Galaxies Astro 530 Prof. Jeff Kenney. CLASS 12 February 21, 2018 EllipDcal Galaxies: KinemaDcs

Transcription

1 Galaxies Astro 530 Prof. Jeff Kenney CLASS 12 February 21, 2018 EllipDcal Galaxies: KinemaDcs 1

2 Midterm exam next Wednesday Feb 28 Covers up thru lecture of Feb 14 and HW5 (thru disk phenomena, no EllipDcals). Concepts and short problems. Bring calculator. Next HW (#6, on E s) will be due Wed, Mar 7 Galaxy talks find & look over journal papers. Talk to me for further guidance on topic and references. 2

3 Key points on EllipDcal galaxies Largest (most massive) galaxies in universe are E s Oldest galaxies in universe most of their stars formed early in universe; the galaxy may have grown or changed since early universe Appear simple but are complex kinemadcs of stars somedmes reveal surprises 3

4 spectrum of ellipdcal 4

5 Spectra of main sequence stars Cool + red low mass star long- lived Warm + yellow Hot + blue high mass star short- lived 5

6 Cool + red spectra of main sequence stars Warm + yellow Hot + blue spectrum of ellipdcal 6

7 model spectra of galaxies (stellar populadons) at different Dmes afer burst of star formadon simple stellar popula-on: bunch of stars of different masses (with pardcular IMF), all with same age during burst lots of gas emission lines (like extreme Spiral) long afer burst absorpdon lines from stars, no gas emission lines (like EllipDcal) Dme in Gyr afer start of burst 7

8 Compare the spectra of K giant star & S0 galaxy star (K giant) S0 galaxy 2 differences: 1. galaxy spectrum is redshifed wrt MW star (expansion of universe) 2. lines broader in galaxy due to velocity smearing. 8

9 The observed profile of a stellar absorpdon line from a galaxy is generally complex because: 1. The distribudon of stellar velocides at any point in a stellar system may be complex. (Stars in galaxies form a collisionless system, so the stars which pass through any point in a galaxy can have very different orbits.)(this is not true for gas, which is collisional, and generally has simpler kinemadcs) 9

10 The observed profile of a stellar absorpdon line from a galaxy is generally complex because: y x z 2. The observed lines are a weighted sum of stars within a cylindrical 3D volume: all stars along the line- of- sight (z direcdon) and within the spadal resoludon element (x+y direcdons) 10

11 Within each resoludon element extended in x & y direcdons and at each depth z there are y stars on various orbits x Slice at depth z only the LOS (z) component of velocity is observed via doppler shif z direcdon = Line of sight (LOS) GALAXY 11

12 IntroducDon to kinemadcs for EllipDcals Ordered mo-ons: v: mean velocity v = v rot + v noncirc measured by peak or mean of line Disordered mo-ons: σ: velocity dispersion, measured by linewidth The rado v/σ is used to compare the reladve importance of ordered and random modons 12

13 IntroducDon to kinemadcs for EllipDcals Ideally would like to measure v, σ at every point in 3D space v/σ ordered/disordered ObservaDonally v los and σ los are luminosity- weighted averages over finite volumes and along line- of- sight v los /σ los ordered/disordered+ordered (since averaging over large volume includes some ordered component which varies within volume) 13

14 Line- of- sight Velocity DistribuDon (LOSVD) Observed spectrum Spectrum of 1 star LOSVD Velocity dispersion σ - - fit LOSVD with gaussian (even if distribudon is not gaussian!) 14

15 Line of sight velocity distribudon (LOSVD) for spiral disk measure spectrum in small area of spiral km/s 0 km/s +200 km/s V los Small area of outer spiral disk Simple case Fast rotadon modon V los = 200 km/s Small random modons σ los = 20 km/s V los /σ los >>1 FWHM V los V los = mean los velocity σ los = los velocity dispersion = FWHM/2.35 (if gaussian) 15

16 Stellar velocity dispersion in general is anisotropic (different in different direcdons) σ R, σ φ, σ z for stars need not be equal this can happens for stars since they are collisionless pardcles that experience many collisions end up with equal random modons in all direcdons 16

17 τ<100 Myr (O,B stars) ~100 pc random modons of stars in disk of Milky Way is different in different direcdons!! more random modon in radial direcdon than azimuthal and verdcal direcdons!

