astro 6590: dwarf galaxieshosting.astro.cornell.edu/.../pdf08/a6590_081105_dwarfs2.pdf · 2008. 11....
TRANSCRIPT
1
Astro 6590: Dwarf galaxies• Overview of dwarf galaxies• Distinctive features of dwarfs
• Dwarfs vs giants• Dwarfs vs globular clusters
• Dwarfs: challenges to CDM• Missing satellite problem: how many dwarfs?• The core-cusp problem: dark matter halos
• Morphological sequence of dwarfs: dE, dSph, dI• Morphological segregation
• Local Group dwarfs• Perseus dwarfs• Virgo dwarfs• Transition dwarfs
• Star formation history• Metallicity-luminosity relation
• Blowout in dwarfs• Stellar streams and disrupted satellites near the Milky Way• DDO 154: an almost dark galaxy
HST Survey of the Perseus Cluster core
12 Orbits centered around the core of Perseus(Conselice (PI), Held, DeRijke)
2
Low-Mass Galaxies in the Perseus Cluster (Conselice, Ringberg conf.)
Dwarf ellipticals exist in clusters and morphologicallyappear similar to Local Group dwarf ellipticals
There are however 5 - 10 asmany low-mass cluster galaxiesper giant galaxy as there are inLooser groups
Discovered in 1950s by G. Reaves througha Virgo cluster imaging survey
Background Spirals?
Perseus Cluster HST survey
(Conselice, Gallagher & Wyse ‘03 sample)
Compact dwarfs
Ambiguous cases
Real Dwarfs
3
Color-Magnitude diagram for Perseus galaxies with LocalGroup dEs plotted.
Cluster dEs deviate from the color-magnitude relation
Conselice, Gallagher & Wyse (2003)
Clues for this deviation are old (e.g., Held & Mould 1994)
Low L galaxies in clusters are a heterogenous population
Both 'blue' and 'red‘ dEs
Conselice et al. (2003)
Cluster dEs likely have a mix of stellar populations – some might be metal rich
Three stellar synthesis-modeled age tracts at constant metallicity (3 different ones) for different age populations.
4
Kinematics of dEs
Kinematics of dwarfs
5
Kinematics of dwarfs
Internal Dynamics of NGC 205
CFHT12K mosaic imag(Demers et al. 2003)
Spectroscopic targetschosen to be likelyRGB stars in NGC 205.
Keck/ DEIMOS slitmasks chosen to lie along tidal extensions.
6
Inner rotation =15 km s-1
Geha, Guhathakurta, Rich & Cooper (2006)
radius (arcmin)
velo
city
(km
/s)
NGC 205ve
loci
ty (k
m/s
)
NGC 185Bender et al (1991)
NGC 205Geha et al. (2006)
NGC 147Geha et al. in prep
radius (arcmin)
vrot = 17 km/s
Local Group rotation curves
7
dSphs in the LG
Rotation of dEs in Virgo
Unlike dEs in the Local Group, some dEsin Virgo are rotation-
dominated.
van Zee et al 2004
Long slit stellar absorption line
spectroscopy from Palomar 5m
Rotation curveVelocity dispersion
8
dE’s as stripped dI’s?
• Rotation-dominated dEsappear to avoid the cluster center
• Non-rotating dEsare found in the core
van Zee et al 2004
Velocity distribution of dEs is large, with moresubstructure than giant cluster galaxies
Conselice et al. (2001)A sign of their formation mechanism?
9
dEs in Virgo
• dEs are widely distributed in Virgo.• Do not cluster around giants• Different dynamics from Es
• dEs versus dE,Ns• At the faint end, MB ~ -12, only 10% of dEs are nucleated.• At the bright end, MB ≤ -16, almost 100% are.
• A significant fraction of dEs in Virgo are rotation dominated.
Binggeli et al. 2000, A&A, 359, 447Conselice et al. 2001, ApJ 559, 791.Oh & lin 2000, ApJ 543, 620van Zee et al. 2004, AJ
Fornax dSph
10
Fornax CMD
Carina Dwarf: Episodic SF
Plots from Smecker-Hane 2004
11
IC1613 – Dolphin et al. LeoA – Dolphin et al.
M32 – Alonso et al.
Observational Constraints
All dwarfs so studied haveancient stars (> 10 Gyr)
LG dwarfs: wide range of SFH
12
Substructure in the Local Group
Diagram from Eva Grebel
Giant spiralsdSph (+dEll)dIrrdIrr/dSph
Galaxies mainly clustered around the two principal galaxies MW &
M31
Transformations??
