Spiral Galaxies

From our study of the Milky Way:

• Rotating disk• Exponential in R an z• Thick and thin disks• Spiral arms and bar• Stellar populations:

• Star Formation• Chemical Evolution

• Stellar halo

What about other spiral galaxies? How can lessons learned in the Milky Way be generalized to spiral galaxies as a whole?

Spiral Galaxies: Structure

Spiral galaxy luminosity profiles are reasonably well described by a varying combination of disk and bulge profiles.

Disk: linear surface brightness profile (mag arcsec−2), so exponential fall-off with radius. Characterized by:• central surface brightness: µ0• radial scale length: hR

Central bulge: profile varied• classical r−1/4 profile (early mergers?)• exponential profile (“pseudo-bulges”,

maybe from scattering of disk stars?)

Disks also typically have stellar halos, butthese are too diffuse to be observed inintegrated light.

NGC 7331, Hager & Grossman

Spiral Galaxies: Structure

But spiral galaxies are not perfectly symmetric. Fit ellipses to isophotes. If the disk is axisymmetric (but inclined), ellipses should be constant with radius.

Changing isophotes ⇒ asymmetry: best fit ellipse as a function of radius can change orientation (“position angle” or PA), ellipticity, and deviation from ellipse (A4).

“Fourier modes” (m): At some radius, think of galaxy surface brightness as a function of angle. If the disk is axisymmetric, this would be a constant.

m=0: constant, axisymmetricm=1: one hump, lopsidedm=2: two humps, bar or grand design spiral armsm=3+: multi-armed spirals, other weirdness

Many different metrics exist to quantify asymmetry.

de Souza+ 04

µ

PA

Ellip

A4

Raw image After axisymmetric model subtraction

Morphology depends on wavelength!

UV/blue: see young stars, spiral arms enhanced.

red/near-IR: dominated by light from old stars. Smoother, emphasizes central bulge

mid-IR: see emission from dust heated by young stars, traces spiral arms.

Spiral Galaxies: Vertical Structure

from Freeman 02 ARAA

NGC 4762

IC 5249

Many disk galaxies observed to have a thick disk, but not all.

Disks also are seen to flare or warp in the outskirts.

Why might this be?• Interactions?• Low density in outskirts: no restoring force for

vertical oscillations𝜎!" ≈ 2𝜋𝐺Σℎ!

Spiral Galaxies: Color Profiles

Disks typically show color gradients:

• Galaxy colors become bluer in the outskirts

• Fitted scale lengths (hR) are shorter in the red.

M101 color gradient(false color, but maps real color)

Mihos+ 13

Color gradients in other spiralsde Jong 1996

Spiral Galaxies: Color Profiles

M106

Watkins+ 16

But sometimes color gradients reverse in the extreme outskirts.

Spiral Galaxies: Metallicity Gradients

Bresolin 07

Spiral galaxies typically also show metallicity gradients: outskirts are more metal poor than inner regions.

Spiral Galaxies: Metallicity

Tremonti+ 04

Low mass galaxies are typically also metal-poor.

Spiral Galaxies: Atomic/Molecular Gas

Spirals are rich in cold atomic and molecular gas – fuel for star formation• Atomic: probe via 21cm neutral hydrogen line• Molecular: no H2 emission, so trace via other molecules (CO, etc)

𝑀#$% = 𝑀&' + 𝑋×𝐼()

whereMHI = HI massICO = CO line fluxX = scale factor (uncertainty!)

𝑓! =𝑀!"#

𝑀!"# +𝑀#$"%#

luminosity → surface brightness → size → (blue) color (red)

McGaugh & de Blok 97

Spiral Galaxies: Atomic/Molecular Gas

log

M(H

2)/M

(HI)

log

M(H

2)/M

(HI)

Young & Knezek 89

Late type spirals are HI-rich but H2-deficient

Spiral Galaxies: Atomic/Molecular Gas

Kennicutt & Evans ARAA 2012

HI typically much more extended thanmolecular gas

star formation

atomic gas molecular gas

Spiral Galaxies: Atomic/Molecular Gas

Hitschfeld+ 09

Molecular Gas Atomic Hydrogen

Recapping…

Disk galaxies show trends

Low luminosity / low surface brightness galaxies: lower in metallicity, bluer, more gas-rich, and their gas is mostly atomic.

