Stellar kinematics encompasses the measurement of stellarvelocities in theMilky Way and itssatellites as well as the internal kinematics of more distantgalaxies. Measurement of the kinematics of stars in different subcomponents of the Milky Way including thethin disk, thethick disk, thebulge, and thestellar halo provides important information about the formation and evolutionary history of our Galaxy. Kinematic measurements can also identify exotic phenomena such as hypervelocity stars escaping from the Milky Way, which are interpreted as the result of gravitational encounters ofbinary stars with thesupermassive black hole at the Galactic Center.
Stellar kinematics is related to but distinct from the subject ofstellar dynamics, which involves the theoretical study or modeling of the motions of stars under the influence ofgravity. Stellar-dynamical models of systems such as galaxies or star clusters are often compared with or tested against stellar-kinematic data to study their evolutionary history and mass distributions, and to detect the presence ofdark matter orsupermassive black holes through their gravitational influence on stellar orbits.
Relation between proper motion and velocity components of an object. At emission, the object was at distanced from the Sun, and moved at angular rateμ radian/s, that is,μ = vt / d withvt = the component of velocity transverse to line of sight from the Sun. (The diagram illustrates an angleμ swept out in unit time at tangential velocityvt.)
The component of stellar motion toward or away from the Sun, known asradial velocity, can be measured from the spectrum shift caused by theDoppler effect. The transverse, orproper motion must be found by taking a series of positional determinations against more distant objects. Once the distance to a star is determined throughastrometric means such asparallax, the space velocity can be computed.[2] This is the star'sactual motion relative to theSun or thelocal standard of rest (LSR). The latter is typically taken as a position at the Sun's present location that is following a circular orbit around theGalactic Center at the mean velocity of those nearby stars with low velocity dispersion.[3] The Sun's motion with respect to the LSR is called the "peculiar solar motion".
The components of space velocity in theMilky Way'sGalactic coordinate system are usually designated U, V, and W, given in km/s, with U positive in the direction of the Galactic Center, V positive in the direction ofgalactic rotation, and W positive in the direction of theNorth Galactic Pole.[4] The peculiar motion of the Sun with respect to the LSR is[5]
(U, V, W) = (11.1, 12.24, 7.25) km/s,
with statisticaluncertainty (+0.69−0.75, +0.47−0.47, +0.37−0.36) km/s and systematic uncertainty (1, 2, 0.5) km/s. (Note that V is 7 km/s larger than estimated in 1998 by Dehnen et al.[6])
Stellar kinematics yields importantastrophysical information about stars, and the galaxies in which they reside. Stellar kinematics data combined with astrophysical modeling produces important information about the galactic system as a whole. Measured stellar velocities in the innermost regions of galaxies including the Milky Way have provided evidence that many galaxies hostsupermassive black holes at their center.[7][8] In farther out regions of galaxies such as within the galactic halo, velocity measurements ofglobular clusters orbiting in these halo regions of galaxies provides evidence fordark matter. Both of these cases derive from the key fact that stellar kinematics can be related to the overallpotential in which the stars are bound. This means that if accurate stellar kinematics measurements are made for a star or group of stars orbiting in a certain region of a galaxy, the gravitational potential and mass distribution can be inferred given that the gravitational potential in which the star is bound produces its orbit and serves as the impetus for its stellar motion. Examples of using kinematics combined with modeling to construct an astrophysical system include:
Rotation of the Milky Way's disc: From theproper motions andradial velocities of stars within the Milky way disc one can show that there is differential rotation. When combining these measurements of stars' proper motions and their radial velocities, along with careful modeling, it is possible to obtain a picture of the rotation of the Milky Waydisc. The local character of galactic rotation in thesolar neighborhood is encapsulated in theOort constants.[9][10][11]
Structural components of the Milky Way: Using stellar kinematics, astronomers construct models which seek to explain the overall galactic structure in terms of distinct kinematic populations of stars. This is possible because these distinct populations are often located in specific regions of galaxies. For example, within theMilky Way, there are three primary components, each with its own distinct stellar kinematics: thedisc,halo andbulge or bar. These kinematic groups are closely related to the stellar populations in the Milky Way, forming a strong correlation between the motion and chemical composition, thus indicating different formation mechanisms. For the Milky Way, the speed of disk stars is and an RMS (Root mean square) velocity relative to this speed of. For bulge population stars, the velocities are randomly oriented with a larger relative RMS velocity of and no net circular velocity.[12] The Galactic stellar halo consists of stars with orbits that extend to the outer regions of the galaxy. Some of these stars will continually orbit far from the galactic center, while others are on trajectories which bring them to various distances from the galactic center. These stars have little to no average rotation. Many stars in this group belong to globular clusters which formed long ago and thus have a distinct formation history, which can be inferred from their kinematics and poor metallicities. The halo may be further subdivided into an inner and outer halo, with the inner halo having a net prograde motion with respect to the Milky Way and the outer a netretrograde motion.[13]
External galaxies: Spectroscopic observations of external galaxies make it possible to characterize the bulk motions of the stars they contain. While these stellar populations in external galaxies are generally not resolved to the level where one can track the motion of individual stars (except for the very nearest galaxies) measurements of the kinematics of the integrated stellar population along the line of sight provides information including the mean velocity and thevelocity dispersion which can then be used to infer the distribution of mass within the galaxy. Measurement of the mean velocity as a function of position gives information on the galaxy's rotation, with distinct regions of the galaxy that areredshifted /blueshifted in relation to the galaxy'ssystemic velocity.
