Gaia Proper Motion Anomalies

• Gaia proper motion anomalies are statistically significant deviations from expected linear motions, indicating potential unseen companions, reference-frame effects, or systematic catalog discrepancies.
• They are measured by comparing long-term Hipparcos–Gaia baselines and by employing vector spherical harmonic analyses to isolate astrophysical signals from systematic errors.
• Assessing these anomalies provides actionable insights for detecting stellar multiplicity, calibrating astrometric catalogs, and refining models of Galactic structure and dynamics.

Searching arXiv for relevant papers on Gaia proper motions, secular aberration, and proper motion anomalies. arxiv_search(query="Gaia proper motion anomaly secular aberration drift Gaia reference frame quasar proper motions", max_results=5, sort_by="relevance") Gaia proper motion anomalies are departures of measured proper motions from a linear, single-object, or frame-stationary expectation, and the term is used in several technically distinct but related senses in modern astrometry. In stellar applications, a proper motion anomaly is the difference between a long-term proper motion and a short-term catalog proper motion, often interpreted as the reflex signature of an unseen companion (Kervella et al., 2019). In reference-frame and cosmological work, the same phrase denotes coherent large-scale patterns such as dipoles, rotations, glides, and secular aberration drift in quasar or galaxy proper motions (Bachchan et al., 2015). In Galactic dynamics, Gaia proper motions test whether spatial alignments are dynamically meaningful or instead kinematic outliers, as in the case of the Vast Polar Structure of the Milky Way (Pawlowski, 2017). Across these uses, the subject connects Gaia’s astrometric reference frame, the propagation of correlated uncertainties, and the distinction between astrophysical signals and catalog-level systematics.

1. Scope of the term

In the most restrictive stellar-dynamical sense, a proper motion anomaly is the offset between a long-baseline proper motion and a short-baseline proper motion. For Hipparcos–Gaia analyses, the long-term proper motion is derived from the Hipparcos and Gaia positions over a baseline of about 24.25–24.75 yr, while the short-term proper motions are those reported by Hipparcos or Gaia over their own observing windows; any significant mismatch indicates non-linear photocenter motion and is interpreted as evidence for a companion (Kervella et al., 2019). This framework was first applied to Classical Cepheids and RR Lyrae stars and then generalized to nearby stars, where the combination of Hipparcos and Gaia DR2 or EDR3 yields tangential-velocity anomaly sensitivities down to the planetary regime for favorable targets (Kervella et al., 2018).

In a broader catalog and reference-frame sense, Gaia proper motion anomalies are systematic patterns in the proper-motion field that arise from frame spin, glide, secular aberration drift, sky non-uniformity, or magnitude-dependent systematics. In this usage, an anomaly is not necessarily unexplained physics; it may be an expected relativistic or kinematic effect in a chosen reference system, such as the dipolar proper-motion field produced by the acceleration of the Solar System barycentre (Brown et al., 5 Mar 2025). The same category includes differential rotations between Gaia and ground-based proper-motion systems at levels from $0.2$ to $2.9$ mas yr$^{-1}$, and the strong magnitude dependence found in HSOY and GPS1 relative to Gaia EDR3 (Akhmetov et al., 2021).

A third usage appears in Galactic structure studies. There, Gaia proper motions are “anomalous” if they violate a dynamical hypothesis motivated by spatial alignment. In the VPOS context, a satellite close to the VPOS plane but with a measured proper motion producing a large orbital misalignment angle would constitute a proper motion anomaly relative to the VPOS hypothesis (Pawlowski, 2017). In halo-star work, the concern is whether Gaia proper-motion systematics at faint magnitudes could mimic anisotropy or substructure; the consistency of Gaia eDR3 proper motions with halo anisotropy measurements and with the ACS/Mon structure argues against large kinematically significant anomalies in that regime (III et al., 2022).

Extragalactic applications add a further sense. True quasars and distant AGN should have negligible bulk proper motions, so significant Gaia proper motions in extragalactic sources are often interpreted as imitations produced by photocenter variability or transient events within Gaia’s resolution element, that is, a VIM effect (Khamitov et al., 2023). This suggests that “Gaia proper motion anomalies” is best treated as an umbrella term for statistically significant non-linear motions, coherent catalog-wide vector fields, or astrophysically induced apparent motions, with interpretation determined by context.