High luminosity / high surface brightness galaxies: higher in metallicity, redder, lower gas fractions, more of their gas is in molecular phase.

Disk galaxies show gradients

Outer regions are lower in metallicity, bluer, more gas-rich, and atomic gas dominated.

Inner parts are higher in metallicty, redder, less gas-rich, and more molecular gas.

Note: many exceptions exist, enough so that there are whole classes of galaxies that defy these relationships. But as a general thumbnail sketch of galaxies, this is reasonable.

Spiral Galaxies: Star Formation rates

Σ*+, ~ Σ#$%-

Global star formation rates (SFR): ~ 0.01 − 100 M☉/yrMilky Way SFR ~ 1 M☉/yr

Gas depletion times: 𝜏 = .!"#*+,

~ 10/ − 1001 yr

SFR often parameterized via Schmidt (1959) law

Kennicutt & Evans ARAA 2012

N>1 means there’s more to SFR than just having fuel.

• Collisions of gas clouds? 𝑁~2

• Gravitational collapse? 𝑁~1.5

𝑡234~1𝜌#$%

→ 𝑆𝐹𝑅 ~ ⁄𝜌#$% 𝑡234 → 𝑆𝐹𝑅 ~𝜌#$%0.6

Σ&'( is SFR per areaM☉/yr/kpc2

Spiral Galaxies: Star Formation efficiency

“Efficiency”: how much of the gas is converted to stars over 108 yrs.

Another way to think of it is a changing efficiency as a function of density:

“efficiency”: fraction of gas converted to stars over a characteristic timescale.

Higher gas density ⇒ higher star formation efficiency

Kennicutt & Evans ARAA 2012

SFR correlates directly (N=1) with densemolecular gas!

Spiral Galaxies: Star Formation efficiency

Density drives star formation, but via the formation of molecular gas.

Low density environments (galaxy outskirts, low surface brightness galaxies) are rich in HI, but cannot easily form molecular gas.

Star formation in these environments is weak and sporadic.

Evolution is mild: low SFRs, low surface brightness, low metallicity.

Characteristic colors are bluer: less integrated SFR to build up older populations.

👈HSB: M101

👉LSB: Malin 1

Measuring Rotation: Longslit spectroscopy

Place a spectrograph slit across the major axis of a galaxy, measure the wavelength of emission lines, get velocities from Doppler shift.

Need to correct rotation velocity for inclination of galaxy.

𝑉7,9:% = 𝑉7,;<=> sin 𝑖

face-on: 𝑖 = 0°edge-on: 𝑖 = 90°

McGaugh+ 01

H⍺

[NII]

[SII]

Inte

nsity

Wav

elen

gth ⟹

Position ⟹

Velo

city

[km

/s]

Position along major axis [arcsec]

Measuring Rotation: 21-cm line / Single dish radio telescopes

Look at HI 21-cm emission line.

Drawback: Individual radio telescopes have relatively poor spatial resolution. So we can’t map the rotation curve in detail, but we can still estimate the rotation speed of a galaxy.

If we add up all the 21-cm emission from a rotating galaxy, what would we expect the emission line profile to look like?

W20

𝑊"1 ≈ 2𝑉7 sin 𝑖

Measuring Rotation: 21-cm line / Single dish radio telescopes

Not every galaxy is so well behaved, though…..

NGC 4351

Measuring Rotation: 21-cm interferometry

Radio telescope arrays use interferometry to get good spatial resolution. Image the galaxy in 21cm, scanning the receiver in wavelength.