Mass distributions: Through measurement of the kinematics of tracer objects such as globular clusters and the orbits of nearbysatellite dwarf galaxies, we can determine the mass distribution of the Milky Way or other galaxies. This is accomplished by combining kinematic measurements with dynamical modeling.
Expected motion of 40,000 stars in the next 400 thousand years, as determined by Gaia EDR3
In 2018, theGaia Data Release 2 (GAIA DR2) marked a significant advancement in stellar kinematics, offering a rich dataset of precise measurements. This release included detailed stellar kinematic andstellar parallax data, contributing to a more nuanced understanding of the Milky Way's structure. Notably, it facilitated the determination of proper motions for numerous celestial objects, including the absolute proper motions of 75globular clusters situated at distances extending up to and a bright limit of.[14] Furthermore, Gaia's comprehensive dataset enabled the measurement of absolute proper motions in nearbydwarf spheroidal galaxies, serving as crucial indicators for understanding the mass distribution within the Milky Way.[15] GAIA DR3 improved the quality of previously published data by providing detailed astrophysical parameters.[16] While the complete GAIA DR4 is yet to be unveiled, the latest release offers enhanced insights into white dwarfs,hypervelocity stars, cosmologicalgravitational lensing, and the merger history of theGalaxy.[17]
Stars within galaxies may be classified based on their kinematics. For example, the stars in the Milky Way can be subdivided into two general populations, based on theirmetallicity, or proportion of elements with atomic numbers higher than helium. Among nearby stars, it has been found thatpopulation I stars with higher metallicity are generally located in the stellar disk while older population II stars are in random orbits with little net rotation.[18] The latter have elliptical orbits that are inclined to the plane of the Milky Way.[18] Comparison of the kinematics of nearby stars has also led to the identification ofstellar associations. These are most likely groups of stars that share a common point of origin in giant molecular clouds.[19]
There are many additional ways to classify stars based on their measured velocity components, and this provides detailed information about the nature of the star's formation time, its present location, and the general structure of the galaxy. As a star moves in a galaxy, the smoothed out gravitational potential of all the other stars and other mass within the galaxy plays a dominant role in determining the stellar motion.[20] Stellar kinematics can provide insights into the location of where the star formed within the galaxy. Measurements of an individual star's kinematics can identify stars that are peculiar outliers such as a high-velocity star moving much faster than its nearby neighbors.
Depending on the definition, a high-velocity star is a star moving faster than 65 km/s to 100 km/s relative to the average motion of the other stars in the star's neighborhood. The velocity is also sometimes defined assupersonic relative to the surrounding interstellar medium. The three types of high-velocity stars are: runaway stars, halo stars and hypervelocity stars. High-velocity stars were studied by Jan Oort, who used their kinematic data to predict that high-velocity stars have very little tangential velocity.[21]
Four runaway stars moving through regions of dense interstellar gas and creating bright bow waves and trailing tails of glowing gas. The stars in these NASA Hubble Space Telescope images are among 14 young runaway stars spotted by the Advanced Camera for Surveys between October 2005 and July 2006.
A runaway star is one that is moving through space with an abnormally highvelocity relative to the surroundinginterstellar medium. Theproper motion of a runaway star often points exactly away from astellar association, of which the star was formerly a member, before it was hurled out.
