Secular Evolution of Disk Galaxies Bruce G. Elmegreen IBM T.J. - - PowerPoint PPT Presentation

Secular Evolution of Disk Galaxies

Bruce G. Elmegreen IBM T.J. Watson Research Center Yorktown Heights, NY

“Dynamics of Disk Galaxies,” Seoul National University, South Korea October 21-24, 2013

Reionization requires a SFR ~ 0.018([1+z]/8)3

(Shull +12)

which is consistent with reionization by z~7

Star Formation History of the Universe

Blue = direct from uv light Yellow = dust corrected Bouwens +11

Saintonge +13: Depletion time decreases as 1/(1+z) and gas becomes more molecule-rich with z

Depletion Time fgas

d Mgas dt = d Mgas dt accretion

• SFR

where SFR = A Mgas solution for time >> d Mgas dt accretion Mgas SFR is SFR ~ n

On cosmological scales with short depletion times, accretion determines the star formation rate

For a galaxy:

Dave, et al. (2012): “equilibrium SF”

Star Formation History of the Universe

Blue = direct from uv light Yellow = dust corrected Bouwens +11 High accretion rates

Ceverino, Dekel & Bournaud 2010

160 kpc x160 kpc Disk

Chaotic galaxy formation in a simulation with cold and hot flows: wild instabilities and big clumps σ/Vrot high because of inflow energy & high Σgas (QToomre~1) clump size/galaxy size ~ (σ/Vrot)2

Newman +12:

Kinematic properties of clumps in a z~2 galaxy

Blue-shifted line wings suggest massive winds σ ~ 85-290 km/s Vwind~400-800 km/s

M/SFR ~ 1-8

see also Genzel +11 for other clump winds; Steidel +10; Pettini +00; Weiner +09 for general galactic outflows

.

Bournaud +13: Do clumps survive?

• realistic feedback
• radiation press.
• HII regions
• supernovae

Clumps always accrete new gas even as they expel gas in a wind and lose stars by tidal forces

Clump 2C: continuous clump merging & accretion along filaments

Bournaud +13

Bournaud +13

Clump accretion burst SF burst & Wind burst + stellar loss

Bournaud +13 youngest clumps contain old field stars that fell in during the clump-forming instability

• ld clumps have formed stars

continuously and also lost some stars that they formed

Persistence of clump torques

• Observed ~200 Myr stellar age (EE05-09, Wuyts +12)

– may correspond to ~400 Myr real age

• Clump mass dominated by gas (5xMstars – Bournaud +13)

– 108 – 109 MO produce strong torques

• The high fraction of clumpy galaxies (40% at z~2 in E05 &

Wuyts +12) implies that clump formation is a long-lived phase (even if individual clumps last 200-400 Myr) High clump torques persist for 1 Gyr or more at z>1

• Cacciato +12 models make clumpy disks until z~0.5

because of continued gas accretion (e.g., Dekel +09, )

– many epochs of clump formation/migration/destruction – disks stabilize when stars dominate gas (clumps spirals)

Star Formation History of the Universe

Blue = direct from uv light Yellow = dust corrected Bouwens +11 clumpy phase

Fathi +12: Disks are exponential even at z=5.3, when the universe is only ~1 Gyr old. Exponential disks need to form quickly. Before spiral arm and bars (which appear a z~1-2) Young disks are also small: 15% today’s size at z=2-5.8

Clump interactions can make exponential disk + bulge

Bournaud, Elmegreen & Elmegreen 07

Clumpy disks should also be thick disks

• In Milky Way, the similarity in α/Fe (Melendez +08), K-giant

abundances (Bensby +10) and ages for bulge and thick disk bulge and thick disk formed at the same time.

• High σgas at z~2 (Forster-Schreiber +06, Weiner +06, ...) makes

clumps big and disk thick for the same reason, the characteristic length of a gravitating object is large (~2 kpc)

• L ~ σ2 / πG Σ

Milky Way thick disk (defined kinematically) has metallicities up to solar. The most metal rich are ~8-9 Gyr old. The thick disk formed from ~12 to ~9 Gyr

The Milky Way’s thick phase lasted until ~9 Gyr ago

9 Gyr ago 12 Gyr ago z=1.3 to... z=3.6 Bensby +07 Redshift Universe Age

Minchev et al. 2013 Disk model with accretion and minor mergers. Stellar dispersion increases with age in the solar vicinity. The rapid rise at 8-12 Gyr is from chaotic structures in the young disk

Star Formation History of the Universe

Blue = direct from uv light Yellow = dust corrected Bouwens +11

Thin disk phase Thick disk & clumpy phase

Question: Is all the z~2 SF going to the thick disk?

