Galaxy Evolution Majority of galaxies belong to clusters and groups of galaxies. Density of galaxies in clusters is roughly 100x greater than that of stars in galaxies. There is a higher probability of interactions and/or mergers between galaxies in regions of higher galaxy number density. Interactions tend to increase the velocity dispersion, and probably destroy gaseous disks in late-type galaxies, causing their stars to become elliptical-like. Interactions also apply tidal effects on gas and stars. Causes gas to compress and form stars, often in immense bursts of star formation which can exceed the quiescent star formation rate by 10-100 x!
NGC4038/4039, Antennae Galaxies
NGC4038/4039, Antennae Galaxies
NGC4038/4039, Antennae Galaxies Z. Wang et al.
Galaxy Evolution What happens when galaxies collide? Stars are very far apart and pass right through each other. Effects are gravitational. As a galaxy of mass M passes through another, if feels the gravity and produces a wake of higher density (because stars in the other galaxy have been compressed along the path of the moving galaxy. This is dynamical friction and is a net gravitational force on M that opposes the galaxy s motion. Kinetic energy was transferred from M to the surrounding material as M s speed is reduced.
Dynamical Friction M Contours show the density enhancement of stars due to the motion of a mass M in the positive z direction. V A high-density wake trailing M causes a net gravitational force on M opposing its motion Mulder 1983, A&A, 117, 9
Dynamical Friction Derivation of Force of dynamical fraction is complicated. It depends on the speed, vm, the density of the surrounding material, ρ, and mass, M, squared. One can estimate the timescale for dynamical friction to halt two colliding objects. Consider the dynamical friction force on Globular clusters within the Milky Way. The dark matter distribution is Insert this into the dynamical friction force equation above, which gives
Dynamical Friction Assume globular clusters orbits are circular, then the angular momentum is just L=M vm r. And, the torque is τ = r fd = (dl / dt). For a flat rotation curve, vm is constant at large radii. Thus, the derivative of L is (dl/dt) = MvM (dr/dt). This gives: Integrating this equation gives an expression describing the time required for the globular cluster to spiral into the center of the host galaxy from an initial radius, r1. Or: Or, you can calculate the maximum distance a globular cluster could have traveled in a time tmax:
Dynamical Friction Example: Consider a globular cluster that orbits in the Andromeda galaxy (M31). Assume the cluster s mass is 5 x 10 6 solar masses with vm = 250 km s -1. Age of M31 is approximately ~13 Gyr. The maximum radius at which a GC could have spiraled into the center of M31 is rmax = 3.7 kpc, whereas the halo of M31 is more like 100 kpc. Note that rmax ~ M 1/2. Clusters with greater masses have higher maximum radii, so this likely explains the lack of massive GCs in M31 today. NOTE! Dynamical friction affects globular clusters, and satellite galaxies around larger galaxies, and other galaxies in clusters of galaxies.
Dynamical Friction Artists rendition of the Sagittarius dwarf galaxy merging with the Milky Way. NOTE! Dynamical friction affects globular clusters, and satellite galaxies around larger galaxies, and other galaxies in clusters of galaxies.
Slide Rapid encounters
How do Galaxies Form? I. Monolithic Collapse. A galaxy forms from a single gas cloud that collapse, fragment and form stars. First serious model proposed by Olin Eggen, Donald Lynden-Bell, and Allan Sandage in 1962 (Eggen, Lynden- Bell, & Sandage - ELS model). ELS noticed that metal-rich stars lie in the Galactic plane with circular rotations. Metal-poor stars have more eccentric orbits, some with highly elliptical orbits in and out of the plane of the Galactic disk Hypothesis: Milky Way (and other galaxies) formed from a rapid collapse of a large proto-galactic nebula. Oldest stars formed early in the collapse, which locked in their orbits. These would be metal-poor (Population II) since the gas had had little enrichment. As proto-galactic cloud continued to fall inward, gas collisions are more frequent, so more stars form, with higher metallicity. Gas settles into disk with higher metal-mass fraction, and continues to form stars (such as our Sun). These would be Population I stars.
How do Galaxies Form? I. Homologous Collapse - ELS model. One can calculate the free-fall time for a gas cloud with the mass of the Milky Way to collapse. Assume a Mass of 5 x 10 11 solar masses with a nearly spherical volume of radius 50 kpc (dark matter halo, and size of stellar halo). Assume mass was initially distributed uniformly. The initial density is ρ0 = 3M/4π r 3 = 8 x 10-23 kg m -3. Ch. 12 works out that the free-fall time, tff, for collapse is tff = [(3π/32) (1/Gρ0) ] 1/2 = 200 Myr. If the inner regions of the Galaxy were more centrally condensed (as the NFW dark-matter distribution is), then the inner parts would collapse faster and the outer parts slower. This could explain the existence of old stellar populations in the bulge.
