[PPT] - A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE STEVE FINKELSTEIN THE PowerPoint Presentation

SLIDE 1

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

STEVE FINKELSTEIN THE UNIVERSITY OF TEXAS AT AUSTIN

IMAGE CREDIT: JASON JAACKS

SLIDE 2

PRIMER: REDSHIFT AND LOOKBACK TIME

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

2 4 6 8 10 12 14 Redshift 6 8 10 11 12 12.5 13 13.5 Lookback Time (Gyr) Emergence of Hubble Sequence Epoch of Galaxy Assembly

Lookback time = 12.9-13.3 Gyr 1-3 Gyr after Big Bang

Epoch of Reionziation

Lookback time = 12.9-13.3 Gyr 0.5-1 Gyr after Big Bang

Epoch of Galaxy Formation

Lookback time = 13.3-13.6 Gyr 200-500 Myr after Big Bang

SLIDE 3

(SOME) QUESTIONS WE HAVE ANSWERED WITH HUBBLE

▸ Galaxies exist in great number between

500 Myr and 1 Gyr after the Big Bang, and the cosmic star-formation rate density evolves smoothly upward from z=8 to z=4 (e.g, work by Bowler+, Bouwens+, Finkelstein+, McLeod+, Oesch+, McLure+, Ishigaki+).

▸ Even the smallest galaxies we can see with

Hubble are still enriched by previous generations of star-formation (e.g., Bouwens+12,14, Finkelstein+12, Dunlop+13, Rogers+14, Smit+15).

▸ Galaxies alone could reionize the universe

if their ionizing photon escape fractions are relatively high, >10% (e.g., Kuhlen 12, Finkelstein+12, Robertson+13,15, Bouwens+15b, Livermore+17).

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

2 4 6 8 10 Redshift −3.5 −3.0 −2.5 −2.0 −1.5 −1.0 −0.5 0.0 log Cosmic SFRD (M yr −1 Mpc −3) Unresolved Galaxies Total SFR D e n s i t y Reference Bouwens+15 Finkelstein+15 McLeod+15 Oesch+13/14 MD+14 10 5 3 2 1.5 1 0.8 0.6 0.5 Time Since Big Bang (Gyr) Reference MD+14 Oesch+14 62D 27 // C9/22:6CCC 1/:072 1 .77D/70/7,/016/:072://79/09/6CCC 1/:072 1:

FINKELSTEIN 16 BOUWENS+14

r r <

r

~ t r

t

t ~ r µ +

~

>

~
r

=

r

r > t =

2

r t ) r > ~

ROBERTSON+15

SLIDE 4

QUESTIONS WE HOPE TO ANSWER WITH JWST

▸ When is the epoch of the first galaxies? ▸ What is the evolution of the cosmic SFR density at

z > 8?

▸ Are the galaxies we can see enriched by

Population II star-formation?

▸ Have we missed anything with our UV-only view of

the distant universe?

▸ How do the conditions for star-formation

and black hole growth evolve with cosmic time (e.g., spectroscopic studies)?

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

Glazebrook+17 Oesch+14 Finkelstein+17

3 4 5 6 7 8 9 10 11 12 13 −4.5 −4 −3.5 −3 −2.5 −2 −1.5 −1 Redshift log SFR Density [M /yr/Mpc3] > 0.7 M /yr GOODS−N/S + HUDF09/XDF previous CDFS (Oesch+13) Ellis+13 (corrected) CLASH ∝ ( 1 + z ) − 3 . 6 ∝(1+z)−10.9 2 1.4 1 0.8 0.6 0.5 0.4 Time [Gyr]

2 4 6 Redshift 23.0 23.5 24.0 24.5 25.0 log ε912 (erg s−1 Mpc−3 Hz−1)

AGN EMISSIVITY

SLIDE 5

WHAT WFIRST BRINGS TO THE TABLE: SCIENCE ENABLED BY A ~100X INCREASE IN FOV

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE 26 28 30 32 Apparent AB Magnitude

2

2 4 log Area (deg2)

WFC3 FOV CANDELS Wide CANDELS Deep HUDF UDF/HFF Parallels HFFs (µ=5)

26 28 30 32 Apparent AB Magnitude

2

2 4 log Area (deg2)