Thick / Thin disk mass ratios versus circular velocity

Assumes conservative ratio of [ M/L ]Thick / [ M/L ]thin ~ 1.2 (gives the least thick disk mass) Star mass only Star + gas mass Comeron, Elmegreen, et al. 2012 (based on Spitzer 3.6 mm survey of vertical disk light profiles)

Thick disks contain about 1/3

• f the total disk mass

The thick disk stellar mass is 1-4 x 1010 MO for Vc = 200 km/s. That is from a high SF rate of 10-20 MO/yr for ~ 1-2 Gyr (the duration of the clumpy phase) The thin disk stellar mass is 4-8 x 1010 MO for Vc = 200 km/s. That is from a low SF rate of 4-8 MO/yr for ~ 10 Gyr (the duration of the spiral phase) Thin disk star mass Thick disk star mass

Vcircular Answer: YES, all SF before z~2 can be in the thick disks today and all SF after z~2 can be in the thin disks

Bars are not formed by, nor drivers of, secular evolution at z>1 Suggests: thick/clumpy phase does not make bars.

Sheth + 2010

80:0:20

Bournaud & Elmegreen 09

The clumpy phase is disk-dominated (>80% stars + gas in disk). Thick disks, bulges and stellar halos (from the clumpy/merger phase), precede the main spiral phase.

When do spirals appear?

varying percentage of disk:bulge:halo Gas Stars

Law +12: z=2.18 spiral galaxy: the oldest spiral found

• probably tidal, thick arms (high velocity dispersion)

Grand Design Multiple Thin Arm “Woolly” “Irregular Long Arm” Flocculent color = ACS B, V, I B/W = WFC3 H band

z=1.24 z=0.92 z=1.03 z=0.29 z=0.60 z=0.53 z=0.50 z=0.78 z=1.11 z=0.25 z=0.03 z=0.50 z=0.12 z=0.63 z=1.37

Elmegreen +13: Looking for thick or irregular arms in the HST UDF

Intermediate?

“Woolly” “Irregular Long Arm”

z=1.32 z=1.41 z=0.78 z=0.95 z=0.69 z=0.47 z=1.40 z=2.58 z=0.51

More examples of the intermediate types:

(Not a resolution difference: new types span a wide range of redshift, and beyond z~1, spatial resolution is about constant anyway.) H-band

• Earliest spiral in the UDF: z~1.8
• Earliest Multiple Thin-Arm at z~0.6
• Multiple Thin-Arm and Woolly galaxies are largest and brightest
• Flocculents are the faintest

Star Formation History of the Universe

Blue = direct from uv light Yellow = dust corrected Bouwens +11

Thin disk & bar/spiral phase Thick disk & clumpy phase

SUM: Cosmological Secular Processes

• clump stirring z > 2

thick exponential disk, bulge, spheroid

• bars and spirals appear z~1-2

– thin disk phenomena – need stabilizing bulge/thick disk/spheroid

• simultaneous with non-secular processes:

– mergers, minor mergers, gas accretion

Modern Secular Processes

• Angular momentum transfer inside disk and

between disk and halo – formation of “pseudobulges” (radial accretion) – bar growth and slowing down

• Stellar scattering*
• disk growth
• formation of “outer disk,”
• formation of exponential profiles
• Bar thickening, vertical resonance
• plus continued accretion and mergers

Modern Secular Processes

• Angular momentum transfer inside disk and

between disk and halo – formation of “pseudobulges” (radial accretion) – bar growth and slowing down

• Stellar scattering*
• disk growth
• formation of “outer disk,”
• formation of exponential profiles
• Bar thickening, vertical resonance
• plus continued accretion and mergers

Hohl 1971 initially uniform disk forms bar forms exponential profile

Sellwood & Binney 2002: spiral growing from groove instability scatters stars around corotation: Stars initially outside CR lose angular momentum and stars inside CR gain angular momentum.

A stellar disk forms many transient spirals that scatter stars around their CRs all over the disk.