Mechanisms that may give rise to a runaway star include:
Gravitational interactions between stars in astellar system can result in large accelerations of one or more of the involved stars. In some cases, stars may even be ejected.[22] This can occur in seemingly stable star systems of only three stars, as described in studies of thethree-body problem in gravitational theory.[23]
A collision or close encounter between stellar systems, including galaxies, may result in the disruption of both systems, with some of the stars being accelerated to high velocities, or even ejected. A large-scale example is the gravitational interaction between theMilky Way and theLarge Magellanic Cloud.[24]
Asupernova explosion in amultiple star system can accelerate both the supernova remnant and remaining stars to high velocities.[25][26]
Multiple mechanisms may accelerate the same runaway star. For example, a massive star that was originally ejected due to gravitational interactions with its stellar neighbors may itself go supernova, producing a remnant with a velocity modulated by the supernova kick. If this supernova occurs in the very nearby vicinity of other stars, it is possible that it may produce more runaways in the process.
An example of a related set of runaway stars is the case ofAE Aurigae,53 Arietis andMu Columbae, all of which are moving away from each other at velocities of over 100 km/s (for comparison, theSun moves through the Milky Way at about 20 km/s faster than the local average). Tracing their motions back, their paths intersect near to theOrion Nebula about 2 million years ago.Barnard's Loop is believed to be the remnant of the supernova that launched the other stars.
Another example is the X-ray objectVela X-1, where photodigital techniques reveal the presence of a typical supersonicbow shock hyperbola.
Halo stars are very old stars that have a low metallicity[27] and do not follow circular orbits around the center of the Milky Way within its disk. Instead, the halo stars travel in elliptical orbits, often inclined to the disk, which take them well above and below the plane of the Milky Way. Although their orbital velocities relative to the Milky Way may be no faster than disk stars, their different paths result in high relative velocities.
Typical examples are the halo stars passing through the disk of the Milky Way at steep angles. One of the nearest 45 stars, calledKapteyn's Star, is an example of the high-velocity stars that lie near the Sun: Its observed radial velocity is −245 km/s, and the components of its space velocity areu = +19 km/s,v = −288 km/s, andw = −52 km/s.
Positions and trajectories of 20 high-velocity stars as reconstructed from data acquired byGaia, overlaid on top of an artistic view of the Milky Way
Hypervelocity stars (designated asHVS orHV in stellar catalogues) have substantially higher velocities than the rest of the stellar population of a galaxy. Some of these stars may even exceed theescape velocity of the galaxy.[28] In the Milky Way, stars usually have velocities on the order of 100 km/s, whereas hypervelocity stars typically have velocities on the order of 1000 km/s. Most of these fast-moving stars are thought to be produced near the center of the Milky Way, where there is a larger population of these objects than further out. One of the fastest known stars in our Galaxy is the O-class sub-dwarfUS 708, which is moving away from the Milky Way with a total velocity of around 1200 km/s.
Jack G. Hills first predicted the existence of HVSs in 1988.[29] This was later confirmed in 2005 by Warren Brown,Margaret Geller,Scott Kenyon, andMichael Kurtz.[30] As of 2008,[update] 10unbound HVSs were known, one of which is believed to have originated from theLarge Magellanic Cloud rather than theMilky Way.[31] Further measurements placed its origin within the Milky Way.[32] Due to uncertainty about the distribution of mass within the Milky Way, determining whether a HVS is unbound is difficult. A further five known high-velocity stars may be unbound from the Milky Way, and 16 HVSs are thought to be bound. The nearest currently known HVS (HVS2) is about 19 kpc from the Sun.
As of 1 September 2017[update], there have been roughly 20 observed hypervelocity stars. Though most of these were observed in theNorthern Hemisphere, the possibility remains that there are HVSs only observable from theSouthern Hemisphere.[33]
It is believed that about 1,000 HVSs exist in the Milky Way.[34] Considering that there are around 100 billion stars in theMilky Way, this is a minuscule fraction (~0.000001%). Results from the second data release ofGaia (DR2) show that most high-velocity late-type stars have a high probability of being bound to the Milky Way.[35] However, distant hypervelocity star candidates are more promising.[36]
In March 2019,LAMOST-HVS1 was reported to be a confirmed hypervelocity star ejected from the stellar disk of the Milky Way.[37]
In July 2019, astronomers reported finding an A-type star,S5-HVS1, traveling 1,755 km/s (3,930,000 mph), faster than any other star detected so far. The star is in theGrus (or Crane)constellation in the southern sky and is about 29,000 ly (1.8×109 AU) from Earth. It may have been ejected from the Milky Way after interacting withSagittarius A*, thesupermassive black hole at the center of the galaxy.[38][39][40][41][42]
Runaway star speeding from 30 Doradus. Image taken by the Hubble Space Telescope.