WFC3 FOV NIRCam FOV CANDELS Wide CANDELS Deep HUDF UDF/HFF Parallels HFFs (µ=5) JWST "CANDELS" JWST UDF JWST FFs

26 28 30 32 Apparent AB Magnitude

2

2 4 log Area (deg2)

WFC3 FOV NIRCam FOV WFIRST WFC FOV CANDELS Wide CANDELS Deep HUDF UDF/HFF Parallels HFFs (µ=5) JWST "CANDELS" JWST UDF JWST FFs WFIRST HLS WFIRST 1 deg2 WFIRST 3 deg2

~500 hr GO program

SLIDE 6

THE KINDS OF NUMBERS WE’RE DEALING WITH

▸ Predictions assume smoothly evolving

Schechter UV LF (Finkelstein 16).

▸ Limiting magnitudes = 26.5 for HLS (except for

z=7, which is limited by z’LSST=26.2 depth), with empirically derived (from HST) magnitude-dependent completeness applied.

▸ GO deg

2

survey is a roughly 500 hr survey

bserving one square degree to m~29.

▸ To survey a sq. deg. with JWST to this

depth would take several 1000’s of hours

f integration, plus extensive overheads.

Redshift

Expected # (HLS) Expected # (deg2 GO)

6 ~3,300,000 ~21,000 7 ~530,000 ~9200 8 ~280,000 ~4000 9 ~75,000 ~1700 10 ~19,000 ~700

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

23
22
21
20
19
18
17

Absolute UV Magnitude 100 101 102 103 104 105 Number per bin

23
22
21
20
19
18
17

Absolute UV Magnitude 1 10 100 1000 10000 Number per bin

HLS GO

z=7 z=10

SLIDE 7

▸ HST and JWST are severely limited in volumes that they can simultaneously

probe. The following are some high priority questions likely to remain open

in ~a decade:

▸ How do the physics which regulate star-formation evolve with cosmic

time?

▸ How has cosmic variance affected our current results, particularly at faint

luminosities?

▸ What is the impact of environment on reionization and galaxy evolution? ▸ What is the large-scale distribution of the detectability of Lyα emission in

the epoch of reionization?

▸ What is the contribution of AGNs to reionization?

OPEN QUESTIONS FOR WFIRST

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

SLIDE 8

OPEN QUESTIONS FOR WFIRST

▸ How do the physics which regulate star-formation

evolve with cosmic time?

▸ A phenomenological model which assumes that the

star-formation rate tracks the halo mass accretion rate predict that the ratio of stellar-mass formed to halo mass (SMHM) increases with increasing redshift at z > 4 (Behroozi+13, Behroozi & Silk+15).

▸ This implies that galaxies are perhaps better at

converting gas into stars at higher redshifts, counter to a variety of theoretical predictions (e.g., lower-Z should reduce SF efficiency). Other factors, such as reduced negative feedback effects, could be at play.

▸ One example - changing the timescale for

converting gas into stars (Somerville+12) by a factor

f four results in large changes at the bright end!

▸ Current volumes probed do not contain

enough galaxies to constrain these physics!

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

10

9

10

11

10

12

Mh [MO

]

0.0001 0.001 0.01 0.1 M* / Mh z = 4, Behroozi et al. (2013) z = 5, Behroozi et al. (2013) z = 6, Behroozi et al. (2013) z = 7, Behroozi et al. (2013) z = 8, Behroozi et al. (2013) z = 9.6, Median Prediction z = 10.8, Median Prediction z = 12.5, Median Prediction z = 15.0, Median Prediction

BEHROOZI+15

−23 −22 −21 −20 −19 MUV 10−7 10−6 10−5 10−4 10−3 ϕ (# Mag−1 Mpc−3)

Currently permitted range of models

Finkelstein+15 Bouwens+15 UltraVISTA (Bowler+14)

z = 7

SLIDE 9

▸ Galaxy clustering results have observationally found a

similar trend - higher SMHM at fixed halo mass (Harikane+16,17).

▸ A similar result was found via abundance matching the

UV luminosity function, and looking at evolution at fixed UV magnitude (~fixed stellar mass; Finkelstein+15b), though this is subject to UV scatter, and nebular contamination in M* estimates.

▸ Stefanon+17 found less evolution via

rest-z’ luminosity function abundance matching, though they were exploring progenitors/descendants, and had small numbers at z > 5.