Sellwood & Binney 2002

• 1. pattern speeds
• 2. stellar L changes
• 3. CR features highlighted

R-final histograms

Minchev et al. 2012: multiple spirals stimulated at resonances, all with different pattern speeds. Stellar angular momentum scatters at CRs, but mostly with the bar. Scattering elsewhere is not CR scattering, but scattering in the time-changing potential of multiple spirals.

• - see also Roskar et al.

Scattering: extends disk, turns an exponential with a sharp edge into double-exponential

Bar CR moving

• ut with time

Bar OLR moving

• ut with time

Outer disk moving

• ut with time

Late type has less disk extension because the extra gas causes more central accretion and a weaker bar, which then scatters less to the main disk, which is therefore less unstable to form spirals.

Early Hubble Type, S0 Intermediate, Sa Late Hubble Type, Sb Minchev +12

Barker + 2012: An extended outer disk in spiral galaxy NGC 2403

(image 24 kpc on a side)

• ld stars
• ld

young

SF to 8 scale lengths

Grossi +11: M33 far outer disk, stars extend further than gas, but SF (200 Myr old) still at 60’ where Σgas ~ 1 MO/pc2 Stars Gas

young stars

• ut to here

young stars still here SF out to ~10 inner disk scale lengths SF to here

Radburn-Smith+12: NGC 7793 radial migrations with spiral scattering can explain the disk profile. Star formation for a small distance into the outer disk, down to ΣHI ~ 1 MO/pc2

Hunter + 11: A deep study of the outer disks of 5 dwarf irregular galaxies

• - no spirals, HI>>H2, gas >> stars, extreme-low SFR

DDO 133 (Barred)

V, Hα, NUV images: DDO 133 (ellipses = breaks & outer limit)

V, Hα, NUV images: DDO 53 (ellipses = breaks & outer limit)

End of SF at ΣHI ~ 1 MO/pc2 or beyond

DDO 53 non-barred

Hunter +11

5 Im type galaxies with extended exponential disks, types I or II: none have spirals there is relatively little shear

• nly one is barred

What causes exponential shapes? All are clumpy: perhaps the clumps make exponential profiles in the absence of spirals and bars (as in the high redshift models).

Hunter +11

Stellar Surface Densities from (V, B-V) using Bell & de Jong 2001 can be remarkably featureless

Hunter +11

Down to 0.01 MO/pc2 ~ 1/6000 solar neighborhood

Hα: SFR in last 10 Myr FUV: SFR in last several 100 Myr V: SFR in over 1 Gyr

(NGC 4163 is odd with disturbed HI velocities)

SF follows (?) or maintains (?) the exponential disk can stellar scattering alone do this? do we need SFR proportional to existing stars (Shi +11)? Ratio of recent to past SFR is close to unity throughout the visible disk

Hunter +11

Elmegreen & Struck 2013: What makes exponential disks in dwarf irregular galaxies? test particles in initially flat disk. scatter off point masses

• near-Flat RC case:

70 clumps @ 0.002MH each

• near-Solid RC case:

6 clumps @ 0.15MH each

near-Flat RC near-Solid RC after 2.7 Gyr after 0.49 Gyr

6 clumps, 0.15MH each 12 clumps, 0.075MH each 24 clumps, 0.0375MH each 12 clumps, 0.075MH each 12 clumps, 0.0375MH each 48 clumps, 0.01875MH each

(7.8 Gyr) (7.8 Gyr) (3.6 Gyr) (3.6 Gyr) (3.6 Gyr)

Exact Solid Body RC Flat Body RC

(7.8 Gyr)

• Solid Body RC
• slow evolution (shear only from asymmetric drift)
• slope keeps increasing
• Flat RC
• faster evolution (shear increases scattering rate)
• fills outer disk with continuing exponential
• slope stops growing

(7.8 Gyr) (3.6 Gyr)

Summary of Secular Processes

• z>2: high accretion rate causes clumpy SF

structures, which drive secular processes of angular momentum transfer (bulge formation) and disk thickening

• z<1: slower accretion causes turbulent speeds

to drop, then clumps get smaller, thin disk forms, bar instability may happen, and new secular processes begin: bar and spiral torques, bar and spiral stellar scattering

• Downsizing: continued “high” accretion rates in

dwarf irregulars, producing SF clumps and clump-related torques & scattering even today