo, and is properly hedged: And V's alignment with the disk is explicitly called not well constrained.\n\nThe main scientific payoff is And VI: its orbit is consistent with co-planar, co-rotating motion around M31's disk, it stays beyond ~90 kpc at pericenter, and the low tidal influence fits with Pickett et al.'s cuspy mass profile. That's a nice, concrete connection.\n\nNow the soft spot, which is real but not fatal. The absolute proper motions rest entirely on background galaxies for the zero point, and the paper gives no systematic error for galaxy-centroid bias between WFPC2's DL model and ACS's hst1pass. Table 2 provides an internal check: for And VI the two ACS-chip zero points in RA are -27±12 and +34±25 μas/yr, a ~2.2σ disagreement. The adopted RA proper motion is only -1.6±12.3 μas/yr, so the zero point essentially sets that value. A 40 μas/yr shift, well within the chip-2 uncertainty, rotates the PM vector and could weaken the co-rotation claim. The declination zero points agree well, which is reassuring, but the paper should acknowledge this directly and add a systematic term. And V is less affected because its orbit is already weakly constrained and the text says so.\n\nAlso minor: using the F555W DL model and PSF library for F450W is a reasonable shortcut but deserves a sentence about possible filter-dependent bias; the authors cite their own previous work for the method, which is fine since the method is published and tested.\n\nBottom line: the paper is not overclaiming, the measurements are the sort of thing the small M31-satellite community needs, and the internal inconsistency is worth a revision, not a rejection. I'd send it to a serious referee. For a reading group, it's a good example of careful HST astrometry and how to handle a delicate zero point.","headline":"First PMs for And V/VI; And VI's zero point has an internal 2.2σ chip discrepancy that should be addressed before relying on the co-rotation claim.","tokens_in":15056,"tokens_out":2524,"would_cite":true,"duration_ms":24436,"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":"The paper measures absolute proper motions for two M31 dwarf satellites and derives their orbits, finding And VI co-orbits the disk while And V counter-orbits.","keywords":["proper motions","dwarf galaxies","Andromeda V","Andromeda VI","HST astrometry","satellite planes","orbit integration","background galaxies"],"falsifier":"Compare the galaxy-based absolute proper motions with those from a future Gaia data release whose faint-end errors reach about 0.05-0.1 mas/yr for the same fields; a discrepancy larger than the quoted ~10-20 microarcsecond-per-year errors would falsify the galaxy calibration, as would a test with synthetic galaxies inserted into the WFPC2 and ACS frames showing systematic epoch- or filter-dependent centroid offsets.","tokens_in":14008,"feed_emoji":"🔭","tokens_out":7257,"duration_ms":54454,"temperature":0.7,"pith_summary":"This paper measures the absolute proper motions of two dwarf satellite galaxies of Andromeda (M31): Andromeda V and Andromeda VI/Pegasus, using Hubble Space Telescope images taken about 20 years apart. The motions are converted from relative to absolute using roughly 100 background galaxies per satellite as a fixed reference frame. From these motions, combined with line-of-sight velocities and updated distances, the paper computes orbits around M31 and finds that And VI is consistent with co-orbiting along M31's disk plane, while And V is consistent with counter-orbiting, though with larger uncertainties. And VI, the farthest M31 satellite with a measured proper motion, appears to stay beyond about 90 kpc from M31, implying weak tidal influence over its orbit.","feed_headline":"Two M31 dwarf orbits: co-spinning and counter-spinning","feed_subtitle":"HST proper motions reveal And VI circling with the disk while And V moves against it.","key_machinery":"The method rests on converting a relative proper motion into an absolute one using background galaxies as zero-point reference objects. For each satellite, early-epoch WFPC2 images from 2000 and late-epoch ACS/WFC images from 2020 are processed with the hst1pass code, and WFPC2 positions are refined by a deep-learning centering model that overcomes undersampling. Chip-to-chip and epoch-to-epoch transformations use polynomial fits, and the absolute proper motion is obtained as the mean motion of satellite members minus the weighted mean motion of roughly 100 galaxies per satellite. Orbit parameters come from integrating backward in time with Monte Carlo draws over distance, line-of-sight velocity, and proper-motion uncertainties, using two M31 potentials.","core_discovery":"The paper reports two absolute proper motions: for And V, $(\\mu_\\alpha, \\mu_\\delta) = (26.1 \\pm 21.5, -74.2 \\pm 19.1)\\ \\mu\\mathrm{as}\\,\\mathrm{yr}^{-1}$, and for And VI, $(\\mu_\\alpha, \\mu_\\delta) = (-1.6 \\pm 12.3, -52.6 \\pm 11.2)\\ \\mu\\mathrm{as}\\,\\mathrm{yr}^{-1}$. These are the first proper motions for M31 satellites that lie far from the Great Plane of Andromeda but close to the M31 disk plane. Orbit integrations using two M31 mass models and two M31 proper motions show that And VI's orbit is well aligned with the M31 disk and co-rotates with it, while And V's orbit is consistent with a counter-rotating alignment, though less tightly constrained. Both satellites are bound to M31, and And VI remains at distances of at least about 90 