Are Young Galaxies Visible?R.B. Partridge & P. J. E. Peebles 1967

(Astrophysical Journal, vol. 147, p.686)

Christian Herenz

Extragalactic Science Club 2010

December 14th, 2010

Citations from ADS

The Authors

R. B. PartridgeP. J. E. Peebles

Outline

I Aim, Method & ResultsI Galaxy Model

I Time of FormationI Luminosity

I Element ProductionI Stellar Luminosity Function

I SpectrumI Predicted ObservabilityI Unsuccessful Hunt for Primeval GalaxiesI Properties of Lyα-Emitters

Are Young Galaxies Visible?

I Aim: To asses possibility of detection of individual galaxiesin an early evolutionary stage.

I Method:I Theoretical model of galaxy formation and evolutionI Model based purely on assumptions, (almost) no

connection to observations. (exc. CMB)I Derivation of observational properties of these “Young

Galaxies”.I Results:

1. “Young Galaxies” go through a stage of high luminosity(after t ∼ 108 years for ∆t ∼ 107 years)

2. For a Milky-Way like Galaxy: L ∼ 1046 erg s−1

3. 6 - 7 % of total L in Lyα-Line

The Model for the Formation of Galaxies

I Model to Adress the 2 Questions:1. Time of formation2. Luminosity of the Galaxy

I Focus on: Milky Way like galaxies(because observational values needed for computationsare available here)

I Monolithical collapse:Galaxies evolve in isolation. No dark matter.

I 2 important consequences in their picture:1. Halo stars preserve memory of the maximum size of

proto-galaxy.2. Onset of star formation = well defined moment at earliest

stages of inital collapse

Time of Formation (1)- Spherically symmetric modell- Newtonian Approximation

R(t) = − GM

(R(t))2 (1)

- Solution to Eq. (1) in parametric form1:

R(η) =

(3M

32πρp

)1/3

(1−cos η) ; t(η) =

(3

32πGρP

)1/2

(η−sin η)

(2)- Material stops expanding (i.e. η = π) at

tP =

√3π

32Gρp(3)

(tP - Protogalaxy formation time)1Multiplication with R leads to DEq. of Cycloid

Time of Formation (2)Input in (3):

I Mass of Protogalaxy:

MP = 1.2× 1011M (4)

I Radius of Protogalaxy:

RP = 20 kpc (5)

Then Eq. (3) gives:

tP = 1.4× 108 years (6)

From Eq. (2) and (3) we see→ free fall time to highly contracted state:

2× tP & 3× 108 years

Minimum time to form a disk.

Luminosity

Two ways to estimate luminosity of the young galaxies:1. Element Production2. Luminosity of the First Generation of Stars

In general:Young galaxies must have been much brighter than theyare now.

Luminosity - Element Production- At the time the disk has formed→ most metals have been produced by 1st gen. stars→ Lower Limit of Luminosity:

Ly ' 0.007Mc2 ∆X∆t' 5× 1047 ∆X

∆t8erg s−1 (7)

(∆t8 in 108 years)

- For MW: M/L = 10×M/L→ Young Galaxy brighter than now by

Ly

L' 1× 104 ∆X

∆t8(8)

→ Put in: ∆X = Z = 0.02 & ∆t8 = 0.3 (c.f.)

Ly & 3× 1046 erg s−1 or Ly & 700L (9)

Luminosity - Stars

Normalize:∑

MVΨ(Mv )M(MV ) = 1.2× 1011M [Eq. (4)] !

Luminosity for 1 MBol. = 0 star: LMBol.=0? = 3× 1035 erg s−1

⇒ Ly = 7.5× 1046 erg s−1 (10)

Assumption here: All Stars created at once...

...but consider timescales (Cols. 7 & 8)

- High M stars contribute most L (τM . 107 years)- Formation: τform. τM (τM ∼ 106)→ Adopt: ∆t = 3× 107 years [as in Eq. (9)]! ∆t & τM for high M, only fraction ε of them radiates in ∆t

with Col. (8) ⇒ Ly = 2.5× 1046 (11)Compare with Eq. (9)

Spectrum- Avg. Teff. of O & B type stars: Teff. = 30000 K- Stellar opacity→ Lyman decrement (stellar atmosphere models):

Fλ(912+)

Fλ(912−)∼ 18 (12)

⇒ 10 % of total flux is shortward of 912 A

Spectrum would look like this if ISM HI is distributed uniformly, but...

...assuming HI distributed in clouds with nHI = 10− 100 cm−3

→ optically thick case, γion.’s get converted to Lyman-γ’s→ resonant Lyα line!⇒ as much as 2/3 of γion. could be converted to Lyα

(6-7 % of total flux!)

s p d f

n=1

n=2

n=3

n=4

n=5

γLyα

(γγ)

...in the most extreme case, the spectrum would look like this:

ObservabilityDepends on Cosmology, here ΩM = 1, ΩΛ = 0 andH0 = 70 km s−1 Mpc−3 are used.

Observational Parameters:

I Redshift: z ∼ 10− 30I → Radiation: λ ∼ 1− 3µI → Angular Diameter: & 5”I → Probability of intersecting

random line of sight: & 5 %

Prediction:

With a D = 2m telescope, a narrow-band (NB) filter (∆λ = 20A) & 5 min integration time a primeval galaxy at z = 7 could be

seperated from sky-background.( mλmax.

AB ∼ 23 ⇒ mNBAB ∼ 19 )

Unsuccessful Hunt for Primeval GalaxiesI Basic problem faced by observers: PGs should be

I relatively bright &I large &I numerous.

I So where are these objects?2

I Monolithical collapse model is not appropiateI Dark Matter based models emphasize gradual assembly of

massive galaxies...I ...moreover, radiation transfer in HI regions is more

complicated (attenuation by dust etc.).

⇒ PGs are less massive!⇒ Flux limits were optimistically bright!

2Pritchet 1994, PASP 106, 1052

Observational Properties of Lyα-Emitters

- Mostly used: Narrowband-Imaging (here: z = 4.55 LAEs3 in afield around a QSO- Typical observed parameters of LAEs:

LLyα ∼ 1042 − 1043 erg s−1, rLyα ∼ 1 kpc,MLAE ∼ 109 − 1010MAvg. age of stars ∼ 106 − 108 years, z ∼ 3− 8

3Hu & McMahon 1996, Nat 382, 231

The End

“Cosmologists are always wrong, but never in doubt.”

Robert P. Kirshner4

I Questions & DiscussionI Compare with contemporary views of galaxy formation...I ...so why is this paper still important today?

I Thanks for your attention!

4PNAS 101, 8-13 (2004)