A purely acceleration-based measurement of the fundamental Galactic parameters

Jo Bovy∗ David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto, 50 St. George Street, Toronto, Ontario, M5S 3H4, Canada (Dated: July 29, 2021) Klioner et al. have used the Gaia EDR3 data to directly measure the solar system’s accelera- tion within the Milky Way using the apparent proper motions of quasars. Here I show that this single absolute acceleration measurement in combination with relative accelerations obtained from pulsar orbital decay allows one to determine all of the parameters describing the dynamics of our −1 local Galactic environment, including the circular velocity at the Sun V0 = 244 ± 8 km s and its 0 −1 −1 derivative V0 = 2 ± 9 km s kpc , the local angular frequency, the Oort constants, and the Sun’s motion with respect to the LSR. This is the first determination of these parameters that only uses the general theory of relativity without the need for additional assumptions.

Introduction—A good determination of the local gravi- where V0 and R0 are the local circular velocity and the tational and velocity fields near the Sun in the Milky Way distance to the Galactic center. This acceleration points is of fundamental importance to our understanding of the towards the Galactic center. local matter distribution [1]—including its dark matter Galactic gravitational fields are often expressed in content [2–4]—for correcting observations for the effects terms of the velocity of a circular orbit and the circu- of the Sun’s velocity and acceleration, for predicting the lar velocity at the Sun, V0, is one of the most funda- expected annual modulation signal in dark-matter de- mental parameters of the Milky Way. It is clear that tection experiments [5], and for many other applications the acceleration measurement in Equation (1) provides in astro(-particle )physics. However, our knowledge of a stringent constraint on V0, but to actually obtain V0 the gravitational field within any galaxy, including our (and all of the other fundamental parameters that I will own, has been fundamentally limited by the fact that discuss below), we need to combine Equation (1) with a gravitational theories (i.e., that of the general theory of measurement of the Sun’s distance to the Galactic cen- relativity, or its usually applicable Newtonian limit) only ter, R0. Luckily, we are fortunate enough to now have directly connect accelerations to the mass distribution, a measurement of R0 that essentially only employs fun- while we cannot measure the absolute galactic accelera- damental physics, because a highly precise measurement tion of any star. of R0 was obtained from post-Newtonian modeling of the This changed recently, because Klioner et al. [6] per- orbit of the star S2 around the Milky Way’s supermassive formed a precise and, more importantly, accurate deter- black hole. Precise measurements of S2’s orbit using the mination of the acceleration of the solar system barycen- GRAVITY instrument on the VLT have allowed its grav- ter with astrometric data from Gaia EDR3 [7, 8]. This itational redshift [12] and Schwarzschild precession [13] measurement uses the apparent proper-motion pattern of to be measured, allowing a precise determination of R0 distant quasars induced by the aberration effect that the [14], because R0 connects the observed motion on the sky acceleration gives rise to. This measurement agrees with with the observed line-of-sight velocity projection of the earlier determinations using Very Long Baseline Interfer- orbit. I use the value for R0 from the recent re-analysis ometry [9, 10], but is based on a set of reference sources of the GRAVITY data [15] that is orders of magnitude larger, allowing an analy- sis that is much freer of systematics. The Milky Way’s R0 = 8.275 0.034 kpc , (2) ± gravitational field is the largest contributor to the total acceleration [11] and at the current measurement pre- but note that the GRAVITY collaboration’s preferred cision, plausible non-axisymmetric contributions to the value for R0 changes by more than its uncertainties be- acceleration are too small to be detected. The largest tween different analyses (a situation addressed in Ref. arXiv:2012.02169v3 [astro-ph.GA] 28 Jul 2021 non-Galactic contribution comes from the Large Magel- [15]) and disagrees with the analysis of the gravitational- lanic Cloud [6], but it is only at the few percent level and redshift of S2 by Ref. [16], who find R0 = 7.95 0.06 kpc. ± almost perpendicular to the Galactic contribution. The Together, the measurements in Equations (1) and (2) al- Klioner et al. measurement is therefore a direct mea- low the local circular velocity to be determined surement of the centripetal acceleration at the Sun, with s magnitude a0, which I express as a proper motiona ˜ −1
R0 V0 = 244 8 km s . (3) 8.275 kpc a V 2 ± a˜ = 0 = 0 = 5.05 0.35 µas yr−1 , (1) c c R0 ± For the value of R0 from Ref. [16], the central value shifts −1 only to V0 = 239 km s . Another parameter that immediately follows is the an- ∗ [email protected] gular frequency of the circular orbit at the Sun, Ω0 = 2

6 )

1 5 − 4 as yr µ ( ˜ a 3

2 R 0 )& (2013) 0 . (R V c Oort constants IAU MWPotential2014McMillan (2017) Irrgang et al Dehnen & Binney (1998)

FIG. 1. Comparison of the Gaia EDR3 measurement of the local acceleration, expressed as a proper motiona ˜ (see Eqn. 1), with previous indirect determinations and popular Milky Way models.

