How do galaxies evolve into the forms they have today? Spheroid - - PowerPoint PPT Presentation

Disk Galaxies Spheroid Galaxies

Susan Kassin (Space Telescope Science Institute) Hubble 1936

Raymond Simons (Johns Hopkins), Camilla Pacifici (STScI), DEEP2 & SIGMA Survey teams, VELA simulation team

How do galaxies evolve into the forms they have today?

Galaxies at a redshift of 1 (8 billion years ago)

credit: AEGIS Survey

NGC 4414, credit: HubbleSite.org NGC 1132; credit: Hubble Heritage

What are galaxies?

How and when are disks assembled? How & when do disks obtain their current well-ordered state?

M63, credit: HubbleSite.org

Disk galaxies rotate under the influence of gravity

Weiner incl. Kassin et al. 2006

+200 km/s

• 200 km/s

Velocity map of a local disk galaxy

Sandage & Bedke 1988

NGC 7171 30”

How and when do quiescent spheroids form?

NGC 1132; credit: Hubble Heritage

• Disk Formation
• “Mass of Disk Formation”
• Assembly of disks over cosmic time
• Spheroid Formation
• Formation timescales vary with galaxy mass and redshift
• JWST observations of nascent galaxies

Outline

Models of Disk Formation and Evolution

Background

• 1. Angular momentum acquisition

e.g., White & Rees 1978; Fall & Efstathiou 1980; Blumenthal et al. 1984; Mo, Mao & White 1998; Dalcanton, Spergel, & Summers 1997

• 2. Baryons collapse to a disk
• 3. Collapse from the inside out (onion skin)

Analytic Model of Disk Collapse

• Disks start off well-ordered
• They grow in radius in an “onion skin” manner

Analytic & numerical theory: Tully-Fisher is tight

magnitude rotation velocity (km/s)

e.g., Mo, Mao, & White 1998; Sales et al. 2017

Luminosity ~ Velocity3

Dynamical theory of isolated disk evolution

Aumer, Binney, & Schönrich 2016 Stellar velocity dispersion (km/s) Time (Gyrs)

• Molecular clouds (e.g., Aumer, Binney, & Schönrich 2016)
• Merging satellites (e.g., Velazquez & White 1999)
• Buckling of bars (e.g., Debattista et al. 2006)
• Spiral structure
• Minor mergers (e.g., Moster et al. 2011)

Stars start off in a well-ordered and thin disk. Their velocity dispersion increases with time via:

Analytic Theory of an isolated disk

• Disks start off well-ordered and thin
• They thicken with time
• Disks lie on a tight Tully-Fisher Relation
• Disks grow in radius in an “onion skin” manner

We examine this picture with observations

• f galaxy kinematics at a range of redshifts.

• Disk Formation
• “Mass of Disk Formation”
• Assembly of disks over cosmic time
• Spheroid Formation
• Formation timescales vary with galaxy mass and redshift
• JWST observations of nascent galaxies

Outline

To study the evolution of galaxy kinematics, we need spectra and Hubble images for a hundreds of representative galaxies over a significant range in redshift and stellar mass.

• Star-forming galaxies on the “main sequence”
• No cuts on morphology
• Spectra from DEEP2 & SIGMA Surveys

(Newman et al. 2012, Simons, Kassin et al. 2017)

• Hubble images from AEGIS & CANDELS (Davis et al. 2007, Grogin et al. 2011)
• Total stellar mass is measured from multi-band photometry

Galaxy kinematics from z=3 to now

• We measure σ, which had not been done before, and Vrot
• We corrected Vrot and σ for the effects of seeing, which had not been done before.
• Slits are aligned to within 45° of galaxy major axes to measure Vrot
• We correct Vrot for inclination using Hubble image

Weiner inc. Kassin et al. 2006, Kassin et al. 2007, Covington, Kassin et al. 2010, Kassin et al. 2014

Hα

Galaxy kinematics are measured from emission lines

Simons, Kassin et al. 2016

z=2

wavelength radius radius velocity velocity dispersion radius

175 km/s 70 km/s

“Local” Tully-Fisher Relation at z=0.2

log rotation velocity Vrot (km/s)

Tully-Fisher “Ridge-line” from Reyes et al. 2011

log stellar mass (M⦿)

Simons, Kassin et al. 2015; also Bekeraite et al. 2016 (CALIFA), Bloom et al. 2017 (SAMI)

Ordered disks lie on ridge-line, Disturbed galaxies lie off of it

“Local” Tully-Fisher Relation (z~0.2)

Simons, Kassin et al. 2015; also Bekeraite et al. 2016 (CALIFA), Bloom et al. 2017 (SAMI)

“Local” Tully-Fisher Relation (z~0.2)

Simons, Kassin et al. 2015; also Bekeraite et al. 2016 (CALIFA), Bloom et al. 2017 (SAMI)

disks will form disks may not form

“Mass of Disk Formation”

Analytic Theory of an isolated disk

• Disks start off well-ordered and thin
• They thicken with time
• Disks lie on a tight Tully-Fisher Relation
• Disks grow in radius in an “onion skin” manner

Tully-Fisher falls apart for local low mass galaxies. Local low-mass star-forming galaxies are often not disks.