18 Stellar velocity dispersion in general is anisotropic σ R, σ φ, σ z need not be equal σ los depends on σ R : σ φ : σ (rados) and which z components happen to be along l- o- s Simplest case: isotropic velocity dispersion σ R = σ φ = σ z = σ los (σ los doesn t depend on viewing angle) 18

19 LOSVD for ellipdcal km/s 20 km/s 0 km/s +300 km/s V los FWHM Small area of ellipdcal V los Simple case slow rotadon modon V los ~ 20 km/s large random modons σ los ~ 100 km/s V los /σ los <<1 measure spectrum in small area of E galaxy V los = mean los velocity σ los = los velocity dispersion = FWHM/2.35 (if gaussian) 19

20 Complex case of LOSVD km/s 30 km/s 0 km/s +200 km/s V los FWHM Few stars with large counter- rotadonal modons and small random modons (accreted small galaxy?) V los Few stars with large Main component of line profile rotadonal modons not symmetric and small random modons not gaussian (disk component?) but V los < σ los For most stars random modons dominate 20

21 Are EllipDcals Oblate Spheroids whose shape is governed by rotadon?.. 21

22 Are EllipDcals Oblate Spheroids whose shape is governed by rotadon? Open circles: mid- sized E s (lower luminosity, M B >19.5) Filled circles: luminous (M B <19.5) E s Dowed line: oblate spheroid flawened by rotadon (O.S.F.B.R.) observed (apparent) ellipdcity ProjecDon effects: both V los and ε vary by appx sin(i) for inclined system observed v los (max)/σ los 22

23 E galaxy rotadon vs. dispersion v<<σ NGC 1399 cd ellipdcal Center of Formax Cluster measure spectra at many posidons across E galaxy V rot ~30 km/s σ~250 km/s à v/σ ~0.1 is low!! Most modon in random direcdons! v and σ ~ constant over much of galaxy, although is ~50% higher in center - > ofen reasonable to characterize E galaxy with single value of v and σ 23

24 Are EllipDcals Oblate Spheroids whose shape is governed by rotadon? 2 oblate spheroids with same σ : small v rot large v rot slightly ellipdcal, only a bit flawened by small v rot highly ellipdcal, flawened by large v rot 24

25 Are EllipDcals Oblate Spheroids whose shape is governed by rotadon? Open circles: mid- sized E s (lower luminosity, M B >19.5) Filled circles: luminous (M B <19.5) E s Dowed line: oblate spheroid flawened by rotadon (O.S.F.B.R.) ProjecDon effects: both V los and ε vary by appx sin(i) for inclined system Most mid- sized E s (& bulges) roughly consistent with O.S.F.B.R. Many luminous E s NOT O.S.F.B.R. but must instead be triaxial bodies flawened by anisotropic distribudon of random velocides. 25

26 Triaxial galaxies: flawened by anisotropic velocity distribudon Random modons dominate ordered modons (e.g. rotadon). But random modons can have different amplitudes in different direcdons (x,y,z). 26

27 80% of E s show deviadons from purely ellipdcal isophotes true isophote shape perfect ellipse b Δr r φ a a 4 equadon of ellipse: x = a cos φ y = b sin φ r = [x 2 +y 2 ] 1/2 DeviaDons from ellipses can be described by a fourier series expansion in azimuth Δr(φ) = a 3 cos(3φ) + b 3 sin(3φ) + a 4 cos(4φ) + b 4 sin(4φ)+ cos(4φ) term generally dominates a 0 term size of ellipse a 1 term offset center a 2 term ellipse of different shape a 3 term egg- shaped distordon, generally small 27

28 Disky & boxy isophotes in EllipDcals SchemaDc diagrams of disky isophotes with a 4 /a=0.1 and boxy isophotes with a 4 /a=- 0.1 Disky a 4 > 0 a a 4 Bender etal 1988 Boxy a 4 < 0 a 4 28

29 stellar kinemadcs (V/σ) vs. Disky- Boxy isophote shapes 1.0 (v/σ * ) 0.1 SG Fig 6.15 (Bender) Open circles: lower lum E s (M B >19.5) Filled circles: luminous E s (M B <19.5) (v/σ * ) = (v max /σ) / (v/σ iso ) measured expected for oblate spheroid boxy disky Disky E s have high v/σ (more ordered modon) These galaxies contain significant disks (o.s.f.b.r.) in addi=on to dynamically ho>er component Most boxy E s have low v/σ (more random modon) dynamically hot formed (at least partly) by mergers 29

30 V/σ vs. Luminosity (~stellar mass) Open circles: mid- sized E s (lower lum., M B >19.5) Filled circles: luminous (M B <19.5) E s (v/σ * ) = (v max /σ) / (v/σ iso ) low mass high mass SG Fig 6.15 (Bender) In low mass E s rotadon important In many high mass E s random modons most important 30

31 V/σ vs. Luminosity and Disky- Boxy Open circles: mid- sized E s (lower lum., M B >19.5) Filled circles: luminous (M B <19.5) E s (v/σ * ) = (v max /σ) / (v/σ iso ) SG Fig 6.15 (Bender) Many high luminosity E s are triaxial bodies with low v/σ and boxy isophotes - > Could form through mergers of gas- poor galaxies (less dissipadon during formadon) or low angular momentum mergers(?) Many low luminosity E s are oblate spheroids with high v/σ and disky isophotes à Could form through mergers of gas- rich galaxies (since gas sewles to a rotadng disk before it forms many stars)(more dissipadon during formadon) or high angular momentum mergers(?) Disky E s contain embedded stellar disks, may be part of condnuous sequence with S0 s. 31