Different types of dwarfs:
dE
dSph
dI
Hybrid/Transition cases (eg, LGS3, Phoenix)
different scaling relations inr, M, L, SB, etc (Kormendy 1985;Kormendy and Freeman 2004).
}
What are the evolutionary relationships?
13
Characteristics of the satellite systems
(Some) simulations suggest that:• abundance and dynamics of the
satallites depend on the host halo• predictions accurate for satellites
with v>50 km/s• VDF significantly different at low v• the Milky Way should have ~50
satellites with v>20km/s with M>3E8 and within 570kpc
Issues:• Resolution in the simulations• Gastrophysics• Reionization
Grebel, Gallagher & Hardbeck 2003Transition type dwarfs
Distance to nearest massive galaxy
dSphs: Filled circlesdEs: Open circlesdIrr/dSph: Filled diamondsdIrr: Open diamonds
•General trend for HI masses to increase with increasing D
•dSphs typically < 105 M•dIrrs typically > 107 M•Transition types intermediate
14
Grebel, Gallagher & Hardbeck 2003Transition type dwarfs
dSphs & dIrrs• Show similar exponential radial surface brightness profiles• Different versions of the same?
dSphs vs dIrrs• dSphs are more metal-rich at same L
•Evident both in PNe and stellar Z• Gas poor• Different histories?
Phoenix, DDO 210, LGS 3, Antlia, KKR 25, Leo T• Locus in L-Z diagram is consistent with dSphs• Have mixed dIrr/dSph morphologies
• Low stellar mass• Low angular momentum• Small HI mass
• Would closely resemble dSphs if gas were removed• Progenitors of dSphs?
The case of Leo T• Discovered by Irwin et al.
2007 MNRAS 384, 535⇒D = 420 kpc⇒V = +35 km/s⇒MT,V = -7.1⇒Rplummer ~ 170pc⇒RHI ~ 300 pc
• CMD has both RGB (6-8 Gyr) and young blue *s (200 Myr)
• Morphologically dSph but SF⇒“transition” dwarf
• Two phase HI(500/6000K)• Stellar mass: 1.2 x 105 M• HI mass: 2.8 x 105 M• Peak NHI = 7 x 1020 cm-2
• Mdyn(R<RHI) = 3.3 x 106 M• M/L > 50 Kopylov et al. 0803.1107
15
The case of Leo TRyan-Weber et al. 2008, MNRAS 384, 535
Cold and warm components seen but no organized rotation
Outer contour 2 x 1019 cm-2
HI radius = 2.5’ = 300 pc
The case of Leo T• Discovered by Irwin et al.
2007 MNRAS 384, 535⇒D = 420 kpc⇒V = +35 km/s⇒MT,V = -7.1⇒Rplummer ~ 170pc ⇒RHI /Rplummer ~ 1.8
• CMD has both RGB (6-8 Gyr) and young blue *s (200 Myr)
• Morphologically dSph but SF⇒“transition” dwarf
• Two phase HI(500/6000K)• Stellar mass: 1.2 x 105 M• HI mass: 2.8 x 105 M• Peak NHI = 7 x 1020 cm-2
• Mdyn(R<RHI) = 3.3 x 106 M• M/L > 50
• Forming stars at slow rate 1.5-2 x 10-5 M /yr
• Gas consumption time ~ 5 x 109 yr
• Low NH but also low central σ ~2 km/s => form *s
Ryan-Weber et al. 2008, MNRAS 384, 535
16
Environmental effectsAstro-ph/0403222
Morphological Segregation• Gas-poor, low-mass dwarfs
(dSphs) tend to cluster around massive galaxies (Cetus & Tucana are exceptions)
• Gas-rich, high-mass dwarfs (dIs) are found to be widely distributed
• Observed in nearby groups and clusters as well as the Local Group
Grebel 2005
Do these trends result from morphological transformations due to the influence of the massive primary galaxy? (i.e. tidal or ram pressure stripping) ⇒ info on orbits of satellites would help
17
1) Not a single dwarf galaxy studied lacked an old population although how dominant that population was varied.
2) Evidence was found for a common episode of star formation (ancient Pop II stars found in Galactic halo and “galactic dSphs”to be same age within 1 Gyr).
3) Even the least massive dSphs showed evidence of some kind of continuous star formation (but with decreasing intensity) over several Gyr - no cessation of star formation during or after reionization.
4) No two histories look the same.Grebel & Gallagher
Dwarf galaxies & metallicity
18
Blow out in Holmberg II
Bureau & Carignan2002, AJ 123, 1316
Blow out in Holmberg II
Rhode et al. 1999
AJ 118, 323
19
Blow out in Holmberg II
Rhode et al. 1999
AJ 118, 323
Extremely low mass galaxies may lose most or all of their ISM after a vigorous star formation episode.