• Disk Formation
• “Mass of Disk Formation”
• Assembly of disks over cosmic time
• Spheroid Formation
• Formation timescales vary with galaxy mass and redshift
• JWST observations of nascent galaxies

Outline

6”

s dominated

z~1 HST

mixed

z~1 HST

z~1 HST

Vrot dominated

Gas kinematics tell us about the physical state of galaxies

75 km/s 55 km/s 29 km/s 59 km/s Vrot Vrot sini = 208 km/s s= 40 km/s spatial, 8” 1” 0.75”

6”

s dominated

z~1 HST

mixed

z~1 HST

z~1 HST

Vrot dominated

Gas kinematics tell us about the physical state of galaxies

σ is a gas velocity dispersion.

• Integrates over velocity gradients on scales below the seeing
• Quantifies disordered motions in galaxies

(Weiner et al. 06, Kassin et al. 2007, Covington, Kassin et al. 2010)

• Does not indicate a thick disk like in the Milky Way (Gilmore & Reid 1983)

1” 0.75”

Tully-Fisher Relation at z~0.2

Kassin et al. 2007

log Vrot (km/s) log stellar mass (M⦿)

log stellar mass (M⦿)

Tully-Fisher Relation from z~0.2 to z~2

Kassin et al. 2007; Simons, Kassin et al. 2016

log Vrot (km/s)

z~0.2 z~0.5 z~0.8 z~1 z~2

log stellar mass (M⦿) log stellar mass (M⦿) log Vrot (km/s) log S0.5 (km/s) S0.5

2  0.5Vrot 2 + s2 z~0.2 z~0.5 z~0.8 z~1 z~2

Do Vrot and σ evolve with time?

Kassin et al. 2012, Simons, Kassin et al. 2017, σ trend also Wisnioski et al. 2015 & Turner et al. 2017

The evolution of galaxy kinematics

redshift Vrot (km/s) σ (km/s) redshift

109 – 1010 M M★ = 1010 – 1011 M

Representative star-forming galaxies, no cut on morphology

NGC 4388, credit: ESA/Hubble & NASA

Let’s characterize disk galaxies:

• Vrot/σ ~ 10 (local massive disks)
• Vrot/σ > 3 (analog of local low mass disks)
• Vrot/σ > 1 (barely rotation supported)

How does the fraction of disk galaxies evolve?

z=1, Hubble/ACS V+I

V/σ = 5.6 V/σ = 4.9 V/σ = 3.4 V/σ = 2.5 V/σg = 1.8 V/σg = 1.6

V/σ > 3 V/σ < 3

V/σ is correlated with visual morphology

6” Ordered kinematics & morphology Disordered kinematics & morphology

Evolution of the fraction of the disk fraction

• f star-forming galaxies

1010 – 1011 M 109.5 – 1010.5 M 109 – 1010 M fraction of star- forming galaxies with V/σ>1 redshift

Kassin et al. 2012; Simons, Kassin et al. 2016 & 17 V/s = 1.6 z~1 HST

Quantitatively, surveys agree

1010 – 1011 M 109.5 – 1010.5 M 109 – 1010 M fraction of star- forming galaxies with V/σ>1 redshift

Kassin et al. 2012; Simons, Kassin et al. 2016 & 17

Disk Assembly

At z~2, most galaxies were not disks

fraction of star- forming galaxies with V/σ>3 1010 – 1011 M 109.5 – 1010.5 M 109 – 1010 M

z~1 HST V/s = 3.4 Kassin et al. 2012; Simons, Kassin et al. 2016 & 17

redshift

Disk Settling

Analytic Theory of an isolated disk

• Disks start off well-ordered and thin
• They thicken with time
• The Tully-Fisher relation has little scatter
• Disks grow in radius in an “onion skin” manner

Disks start off with lots of disordered motions. They lose disordered motions and “thin out” with time. Disks grow in radius, but not in an orderly manner. Tully-Fisher has large scatter to low rotation velocity.

My observations

Nbody Shop at U. Washington (courtesy Governato)

• Disk Formation
• “Mass of Disk Formation”
• Assembly of disks over cosmic time
• Spheroid Formation
• Formation timescales vary with galaxy mass and redshift
• JWST observations of nascent galaxies

Outline

NGC 1132; credit: Hubble Heritage

How and when do quiescent spheroids form?

• Classical picture that dominated for decades:
• Formed in a single burst in the early universe

(Partridge & Peebles 1967; Larson 1975)

• Mergers of disks can also form an early type

(Toomre & Toomre 1972; Toomre 1977)

• Breakthrough (Bell et al. 2006, Faber incl. Kassin et al. 2007) :
• Quiescent galaxies increase in numbers by a factor of 2-4 over the last 8 billion

years since z~1, so all could not have formed in a single burst in the early universe (see also e.g., Faber et al. 1995)

• Quiescent spheroids must form from star-forming disks

Star-forming galaxies continuously add to the quiescent population, but it’s still unclear how.

How do quiescent spheroids form?