V0/R0, which is constrained to be This discrepancy, if real, is an indication that the solar neighborhood as a whole may be perturbed away from a r −1 −1
8.275 kpc circular orbit by non-axisymmetric forces, at the level of Ω0 = 29.4 1.0 km s kpc . (4) 10 km s−1. ± R0 ≈ By combining the measurements of R0 and V0 in the However, we can do much more with these mea- introduction with the observed proper motion µSgr A∗ of surements if we combine them with a few other key the supermassive black hole at the center of the Milky measurements of velocities and relative accelerations in Way, Sgr A∗—assumed to be at rest at the Galactic the Milky Way and in fact determine essentially all fun- center—we can now directly determine the Sun’s peculiar damental parameters describing the local gravitational velocity, because and velocity field. I explore this in this brief paper. V While the resulting measurements do not currently µSgr A∗ = Ω0 + . (5) have high precision, they stand out as being the only R0 ones that do not assume dynamical equilibrium or any Combining the latest measurement of µSgr A∗ = 6.411 other method for connecting observed velocities to the 0.008 mas yr−1 [20] with the measurements from− the in-± gravitational field. troduction, we get

−1 The Sun’s motion with respect to the LSR—Generally V = 8.0 8.4 km s . (6) a nuisance parameter, the Sun’s peculiar velocity with re- ± spect to the circular orbit through its location (the LSR) This measurement is at the lower end of the previ- is an important parameter when studying the local ve- ously considered range and it agrees with the accepted −1 locity distribution and its gradients and is, for better or V = 12 2 km s from Ref. [18]. It is about 2σ ± −1 worse, often used as part of the correction of observed ve- off from the global value of V 25 km s from Ref. ≈ locities to the Galactocentric reference frame (this is, in [2, 19]. The uncertainty in V is largely driven by fact, unnecessary; however, it is often done). While in the the uncertainty ina ˜ and it should therefore improve radial and vertical directions, the assumption of axisym- significantly with future Gaia data releases, which will metry allows the Sun’s motion to be determined simply allow us to get to the root of the discrepancy between as the opposite of the mean velocity of local stars in those local and global values. directions—because on average these velocities should be zero in the Galactocentric frame—the asymmetric drift The local acceleration compared to previous constraints makes the determination of the Sun’s peculiar velocity in and models—Before continuing the determination of the the direction of Galactic rotation, V , much trickier [17]. local fundamental dynamical parameters, we briefly dis- The generally accepted value derived from observations cuss how the observed acceleration compares to that −1 of local stars is V 12 km s faster than Galactic ro- derived from earlier determinations of V0 and R0 and tation [18]. However,≈ stellar velocities on larger scales also how often-used models for the Milky Way’s grav- −1 are most easily understood if V 25 km s [2, 19]. itational potential predicted the observed acceleration. ≈ 3

300

) 250 1 −

200

) (km s Observed pulsar data R (

c Gaia EDR3 a˜ V 150 MWPotential2014 McMillan (2017) 1 1 1 100 V = 242 7 km s− + (2 9 km s− kpc− )(R R ) c ± ± − 0 7.0 7.5 8.0 8.5 9.0 R (kpc)

FIG. 2. Local circular velocity curve determined solely from observed absolute accelerations for the solar system (Gaia EDR3; black point) and 6 pulsars relatively close to the Galactic mid-plane whose relative acceleration obtained from pulsar timing can be turned into an absolute determination using the solar system’s acceleration. The rotation curve is approximately flat over the 2 kpc surrounding the solar radius. Also shown is a linear fit (red curve) and the predicted rotation curve for two popular models for the Milky Way’s gravitational field.

