Hubble Science Briefing

CANDELS - Observing Galaxy Assembly with the Hubble Space Telescope

Henry Ferguson 1 August 2013

1

Cosmology in a nutshell

• • The universe had a beginning In the beginning it was small and hot and nearly uniform in density • • It is filled mostly with dark matter and dark energy We know how they behave, but don’t know what they are.

• As it expanded, the high-density regions collapsed and formed galaxies.

2

Motivation

• The standard cosmological model is a huge success on large scales Predicted fluctuations in the Cosmic Microwave background match observations amazingly well These fluctuations are the seeds of galaxies. Theoretical predictions (red) reproduce the observed clustering of galaxies (blue) amazingly well.

3

Motivation

Galaxy evolution within this cosmology is complicated!

Governato et al. Simulations of a Milky-Way-like galaxy (blastwave feedback) http://www.youtube.com/watch?v=n0jRObc7_xo 4

Cosmology, power spectrum Merging of dark-matter halos

Ingredients in theoretical models of galaxies

Gas heating and cooling Star formation, black-hole formation Stellar winds Supernovae, feedback from black holes chemical enrichment Multiple generations of stars dust absorption & emission

galaxy observables

5

Theoretical models tell us where and when galaxies should form.

But are these models correct?

GIF Simulation Kauffmann, Colberg, Diaferio & White 1999 6

A crude way to make the connection between galaxies and dark-matter halos

• • Pick a slice of the universe Estimate the stellar mass of each galaxy in that slice • Rank order the galaxies from most to least massive • • Rank order the dark-matter halos in a cosmological simulation of the same volume Assign the most-massive galaxy to the most massive dark-matter halo, and so on down the list 7

Result: The mass in stars compared to the mass in dark matter

Mass of the Milky Way 10% 1% Mass of the dark-matter halo (relative to the sun) The conversion of gas into stars is inefficient for high-mass and low-mass dark-matter halos 8

Some kind of feedback is needed to prevent all the gas from turning into stars

Feedback from supernovae?

Feedback from black holes?

The conversion of gas into stars is inefficient for high-mass and low-mass dark-matter halos 9

Open Questions

• • Do galaxies form at dark-matter peaks?

• Really? All of them?

• • • Which feedback mechanisms are most important in governing star formation?

Heating of gas by radiation from other galaxies?

Expelling gas explosively?

Injecting energy into the gas from a central black hole?

• What is needed to make the  CDM cosmological model consistent with all the observables?

10

The need for large surveys

• • Galaxy formation/evolution is inherently statistical • • • Making the link between galaxy-scale physics and cosmology is all about measuring distributions: Need to know what kinds of galaxies are common and what kinds are rare. Need to measure correlations between the structure of galaxies and their stellar content.

Need to measure correlations between galaxy properties and their environment.

11

CANDELS

C

osmic

A

ssembly

N

ear-infrared

D

eep

E

xtragalactic

L

egacy

S

urvey ~175 team members ~45 institutions 12 countries candels.ucolick.org

Builders: Harry Ferguson, Sandra Faber, Adam Riess, Steve Rodney Norman Grogin, Dale Kocevski Anton Koekemoer 12

CANDELS in a nutshell

• • • • • CANDELS observes representative slices of the universe.

These slices include Hubble’s deep fields, but cover more area These are the reference fields for studies of the distant universe • Near-infrared images provide crucial insight into galaxy assembly CANDELS works in concert with other surveys to provide as much information as possible on each galaxy.

Searching for distant supernovae as we go along.

13

Looking back in time

Redshift (

z

) • • • CANDELS sees about 250,000 galaxies • Most are at high redshift (very distant) But for many of them, it’s devilishly hard to determine exact distances.

Light reaching us today left these galaxies billions of years ago, providing a historical record of galaxy evolution.