Each spot in the galaxy yields a 21-cm line spectrum, which tells you HI intensity, velocity, velocity width (dispersion). Construct a 2D map of the projected velocity field.

Fit to a 2D rotating disk model to derive rotation curve.

Measuring Rotation: Integral Field Spectroscopy

Instead of a spectrograph slit, use densely packed bundle of optical fibers to put light into the spectrograph.

Provides 2D mapping, like interferometry, but gives full optical spectrum at every position : emission lines (gas kinematics, metallicity), absorption lines (stellar kinematics, metallicity), spectral shape (stellar populations), etc!

Chun and Park 17

Velo

city

Disp

ersio

n

Rotation Curves of Spiral Galaxies

Systematic Trends

More luminous galaxies rotate faster: Tully-Fisher Relation

Rotation curves are generally flat in their outskirts. Falling rotation curves almost always due to• Bright inner disk/bulge (high stellar density)• Disturbances in outer disk (non-rotational motion)

Rotation curves always rotate too fast for their total stellar mass ⇒ dark matter or modified gravity.

Persic+ 96

Tully Fisher Relation

Sakai+ 00

Calibrated TF relations from HST Cepheid distances

Connection between circular velocity and total luminosity

Slope and scatter both provide important astrophysical constraints.

As we measure the TF relation in redder bands:• Slope gets steeper• Scatter goes down

In near-IR, scatter in well-defined samples is consistent with purely observational error: no intrinsic scatter!?!

Systematic uncertainties:• Dust correction• Inclination correction• Vrot measurement (W50, W20, Vmax, Vflat, etc)

𝜎 ≈ 0.3 − 0.5

𝜎 ≈ 0.1

Tully Fisher Relation: Implications

Luminosity traces stars, velocity traces total mass. The TF relation demands a tight coupling between baryons and dark matter.

Mass couples with velocity and size: 𝑀;9; ∝ 𝑉7"𝑅Luminosity couples with luminosity density and size: 𝐿;9; ∝ 𝐼1𝑅"

Plotted as absolute magnitude (−2.5 log 𝐿) versus log 𝑉7, this would yield a TF slope of −10, close to what we see in the near-IR.

TF does not depend on surface brightness (I0). So that means

𝑀𝐿 ∝ 𝐼1?1.6

M is dominated by dark matter.L is determined by stars.

The dark matter knows about the baryons.

Sprayberry+ 95

• = LSBx = normal

So we would expect 𝐿;9; ∝@$%

'&'(

)

The Baryonic Tully-Fisher Relation (McGaugh 05, etc)

Instead of correlations between light and velocity, look at the connection between baryonic mass and velocity.

Convert light to stellar mass using stellar population models 👉

𝑀∗ =𝑀𝐿 ∗

𝐿

Low luminosity galaxies deviate from TF, they are too low.

Add in gas mass 👉

𝑀: = 𝑀∗ +𝑀#$%

Wow.

slope= 3.94± 0.07 (ran)± 0.08 (sys)

scatter consistent with pure observational uncertainties.

Why Tully-Fisher is so important:

It demonstrates a very tight connection between baryonic matter (normal stuff) and gravitational motion.

• Dark matter models: very tight coupling between baryonic and non-baryonic matter. Doesn’t come naturally from models. Solutions involve a lot of model-tuning.

• Alternative gravity models: more than just Newton/Einstein gravity.

It can be a useful tool for getting distances.

Circular velocity is distant-independent. Measure it for a galaxy, use TF to get the galaxy’s absolute magnitude, couple that with the measured apparent magnitude, get a distance.

𝑀 = 𝑎 log𝑉7 + 𝑏

It can be a useful tool for studying galaxy evolution.

When galaxies deviate from the mean, it tells you their kinematics are screwy or their mass-to-light ratio is different. If you see this systematically as a function of redshift, environment, or galaxy type, you learn about how galaxies differ in these respects.