▸ Most of these studies are limited by small

sample sizes (the clustering study used HSC, so had large samples, but potentially much higher sample contamination), so conclusions remain difficult.

OBSERVATIONAL EVIDENCE?

HARIKANE+17

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

2 4 6 8 10 Redshift 0.00 0.01 0.02 0.03 0.04 0.05 0.06 Stellar Mass/Halo Mass (MUV=-21)

Unknown Evolution at z > 6

Observed Data Observed Fit (4<z<7) Observed Fit (z>6) Predicted

FINKELSTEIN+15B STEFANON+17

SLIDE 10

▸ There is now some evidence

that the bright end of the UV luminosity function may be “super”-Schechter, e.g., a double power law (e.g., Bowler+14, 15; Ono+17, Stefanon+17, Stevans in prep).

▸ Interesting physics?

Dust attenuation? Contamination by AGNs?

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

z=4 UV luminosity function BY UT student Matt Stevans, using ~20 deg2 SHELA survey data, and simultaneously fitting AGN and star- forming galaxy luminosity functions.

FURTHER OBSERVATIONAL EVIDENCE?

Stevans+17

SLIDE 11

COSMIC VARIANCE

▸ Fractional uncertainty due to cosmic

variance is ~40% in the HUDF.

▸ Will be similar in a JWST UDF-

style observation due to small volume probed.

▸ How much are our conclusions on

faint galaxies biased by cosmic variance?

▸ Lensing helps provide

independent lines-of-slight, though volumes are tiny, so still CV issues.

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE s 2 ¢

1

¢ ¢ = - ¢ ´ ¢ s = -

s 2 ¢

1

¢ ¢ = - ¢ ´ ¢ s = -

Finkelstein+15

SLIDE 12

ROBUSTNESS OF BRIGHT-END ABUNDANCES TO COSMIC VARIANCE

5 6 7 8 9 10 11 Redshift 1 10 100 1000 N / σ(N) for MUV=−22

WFIRST HST JWST

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

CV estimated by combining observed biases (Barone- Nugent+14) with dark matter CV estimates (Newman+Davis 02, Moster+11).

▸ WFIRST HLS will

allow measurements

f the abundance of

bright galaxies at z=6-8 with S/N > 100 (S/N > 10 at z=9-10).

z=7

SLIDE 13

▸ How do environmental factors

affect star-formation in the epoch

f reionization?

▸ Current volumes probed at z >

6 do not yet allow robust measures of environment to be made.

▸ The WFIRST HLS will probe 10-20

cGpc

3 volumes in unit-redshift

bins at z=10-6, observing galaxies in the full range of cosmic environments.

▸ Will also allow measurements

f the cosmic SFR density both

robust against CV, and as a function of environment.

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

Madau & Dickinson 2014

ENVIRONMENT AND STAR FORMATION

m=26 galaxies

SLIDE 14

ENVIRONMENT AND REIONIZATION

▸ Environment is likely linked to

reionization - the most overdense regions formed stars first, and likely ionized first.

▸ SKA 21cm line tomography should be

able to resolve neutral/ionized regions

n the order of a few pMpc,

comparable to the expected size of ionized bubbles in the early universe (e.g., Furlanetto 06; Iliev+14).

▸ Could then target ionized regions with a

deeper GO survey to explore whether reionization “feedback” is suppressing star formation (though this may be more applicable to a LUVOIR-type mission). A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

High spatial resolution will also allow improved searches for AGNs at high redshift, better constraining their contribution to reionization.

SLIDE 15

PROBING REIONIZATION WITH LYMAN ALPHA EMISSION

▸ Lyα is resonantly scattered by neutral hydrogen,

so if it is emitted from a galaxy with a surrounding neutral IGM, it will be significantly spatially diffused, well beyond detectable levels.

▸ Also, it is relatively “abundant” at z=6, just

after the end of reionization.

▸ Simulations show that a patchy IGM should be

directly traceable by the patchiness of Lya emission.

▸ Real galaxies make this more complicated, as

they create HII regions, and they can impart a kinematic offset to Lyα photons, escaping from even modestly neutral regions.