HVSs are believed to predominantly originate by close encounters ofbinary stars with thesupermassive black hole in the center of theMilky Way. One of the two partners is gravitationally captured by theblack hole (in the sense of entering orbit around it), while the other escapes with high velocity, becoming a HVS. Known as theHills mechanism, this is analogous to the capture and ejection ofinterstellar objects by a star.
Supernova-induced HVSs may also be possible, although they are presumably rare. In this scenario, a HVS is ejected from a close binary system as a result of the companion star undergoing a supernova explosion. Ejection velocities up to 770 km/s, as measured from the galactic rest frame, are possible for late-type B-stars.[43] This mechanism can explain the origin of HVSs which are ejected from the galactic disk.
Known HVSs aremain-sequence stars with masses a few times that of the Sun. HVSs with smaller masses are also expected and G/K-dwarf HVS candidates have been found.
Some HVSs may have originated from a disrupted dwarf galaxy. When it made its closest approach to the center of the Milky Way, some of its stars broke free and were thrown into space, due to theslingshot-like effect of the boost.[44]
Someneutron stars are inferred to be traveling with similar speeds. This could be related to HVSs and the HVS ejection mechanism. Neutron stars are the remnants ofsupernova explosions, and their extreme speeds are very likely the result of an asymmetric supernova explosion or the loss of their near partner during the supernova explosions that forms them. The neutron starRX J0822-4300, which was measured to move at a record speed of over 1,500 km/s (0.5% of thespeed of light) in 2007 by theChandra X-ray Observatory, is thought to have been produced the first way.[45]
One theory regarding the ignition ofType Ia supernovae invokes the onset of a merger between two white dwarfs in a binary star system, triggering the explosion of the more massive white dwarf. If the less massive white dwarf is not destroyed during the explosion, it will no longer be gravitationally bound to its destroyed companion, causing it to leave the system as a hypervelocity star with its pre-explosion orbital velocity of 1000–2500 km/s. In 2018, three such stars were discovered using data from the Gaia satellite.[46]
A set of stars with similar space motion and ages is known as a kinematic group.[48] These are stars that could share a common origin, such as the evaporation of anopen cluster, the remains of a star forming region, or collections of overlapping star formation bursts at differing time periods in adjacent regions.[49] Most stars are born withinmolecular clouds known asstellar nurseries. The stars formed within such a cloud compose gravitationally boundopen clusters containing dozens to thousands of members with similar ages and compositions. These clusters dissociate with time. Groups of young stars that escape a cluster, or are no longer bound to each other, form stellar associations. As these stars age and disperse, their association is no longer readily apparent and they become moving groups of stars.
Astronomers are able to determine if stars are members of a kinematic group because they share the same age,metallicity, and kinematics (radial velocity andproper motion). As the stars in a moving group formed in proximity and at nearly the same time from the same gas cloud, although later disrupted by tidal forces, they share similar characteristics.[50]
A stellar association is a very loosestar cluster, whose stars share a common origin and are still moving together through space, but have become gravitationally unbound. Associations are primarily identified by their common movement vectors and ages. Identification by chemical composition is also used to factor in association memberships.
Stellar associations were first discovered by theArmenian astronomerViktor Ambartsumian in 1947.[51] The conventional name for an association uses the names or abbreviations of theconstellation (or constellations) in which they are located; the association type, and, sometimes, a numerical identifier.
Viktor Ambartsumian first categorized stellar associations into two groups, OB and T, based on the properties of their stars.[51] A third category, R, was later suggested bySidney van den Bergh for associations that illuminatereflection nebulae.[52] The OB, T, and R associations form a continuum of young stellar groupings. But it is currently uncertain whether they are an evolutionary sequence, or represent some other factor at work.[53] Some groups also display properties of both OB and T associations, so the categorization is not always clear-cut.
Young associations will contain 10 to 100 massive stars ofspectral classO andB, and are known asOB associations. In addition, these associations also contain hundreds or thousands of low- and intermediate-mass stars. Association members are believed to form within the same small volume inside a giantmolecular cloud. Once the surrounding dust and gas is blown away, the remaining stars become unbound and begin to drift apart.[54] It is believed that the majority of all stars in the Milky Way were formed in OB associations.[54]O-class stars are short-lived, and will expire assupernovae after roughly one million years. As a result, OB associations are generally only a few million years in age or less. The O-B stars in the association will have burned all their fuel within ten million years. (Compare this to the current age of theSun at about five billion years.)