This is shown in Fig. 1. I compare there to the pa- Using pulsar orbital decay to directly measure the radial rameters still recommended by the IAU [21], which have dependence of the Galactic acceleration—So far I have −1 V0 = 220 20 km s and R0 = 8.5 1.0 kpc, but also to used only the single, new local acceleration measurement the value± implied by the measurement± of the local Oort from Gaia EDR3. However, while no absolute measure- constants using Gaia DR1 [22] after combining it with ments of the acceleration of any star existed before, mea- the measurement of R0 from Equation (2); these both surements of the relative acceleration of a small number imply much lower values ofa ˜ 4 to 4.3 µas yr−1 com- of stars with respect to the solar system barycenter did pared to the observed value, indicating≈ for the Oort con- exist. These derive from the fact that the observed period stants that they are likely affected by non-axisymmetric changes of binary pulsars have a contribution from the streaming motions. There have been, of course, dozens, relative acceleration of the pulsar and observer reference if not more, determinations of V0 in the recent past that frames (in addition to the, often considered more inter- could be cast as measurements ofa ˜ and I do not intend esting, gravitational-wave contribution) [28, 29]. As re- to review all of them here. Suffice it to say that the ob- cently emphasized by Chakrabarti et al. [30] and Philips served value ofa ˜ is at the high end of that implied by et al. [31], assuming that the general theory of relativity previous measurements of V0. holds and, thus, that its contribution to the observed pe- riod changes can be determined (albeit with uncertain- We also compare to some standard models of the Milky ties), the observed period changes can be turned into Way’s gravitational field: the MWPotential2014 model −1 measurements of the relative Galactic acceleration at the from Ref. [23] (assuming an uncertainty of 10 km s on binary pulsar’s location and the Sun. Chakrabarti et al. V0 and 0.1 kpc in R0), the model from McMillan (2017) used such relative accelerations derived from observations [24], model I from Irrgang et al. (2013) [25] (which is sim- of 14 binary pulsars to constrain the local gravitational ilar to the often used model of Allen & Santillan [26]), field, while Philips et al. use both spin and orbital peri- and the range predicted by models I though IV of Dehnen ods to do the same. Relative accelerations are a useful & Binney (1998) [27]. All of these models, except for that constraint on their own, but Gaia’s measurement of the of Irrgang et al., predict values fora ˜ that are smaller than local acceleration itself allows these relative accelerations the observed value, although within the current level of to be turned into a set of absolute accelerations covering uncertainty, the tension is not very large. We can con- a range of distances from the Sun. Complementary to clude that for the current measurement, Galactic models the analysis of Chakrabarti et al., who focus largely on are at minor variance with the directly determined ac- constraining the vertical gravitational field because they celeration and that future improvement in the direct de- are limited to relative accelerations (and the Sun’s ver- termination will be a valuable constraint on global Milky tical acceleration is approximately zero), I use the abso- Way models. 4 lute accelerations here to determine the radial behavior of distance, because its sign is wrong. Determining a more the acceleration or, if you will, the local circular velocity precise distance to the Hulse-Taylor pulsar would be use- curve. ful, because as the binary pulsar at the largest distance, Chakrabarti et al. [30] compiled observed periods and it would provide a strong constraint on the Galactic po- period changes, parallaxes and proper motions (neces- tential. sary for a purely kinematic contribution to the observed To determine the local rotation curve, I focus on those period change [32]), and period changes expected from pulsars for which the contribution from the vertical accel- the general theory of relativity, for a sample of field bi- eration to the line-of-sight projection is negligible com- nary pulsars. I use their compilation, but remove four pared to that from the radial acceleration. Specifically, pulsars with such large measurement uncertainties that I select those quasars for which the relevant term in they do not meaningfully constrain Galactic radial accel- the projection involving the vertical acceleration is at erations (J0737-3039A/B, J1012+5307, J1603-7202, and least five times smaller