The Formation History of the
Ultra-Faint Dwarf Galaxies
Tom Brown
Space Telescope Science Institute
Jason Tumlinson (STScI), Marla Geha (Yale), Evan N. Kirby (UC Irvine), Don A. VandenBerg (U of Victoria), Ricardo R. Munoz (U de Chile),
Jason S. Kalirai (STScI), Josh D. Simon (Carnegie),
Rorberto J. Avila (STScI), Puragra Guhathakurta (UCO/Lick), Alvio Renzini (INAF), Henry C. Ferguson (STScI),
Luis Vargas (Yale), Mario Gennaro (STScI)
The Formation History of the
Ultra-Faint Dwarf Galaxies
Overview
•
Background: ultra-faint dwarf galaxies•
Hubble program overview•
Ages in ultra-faint dwarfs•
IMF in ultra-faint dwarfs•
SummaryTumlinson (2010)
Dark Matter Distribution
350 kpc
Tumlinson (2010)
Subhalos with star formation continuing past reionization
Tumlinson (2010)
Fossil subhalos - star formation truncated by reionization
Tumlinson (2010)
Most subhalos never form stars at all
SDSS Field of Streams
Belokurov et al. (2007)
Luminosity vs Size
1 10 100 1000
rh (pc) -2
-4 -6 -8 -10 -12 -14
M V (mag)
Globular Clusters
classical dSphs
ultra-faint dwarfs
Harris (1996) Mateo (1998) Martin et al. (2008)
Luminosity vs Size
Harris (1996) Mateo (1998) Martin et al. (2008)
1 10 100 1000
rh (pc) -2
-4 -6 -8 -10 -12 -14
M V (mag)
Globular Clusters
classical dSphs
ultra-faint dwarfs
Ground observations imply old populations in UFDs - but how old?
Leo IV (Sand et al. 2010)
MMT
Coma Berenices (Munoz et al. 2010)
CFHT
16
18
20
22
24
-0.5 0 0.5 1 1.5 2
I0
(V-I)0 UMa IUrsa Major I (Okamoto et al. 2008)
Subaru Bootes I
(Okamoto 2010) Subaru
Hercules (Sand et al. 2009)
LBT
Canes Venatici II (Greco et al. 2008)
WHT
0.0 0.5 1.0 1.5 B-V
Hubble Program Overview
•
Relative ages to <1 Gyr (SNR~100 at main sequence turnoff)-
Relative to each other (synchronized star formation histories?)-
Relative to ancient populations (e.g., globular clusters)•
IMF in old dynamically un-evolved populations•
Advanced Camera for Surveys (ACS) prime - centered on galaxy•
Wide Field Camera 3 (WFC3) parallel - in galaxy outskirts•
113 orbits - F606W (broad V) and F814W (I)•
Comparison to ACS programs in the Local Group with same bands-
Dozens of globular clusters (empirical ancient templates)-
Many pencil-beam surveys of Local Group galaxies•
Victoria-Regina Isochrones updated to latest physics•
Leverage Keck spectroscopy to fix metallicities while fitting agesUrsa Major I
ACS
WFC3
202’’ dist = 97 kpc
MV = -5.5 rh = 11.3’
rh = 318 pc 27 orbits
N E
Leo IV
ACS
WFC3
202’’
dist = 154 kpc MV = -5.8
rh = 4.6’
rh = 206 pc 16 orbits
N E
Leo IV 16 orbits faint limit
V~28.5
Leo IV 16 orbits faint limit
V~28.5
MSTO star V=24.8
MSTO star V=24.9
Leo IV 16 orbits faint limit
V~28.5
M92
(NGC 6341)
[Fe/H]=-2.3 (m-M)o=14.62
E(B-V)=0.023
M92
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 -0.0 0.2
m606 - m814 (STMAG) 24
22 20 18 16 14 12
m 814 (STMAG)
M92
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 -0.0 0.2
m606 - m814 (STMAG) 24
22 20 18 16 14 12
m 814 (STMAG)
"Empirical" Isochrone (Ridge Line)
Brown et al. (2005)
M92
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 -0.0 0.2
m606 - m814 (STMAG) 24
22 20 18 16 14 12
m 814 (STMAG)
Victoria-Regina Theoretical Isochrone Age = 13.0 Gyr [Fe/H]=-2.3
VandenBerg et al. (2012)
Notes:
[O/Fe] higher than assumed previously Teff-color relation empirically calibrated
Ages
-0.8 -0.6 -0.4 -0.2 0.0 m606-m814 (STMAG) 28
26 24 22 20 18
m 814 (STMAG)
Hercules
-0.8 -0.6 -0.4 -0.2 0.0 m606-m814 (STMAG) Leo IV
-0.8 -0.6 -0.4 -0.2 0.0 m606-m814 (STMAG) UMa I
28 26 24 22 20 18
m 814 (STMAG)
Boo I CVn II ComBer
Brown et al.