OB associations have also been found in theLarge Magellanic Cloud and theAndromeda Galaxy. These associations can be quite sparse, spanning 1,500 light-years in diameter.[19]
Young stellar groups can contain a number of infantT Tauri stars that are still in the process of entering themain sequence. These sparse populations of up to a thousand T Tauri stars are known asT associations. The nearest example is theTaurus-Auriga T association (Tau–Aur T association), located at a distance of 140parsecs from the Sun.[57] Other examples of T associations include theR Corona Australis T association, theLupus T association, theChamaeleon T association and theVelorum T association. T associations are often found in the vicinity of the molecular cloud from which they formed. Some, but not all, include O–B class stars. Group members have the same age and origin, the same chemical composition, and the same amplitude and direction in their vector of velocity.
Associations of stars that illuminate reflectionnebulae are calledR associations, a name suggested by Sidney van den Bergh after he discovered that the stars in these nebulae had a non-uniform distribution.[52] These young stellar groupings contain main sequence stars that are not sufficiently massive to disperse the interstellar clouds in which they formed.[53] This allows the properties of the surrounding dark cloud to be examined by astronomers. Because R associations are more plentiful than OB associations, they can be used to trace out the structure of the galactic spiral arms.[58] An example of an R association isMonoceros R2, located 830 ± 50parsecs from the Sun.[53]
If the remnants of a stellar association drift through the Milky Way as a somewhat coherent assemblage, then they are termed amoving group orkinematic group. Moving groups can be old, such as theHR 1614 moving group at two billion years, or young, such as theAB Dor Moving Group at only 120 million years.
Moving groups were studied intensely byOlin Eggen in the 1960s.[59] A list of the nearest young moving groups has been compiled by López-Santiagoet al.[48] The closest is theUrsa Major Moving Group which includes all of the stars in thePlough / Big Dipperasterism except forDubhe andAlkaid. This is sufficiently close that theSun lies in its outer fringes, without being part of the group. Hence, although members are concentrated atdeclinations near 60°N, some outliers are as far away across the sky asTriangulum Australe at 70°S.
The list of young moving groups is constantly evolving. The Banyan Σ tool[60] currently lists 29 nearby young moving groups[62][61] Recent additions to nearby moving groups are theVolans-Carina Association (VCA), discovered withGaia,[63] and theArgus Association (ARG), confirmed with Gaia.[64] Moving groups can sometimes be further subdivided in smaller distinct groups. The Great Austral Young Association (GAYA) complex was found to be subdivided into the moving groupsCarina,Columba, andTucana-Horologium. The three Associations are not very distinct from each other, and have similar kinematic properties.[65]
Astellar stream is an association ofstars orbiting agalaxy that was once aglobular cluster ordwarf galaxy that has now been torn apart and stretched out along its orbit by tidal forces.[67]
The general motion and orientation of the sun, with Earth and the Moon as its Solar System satellites
The sun, taking along the whole Solar System, orbitsthe galaxy's centre of mass at an average speed of 230 km/s (828,000 km/h),[71] taking about 220–250 millionEarth years to complete a revolution (agalactic year), having done so about 20 times since the sun's formation.[72][73] The direction of the sun's motion, theSolar apex, is roughly in the direction of the starVega.[74] In the past the sun likely moved through theOrion–Eridanus Superbubble, before entering the Local Bubble.[75]
The sun's idealised orbit around the Galactic Centre in an artist's top-down depiction of the current layout of the Milky WayPath of sun around the Milky Way from 250 million BC to AD 250 million, according to a simple model, and assuming we are 8 kiloparsecs from the centre. The scales are in kiloparsecs. At present we are at (0, 8) and moving to the right. The points are spaced 25 million years apart. TheAndromeda Galaxy is towards the upper right 765 kiloparsecs away.