32 Stellar velocity fields of slow & fast rotator ETGs full velocity fields (from IFUs) give clearer picture of kinemadcs than single pixel or slit V ETG = Early Type Galaxies = E + S0 intensity V σ Maps show central ~40 (SAURON FOV 33 x41 ) Emsellem etal 2011 ATLAS3D

33 kinemadcs work bewer than isophote shapes to separate 2 classes of E s (depends less on viewing angle) angular momentum parameter λ Re ε true = apparent ellipdcity oblate spheroid line ATLAS3D Emsellem+2011 fast rotators kinemadcs and shape consistent with oblate spheroids (~86% of nearby ETGs) true ellipdcides slow rotators kinemadcs and shape inconsistent with oblate spheroids so triaxial (~14% of nearby ETGs) most massive Es angular momentum parameter (measured within effecdve radius) 33

34 EllipDcal galaxies cd L>>L* (L~2-10L*) generally triaxial, boxy, slow rotators luminous L~L* (L~0.5-2L*) mix midsized L<L* (L~ L*) generally oblate, disky, fast rotators No sharp cutoffs L* = 2x10 10 L sun L MW = knee (break) in galaxy luminosity funcdon why? the most massive galaxies have (on average) experienced the most mergers and mergers of big galaxies without much gas make boxy, triaxial galaxies 34

35 Modified version of Hubble classification scheme showing disky & boxy ellipticals more boxy more disky Normalized SFR Direction of evolution Stellar Bulge/disk ratio Cappellari+2011 Disky (fast rotadng) ellipdcals are a natural extension of the sequence of galaxies with bulges & disks but liwle cold gas or star formadon. Disky E s have higher B/D than S0 s. Their bulges are rotadng. Boxy (slow rotadng) ellipdcals are different. No disks. Not much rotadon. 35

36 Which EllipDcal has odd core kinemadcs? A B C D Franx, Illingworth & Heckman 1989

37 Which EllipDcal has odd core kinemadcs? A B C D Franx, Illingworth & Heckman 1989

38 Which EllipDcal has odd core kinemadcs? Franx+1989 ATLAS 3D Krajnovic+2011 In NGC 4406 (M86), the inner and outer galaxy rotate about different axes! They differ by 90 deg. - > Evidence of merger Stellar velocity map of central +/- 30 in M86

39 KinemaDcally decoupled cores gas driven to center in merger sewles to disk in plane different from outer galaxy undergoes star formadon to form central stellar disk 39

40 KinemaDcally decoupled core in ellipdcal NGC 4365 Davies etal 2001 Inner 7 rotates orthogonally to main body of galaxy Inner 4 disky, outer part boxy Inner part v/σ =1.3 (disklike) 2% of total mass in central decoupled disk Shallow central cusp Overall galaxy is triaxial No shells or other morphological peculiarides No sign of dust Age of stellar populadon is ~14 Gyr in both decoupled disk and main body Formed through old merger(s) NGC 4365 SDSS image

41 KinemaDcally Decoupled Cores in ~30% of E s Abrupt change >20 deg in kinemadc PA Exist in large fracdon of slow rotator E s (more luminous, boxy, triaxial), and in smaller fracdon of fast rotator E s (less luminous, disky, oblate) Most are in slow rotator E s their KDCs are the largest (kiloparsec- scale KDCs) and oldest (older than 8 Gyr) A few fast rotator E s have KDCs whch are more compact (less than a few hundred parsec), and younger (a range of stellar ages from 0.5 to 15 Gyr with 5/6 younger than 5 Gyr). Most are counter- rotadng (McDermid+2006). Krajnovic etal 2008

42 RotaDng both ways at once! the amazing galaxy NGC 4550 NGC 4550 Image looks featureless no dust and no ongoing star formadon an EllipDcal E7 (flawest known E) But the modons of the stars reveal bizarre behavior WIYN BRI

43 Measuring Spectra in NGC 4550 WIYN Sparsepak fiber array spectrograph Obtain a spectrum at each of 90 posidons Measure modons of stars from the Doppler shifs of their spectral lines Fiber posidons on image of NGC 4550

44 Spectra showing modons of stars in NGC 4550 Each spectral line shows the modons of stars, via the doppler shif Lines show 2 peaks one bunch of stars rotadng toward us (blueshifed), the other bunch of stars rotadng away from us (redshifed)

45

46 How did NGC 4550 form? best guess: Merger of very gas- rich galaxy with pre- exisdng stellar disk CounterrotaDng gas sewles to plane of rotadng stellar disk Gas undergoes star formadon once it sewles to disk plane, forming a counterrotadng stellar disk! galaxies with very extended counterrota<ng stellar disks are rare 46

47 Evidence for accredon & mergers in E s Disturbed dust & gas Shells & ripples & outer Ddal features in stellar distribudons KinemaDcally disdnct cores Major merger remnants resemble E s Large fracdon (>50%) of E s show evidence for merger/accredon more in boxy than disky, but also many disky à E s condnue to increase their mass with age