Starbursting and Blue Compact Dwarfs (BCDs)
• Small size
• Very blue
• Optical light dominated by line emission
• Young stellar population (and old stars?)
• Low metallicity => 1/40th solar
20
IZw 18: A Young Galaxy?
I Zw 18van Zee et al. 1998
• Small size
• Very blue
• Optical light dominated by line emission
• Young stellar population (and old stars?)
• Low metallicity => 1/40th solar
Not a young galaxy just one which did not form many/any stars until recently
The Magellanic Clouds
LMC SMC
• The Magellanic Clouds are contained within a common HI envelope.
• The Magellanic Stream traces their interaction with the MW.
• Best fit by tidal encounter with MW (but parameters still uncertain)
21
The Sagittarius Dwarf
• Discovered in 1994 by Ibata et al.
• Small, relatively metal rich
• About 14 kpc away• Being deformed
by the Milky Way• In 1998- 2MASS
shows that stream circles whole sky
David Martinez-Delgado (MPIA) & Gabriel Perez (IAC), The Sagittarius Dwarf Tidal Stream Drawing. From: http://antwrp.gsfc. nasa.gov/apod/ap050529.html
Sagittarius streams in 2MASS
• Used M-Giant Stars from the 2MASS database• Computer modeling: N-body simulations
– Mass and morphology of the Milky Way, mass and location of Sgr
• Used physical constraints to rule out certain models with out a full trial
• • Picked the best fitting oblate, prolate, and spherical models
Law, Johnston, Majewski, 1998
22
Tidal Tidal debris is usually assumed to contribute to the spheroidalcomponent of the galaxy—like the Sagittarius stream—but it may also contribute to the Galactic disk(s
Tidal debris in the Milky Way?
Tidal debris in the disk of the Milky Way
ω Cen
The Jz distribution of metal poor stars in the vicinity of the Sun suggests the presence of distinct kinematical groups.
23
The Milky Way “Field of Streams”
SDSS SEGUE and other surveys designed to measure spectrum of millions of stars in MW to get velocities, metallicities => fossil record of MW history
Ring around the Galaxy: its progenitor?
Ibata et al. 2003
24
Tidal Remnants
HI traces event history even in
absence of stellar debris.
Durrell & DeCesar;
+Yun 1994
HI in the M81/M82 system
Yun et al 1994 VLA-D
De Mello et al. 2007, astro-ph/0711.2685: GALEX +ACS=> modest SF intermediate between low levels in Magellanic Bridge and SF tails in major mergers like Antennae.
25
HI in the M81/M82 system
Chynoweth et al 2008 astro-ph/0803.3631
5 previously unidentified cloudsMHI = (0.7-8.4) x 107 M
A few examples ofTDGs
Duc et al., 1996,1997,1998,1999; Hibbard etal., 1996;Malphrus et al., 1997
These are young, massive,gaseous, TDG candidates
26
Hibbard et al. 1994
• Expected to form in dynamically cool, gas-rich tidal tails• Should be a mix of pre-existing stars from tidally disrupted material
and a “new generation” of young stars produced as HI gas condenses => fall off the metallicity-luminosity relation
• Should contain little dark matter (assuming no DM in galaxy disks)• Although young TDGs may be prominent due to the burst of star
formation, their long-term fate remains unclear
Tidal Dwarf Galaxies
Tadpole galaxy (HST/ACS)
A1367 - Gavazzi et al. 2003, Sakai et al., 2003
Other types of tidal objectsGiant star-forming regions along
tidal tails
Detached « intergalactic »star forming regions
All called« Tidal Dwarf Galaxies candidates »
27
VCC 2062: an old tidal dwarf?
VLA map (blue) superposed on true color (BVR) image of N4694 (left) and VCC 2062 (right) with GALEX (red) overlaid.
Duc et al. 2007, A&A 475, 187
• Faint, low SB dwarf
• Kinematically detached rotating condensation within HI tail
• Linked to disturbed N4694 (SB0 pec)
• VCC 2062 has strong CO and high O/H => deviates from metallicity -luminosity relation
Ultracompact Dwarf Galaxies• Cluster UCD’s (Hilker et al. 1999; Drinkwater et al. 1999, 2000)
• First identified by spectroscopic survey of Fornax; also found in Virgo• To date, found only in cluster centers (but see last slide)• -13.5 < MB < -11.5 (intermediate between GCs and dE/dSph)
• Unresolved in ground-based images• so r½ ≤ 50 pc versus normal dwarfs (r½ ~ 300pc) or typical MW GCs
(r½ ~3 pc)• Formation scenarios:
• Massive (intra-cluster) GCs• Remnant nuclei of stripped (“threshed”) dE,N• Evolved products of YMGCs, massive superstar clusters formed in
galaxy interactions
• “Ultracompact BCDs” identified in SDSS (Corbin et al. 2006)• D(r band) < 6” => D < 1 kpc• z < 0.01 (to allow for high resolution imaging) • Only 9 in SDSS DR2 (rare!)