(2012, 2013 in prep)
-0.8 -0.6 -0.4 -0.2 0.0 m606-m814 (STMAG) 28
26 24 22 20 18
m 814 (STMAG)
Hercules M92
-0.8 -0.6 -0.4 -0.2 0.0 m606-m814 (STMAG)
Leo IV M92
-0.8 -0.6 -0.4 -0.2 0.0 m606-m814 (STMAG)
UMa I M92
28 26 24 22 20 18
m 814 (STMAG)
Boo I M92 CVn II M92 ComBer M92
Brown et al.
(2012, 2013 in prep)
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 m606-m814 (STMAG)
29 28 27 26 25 24
m 814 (STMAG)
Hercules
CMDs of the UFDs, all look very similar
Composite UFD CMD looks like a single-age
population (to first order)
Brown et al. (2012, 2013 in prep)
CMDs of the UFDs, all look very similar
Composite UFD CMD looks like a single-age
population (to first order)
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 m606-m814 (STMAG)
29 28 27 26 25 24
m 814 (STMAG)
Hercules Leo IV
Brown et al. (2012, 2013 in prep)
CMDs of the UFDs, all look very similar
Composite UFD CMD looks like a single-age
population (to first order)
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 m606-m814 (STMAG)
29 28 27 26 25 24
m 814 (STMAG)
Hercules Leo IV CVn II
Brown et al. (2012, 2013 in prep)
CMDs of the UFDs, all look very similar
Composite UFD CMD looks like a single-age
population (to first order)
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 m606-m814 (STMAG)
29 28 27 26 25 24
m 814 (STMAG)
Hercules Leo IV CVn II UMa I
Brown et al. (2012, 2013 in prep)
CMDs of the UFDs, all look very similar
Composite UFD CMD looks like a single-age
population (to first order)
Brown et al. (2012, 2013 in prep)
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 m606-m814 (STMAG)
29 28 27 26 25 24
m 814 (STMAG)
Hercules Leo IV CVn II UMa I Boo I
Brown et al. (2012, 2013 in prep)
CMDs of the UFDs, all look very similar
Composite UFD CMD looks like a single-age
population (to first order)
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 m606-m814 (STMAG)
29 28 27 26 25 24
m 814 (STMAG)
Hercules Leo IV CVn II UMa I Boo I Com Ber
-0.6 -0.5 -0.4 -0.3 m606-m814 (STMAG)
25.5 25.0 24.5 24.0
m 814 (STMAG)
Hercules
12.5 Gyr isochrone
-0.6 -0.5 -0.4 -0.3
m606-m814 (STMAG)
11.5 Gyr Model
12.5 Gyr isochrone
Brown et al.
(2012, 2013 in prep)
-0.6 -0.5 -0.4 -0.3 m606-m814 (STMAG)
25.5 25.0 24.5 24.0
m 814 (STMAG)
Hercules
12.5 Gyr isochrone
-0.6 -0.5 -0.4 -0.3
m606-m814 (STMAG)
12.5 Gyr Model
12.5 Gyr isochrone
Brown et al.
(2012, 2013 in prep)
-0.6 -0.5 -0.4 -0.3 m606-m814 (STMAG)
25.5 25.0 24.5 24.0
m 814 (STMAG)
Hercules
12.5 Gyr isochrone
-0.6 -0.5 -0.4 -0.3
m606-m814 (STMAG)
13.5 Gyr Model
12.5 Gyr isochrone
Brown et al.
(2012, 2013 in prep)
-0.6 -0.5 -0.4 -0.3 m606-m814 (STMAG)
25.5 25.0 24.5 24.0
m 814 (STMAG)
Hercules
12.5 Gyr isochrone
-0.6 -0.5 -0.4 -0.3
m606-m814 (STMAG)
12.5 Gyr Model
12.5 Gyr isochrone
Brown et al.