2. Astrometric formalism and diagnostic quantities

For resolved or unresolved Galactic objects, proper motions become physically interpretable when combined with distance and line-of-sight velocity. The standard conversion from angular to tangential velocity is

$$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$

with $\mu$ in mas yr$^{-1}$ and $D$ in kpc (Pawlowski, 2017). In equatorial coordinates, the transverse velocity components are built from $\mu_\alpha \cos\delta$ and $\mu_\delta$, then combined with $v_{\rm los}$ to yield the full 3D velocity. In Galactic or spherical analyses this enables computation of cylindrical or spherical velocity moments, anisotropy parameters, or orbital poles (III et al., 2022).

In companion searches, the key observable is the proper motion anomaly vector,

$2.9$0

or, in the EDR3 implementation,

$2.9$1

where $2.9$2 is the long-term Hipparcos–Gaia proper motion (Kervella et al., 2018, Kervella et al., 2021). Converting this vector to a tangential-velocity anomaly gives

$2.9$3

with parallax $2.9$4 in mas (Kervella et al., 2019). The significance of the anomaly is expressed through a signal-to-noise ratio, written in companion surveys as $2.9$5 or $2.9$6, and secure detections are typically associated with thresholds such as $2.9$7 or $2.9$8 (Kervella et al., 2019, Kervella et al., 2021).

For dynamical alignments in the Milky Way, the central quantity is the orbital angular momentum,

$2.9$9

and the alignment angle with a preferred plane is obtained from

$^{-1}$0

This permits direct comparison between predicted minimum alignment angles and measured angles from proper motions, making it possible to define a proper motion anomaly relative to a specific orbital-plane hypothesis (Pawlowski, 2017).

In reference-frame work, the proper-motion field of quasars is modeled by frame spin plus acceleration. With tangent vectors $^{-1}$1 and $^{-1}$2, the proper motions take the form

$^{-1}$3

so the fitted field is a degree-1 vector spherical harmonic expansion comprising a rigid rotation and a dipole (Bachchan et al., 2015). In catalog-comparison analyses, systematic differences between proper-motion systems are parameterized as a solid-body rotation plus a glide, using

$^{-1}$4

and

$^{-1}$5

expanded in $^{-1}$6 and $^{-1}$7 (Akhmetov et al., 2021). This formalism distinguishes object-level anomalies from frame-level ones.

3. Companions, multiplicity, and acceleration signatures

The most developed use of Gaia proper motion anomalies is the detection of stellar and substellar companions. In nearby stars within 50 pc, combining Hipparcos and Gaia DR2 yielded tangential-velocity anomaly measurements with a median accuracy of $^{-1}$8 m/s per parsec of distance, and $^{-1}$9 of the stars studied showed a PMa greater than $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$0 sigmas (Kervella et al., 2018). The PMa constrains the normalized companion mass $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$1, and after accounting for inclination and temporal smearing it maps into a locus in companion-mass–separation space rather than a unique solution (Kervella et al., 2018).

The same method applied to Classical Cepheids and RR Lyrae stars established proper motion anomaly as a large-scale multiplicity diagnostic. Among 254 tested Classical Cepheids, 57 binaries were identified from PMa and 75 additional candidates were found; among 198 tested RR Lyrae stars, 13 showed a significant PMa and 61 additional candidates were identified (Kervella et al., 2019). This implied that the binary fraction of Cepheids is likely above $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$2, while that of RR Lyrae stars is at least $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$3 (Kervella et al., 2019). For systems with spectroscopic orbits, combining PMa with radial velocities yielded inclinations, nodes, and companion masses, as demonstrated for V1334 Cyg and TU UMa (Kervella et al., 2019).

With Gaia EDR3, the method was extended to a much larger multiplicity census. For 117,955 Hipparcos stars, 37,515 showed a significant PMa with $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$4, corresponding to $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$5, and 12,914 hosted common-proper-motion bound candidate companions (Kervella et al., 2021). Including RUWE$$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$6 as an additional binarity indicator, 50,720 Hipparcos stars, or $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$7, exhibited at least one signal of multiplicity (Kervella et al., 2021). The median tangential-velocity anomaly accuracy improved to $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$8 cm/s per parsec of distance, a factor $$v_{\rm tan} = 4.74\,\mu\,D \quad \text{(km s}^{-1}\text{)},$$9 better than DR2, pushing the PMa sensitivity into the planetary-mass regime for many nearby stars (Kervella et al., 2021).