How do Galaxies Form? I. Homologous Collapse - ELS model. Problems with ELS model 1. Given initial rotation of the proto-galactic cloud, all halo stars and globular clusters should orbit in the same net direction. Observations show that the net rotation in the halo is 0 km/s. 2. The age spread in globular clusters is about 2 Gyr. An order of magnitude larger than the ELS free-fall time. 3. ELS model does not explain multiple disk components (thin and thick disks). 4. Globular clusters located nearest the Galaxy center are generally the most metal-rich and oldest, while clusters in the outer halo have a wide range of metallicity and tend to be younger. This is not naturally explained in the ELS model.
How do Galaxies Form? 2. Hierarchical Merger Model Galaxies probably have some formation similar to ELS, but their formation also has a bottom-up hierarchical process of mergers. Realization in the 1970s and 80s that the Big Bang would have left small matter fluctuations that would grow through gravity and merge to form larger mass objects. Our understanding is that there are many, many more low-mass fluctuations than large-mass fluctuations. Density fluctuations of 10 6-10 8 solar masses were much more common than those of 10 12 solar masses. Consider that the Milky Way grew from fragments of mass 10 6 to 10 8 solar masses. Initially these fragments are in isolation, forming stars and globular clusters. They have their own metal enrichment and starformation histories. They then merge. In this model, the inner regions of the growing spheroid where the density of matter was greater, evolution would be fastest. This produces things like metal-rich, old bulges.
Aq-A-2-evolv.mp4 Courtesy V. Springel
How do Galaxies Form? 2. Hierarchical Merger Model Collisions and tidal interactions between merging fragments would disrupt some globular clusters (remember dynamical friction?) and left some intact. In this model, the disrupted systems would have led to the present distribution of field halo stars, while leaving intact globular clusters distributed throughout the spheroid. Through many random mergers, there should be no net rotation of objects in the halo. Model also predicts that some proto-galactic fragments should still be out there. This explains significant number of small galaxies orbiting the Milky Way and nearby Andromeda. These are surviving proto-galactic fragments. Some are merging, like the Sagittarius dwarf.
Andromeda galaxy and dwarfs
Tidal streams of previously merged bits Andromeda galaxy and dwarfs
Galaxy Interactions
Simulation by Chris Mihos (ca. 1996)
Simulation by Volker Springel (ca. 2006)
T = 0 Myr 150 Myr 300 Myr 750 Myr Elliptical Galaxies can form from gas-rich spiral galaxy mergers. 450 Myr 900 Myr 600 Myr 1 Gyr 1.2 Gyr These are simulations of a merger of 2 spiral galaxies with 16,384 particles in each disk and 4096 particles in each bulge. The result is an elliptical galaxy. 1.4 Gyr 1.5 Gyr 1.7 Gyr See Hernquist 1993, ApJ, 409, 548.
Elliptical Galaxies can also grow from mergers of elliptical galaxies. van Dokkum 2005, AJ, 130, 2647
Elliptical Galaxy Formation
Elliptical Galaxy Formation Rings may be left over remnants of merged galaxies NGC 3923 Malin & Carter, Nature, 285, 643, 1980
Ring galaxies are the result of high speed collisions of galaxies in which the smaller one passes through another galaxy almost perpendicular to its disk.
How do Galaxies Form? Remember that galaxies with high redshifts are very far away: cz ~ v = H0 d (for z << 1) z = (λobs - λrest) / λrest Because it takes up to billions of years for the light from distant galaxies to reach us. We see them not as they are, but as they billions of years ago. We can study high-redshift galaxies to learn about galaxy evolution.