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

XHI=0.3 XHI=0.5 XHI=0.7

McQuinn+2007 Stark+11

SLIDE 16

▸ This has led to a booming industry of attempted Lya measurements at z > 6.5, with

some notable successes (e.g., Shibuya+12, Finkelstein+13, Oesch+15, Zitrin+15, Roberts-Borsani+16, Song+16, LaPorte+17).

1.025 1.030 1.035 1.040 1.045 Wavelength (μm) −0.0010 −0.0005 0.0000 0.0005 0.0010 0.0015 0.0020 z=7.5078+/−0.0004 S/N=7.8 z = 7.5078 ± 0.0004 S/N = 7.8 7.42 7.45 7.48 7.51 7.54 7.57 7.60 Redshift from Lyα 8.8 arcsec

F I N K E L S T E I N + 2 0 1 3 O E S C H + 2 0 1 5 Z I T R I N + 2 0 1 5

z=7.51 z=7.73 z=8.68

PROBING REIONIZATION WITH LYMAN ALPHA EMISSION

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

▸ However, the majority of galaxies go

undetected with spectroscopic followup, leading to the inference that the IGM at z ≥ 7 is highly neutral (e.g., Pentericci+11, 14, Treu+13, Fontana+10, Tilvi+14).

Pentericci+14

SLIDE 17

WORKING ON IMPROVEMENTS FROM THE GROUND AND SPACE

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

2016B

Intae Jung

1.095

1.035

1.040 1.045 1.050 1.055 1.060 Wavelength (m) 7.49 7.52 7.55 7.58 7.61 7.64 7.67 7.70 Redshift (Ly)

α

zLyα=7.6013 S/N=5.6

α

We in the midst of a multi-pronged survey for Lyα in the epoch
f reionization, using ~300 HST orbits (FIGS: PI Malhotra; CLEAR:

PI Papovich), and 20+ nights of Keck DEIMOS+MOSFIRE

bservations (PI Finkelstein, primarily through NASA).
Work currently being led by UT student Intae Jung, first

paper out by end of summer on our optical dataset.

SLIDE 18

WORKING ON IMPROVEMENTS FROM THE GROUND AND SPACE

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

2016B

Intae Jung

1.095

1.035

1.040 1.045 1.050 1.055 1.060 Wavelength (m) 7.49 7.52 7.55 7.58 7.61 7.64 7.67 7.70 Redshift (Ly)

α

zLyα=7.6013 S/N=5.6

α

We in the midst of a multi-pronged survey for Lyα in the epoch
f reionization, using ~300 HST orbits (FIGS: PI Malhotra; CLEAR:

PI Papovich), and 20+ nights of Keck DEIMOS+MOSFIRE

bservations (PI Finkelstein, primarily through NASA).
Work currently being led by UT student Intae Jung, first

paper out by end of summer on our optical dataset.

Ground-based near-IR spectroscopy is hard! The night sky lines block ~half of our accessible wavelength range, and atmospheric absorption windows prevent probing the full likely redshift range for a given object.

Rebecca Larson

Likely Lyα emission at z=7.4 from the HST FIGS grism survey (see also Tilvi+16) Proof of concept for Lyα detections with WFIRST!

SLIDE 19

WFIRST CAN PROVIDE THE ABILITY TO PROBE LYA EMISSION OVER LARGE SCALES

▸ JWST will make progress, but the

small FoV of NIRSpec/NIRISS and relatively low sensitivity at λ < 1.3 μm will limit results.

▸ Our Cosmic Dawn SIT (PI Rhoads)

has been advocating for extending the WFIRST grism spectral range down to 0.9/1.0 μm (from baseline 1.3 μm) to allow wide-field Lya studies throughout the EoR.

▸ We’re working now to spec out

potential GO spectroscopic programs spanning 10’s-100 arcmin in length (exploring issues such as depth, # of PAs, etc.)

W F I R S T D A W N

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

Simulated WFIRST 0.9-1.3μm grism data

Lyα from data-cube reconstruction

Simulations led by postdocs Vithal Tilvi (ASU) and Isak Wold (UT) See poster by Tilvi

SLIDE 20

WFIRST/HST JWST LUVOIR WFIRST Lyα SKA/21 cm

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE

SLIDE 21

THANK YOU!

A LARGE-SCALE VIEW OF THE DISTANT UNIVERSE Questions?