CANDELS 14

What The Hubble WFC3 Can Do 15

What WFC3 Can Do 16

Multi-wavelength slices of the universe…

17

Hubble 0.6, 1.25, 1.6 microns

18

Chandra X-ray observatory

0.5-2, 2-8, 5-8 keV (0.0002-0.002 microns) 19

Spitzer observatory 3.6+4.5, 5.6, 8 microns

20

Spitzer and Herschel observatories 24, 100, 160 microns

21

Herschel Observatory 250, 350, 500 microns

22

33 CANDELS submitted papers

Lead Author Paper Title

Yicheng Guo David O. Jones A. Galametz

Discovery of the Most Distant Type Ia Supernova at Redshift 1.914 -- Telecon 12/10 The CANDELS UDS Multiwavelength catalog -- Telecon 10/11/12

A. Cooray J. Lotz R. Bassett V. Tilvi J. Herrington T. Targett

CANDELS Multi-Wavlength Catalogs: Source Detection and Photometry in the GOODS-South Field CANDELS: Strong Lensing Galaxies In HST/WFC3 Imaging Data Of UDS AND GOODS-S The Assembly of Massive Cluster Galaxies at z=1.62

CANDELS Observations of the Color-Morphology Relation at z = 1.6 and its Dependence on Mass and Environment LBGs at z~7 from the zFourGE Survey No Significant Evolution of the Bar Fraction in Large Disk Galaxies from z=1.8 to z=0.6

The properties of (sub)millimetre selected galaxies as revealed by CANDELS WFC3/IR imaging in GOODS-South

G. Barro E. Curtis-Lake D. Rosario V. Bruce H. Yan

The progenitors of red nuggets at z>2 as seen by CANDELS The stellar populations of spectroscopically confirmed z~6 galaxies in the CANDELS UDS/GOODS-S field X-ray selected AGN Hosts are Similar to Inactive Galaxies out to z=3: Results from CANDELS/CDF S Morphologies of Massive Galaxies at 1

CANDELS submitted papers

Lead Author

A. van der Wel

Paper Title

Galfit Structural Parameters of Galaxies from CANDELS

S. Finkelstein A. Grazian K. I. Caputi J. S. Kartaltepe S. Wuyts S. Rodney E. Bell T. Wang S. Wuyts E. Vanzella C. Papovich S. Finkelstein D. Kocevski J.R. Trump A. van der Wel N. A. Grogin

CANDELS: The Contribution of the Observed Galaxy Population to Cosmic Reionization The size-luminosity relation at z=7 in CANDELS and its implication on reionization The nature of H-[4.5]>4 galaxies revealed with SEDS and CANDELS Morphology of Herschel Selected ULIRGs at z~1-3 Smooth(er) Stellar Mass Maps in CANDELS: Constraints on the Longevity of Clumps in High-redshift Star-forming Galaxies A Type Ia Supernova at Redshift 1.55 in Hubble Space Telescope Infrared Observations from CANDELS What turns galaxies off? The morphologies of intermediate-mass and massive quiescent galaxies during the last ten billion years using the CANDELS Survey CANDELS: Correlations of SEDs and Morphologies with Star-formation Status for Massive Galaxies at z ~ 2 Galaxy Structure and Mode of Star Formation in the SFR-Mass Plane from z~2.5 to z~0.1

On The Detection Of Ionizing Radiation From Star-Forming Galaxies At Redshift z~3-4 The Structural Properties and Evolution of Galaxies in a Cluster at z=1.62

Evolution of UV Spectral Slope from z=4-8 CANDELS: Investigating the AGN-Merger Connection at z~2 A CANDELS WFC3 Grism Study of Emission Line Galaxies at z~2: A Mix of Nuclear Activity and Low Metallicity Star Formation Extreme Emission Line Galaxies in CANDELS CANDELS: The Cosmic Assembly Near-Infrared Deep Extragalactic Legacy Survey

A. M. Koekemoer

CANDELS: HST Imaging Data Products and Mosaics 24

CANDELS/CLASH high-z supernovae program

• Improve constraints on cosmic acceleration & Dark Energy • • Observe evolution of SNIa population Test for systematic biases as standard candles • Measure rates of supernovae at high-z: test supernova theory Riess, Rodney, Strolger, Dahlen, Graur, Hayden, Frederiksen … 25

Supernova Follow-up Program for CANDELS Discovery Days 26

CANDELS: Cosmic Dawn

• • Finding some of the most distant galaxies Within 700 million years of the big bang • Tracing the evolution of galaxies with larger samples than possible from deep fields • Finding galaxies bright enough to detect with other observatories 27

One highlight: evolution of UV slopes High-mass galaxies already metal-rich Finkelstein+ 2011 Low-mass galaxies growing dust and metals 28

Galaxy Morphology at cosmic high noon

29

30

31

32

Morphology is important… 33

“Elliptical” or Spheroidal ”

Galaxy Morphology

“Spiral” “Irregular” Nearby galaxies 34

Highlights from “Cosmic high noon”

• • Quiescent vs. star-forming galaxies • Compact spheroids Where did they come from, where did they go?