than the relevant terms involving J2129-5721) and I add the famous Hulse-Taylor pulsar, the radial acceleration, assuming the MWPotential2014 B1913+16, which is the only pulsar useful for this exer- model (however, the cut does not significantly depend on cise at multiple-kpc distances from the Sun. Its parame- what potential model is assumed). This selects 6 pulsars ters are obtained from Ref. [29], but I use the geometric for which I can then determine the radial acceleration and parallax and proper motion from Ref. [33]. However, its implied circular velocity at the radius of the pulsar. we will find that for the current value and uncertainty To account for the slight decrease of the circular velocity of the parallax, the Hulse-Taylor pulsar is in strong dis- with height for these pulsars that are a few hundred par- agreement with the expected Galactic acceleration (to a sec from the plane, I correct the circular velocities using degree that cannot be explained any plausible model), so Eqn. (15) from Ref. [2] using a scale length of 2.5 kpc it will not be useful for constraining the local dynamical (note, however, that this correction amounts to only a parameters. But it’s fun and illustrative to include it. few km s−1). The resulting values are shown in Fig. 2. The observed period change P˙b/Pb (where Pb is the It is clear that while the uncertainties are large, the ro- orbital period) is given by [28] tation curve is consistent with being flat over the 2 kpc range surrounding the solar circle. ! 2 ! P˙b 1 µ D P˙b We can determine the local slope of the rotation curve = ~x (~a1 ~a0) + + , (7) by fitting a simple linear model to these data, which is Pb c · − c Pb obs GR shown by the red line in Fig. 2. This gives a local slope of where D is the distance to the binary pulsar, µ is the mag- nitude of its proper motion, ~x is the unit vector pointing dV V 0 c = 2 9 km s−1 kpc−1 , (8) to the pulsar from the solar system barycenter, and ~a1 0 ≡ dR ± and ~a0 are the Galactic acceleration vectors at the loca- tion of the binary pulsar and the Sun, respectively. which is consistent with being flat. The dependence We can then compute the line-of-sight acceleration of this measurement on the assumed value of R0 from ~x ~a1 from the pulsar observations (I additionally have to Equation (2) is very small; even changing R0 to 7.7 assume· a value for the Sun’s offset from the Galactic mid- kpc does not change it. With this measurement, we plane, for which I use Z = 20.8 pc from Ref. [34]), GR can complete our determination of the fundamental effects, and using the measurement of ~a0/c from Equa- Galactic parameters describing the local gravitational tion (1) (note that I assume axisymmetry, see above). field. Traditionally, the local velocity field has been Comparing the computed line-of-sight accelerations for described using the so-called Oort constants A and B the 11 pulsars to the predicted line-of-sight accelerations [22, 35], which describe the local shear and vorticity of in two popular models for the Milky Way’s gravitational the velocity field and which are directly related to the field, the MWPotential2014 model from Ref. [23] and local angular frequency Ω0 and the slope of the rotation 0 the model McMillan (2017; [24]) reveals excellent agree- curve V0 . The implied measurements for the Oort ment. The sole exception is the Hulse-Taylor pulsar constants are shown together with all of the parameters (B1913+16), which deviates significantly from the ob- that I discussed in Table I. served value. However, the observed line-of-sight accel- eration is strongly influenced by the assumed distance to Conclusion—Gaia EDR3’s direct measurement of the pulsar and this distance is highly uncertain. I have the acceleration of the solar system within the Milky used the recent distance derived from a trigonometric Way galaxy from the apparent proper motion pattern parallax [33], which placed the pulsar at a surprisingly of distant quasars that it induces, is a revolutionary small distance, and the discrepancy between the Galac- moment for Galactic astrophysics. As I demonstrated tic acceleration derived using this distance (taking into here, this single measurement allows us to directly infer account its uncertainty) and the expected value shows most of the fundamental parameters describing the local that the distance is likely much larger. Note that there gravitational and velocity fields using only geometric isn’t any plausible gravitational field that would make measurements or measurements that only assume the determined acceleration make sense at its assumed fundamental physics (essentially, the general theory 5