kpc, supporting the interpretation that its cuspy central mass profile is preserved by a weak tidal field.","pith_inferences":["If And VI genuinely co-orbits the disk plane at about 280 kpc, the disk-aligned population of M31 satellites extends far beyond the Great Plane; a testable prediction is that other distant dwarf spheroidals near the disk plane will show the same co-rotation.","The deep-learning centering of undersampled WFPC2 images could be applied to other archival HST two-epoch pairs, potentially yielding proper motions for many faint Local Group systems.","A confirmed counter-orbiting And V would imply that disk-plane orbits around M31 are not all coherent in sense, complicating the single-rotating-plane picture.","The zero-point precision is limited by galaxy centroid scatter; using a larger or morphologically curated galaxy sample could push the absolute proper motions below the current 10-20 microarcsecond-per-year level."],"forward_implications":["The sample of M31 satellites with measured proper motions grows to six, including two objects outside the Great Plane of Andromeda but near the M31 disk.","And VI's orbit, which keeps it beyond about 90 kpc from M31, corroborates the idea that its steep central mass profile results from a weak tidal history.","Both satellites are well bound to M31, so they do not directly constrain M31's total mass, but they can be used in satellite-based mass constraints.","The last pericenters of both satellites occurred on the far side of M31, offering a possible clue to the observed lopsidedness of the M31 satellite system.","A future Gaia data release with improved faint-end precision could check the galaxy-based zero-point against the EDR3 stars identified in these fields."],"supporting_citations":[{"why":"Supplies the full processing and proper-motion methodology, including the prior measurement for And III.","marker":"Casetti-Dinescu et al. (2024b)"},{"why":"Provides the hst1pass code used for detections and positions in both cameras.","marker":"Anderson (2022)"},{"why":"Develops the deep-learning centering model used to refine WFPC2 positions.","marker":"Casetti-Dinescu et al. (2024a)"},{"why":"Gives the HST+Sats estimate of M31's proper motion used as one reference frame.","marker":"Sohn et al. (2020)"},{"why":"Provides the EDR3 measurement of M31's proper motion used in the weighted average.","marker":"Salomon et al. (2021)"},{"why":"Computes the weighted HST+Sats+EDR3 proper motion of M31 adopted as the fiducial frame.","marker":"Pawlowski & Sohn (2021)"},{"why":"Supplies the RR Lyrae-based distances for M31, And V, and And VI.","marker":"Savino et al. (2022)"},{"why":"Supplies the two M31 mass models used in the orbit integrations.","marker":"Patel et al. (2017)"},{"why":"Supplies the line-of-sight velocities of And V and And VI.","marker":"Collins et al. (2013)"},{"why":"Provides the adopted centers and structural parameters for the two satellites.","marker":"McConnachie (2012)"}],"fun_headline_variants":["M31 dwarfs: And VI co-rotates, And V counters","HST pinpoints orbits of Andromeda V and VI dwarfs","And V and VI: counter and co-orbiting M31 satellites","20-year HST data reveal dwarf orbits around M31"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Background galaxies have zero net motion and their centroids, measured with a stellar point-spread-function model, are unbiased between the 2000 WFPC2 and 2020 ACS epochs and filters.","fun_headline_variants_meta":{"raw":{"variants":["M31 dwarfs: And VI co-rotates, And V counters","HST pinpoints orbits of Andromeda V and VI dwarfs","And V and VI: counter and co-orbiting M31 satellites","20-year HST data reveal dwarf orbits around M31"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000552,"raw_usage":{"total_tokens":2726,"prompt_tokens":1130,"completion_tokens":1596,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":746,"completion_tokens_details":{"reasoning_tokens":1519}},"tokens_in":746,"tokens_out":1596,"duration_ms":10710,"temperature":1.0,"reasoning_tokens":1519,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:18:42.010413+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the galaxy-based absolute proper motions with those from a future Gaia data release whose faint-end errors reach about 0.05-0.1 mas/yr for the same fields; a discrepancy larger than the quoted ~10-20 microarcsecond-per-year errors would falsify the galaxy calibration, as would a test with synthetic galaxies inserted into the WFPC2 and ACS frames showing systematic epoch- or filter-dependent centroid offsets.","supporting_citations":[{"cited_title":"2022, One-Pass HST Photometry with hst1pass, Instrument Science Report ACS 2022-02, ,","cited_arxiv_id":null,"evidence_quote":"Provides the hst1pass code used for detections and positions in both cameras."},{"cited_title":"T., Patel, E., Fardal, M","cited_arxiv_id":null,"evidence_quote":"Gives the HST+Sats estimate of M31's proper motion used as one reference frame."},{"cited_title":"B., Ibata, R., Reyl´ e, C., et al","cited_arxiv_id":null,"evidence_quote":"Provides the EDR3 measurement of M31's proper motion used in the weighted average."},{"cited_title":"S., & Sohn, S","cited_arxiv_id":null,"evidence_quote":"Computes the weighted HST+Sats+EDR3 proper motion of M31 adopted as the fiducial frame."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the line-of-sight velocities of And V and And VI."}],"review_version":1}