• Reside in voids• Very metal poor
28
UltraCompact BlueDwarf Galaxies
Corbin et al. 2006, ApJ 651, 861Sample identified from SDSS
Color images (U, Vnarrow, I)• Composite populations• Starburst regions embedded in
envelopes of red stars
UltraCompact BlueDwarf Galaxies
• Young massive cluster at its center• Number of red supergiants• Underlying, low surface brightness component
UM 463M = -14.2D = 27 MpcD(neigh) ~ 2.4 Mpc
Corbin et al. 2006, ApJ 651, 861Sample identified from SDSS
29
Ultracompact BlueDwarf Galaxies
Left: fraction of light at 4020Å from stellar components by age
Right: fraction of mass contained in different stellar population components
Corbin et al. 2006, ApJ 651, 861Sample identified from SDSS
• All UBCDs contain significant population of stars ~10 Gyr old.
• HI content shows no clear pattern; some rich, some not
Compact objects near M87Hasegan et al 2005, ApJ 627, 203 : HST-ACS Virgo Survey
10 compact, high-L (-11.8<MV<-10.8) “dwarf globular transition objects”
30
Compact objects near M87Hasegan et al 2005, ApJ 627, 203 : HST-ACS Virgo Survey
10 compact, high-L (-11.8<MV<-10.8) “dwarf globular transition objects”
Circles indicated inferred tidal radii
Compact objects near M87Hasegan et al 2005, ApJ 627, 203 : HST-ACS Virgo Survey
10 compact, high-L (-11.8<MV<-10.8) “dwarf globular transition objects”
Get r½, σ, M/L =>
• 2 objects like predictions of population synthesis models for old, metal-rich, high L GCs
• 3 are much larger with r½ ~ 20 pc and 6 < M/LV < 9.
• Resemble nuclei of nucleated dEs in Virgo
• 1 uncertain: stellar supercluster?
31
Compact objects near M87Hasegan et al 2005, ApJ 627, 203 : HST-ACS Virgo Survey
10 compact, high-L (-11.8<MV<-10.8) “dwarf globular transition objects”
Half-light radius r½ vsmagnitude for 2000 sources in M87 field.
DGTO’s near M87Hasegan et al 2005, ApJ 627, 203 : HST-ACS Virgo Survey
10 compact, high-L (-11.8<MV<-10.8) “dwarf globular transition objects”
Central velocity dispersion σ vs M for hot stellar systems
32
DGTO’s near M87Hasegan et al 2005, ApJ 627, 203 : HST-ACS Virgo Survey
10 compact, high-L (-11.8<MV<-10.8) “dwarf globular transition objects”
Representation of the Virial Theorem for hot stellar systems
Lower dashed line shows VT for M/LV =1.45.
Upper line shows VT for M/LV =5
Compact objects near M87Hasegan et al 2005, ApJ 627, 203 : HST-ACS Virgo Survey
10 compact, high-L (-11.8<MV<-10.8) “dwarf globular transition objects”
Scaling relations for low-mass hot stellar systems
33
Compact objects near M87Hasegan et al 2005, ApJ 627, 203 : HST-ACS Virgo Survey
10 compact, high-L (-11.8<MV<-10.8) “dwarf globular transition objects”
• DGTOs: • Some similar to old, metal-rich, high L globular clusters• Some are larger: ultra compact dwarfs• Transition in scaling relations between two types ~ 2 X 106 M• Three are embedded in low surface brightness envelopes.