HST images of... galaxies from 10-12 Gyr ago. galaxies from 6-9 Gyr ago. from Papovich et al. 2005, ApJ, 631, 101
Slide Supermassive black holes are found in most galaxies
Slide Cores of other galaxies show an accretion disk with a possible black hole
Active Galactic Nuclei In 1908, Edward Fath (1880-1959) observed NGC 1068 with his spectroscope, which displayed odd (and very strong) emission lines. In 1926 Hubble recorded emission lines of this and two other galaxies. In 1943 Carl K. Seyfert (1911-1960) reported that a small fraction of galaxies have very, very bright nuclei that show broad emission lines produced by atoms in high ionization states. OIII NGC 1068
Active Galactic Nuclei Today, such objects are called Seyfert galaxies. Seyfert I galaxies have broad emission lines (1000-5000 km/s) Seyfert II galaxies have narrow lines (<500 km/s) NGC 1068 is a Seyfert II OIII NGC 1068
Seyferts Consider NGC 4151, a spiral galaxy 15 Mpc away. Photographs by Carl Seyfert in the 1940s showed a very bright point-like nucleus. Its spectrum is very unusual: in addition to continua + absorption lines from normal stars, Seyfert galaxy nuclei have very strong emission lines. Some are common lines (e.g. H-alpha, H-beta) but others are weird (e.g. twice-ionized oxygen lines), requiring hot gas far out of equilibrium. The lines are very broad, requiring that the gas be Doppler shifted in all directions up to ~20,000 km/s. The nuclei vary in brightness on timescales of months, requiring them to be < 1 parsec in size. The total luminosity of the nucleus can be equivalent to 10 10 L sun!! What is this bizarre object in the center of Seyfert's spiral galaxies? Slide
Flux Flux Wavelength [Angstroms] Mrk 1243 - Seyfert I (Osterbrock 1984, QJRAS, 25, 1)
Flux Flux Wavelength [Angstroms] Mrk 1157 - Seyfert II (Osterbrock 1984, QJRAS, 25, 1)
Active Galactic Nuclei Mrk means from the catalog of E. B. Markarian (1913-1985) who produced a catalog of Seyfert galaxies in 1965. Galaxies known to emit strongly in X-rays are Seyferts (type I s have more X-rays than type II s). Other types of Active Galaxies include radio galaxies, quasars, and blazars. Radio Galaxies After WWII, science of radio astronomy took off. First discrete source of radio waves (other than the Sun) was Cygnus A. Below is a VLA image. ~2 arcmin Redshift of Cygnus A is z=0.057, which from Hubble s Law gives a distance of 240 Mpc. Brightest radio source is well beyond the Milky Way!
Radio Galaxies Later in the 1940s, astronomers began scanning the skies with radio telescopes. They found strange radio structures on opposite sides of radio galaxies, plus a tiny source of radio emission at the nucleus. The nuclei of these radio galaxies shoot out narrow beams of extremely energetic electrons, producing synchrotron radiation. The radio components include: the compact core at the galaxy nucleus, jets, lobes, and a hot spot where the jet slams into the interstellar medium. Slide
Cosmic Jets and Radio Lobes Many active galaxies show powerful radio jets Radio image of Cygnus A Hot spots: Energy in the jets is released in interaction with surrounding material Material in the jets moves with almost the speed of light ( Relativistic jets ). Slide
Slide
Slide Centaurus A
Blazars 1) high radio-brightness accompanied by flatness of the radio spectrum, 2) Strong gamma-ray emission, 3) strong optical variability on very short timescales (less than few days). Slide
Slide Blazar spectrum
AGN variability To change in one hour, the source needs to have a size less than Velocity of light x 1 hour ~ 7 AU. AGNs are very compact!! Slide
What engine powers observed AGNs??? A supermassive black hole?! Slide
Formation of Radio Jets Accretion Disk Jets are powered by accretion of matter onto a supermassive black hole Black Hole Slide Twisted magnetic fields help to confine the material in the jet and to produce synchrotron radiation.
Active Galactic Nuclei
Magnetic field lines Active Galactic Nuclei
Evidence for Black Holes in AGNs Elliptical galaxy M 84: Spectral line shift indicates high-velocity rotation of gas near the center. Visual image NGC 7052: Stellar velocities indicate the presence of a central black hole. Slide
Slide The Mystery of Quasars
Active Galactic Nuclei Quasars As radio telescopes increased numbers of sources in late 1950s, astronomers began identifying them in optical images. In 1960 Thomas Matthews and Allan Sandage found a m=16 mag object matching 3C 48 (3C= Third Cambridge Catalog of radio sources) with an emission line spectrum that could not be identified. Sandage said, The thing was exceedingly weird. In 1963, a similar spectrum was seen in 3C 273. Optically, they looked like point sources (like stars?!) and not like galaxies. The became known as quasi-stellar radio sources = Quasars.
Slide Quasars look like stars, very different from galaxies
The spectral lines of quasar 3C 273 has z = 0.158. This is one of the nearest and brightest quasars (as far as apparent magnitude goes). Later astronomers recognized the emission lines as Balmer Hydrogen lines, but redshifted to incredible velocities, z=0.158 for 3C 273, or v ~ cz = 47,000 km /s! 3C 48 has z=0.367, or a radial velocity of 0.303 c!
Active Galactic Nuclei Currently, Quasars have been identified with redshifts z > 6! Many of these come from the Sloan Digital Sky Survey (SDSS). You book quotes that there are 520 quasars with z > 4. At z = 4 the recessional velocity is 0.92 c! To determine distances at such large redshifts requires geometrical considerations (more on this when we do cosmology). Effectively, the fractional change in wavelength due to the redshift is the same as the fraction change in the size of the Universe (recall the Universe is expanding!) z = (λobs - λrest) / λrest = (Robs - Remitted)/Remitted Where Robs is the size of the Universe when the photon is observed and Remitted is the size of the Universe when the photon was emitted.