The method has also become a pre-selection tool in exoplanet host characterization. A volume-limited sample to 300 pc of 66 stars hosting planets and planet candidates from Kepler, K2, and TESS with significant Hipparcos–Gaia proper motion anomalies was assembled to study S-type planetary architectures in binaries (Zhang et al., 2023). In that sample, TOIs with significant proper motion anomalies showed nearly four times more false positives due to eclipsing binaries than TOIs with marginal proper motion anomalies, and TOIs with significant anomalies exhibited lower Gaia differential velocities than field stars with significant anomalies (Zhang et al., 2023). This suggests that planets are more likely to form or survive in binaries with low-mass substellar companions or stellar companions at wider separation (Zhang et al., 2023). A detailed orbit fit for LTT 1445 ABC further showed a mutual inclination of $\mu$0 deg between the orbit of BC around A and that of C around B, implying a nearly flat architecture (Zhang et al., 2023).

4. Galactic dynamics, halo structure, and satellite systems

In Galactic-structure work, Gaia proper motions determine whether positional alignments correspond to coherent dynamics. The VPOS provides a prominent example. The Milky Way satellites, distant globular clusters, and about half of distant stellar and gaseous streams define a highly flattened, approximately polar configuration whose orbital poles cluster near the normal of the fitted plane (Pawlowski, 2017). Proper motions of the 11 classical satellites indicate that 8 of the 11 are consistent with orbiting along the VPOS, with Sculptor as a notable counter-rotator (Pawlowski, 2017). In this setting, a proper motion anomaly means a large mismatch between the measured orbital-plane angle $\mu$1 and the predicted minimum angle $\mu$2 under the hypothesis that satellites aligned with the VPOS also orbit within it (Pawlowski, 2017).

Gaia’s contribution in this context is mainly statistical. Using post-launch performance estimates, the paper adopted $\mu$3 for a single star at $\mu$4 mag, implying that for nearby systems such as Reticulum II at $\mu$5 kpc, combining a few to tens of member stars should constrain the systemic motion well enough for an object-by-object comparison with VPOS predictions (Pawlowski, 2017). For intermediate-distance satellites like Horologium I at $\mu$6 kpc, Gaia proper motions should enable statistical tests of alignment even if each individual orbit is poorly constrained, whereas for very distant systems like Eridanus II at $\mu$7 kpc Gaia should not provide meaningful constraints (Pawlowski, 2017). This suggests that Gaia proper motion anomalies in satellite systems are best interpreted collectively: either as reinforcement of coherent polar co-rotation, or as evidence that an apparent planar arrangement is transient or selection-driven.

Halo-star work addresses a different concern: whether Gaia proper-motion systematics can masquerade as anisotropy or substructure. Using 17,221 halo F-type stars after a $\mu$8 kpc cut, and Gaia eDR3 proper motions with full covariance propagation, the stellar halo was found to have $\mu$9 for $^{-1}$0 kpc, with mean radial and longitudinal velocities consistent with zero, $^{-1}$1 km s$^{-1}$2, velocity covariances consistent with zero, and tilt angles aligned with spherical coordinates (III et al., 2022). A coherent population with $^{-1}$3 km s$^{-1}$4 near $^{-1}$5 kpc was identified as the ACS/Mon structure, showing that what had previously looked like a dip in anisotropy is better understood as contamination by real substructure rather than a Gaia proper-motion artifact (III et al., 2022). The same paper noted that at $^{-1}$6 mag the mean proper-motion error corresponds to a tangential-velocity error of $^{-1}$7 km s$^{-1}$8, and that the agreement of the measured $^{-1}$9 with both simulations and other tracers argues against large, unmodeled Gaia proper-motion anomalies in this regime (III et al., 2022).

For ultra-faint dwarf galaxies, Gaia DR2 showed that mean proper motions can be recovered even in sparse systems if membership and covariance handling are done properly. Seven UFDs yielded mean motions with typical random uncertainties of $D$0 mas yr$D$1, and the analysis stressed that ignoring the full covariance matrix can induce systematic errors as large as $D$2 mas yr$D$3 in the mean proper motions (Massari et al., 2018). A recognized systematic floor of $D$4 mas yr$D$5 on small angular scales was added to the uncertainties (Massari et al., 2018). In this setting, an apparent proper motion anomaly may be produced not by Gaia itself but by small-number statistics or biased membership, as illustrated by Segue 2.