(2012, 2013 in prep)
Issues still being worked out
•
Distance & reddening to each UFD-
Can affect age by ~0.5 Gyr•
Field contamination in HST CMDs-
Can mask trace sub-populations•
Field contamination in Keck spectroscopy-
Relatively metal-rich contaminants can drive a minority population that is ~1 to 2 Gyr younger in fit•
[O/Fe] to assume in models at low [Fe/H]-
Can affect age by ~0.5 GyrInitial Mass Function
Limited test in ideal population
•
Mass range is not large:~0.5 - 0.8 MSun in 4: Hercules, Leo IV, UMa I, CVn II
~0.35 - 0.8 MSun in 2: Boo I & Com Ber
•
Old but dynamically un-evolved (unlike globular clusters)•
IMF is a free parameter in many high-redshift studies•
Test theoretical predictions IMF evolution(e.g., Tumlinson 2010, Munoz et al. 2009, Bovill et al.
2009, Koposov et al. 2009, Li et al. 2010)
-
Larger characteristic mass at high z?(due to higher CMB temperature)
24 25 26 27 28 m814 (STMAG)
10 100
number of stars
0.764 0.747 0.698 0.617 0.521 mass (MO .)
Hercules
25 26 27 28
m814 (STMAG) 10
100
0.766 0.755 0.716 0.643 0.548 mass (MO .)
Leo IV
Geha et al. (2013)
24 25 26 27 28 m814 (STMAG)
10 100
number of stars
0.764 0.747 0.698 0.617 0.521 mass (MO .)
Hercules
Best-fit power law (_=1.16)
25 26 27 28
m814 (STMAG) 10
100
0.766 0.755 0.716 0.643 0.548 mass (MO .)
Leo IV
Best-fit power law (_=1.31)
Geha et al. (2013)
24 25 26 27 28 m814 (STMAG)
10 100
number of stars
0.764 0.747 0.698 0.617 0.521 mass (MO .)
Hercules
Best-fit power law (_=1.16) Salpeter (1955; _=2.35)
25 26 27 28
m814 (STMAG) 10
100
0.766 0.755 0.716 0.643 0.548 mass (MO .)
Leo IV
Best-fit power law (_=1.31) Salpeter (1955; _=2.35)
Geha et al. (2013)
24 25 26 27 28 m814 (STMAG)
10 100
number of stars
0.764 0.747 0.698 0.617 0.521 mass (MO .)
Hercules
Best-fit power law (_=1.16) Salpeter (1955; _=2.35)
Bottom-light power law (_=0.5)
25 26 27 28
m814 (STMAG) 10
100
0.766 0.755 0.716 0.643 0.548 mass (MO .)
Leo IV
Best-fit power law (_=1.31) Salpeter (1955; _=2.35)
Bottom-light power law (_=0.5)
Geha et al. (2013)
Summary
•
Ultra-faint dwarf galaxies formed bulk (all?) of their stars 12 to 13 Gyr ago (z ~ 4 to 8)•
No intermediate-age (< 10 Gyr) stars-
Unlike most classical dwarf galaxies•
Within current uncertainties, star formation is synchronized-
Unclear if it predates reionization•
IMF is flatter than Salpeter & other populations where directly measured (SMC, U Minor, MW)Backup Slides
-4 -3 -2 -1 [Fe/H]
-1 0 1 2
[O/Fe]
Frebel (2009)
[O/Fe]
significantly
affects the age of the isochrone one would fit to a
color magnitude diagram
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2
m606-m814 (STMAG) 8
6 4 2 0
m 814 (STMAG)
[O/Fe]
significantly
affects the age of the isochrone one would fit to a
color magnitude diagram
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2
m606-m814 (STMAG) 8
6 4 2 0
m 814 (STMAG)
[Fe/H]=-2.35 [O/H]=-1.95 age=13.40 Gyr
[O/Fe]
significantly
affects the age of the isochrone one would fit to a
color magnitude diagram
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2
m606-m814 (STMAG) 8
6 4 2 0
m 814 (STMAG)
[Fe/H]=-2.35 [O/H]=-1.95 age=13.40 Gyr
[Fe/H]=-2.35 [O/H]=-1.71 age=12.75 Gyr