• The emergence of bulges within galaxy disks • • Clumpy star formation AGN host galaxies 35

Massive z~2 galaxies: morphology vs. SFR

Passive (24 μ m faint) galaxies tend to be compact spheroidals Star-forming galaxies tend to be more extended and more disk like Also, the correlation of star-formation with profile shape is much stronger than with stellar mass. (Bell+ 2012)

Wang+ 2012 36

Compact quiescent galaxies

• At redshift z>1.5, most passive galaxies are compact.

• These compact galaxies mostly appear at 1<z<3.

• Many of these compact galaxies disappear by z=0 Cassata+13 37

Blue and red “nuggets”

Relative space densities evolve Densities suggest rapid quenching of blue nuggets (< 1 Gyr)

Barro+13 How tightly packed are the stars?

loosely tightly 38

Quiescent Galaxy Summary

• • • • Morphology is strongly correlated with specific star-formation rate at z~2.

• Quiescent galaxies at this era are small: Red nuggets • Small quiescent galaxies are disappearing by z=0 Reactivating or growing by mergers?

There is a sufficient reservoir of small active galaxies at z~2 – blue nuggets – to form the red nuggets .

39

The Hubble Sequence at z~2

Bulge fractions

• • • •

Massive galaxies M*>10 11 M

 Bulges become dominant in massive galaxies at z~2 Bulges are smaller at fixed mass at z~2 than today.

• While most passive galaxies are bulge dominated, a few passive galaxies appear to be pure disks.

Implications for quenching models?

Era of massive disks at 2

Emergence of clumpy disks

• Gas instabilities in gas-rich disks?

• • Mergers?

The seeds of bulges?

Guo et al. 2011 Ravindranath et al. in prep.

Mozena et al. in prepration 41

42

Star-formation histories pixel by pixel Clumps in massive galaxies at 1

More massive clumps are redder

• • Mass metallicity relation for clumps?

Winds only partially effective Clumps at z~3 More massive clumps Ravindranath+13 44

Radial Trends

• • • • Clumps near the center tend to be redder and higher mass.

Preference for more massive clumps to form near the center?

Migration coupled with aging of the clumps?

Trends roughly agree with predictions Guo+11 Ravindranath+13 Closer to galaxy center 45

Visual classifications

• Inspecting thousands of galaxies by eye • Help from Galaxy Zoo: 40 million classifications!

46

The AGN-Merger connection

• • • Are supermassive black holes part of the feedback loop that regulates star formation in galaxies?

Strong feedback is needed to explain the inefficiency of star formation.

Emerging feedback:

testable

paradigm – AGN • • • • • Mergers drive gas to galaxy centers Trigger starbursts (often IR luminous) Funnel gas onto central black hole Energy injection from BH quenches further star formation Resulting galaxy ends up as a red spheroid 47

Merger AGN connection: theory

Bright

Chandra

source Bright

Spitzer/Herschel

source Distinct

Hubble

Morphology 48

One early result: Morphologies of X-ray-bright galaxies at z~2 X-ray-bright galaxies have hot gas flowing into a central super-massive black hole Kocevski+12 49

Morphologies of X-ray AGN hosts at z~2

AGN hosts are NOT disturbed.

Kocevski+12 50

Morphologies of X-ray AGN hosts

AGN hosts are mostly spheroids.

Kocevski+12 51

Morphologies of X-ray AGN hosts

AGN hosts have many disks.

Kocevski+12 52

Morphologies of X-ray AGN hosts The lack of disturbances and high frequency of disks challenges the standard merger-driven AGN paradigm.

AGN hosts have many disks.

AGN demographics at z ~ 2 look like those at z ~ 1 

internally driven

BH growth and AGN triggering.

Kocevski+ 2012 53

Unexpected Discovery

54

Bursting dwarfs at z ~ 1.7

van der Wel+ 2011 55

Bursting dwarfs at z ~ 1.7?

 Ubiquitous low-metallicity extreme starbursts.

 Can produce most of stellar mass in today’s dwarf galaxies in only 4 billion years.

56

CANDELS is really just getting started

• • • Last observations complete August 10 About 1 year of work to get reliable catalogs Stay tuned!

Follow our research blog at candels-collaboration.blogspot.com

57