5. Extragalactic proper-motion fields and apparent motions

For quasars and galaxies, Gaia proper motion anomalies are usually interpreted as large-scale coherent fields or as apparent motions caused by source structure and variability. The dominant predicted large-scale field is the secular aberration drift arising from the acceleration of the Solar System barycentre. For a circular Galactic orbit with $D$6 km s$D$7 and $D$8 kpc, the acceleration amplitude is

$D$9

and the dipole amplitude at angular separation $\mu_\alpha \cos\delta$0 is

$\mu_\alpha \cos\delta$1

(Bachchan et al., 2015). Simulations with 500,000 quasars showed that Gaia can recover this acceleration to a few percent, with Case A yielding $\mu_\alpha \cos\delta$2as yr$\mu_\alpha \cos\delta$3 for an input of $\mu_\alpha \cos\delta$4as yr$\mu_\alpha \cos\delta$5, and that even extreme quasar photocenter variability of $\mu_\alpha \cos\delta$6as with $\mu_\alpha \cos\delta$7–10 yr produces negligible bias and only a modest increase in noise (Bachchan et al., 2015).

A related controversy concerns whether these effects should be removed from the Gaia catalog itself. Brown, Bastian, and Klioner argued that Gaia astrometry is, and should be, constructed in the Barycentric Celestial Reference System, that the secular aberration drift is a frame-dependent effect rather than a catalog error, and that transforming the catalog into a Galactic-rest or CMB-rest frame is neither necessary nor practically well-defined for stellar kinematics (Brown et al., 5 Mar 2025). In their treatment, the dipole in quasar proper motions is not an anomaly but the expected consequence of the acceleration of the Solar System barycentre in an SSB-based reference system (Brown et al., 5 Mar 2025).

The amplitude of Galactic aberration itself has also been treated as a calibration issue. An independent estimate from stellar astronomy gave a Galactic aberration constant of $\mu_\alpha \cos\delta$8as yr$\mu_\alpha \cos\delta$9, closer to Gaia EDR3’s $\mu_\delta$0as yr$\mu_\delta$1 than to the geodetic VLBI value $\mu_\delta$2as yr$\mu_\delta$3 (Malkin, 2023). This suggests that using the larger VLBI value to correct Gaia would over-correct the dipole and create artificial residuals (Malkin, 2023).

At low redshift, a second coherent extragalactic pattern is expected from the Solar System velocity relative to the CMB. Its amplitude is

$\mu_\delta$4

and is about $\mu_\delta$5–$\mu_\delta$6as yr$\mu_\delta$7 at $\mu_\delta$8 (Bachchan et al., 2015). With $\mu_\delta$9 low-redshift galaxies in $v_{\rm los}$0, simulations showed that Gaia could in principle recover $v_{\rm los}$1 with $v_{\rm los}$2 km s$v_{\rm los}$3 precision and $v_{\rm los}$4 with $v_{\rm los}$5–12 km s$v_{\rm los}$6 Mpc$v_{\rm los}$7 scatter under optimistic centroiding assumptions (Bachchan et al., 2015). A related DR2 study of secular extragalactic parallax obtained only an insignificant upper limit of $v_{\rm los}$8as yr$v_{\rm los}$9 Mpc, but predicted a $2.9$00–$2.9$01 detection by Gaia’s end of mission if nearby, bright galaxies are available (Paine et al., 2019). That same analysis found that peculiar-velocity fields induce low-multipole correlated proper motions on the order of $2.9$02as yr$2.9$03, implying a cosmological noise floor for future quadrupole searches (Paine et al., 2019).