NGC 4414, credit: HubbleSite.org NGC 1132; credit: Hubble Heritage

What is the path from star-forming disks to quiescent spheroids (“quenching”)?

?

stellar mass color What is the path from star-forming disks to quiescent spheroids (“quenching”)?

Pacifici, Kassin et al. 2016

We investigate this with measurements of the star-formation histories of quiescent galaxies.

• Fit the photometry of ~850 quiescent galaxies with a library of galaxy models.

Models vary in terms of SFH, metallicity, stellar mass, age, etc.

• The model SFHs are physically motivated
• Determine the best-estimate SFH for each galaxy by averaging the model

SFHs weighted by their likelihoods

U HST/ACS HST/WFC3 Spitzer/IRAC

data model

magnitude residuals wavelength (microns) Pacifici, Kassin et al. 2016

Measurement of Star-Formation Histories (SFHs) via model fits to galaxy photometry

Realistic star-formation histories are adopted from from simulations

Pacifici, Kassin et al. 2016

Classic SFHs Our SFHs

Delayed Declining Rising Exponential functions of time

Galaxies form hierarchically. This is reflected in their star-formation histories

log specific star-formation rate (10-10 yr-1)

SFHs of nearby galaxies are stochastic

Weisz et al. 2013 Wechsler et al. 2002

SFR (M◉ yr-1) lookback time (Gyr) today past

Best-estimate SFH

Pacifici, Kassin et al. 2016

Best-estimate SFH for a single galaxy: Average over library SFHs weighted by likelihoods

25 best-fit models for the galaxy (greyscale)

Median SFH of galaxies in a bin of stellar mass & z

Median SFH for M* & z bin Interquartile range Individual galaxies (greyscale) Pacifici, Kassin et al. 2016

z~0.3 log M*(M◉) ~ 11 24 galaxies

SFR (M◉ yr-1) lookback time (Gyr)

Pacifici, Kassin et al. 2016

SFR (M◉ yr-1) lookback time (Gyr)

z~0.3 z~0.6 z~0.8 z~1.1 z~1.5 z~2.0

log(M/M)~9 ~9.5 ~10 ~10.5 ~11 ~11.5

stellar mass color What is the path from star-forming disks to quiescent spheroids (“quenching”)?

Pacifici, Kassin et al. 2016

Quiescent galaxies show average trends in SFHs with mass and redshift

Star- formation rate Pacifici, Kassin et al. 2016 time stellar mass color

Dynamical timescales are short (~1 Gyr at z=2), so consistent with gas accretion being stopped quickly and residual gas being used up. Gas accretion is stopped, and galaxies use up their residual gas slowly. Catastrophic event such as stripping or strong feedback.

high mass, high z: form fast, quench fast high mass, low z: form fast, quench slowly low mass, low z: form slowly, quench fast

• Disk Formation
• “Mass of Disk Formation”
• Assembly of disks over cosmic time
• Spheroid Formation
• Formation timescales vary with galaxy mass and redshift
• JWST observations of nascent galaxies

Outline

What will we learn about kinematics from JWST?

Vrot (km/s) σ (km/s)

How do the kinematics of the earliest galaxies evolve?

Kassin Cycle 1 Science

redshift redshift

1010 – 1011 M 109 – 1010 M ?

?

ScienceWise, Magazine, ANU

The JWST/NIRSpec IFU

IFU observations give a spectrum in each pixel. We can create kinematic maps of galaxies from this.

3”

Individual shutters: 0.2" x 0.46"

The JWST/NIRSpec Micro-Shutter Array (MSA)

4 x 365 x 171 = 250k Shutters up to ~100 objects observable simultaneously

3.4’ 3.6’

“Slitlet Stepping” to create IFU-like maps

z=3

Simons, Kassin, Snyder submitted

VELA simulation (Ceverino et al. 2014, 16)

+200 km/s

• 200 km/s

KMOS: 0.2”/pixel, 0.7” seeing NIRSpec: 0.1”/pixel, 0.05” resolution

Mock observations of a simulated merger at z=2

Simons, Kassin, Snyder et al. submitted

Mis-identifying disk galaxies in mock observations of mergers at z~2 Fraction of sight lines where a merger is taken for a disk average angular separation (arcsec)

Criteria 1-5 (Wisnioski+15) Criteria 1-5 + one additional all misidentified none misidentified

Nbody Shop at U. Washington (courtesy Governato)

Cycle 1 proposal: PI Susan Kassin 70 Co-Is JWST NIRSpec R~2700 MSA Survey:

• ~80 Galaxies in 2 areas in the CANDELS

fields

• z = 2 - 6
• ~5-20 hours per “slitlet step”

JWST: Kinematics from z=2 to the end of reionization

This is why I came here!

What our work has shown

• Disk Formation

At high z, galaxies are highly disordered and only just assembling. They later settle to disks. Higher mass galaxies are always the most evolved. There is a “Mass of Disk Formation” in the local universe, above which disks always form, but below which they find it hard to.

• Spheroid Formation

Galaxy quenching time evolves with mass and redshift, its faster at low mass and high redshift.

• JWST kinematic maps will enable us to determine the dynamical state
• f galaxies soon after reionization.