• Presence of DM is fundamental property distinguishing globular clusters from UCDs
• 5/13 DGTOs in Virgo are associated with M87• Proximity to Virgo center may be critical
• UCDs form through tidal stripping of nucleated dwarfs?• But Mieske et al. 2006 AJ 131, 244 found that Fornax dE
nuclei have lower metallicity than UCD’s…• Fornax vs Virgo
UltraCompact Dwarf Galaxies• Most luminous UCDs: -14<MV<-12• Most/all have shallow or steep cusps;
only 1 shows flat “King” core• None show tidal cutoffs (to limit)
Evstigneeva et al. 2007, A.J. 133, 1722Virgo and Fornax images+spectra
ACS images and residuals from fits
34
UltraCompact Dwarf GalaxiesComparison of internal dynamics of UCDs and GCs
Evstigneeva et al. 2007, A.J. 133, 1722Virgo and Fornax images+spectra
•No gap between bright GCs and UCDs
•Virgo UCDs have velocity dispersions and luminosities similar to those of Fornax UCDs
UltraCompact Dwarf GalaxiesFundamental plane for dynamically hot stellar systems
Evstigneeva et al. 2007, A.J. 133, 1722Virgo and Fornax images+spectra
κ-space: axes are combination of central velocity dispersion, effective radius and mean intensity within reff
κ1 ∝ log Massκ2 ∝ log SB3 x M/Lκ3 ∝ log M/L
• Top: UCDs lie on same line as bright GCs but fainter GCs show large spread
• Bottom: UCDs not on main GC relation
35
UltraCompact Dwarf GalaxiesComparison of data with model grids for different samples.
Evstigneeva et al. 2007, A.J. 133, 1722Virgo and Fornax images+spectra
•Majority of dE nuclei are consistent with solar [α/Fe] ~ 0.0
•Majority of UCDshave supersolar ratios (+0.3 to +0.5)
UltraCompact Dwarf Galaxies
Comparison of data with model grids
Evstigneeva et al. 2007, A.J. 133, 1722Virgo and Fornax images+spectra
Hβ index: age sensitive[MgFe] index: total
metallicity
•Virgo UCDs are old•Stellar pops older than present-day dE,Nnuclei
36
Ultra Compact Dwarf GalaxiesKeck Echelle spectroscopy: v(helio), σ, Lick indices (stellar pops)
• Virgo UCDs are old (older than 8 Gyr) and have [Z/H] = -1.35 to +0.35 dex.
• 5 Virgo UCDs have supersolar [α/Fe] abundances• 1 Virgo UCD has a solar abundance ratio
• Virgo UCDs have structural and dynamical properties similar to Fornax UCDs
• Masses ~ (2 to 9) x 107 M• M/LV ~ (3 to 5)• Virgo UCD’s M/L consistent with simple stellar model
predictions => do NOT require dark matter• Origin?
• Could be M/L extreme of known globular cluster population• Could result from simple removal of halo from the nuclei of
nucleated dwarf galaxies => But ages/metallicities argue against simple stripping
Evstigneeva et al. 2007, A.J. 133, 1722Virgo and Fornax images+spectra
Ultra Compact Dwarf GalaxiesPrevious studies have found UCDs only in the centers of rich
clusters; do they exist elsewhere?
• Survey of 6 galaxy groups using 2dF spectrograph on AAT:• Dorado, N1400, N681, N4038 (Antennae), N4697, N5084• Only 1 UCD candidate found; 2 others maybe.
• Compare to simulations of Kravtsov & Gnedin (2005, ApJ623, 650) who give relation between total mass of the cluster population and mass of the host galaxy:
Mmax = 2.9 x 106 M (Mh/1011 M )1.29 ±0.12
• Find that this relation works for Fornax UCDs/N1399 and Virgo UCDs/M87
• What about dominant galaxies in these groups? Would expect to find UCDs in Dorado, N1400 and N5084 group but not others
• Need deeper observations to provide firm test
Evstigneeva et al. 2007
37
DDO 154
Carignan & Beaulieu 1989VLA D HI
Arecibo map outer extent [Hoffman et al. 1993]
Extent ofOptical image
MH = 3.0 x 108 MMstars = 5.0 x 107 MMDyn = 4.0 x 109 M
DDO154
Are there totally “dark” galaxies?
38
Optical galaxy
Giovanelli, Williams & Haynes 1989
HI 1225+01
ML > 200
Dwarf galaxies: summary“Dwarf galaxies of the Local Group”: Mateo 1998 ARAA 36, 435
• Dwarf galaxies exist in large numbers, especially in large clusters
• The Local Group contains a mix of dIrr and dSph/dE’s among its low L galaxies; they are distributed very differently.
• Low L dwarfs tend to be metal poor; the lowest L dwarfs are still composed largely of primordial material.
• Dwarfs are the simplest galactic systems known. But the dwarfs in the LG show complex star formation and chemical enrichment histories.
• Dwarf galaxies may be among the darkest single galaxies known; they provide important constraints on the distribution and nature of DM.
• Tides/interactions are important to the evolution of dwarfs and,possibly, their formation.