Transcript Galactic Structure

Chemical Evolution of
Dwarf Spheroidal Galaxies
Probes of internal and external ‘Feedback’
Rosemary Wyse
Ann Arbor, August 2007
Necessary ingredients include:





Star formation rate
Smooth? Stochastic?
Gas flows
Out/In flows a priori or self-consistent?
Prescription for energy and momentum injection
from supernovae etc
Phase transitions? Radiative transfer..
Stellar IMF
 Fixed at Kroupa, Tout & Gilmore function -- no
evidence for significant variations, but NOT Salpeter
Model of Type Ia supernovae -- rates
Yields
Zeroth order model: Shallow potential well
 Cannot sustain a burst of star formation + Supernovae:
 B.E ~ 3GM2/5R ~ 5x1051erg (M/107 M)2 (1kpc/R)
equivalent to only several SNe
In extreme case of rapid, extensive mass loss, systems
expands to a new, diffuse, equilibrium leading to dSph
(Saito 1979; Dekel and Silk 1986)
Binding energy per unit mass increases with mass,
leading to decreasing efficiency of winds and massmetallicity relationship (Larson 1974; Vader 1986; Arimoto
& Yoshii 1987; Kauffmann & Charlot 1998…etc etc)
 ~10km/s velocity dispersion also corresponds to 104K,
sensitive to (re)ionization (Efstathiou 1990; Bullock et al 01)
 Simplest models derived early SNe-driven winds
dSph should have simple stellar population!
 Also if reionization suppresses further star
formation
Now deep CMDs reveal extended star formation
histories
Hernandez, Gilmore & Valls-Gabaud 2000
Carina dSph
Leo I dSph
Intermediate-age population dominates in typical
dSph satellite galaxies, and star formation rate is low
< 0.0001M/yr
cf. Smecker-Hane et al 1994
Outflows, Inflows
 Extended star formation history obviously means
either managed to retain significant gas after onset of
star formation, or the gas went out, but came back in.
 Gas-free now – why? Ram pressure stripping?
 Low mean stellar metallicity, typically less than a
tenth solar, combined with invariant IMF, means gas
removal from star formation, not by star formation
 Why are many (most?) stars a few Gyr old? Special
epoch of inflow/interactions?
 Tie to formation of the Local Group (Silk, Wyse & Shields
1987) ?
Self-Regulation?
 Recent SPH simulations with self-regulating
SNe heating – cooling – star formation cycle
have faster variations in SFR than observed, and
overall declining SFR (Stinson et al 07)
Regulation of gas phase is promising (Robertson
talk)
Extended, low-rate star formation and slow enrichment
with gas retention, leads to expectation of ~solar (or below)
ratios of [/Fe], such as in LMC stars
Hiatus then burst
Smith et al 2003
Local disk
Gilmore & Wyse 1991
Observational constraints from dSph:
 Want to develop self-consistent models of dSph
evolution, with any SNe-driven outflows
constrained by star formation rate and potential
well, and with inflows motivated
cosmologically, rather than totally ad hoc
 Multi-object medium-resolution spectroscopy
provides large samples with radial velocities
accurate to few km/s and metallicity to ~0.2 dex
We can now obtain even high-resolution spectra
of stars in dSph, and can thus derive detailed
elemental abundances
 HST CMDs for star formation histories
Carina dSph
 Part of Large Program on VLT (PI Gilmore,
co-Is Wyse, Grebel, Wilkinson, Kleyna, Koch,
Evans).
 Determine radial velocities and metallicities
for large sample of candidate members, across
the face of several dSph, plus elemental
abundances for a smaller sample of known
members: FLAMES/GIRAFFE and
FLAMES/UVES
 Constrain potential well from internal
kinematics (Gilmore talk; Walker talk)
Carina dSph: Broad Metallicity
Distribution
437 RGB stars, Ca T spectra
Mean -1.7dex; σ=0.9
Koch et al 2006
Weak dependence of metallicity on galactocentric distance
Closed-box, IRA, outflow proportional
to SFR (Hartwick 1976):
K-giant problem:
Pre-enrichment?
Wind-dominated models:
Lanfranchi & Matteucci (2004); parameter is
star formation efficiency
Tuned wind-dominated model:
Lanfranchi et al. 07
Stochastic Enrichment model
Searle 1977
 Independent star-forming regions, each with
identical ‘enrichment events’ occuring at a
fixed mean rate
 Metallicity of each fragment proportional to
the number of enrichment events in it
 Enrichment is then a Poisson process
 Each fragment assumed to be well-mixed
Model parameter is the mean number of
enrichment events per region – changes shape.
Adopted value of 3 here.
Inefficient stochastic enrichment:
Elemental abundances with bursts of star formation
Gilmore & Wyse 1991
Carina data: bursts + inhomogeneous
star formation
Massive star IMF invariant
Koch et al 2007
Main sequence luminosity functions of UMi dSph
and of globular clusters are indistinguishable.
 normal stellar M/L
HST star counts
Wyse et al 2002
M92 
M15 
0.3M
Massive-star
IMF constrained
by elemental
abundances –
also normal
Stars in satellite galaxies have
different elemental ratios than do
field halo stars…
Milky Way field stars
satellites
Geisler et al 07
Bulk of stellar halo is
OLD, as is bulge:
Did not form from
typical satellites
disrupted later than a
redshift of 2
Unavane, Wyse
& Gilmore 1996
Scatter plot of [Fe/H] vs B-V for local high-velocity
halo stars (Carney): few stars bluer (younger)
than old turnoffs (5Gyr, 10Gyr, 15Gyr Yale)
 Satellites/star forming regions that formed the
field halo must have been accreted/disrupted
prior to self-enrichment by Type Ia supernovae
 And also formed stars only a long time ago, so
if similar to surviving satellites and would have
extended SFH, need to have been accreted a
long time (~10Gyr) ago
 (Λ)CDM predicted much late merging,
incompatible with the Unavane et al limits
 Can newer models be fine-tuned?
State-of-the art ΛCDM prediction, constrained to fit
observations such as total field halo luminosity:
Peak of stellar halo

Robertson et al 2006
Curve: Model predictions from one progenitor,
with 68% of stars to lie on bold solid curve
Crosses: Venn et al 04 compilation of field halo
Concluding remarks:
Data ahead of models…
 Need to look closely at inhomogeneous
chemical evolution models for dSph
 Stars form in clusters/associations,
natural?
 Bulk of the stellar halo of the Milky Way
does not look like the predictions from
ΛCDM model so far..

Type II
Supernova
yields
Salpeter IMF
gives
[/Fe] ~ 0.4
Gibson 1998
Progenitor mass
Schematic [O/Fe] vs [Fe/H]
Wyse & Gilmore 1993
IMF biased to most massive stars
Type II only
Plus Type Ia
Slow enrichment
SFR, winds..
Fast
Self-enriched star forming region.
Assume good mixing so IMF-average yields
M15
Cannot use color on
RGB to estimate
metallicity – there is
presumably an
age-metallicity
relationship that
produces a narrow RGB
(cf. Smecker-Hane et al 1999)
47 Tuc
despite large metallicity
spread.
Blue/red symbol for
[Fe/H] > < –1.7
Targets selected from ESO Imaging Survey
5 fields, each 25 diameter, ~110 stars per pointing