Extragalactic “proper motion anomalies” can also be source-intrinsic in the sense of apparent photocenter motion. A sample of 248 eROSITA X-ray sources with spectroscopically confirmed extragalactic nature and significant Gaia eDR3 proper motions was interpreted as showing VIM-type imitations caused by transient events within Gaia’s resolution element (Khamitov et al., 2023). The catalog includes Seyfert 1, Seyfert 2, LINER, quasar, radio-galaxy, and star-forming-galaxy classes, and the modeled explanation is that flares within $2.9$04 mas shift the photocenter enough to mimic proper motions of $2.9$05 to $2.9$06 mas yr$2.9$07 (Khamitov et al., 2023). A FRED model with parameters such as $2.9$08, $2.9$09 days, and $2.9$10 days reproduces the observed $2.9$11 distribution for many sources, while the most luminous cases, such as a quasar requiring $2.9$12 erg s$2.9$13, suggest TDE-like events rather than ordinary supernovae (Khamitov et al., 2023).

6. Catalog systematics, cross-catalog anomalies, and mitigation

A major source of Gaia proper motion anomalies is not astrophysical but catalog-to-catalog inconsistency. HSOY, a hybrid Gaia DR1–PPMXL catalog with 583,001,653 entries, provided proper motions with formal precisions from significantly less than $2.9$14 mas/yr to $2.9$15 mas/yr, but explicitly inherited some of the zonal errors of PPMXL (Altmann et al., 2017). It dramatically reduced the spurious high-proper-motion population of PPMXL, yet still contained many artifacts, especially among high-$2.9$16 sources and in crowded regions (Altmann et al., 2017). This suggests that anomalies seen in Gaia–HSOY comparisons can reflect long-baseline astrophysical accelerations, but can also arise from mismatches, crowding, and inherited systematics.

A direct comparison of Gaia EDR3 with HSOY, UCAC5, GPS1, PMA, and TGAS formalized these discrepancies as differential rotation plus glide. The derived components ranged from $2.9$17 to $2.9$18 mas yr$2.9$19 depending on catalog and magnitude, and HSOY and GPS1 showed clear magnitude-dependent terms reaching several mas yr$2.9$20 at $2.9$21 (Akhmetov et al., 2021). By contrast, Gaia EDR3 itself was shown to have no rotation and glide relative to LQAC-5, ALLWISEAGN, and Milliquas extragalactic sources at the level of $2.9$22 mas yr$2.9$23 over $2.9$24, while PMA was the closest ground-based catalog to Gaia EDR3 over $2.9$25 at the level of $2.9$26 to $2.9$27 mas yr$2.9$28 (Akhmetov et al., 2021). This means that a Gaia–external-catalog proper-motion discrepancy of $2.9$29–$2.9$30 mas yr$2.9$31 at faint magnitudes is often better interpreted as a reference-frame anomaly than as an astrophysical acceleration.

At the bright end, comparisons of Hipparcos and Gaia reveal a different structure. A VSH decomposition of the Hipparcos–Gaia proper-motion difference field showed that Hipparcos proper motions are the dominant source of sky-correlated distortions in the multi-epoch optical CRF, with a median value of $2.9$32as yr$2.9$33 and a global spin of $2.9$34as yr$2.9$35, whereas Hipparcos positions and Gaia EDR3 proper motions are explicitly consistent by construction at a level of $2.9$36as yr$2.9$37 (Makarov, 2022). However, the Gaia–HG field still contains multiple higher-degree distortions, especially magnetic modes up to degree 7, which should be taken into account when using Hipparcos–Gaia long-term proper motions for astrometric binary searches (Makarov, 2022). This suggests that “Gaia proper motion anomalies” in Hipparcos–Gaia acceleration work are a combination of real orbital signals and residual medium- and small-scale structure in the bright-star proper-motion frame.

Mitigation strategies follow from these diagnoses. In UFD work, using the full covariance matrix and adding a systematic floor of $2.9$38 mas yr$2.9$39 avoids user-induced anomalies in mean proper motions (Massari et al., 2018). In halo-star analyses, bootstrap propagation with the full Gaia covariance matrix helps distinguish genuine substructure from correlated measurement errors (III et al., 2022). In HST–Gaia combinations, GaiaHub uses a long Gaia–HST baseline to reduce random errors at $2.9$40, and comparison of GaiaHub PMs with Gaia-only PMs exposes local systematic floors of order $2.9$41 km s$2.9$42 in $2.9$43 at the distances of the tested clusters, consistent with Gaia small-scale systematics on HST field scales (Pino et al., 2022). A plausible implication is that Gaia proper motion anomalies are most robustly interpreted when framed as a hierarchy: first reference-frame effects, then source-class-specific astrometric behavior, and only then novel astrophysics.