A simple model of the motion of a star in the galaxy gives thegalactic coordinatesX,Y, andZ (rotating coordinates so that the centre of the galaxy is always in theX direction) as:
whereU,V, andW are the respective velocities with respect to thelocal standard of rest,A andB are theOort constants, is the angular velocity of galactic rotation for the local standard of rest, is the "epicyclic frequency", and ν is the vertical oscillation frequency.[76] For the sun, the present values ofU,V, andW used in the reference were km/s, although more recent estimates are and estimates for the other constants areA = 15.5 km/s/kpc,B = −12.2 km/s/kpc, κ = 37 km/s/kpc, and ν=74 km/s/kpc. We takeX(0) andY(0) to be zero andZ(0) is estimated to be 17 parsecs.[77] This model implies that the sun circulates around a point that is itself going around the galaxy. The period of the sun's circulation around the point is. which comes to 166 million years, shorter than the time it takes for the point to go around the galaxy. In the (X, Y) coordinates, the sun describes an ellipse around the point, whose length in theY direction, using the more recent values of U, V, and W, is
and whose width in theX direction is
(Compare this to the distance of the sun from the centre of the galaxy, around 7 or 8 kiloparsecs.)The ratio of length to width of this ellipse, the same for all stars in our neighborhood, isThe moving point is presently at
The oscillation in theZ direction takes the sun
"north" of the galactic plane and the same distance "south" of it, with a period of or 83 million years, approximately 2.7 times per orbit.[78] Although is 222 million years, the value of at the point around which the sun circulates, assuming 8 kiloparsecs for the distance to the centre of the galaxy is
corresponding to around 240 million years (seeOort constants), and this is the time that the point takes to go once around the galaxy. Other stars with the same value of have to take the same amount of time to go around the galaxy as the sun and thus remain in the same general vicinity as the sun.
Velocity components and speed of sun relative to surrounding stars over 500 million years
By this model, the velocity components of the sun with respect to the local standard of rest is:
The speed
is thus aquasiperiodic function with values in the range (around 14 km/s) to (around 23 km/s), with a present value around 18 km/s.
Note that this model does not take into consideration that the Oort "constants" and Ω actually decrease with distance from the centre of the galaxy.
The Sun's orbit around the Milky Way is perturbed due to the non-uniform mass distribution in Milky Way, such as that in and between the galactic spiral arms. It has been argued that the Sun's passage through the higher density spiral arms often coincides withmass extinctions on Earth, perhaps due to increasedimpact events.[79] It takes the Solar System about 225–250 million years to complete one orbit through the Milky Way (agalactic year),[73] so it is thought to have completed 20–25 orbits during the lifetime of the Sun. Theorbital speed of the Solar System about the centre of the Milky Way is approximately 251 km/s (156 mi/s).[80] At this speed, it takes around 1,190 years for the Solar System to travel a distance of 1 light-year, or 7 days to travel1 AU.[81]
^"Stellar Motions (Extension)".Australia Telescope Outreach and Education. Commonwealth Scientific and Industrial Research Organisation. 2005-08-18. Archived fromthe original on 2013-06-06. Retrieved2008-11-19.
^Johnson, Dean R. H.; Soderblom, David R. (1987). "Calculating galactic space velocities and their uncertainties, with an application to the Ursa Major group".Astronomical Journal.93 (2):864–867.Bibcode:1987AJ.....93..864J.doi:10.1086/114370.
^Blaauw, A. (1961). "On the origin of the O- and B-type stars with high velocities (the run-away stars), and some related problems".Bulletin of the Astronomical Institutes of the Netherlands.15: 265.Bibcode:1961BAN....15..265B.
^Tauris, T.M.; Takens, R.J. (1998). "Runaway velocities of stellar components originating from disrupted binaries via asymmetric supernova explosions".Astronomy and Astrophysics.330:1047–1059.Bibcode:1998A&A...330.1047T.
^abBrown, Warren R.; Geller, Margaret J.; Kenyon, Scott J.; Kurtz, Michael J.; Bromley, Benjamin C. (2007). "Hypervelocity Stars. III. The Space Density and Ejection History of Main-Sequence Stars from the Galactic Center".The Astrophysical Journal.671 (2):1708–1716.arXiv:0709.1471.Bibcode:2007ApJ...671.1708B.doi:10.1086/523642.S2CID15074398.
^abIsraelian, Garik (1997). "Obituary: Victor Amazaspovich Ambartsumian, 1912 [i.e. 1908] –1996".Bulletin of the American Astronomical Society.29 (4):1466–1467.Bibcode:1997BAAS...29.1466I.
^Eggen, O.J. (1965). "Moving groups of stars". In Blaauw, Adriaan & Schmidt, Maarten (eds.).Observational Aspects of Galactic Structure: Lecture notes reported by participants. Chicago: University of Chicago Press. p. 111.Bibcode:1965gast.book..111E.