[34, 36–39] and thus that any assumption-based inference TABLE I. Values of the fundamental Galactic parameters de- rived from geometric or fundamental-physics measurements of the local dark and visible matter distribution based of distances, velocities, and accelerations. on it is suspect [40], such assumption-free, fundamental- physics measurements of the local gravitational field will Parameter (unit) Measurement Reference be of great importance in furthering our understand- −1 −1 a0 (km s Myr ) −7.34 ± 0.51 Klioner et al. (2020) ing of the local matter distribution and its perturbations. R0 (kpc) 8.275 ± 0.034 Gravity collab. et al. −1 V0 (km s ) 244 ± 8 Bovy (2020; this paper) −1 −1 Ω0 (km s kpc ) 29.4 ± 1.0 ” 0 −1 −1 V0 (km s kpc ) 2 ± 9 ” A (km s−1 kpc−1) 13.8 ± 4.7 ” B (km s−1 kpc−1) −15.6 ± 4.7 ” −1 V (km s ) 8.0 ± 8.4 ” ACKNOWLEDGMENTS

JB acknowledges financial support from NSERC (fund- of relativity), without the usual need for additional ing reference number RGPIN-2020-04712) and an On- astrophysical assumptions. These determinations are tario Early Researcher Award (ER16-12-061). This work summarized in Table I. While the current precision in has made use of data from the European Space Agency these measurements is not that high, this should improve (ESA) mission Gaia (https://www.cosmos.esa.int/ with future measurements of the local acceleration using gaia), processed by the Gaia Data Processing and Analy- Gaia data, which should get better by a factor of 3 in sis Consortium (DPAC, https://www.cosmos.esa.int/ Gaia’s fourth data release and 8 in its fifth [6].≈ Given web/gaia/dpac/consortium). Funding for the DPAC that the analysis of Gaia DR2≈ paper has conclusively has been provided by national institutions, in particu- demonstrated that the local velocity field is strongly lar the institutions participating in the Gaia Multilateral affected by non-axisymmetry in the Galactic potential Agreement.

[1] J. Bland-Hawthorn and O. Gerhard, The Galaxy in (2020). Context: Structural, Kinematic, and Integrated Prop- [8] Lindegren, Lennart, Klioner, S. A., Hern´andez, J., et al., erties, Ann. Rev. Astron. Astrophys. 54, 529 (2016), Gaia early data release 3. the astrometric solution, As- arXiv:1602.07702 [astro-ph.GA]. tron. Astrophys. 10.1051/0004-6361/202039709 (2020). [2] J. Bovy, C. Allende Prieto, T. C. Beers, D. Bizyaev, L. N. [9] O. Titov, S. B. Lambert, and A. M. Gontier, VLBI mea- da Costa, K. Cunha, G. L. Ebelke, D. J. Eisenstein, P. M. surement of the secular aberration drift, Astron. Astro- Frinchaboy, A. E. Garc´ıaP´erez,L. Girardi, F. R. Hearty, phys. 529, A91 (2011), arXiv:1009.3698 [astro-ph.CO]. D. W. Hogg, J. Holtzman, M. A. G. Maia, S. R. Majew- [10] O. Titov and H. Kr´asn´a,Measurement of the solar system ski, E. Malanushenko, V. Malanushenko, S. M´esz´aros, acceleration using the Earth scale factor, Astron. Astro- D. L. Nidever, R. W. O’Connell, C. O’Donnell, phys. 610, A36 (2018), arXiv:1802.05347 [astro-ph.GA]. A. Oravetz, K. Pan, H. J. Rocha-Pinto, R. P. Schi- [11] R. K. Bachchan, D. Hobbs, and L. Lindegren, Gaia ref- avon, D. P. Schneider, M. Schultheis, M. Skrutskie, V. V. erence frame amid quasar variability and proper mo- Smith, D. H. Weinberg, J. C. Wilson, and G. Zasowski, tion patterns in the data, Astron. Astrophys. 589, A71 The Milky Way’s Circular-velocity Curve between 4 and (2016), arXiv:1512.08917 [astro-ph.IM]. 14 kpc from APOGEE data, Astrophys. J. 759, 131 [12] Gravity Collaboration, R. Abuter, A. Amorim, (2012), arXiv:1209.0759 [astro-ph.GA]. N. Anugu, M. Baub¨ock, M. Benisty, J. P. Berger, [3] L. E. Strigari, Galactic searches for dark matter, Phys. N. Blind, H. Bonnet, W. Brandner, A. Buron, C. Collin, Rep. 531, 1 (2013), arXiv:1211.7090 [astro-ph.CO]. F. Chapron, Y. Cl´enet, V. Coud´eDu Foresto, et al., [4] J. I. Read, The local dark matter density, Jour- Detection of the gravitational redshift in the orbit of nal of Physics G Nuclear Physics 41, 063101 (2014), the star S2 near the Galactic centre massive black hole, arXiv:1404.1938 [astro-ph.GA]. Astron. Astrophys. 615, L15 (2018), arXiv:1807.09409 [5] K. Freese, M. Lisanti, and C. Savage, Colloquium: An- [astro-ph.GA]. nual modulation of dark matter, Reviews of Modern [13] Gravity Collaboration, R. Abuter, A. Amorim, Physics 85, 1561 (2013), arXiv:1209.3339 [astro-ph.CO]. M. Baub¨ock, J. P. Berger, H. Bonnet, W. Brand- [6] Gaia Collaboration, S. A. Klioner, F. Mignard, et al., ner, V. Cardoso, Y. Cl´enet,P. T. de Zeeuw, J. Dexter, Gaia Early Data Release 3. Acceleration of the Solar Sys- A. Eckart, F. Eisenhauer, N. M. F¨orster Schreiber, tem from Gaia astrometry, Astron. Astrophys. 649, A9 P. Garcia, F. Gao, E. Gendron, R. Genzel, S. Gillessen, (2021), arXiv:2012.02036 [astro-ph.GA]. et al., Detection of the Schwarzschild precession in [7] Gaia Collaboration, Brown, Anthony G.A., Vallenari, A., the orbit of the star S2 near the Galactic centre mas- Prusti, T., de Bruijne, J. H.J., et al., Gaia early data sive black hole, Astron. Astrophys. 636, L5 (2020), release 3. summary of the contents and survey prop- arXiv:2004.07187 [astro-ph.GA]. erties, Astron. Astrophys. 10.1051/0004-6361/202039657 [14] Gravity Collaboration, R. Abuter, A. Amorim, 6