Quasar Red Shifts z = 0.178 z = 0 Quasars have been detected at the highest red shifts, up to z ~ 6 z = 0.240 z = Δλ/λ 0 z = 0.302 z = 0.389 Our old formula Δλ/λ 0 = v r /c is only valid in the limit of low speed, v r << c Slide
Redshift z = (Observed wavelength - Rest wavelength) (Rest wavelength) Doppler effect: How come that z > 1?? First, relativistic Doppler effect is described by a different formula: Slide
Slide
However, cosmological redshift is not a Doppler effect! The redshift is due to the expansion of the Universe: as a light wave travels through space, the universe expands and the light wave gets stretched and therefore redshifted. z = R(t obs) R em ) 1 Slide
t = t em Slide p.308
Slide p.308
Slide p.308
t = t obs Slide p.308
Two galaxies permanently located at positions (x1, y1, z1 ) and ( x2, y2, z2 ) at one time find themselves one billion light years apart. Then a few billion years later while located at the same coordinates, they find themselves 3 billion light years apart. The galaxies have not 'moved', nevertheless, their separations have increased. Slide z = R(t obs) R em ) 1
Slide Another evidence of cosmological distance to quasars: gravitational lensing
Slide
Slide
Slide
Active Galactic Nuclei Quasars Luminosities Calculate the luminosity of 3C 273. The apparent magnitude is V=12.8 mag. The modern day distance for its redshift is 620 Mpc. MV = V - 5 log10(d / 10 pc) = -26.2 mag. Using MSun = +4.82 for the absolute magnitude, we can estimate 3C273 s visual luminosity: LV = 100 (M sun -M V )/5 L = 2.6 x 10 12 L = 10 39 W. Bolometric luminosities of Quasars range from 10 38 to 10 39 W, this is more than 100 times the output of a galaxy like the Milky Way!
Quasar sizes Quasars are point sources even in HST images, implying the regions emitting the intense luminosity are < 0.1 kpc. HST studies of QSOs show that they have host galaxies with tell-tale signs of mergers. Variability studies show that the luminosity of Quasars can vary by a factor of 2 within days or even hours. Variability gives as estimate of the size of the emission region because the region must be connected by the speed of light
Variability Variability gives as estimate of the size of the emission region because the region must be connected by the speed of light. R = c Δt (1 - v 2 /c 2 ) 1/2 = c Δt /γ For Δt = 1 hr, taking γ=1 (can only be larger, making R smaller) R = 1.1 x 10 12 m = 7.2 AU (between Jupiter and Saturn). Considering that the luminosity is >100 than the Milky Way, this is an incredibly small size! Recall that there is a maximum luminosity before an object will blow itself apart due to radiation pressure. This the Eddington Limit. L < LEd (1.5 x 10 31 W) x (M /M ) For L = 5 x 10 39 W you can solve for the mass, which would be M > 3.3 x 10 8 M. Finding such a large mass in such a small space is clear evidence for a supermassive black hole. The mass for an object with a Schwarzschild radius =7.2 AU (above) is M = 3.7 x 10 8 M.
Jets and host galaxies have been resolved for nearby quasars Slide 3C273
Active Galactic Nuclei Bahcall 1995, ApJ, 479, 642
Superluminal Motion Individual radio knots in quasar jets: Light-travel time effect: Sometimes apparently moving faster than speed of light! Material in the jet is almost catching up with the light it emits Slide
Evidence for Quasars in Distant Quasar 0351+026 at the same red shift as a galaxy evidence for quasar activity due to galaxy interaction Slide
Galaxies Associated with Quasars Two images of the same quasar, 1059+730 New source probably a supernova in the host galaxy of the quasar Slide
Host Galaxies of Quasars Host galaxies of most quasars can not be seen directly because they are outshined by the bright emission from the AGN. Slide Blocking out the light from the center of the quasar 3C 273, HST can detect the star light from its host galaxy.
Gallery of Quasar Host Galaxies Slide Elliptical galaxies; often merging / interacting galaxies
Quasars 1) Spectra contain strongly redshifted lines indicating large cosmological distances to the objects Gravitational lensing also indicates huge distances This means that quasars are most luminous objects in the Universe! L ~ 10 12 10 14 L sun 2) Broad emission line as in Seyferts, indicating rapid motion 3) Jets, intense radiation from radio waves to gamma-rays observed 4) Host galaxies are found around nearby quasars 5) Rapid variability on the scale of days is observed 1)-5) indicate that quasars sit in the centers of galaxies, are extremely compact and super-luminous. They are similar to AGN! Slide
Slide
Quasars were much more numerous in the early Universe than now Galaxy collisions were more frequent; they supplied more stars and gas to the central black holes Collisions and mergers play crucial role in the AGN activity Modern galaxies with central black holes are sleeping quasars?! Slide