M. Baub¨ock, J. P. Berger, H. Bonnet, W. Brand- from Timing the Binary Pulsar PSR B1913+16, Astro- ner, Y. Cl´enet,V. Coud´eDu Foresto, P. T. de Zeeuw, phys. J. 829, 55 (2016), arXiv:1606.02744 [astro-ph.HE]. J. Dexter, G. Duvert, A. Eckart, F. Eisenhauer, [30] S. Chakrabarti, P. Chang, M. T. Lam, S. J. Vigeland, and N. M. F¨orsterSchreiber, et al., A geometric distance A. C. Quillen, A Measurement of the Galactic Plane Mass measurement to the Galactic center black hole with Density from Binary Pulsar Accelerations, Astrophys. J. 0.3% uncertainty, Astron. Astrophys. 625, L10 (2019), Lett. 907, L26 (2021), arXiv:2010.04018 [astro-ph.GA]. arXiv:1904.05721 [astro-ph.GA]. [31] D. F. Phillips, A. Ravi, R. Ebadi, and R. L. Walsworth, [15] Gravity Collaboration, R. Abuter, A. Amorim, et al., Im- Milky Way Accelerometry via Millisecond Pulsar Timing, proved GRAVITY astrometric accuracy from modeling Phys. Rev. Lett. 126, 141103 (2021), arXiv:2008.13052 optical aberrations, Astron. Astrophys. 647, A59 (2021), [astro-ph.GA]. arXiv:2101.12098 [astro-ph.GA]. [32] I. S. Shklovskii, Possible Causes of the Secular Increase [16] T. Do, A. Hees, A. Ghez, et al., Relativistic redshift of the in Pulsar Periods., Soviet Astronomy 13, 562 (1970). star S0-2 orbiting the Galactic Center supermassive black [33] A. T. Deller, J. M. Weisberg, D. J. Nice, and hole, Science 365, 664 (2019), arXiv:1907.10731 [astro- S. Chatterjee, A VLBI Distance and Transverse Veloc- ph.GA]. ity for PSR B1913+16, Astrophys. J. 862, 139 (2018), [17] J. Binney and S. Tremaine, Galactic Dynamics: Second arXiv:1806.10265 [astro-ph.SR]. Edition (2008). [34] M. Bennett and J. Bovy, Vertical waves in the solar [18] R. Sch¨onrich, J. Binney, and W. Dehnen, Local kinemat- neighbourhood in Gaia DR2, Mon. Not. Roy. Astron. ics and the local standard of rest, Mon. Not. Roy. Astron. Soc. 482, 1417 (2019), arXiv:1809.03507 [astro-ph.GA]. Soc. 403, 1829 (2010), arXiv:0912.3693 [astro-ph.GA]. [35] J. H. Oort, Observational evidence confirming Lindblad’s [19] J. Bovy, J. C. Bird, A. E. Garc´ıaP´erez,S. R. Majewski, hypothesis of a rotation of the galactic system, Bulletin D. L. Nidever, and G. Zasowski, The Power Spectrum of the Astronomical Institutes of the Netherlands 3, 275 of the Milky Way: Velocity Fluctuations in the Galac- (1927). tic Disk, Astrophys. J. 800, 83 (2015), arXiv:1410.8135 [36] T. Antoja, A. Helmi, M. Romero-G´omez, D. Katz, [astro-ph.GA]. C. Babusiaux, R. Drimmel, D. W. Evans, F. Figueras, [20] M. J. Reid and A. Brunthaler, The Proper Motion of E. Poggio, C. Reyl´e, A. C. Robin, G. Seabroke, Sagittarius A*. III. The Case for a Supermassive Black and C. Soubiran, A dynamically young and perturbed Hole, Astrophys. J. 892, 39 (2020), arXiv:2001.04386 Milky Way disk, Nature (London) 561, 360 (2018), [astro-ph.GA]. arXiv:1804.10196 [astro-ph.GA]. [21] F. J. Kerr and D. Lynden-Bell, Review of galactic con- [37] D. Kawata, J. Baba, I. Ciucˇa, M. Cropper, R. J. J. stants., Mon. Not. Roy. Astron. Soc. 221, 1023 (1986). Grand, J. A. S. Hunt, and G. Seabroke, Radial distri- [22] J. Bovy, Galactic rotation in Gaia DR1, Mon. Not. Roy. bution of stellar motions in Gaia DR2, Mon. Not. Roy. Astron. Soc. 468, L63 (2017), arXiv:1610.07610 [astro- Astron. Soc. 479, L108 (2018), arXiv:1804.10175 [astro- ph.GA]. ph.GA]. [23] J. Bovy, galpy: A python Library for Galactic Dynam- [38] J. A. S. Hunt, M. W. Bub, J. Bovy, J. T. Mackereth, ics, Astrophys. J. Suppl. 216, 29 (2015), arXiv:1412.3451 W. H. Trick, and D. Kawata, Signatures of resonance [astro-ph.GA]. and phase mixing in the Galactic disc, Mon. Not. Roy. [24] P. J. McMillan, The mass distribution and gravitational Astron. Soc. 490, 1026 (2019), arXiv:1904.10968 [astro- potential of the Milky Way, Mon. Not. Roy. Astron. Soc. ph.GA]. 465, 76 (2017), arXiv:1608.00971 [astro-ph.GA]. [39] J. Bland-Hawthorn, S. Sharma, T. Tepper-Garcia, [25] A. Irrgang, B. Wilcox, E. Tucker, and L. Schiefelbein, J. Binney, K. C. Freeman, M. R. Hayden, J. Kos, G. M. Milky Way mass models for orbit calculations, Astron. De Silva, S. Ellis, G. F. Lewis, M. Asplund, S. Buder, Astrophys. 549, A137 (2013), arXiv:1211.4353 [astro- A. R. Casey, V. D’Orazi, L. Duong, S. Khanna, J. Lin, ph.GA]. K. Lind, S. L. Martell, M. K. Ness, J. D. Simpson, D. B. [26] C. Allen and A. Santillan, An improved model of the Zucker, T. Zwitter, P. R. Kafle, A. C. Quillen, Y.-S. Ting, galactic mass distribution for orbit computations., Re- and R. F. G. Wyse, The GALAH survey and Gaia DR2: vista Mexicana de Astronomia y Astrofisica 22, 255 dissecting the stellar disc’s phase space by age, action, (1991). chemistry, and location, Mon. Not. Roy. Astron. Soc. [27] W. Dehnen and J. J. Binney, Local stellar kinematics 486, 1167 (2019), arXiv:1809.02658 [astro-ph.GA]. from HIPPARCOS data, Mon. Not. Roy. Astron. Soc. [40] N. Banik, L. M. Widrow, and S. Dodelson, Galactoseis- 298, 387 (1998), arXiv:astro-ph/9710077 [astro-ph]. mology and the local density of dark matter, Mon. Not. [28] T. Damour and J. H. Taylor, On the Orbital Period Roy. Astron. Soc. 464, 3775 (2017), arXiv:1608.03338 Change of the Binary Pulsar PSR 1913+16, Astrophys. [astro-ph.GA]. J. 366, 501 (1991). [29] J. M. Weisberg and Y. Huang, Relativistic Measurements