Metallicity and age of selected nearby GK

1
Metallicity and age of selected nearby G-K
Giants

• L. Pasquini (ESO)
• M. Doellinger (ESO-LMU)
• J. Setiawan (MPIA)
• A. Hatzes (TLS), A. Weiss (MPA), O. von der Luhe
  (KIS)
• L. da Silva (ON), R. de Medeiros (UFRN)
• L. Girardi (INAF-OAP), M.P. di Mauro (INAF-OAR)

Setiawan et al. 2003a,b, 2004, 2005, da Silva et
al. 2006, Doellinger et al. 2006a,b
2
Why Selected ??
These G and K giants have been selected to be
studied for accurate radial velocity
measurements, to understand the nature and
evolution of RV variations among evolved
late-type stars Many Giants are known to be RV
variables, with short timescales (hours to
days) ? ? Pulsations ? Solar-type
oscillations? long timescales (hundreds of
days) ? ? Planets? Inhomogeneities
(rotational modulation)? How frequent are they
? Which amplitude? How do they evolve along the
H-R diagram with mass, age etc.. ??

3
Sample(s)
Stars G-K giants with accurate HIPPARCOS
parallaxes (10) not known binaries,
distributed along the H-R diagram
2 groups 77 stars from the South, 62 from the
North Southern Observations FEROS-La Silla,
1999-2003 (ongoing) R50000 ?RV 22 m/sec
(Setiawan et al. 2004) Northern Observations
TLS H-R spectro. 2004-2006 (ongoing) R67000
?RV 5 m/sec (Döllinger et al. 2006a)
(circumpolar objects)

4
The Sample H-R diagrams

Northern Sample
Southern Sample
Subgiants, RGB stars, Clump Stars, early AGB
5
RV Statistics
Southern Stars 13 Binaries (17) 21 Constant
(27) 43 Variable (56) 3 Planets confirmed 8
Activity Modulation or pulsation Ca II and/or
bisector variations with Period ?RV22 m/sec
Northern Stars 11 Binaries (16) 6 Constant
(9) 9 Long Period (15) 1 Planet confirmed 36
Short Period (60) ?RV5 m/sec

6
?RV along the H-R diagram
Does ?RV depends on the absolute stellar
magnitude (Southern Sample) ?
BD comp. ?Planets

Apparent trends are contaminated by different
physical causes
7
Further causes of ?RV
The apparent trend requires a close scrutiny.

Indications that the most luminous stars show
inhomogeneity modulation
8
Chemical analysis
The well known age-metallicity degeneracy
requires determination of Fe/H to derive more
precise stellar characteristics
FEROS and TLS spectra (templates) for analysis
S/N gt150, R50000-67000, large spectral
coverage Initial Teff from Photometry, log(g)
assuming M1M? Spectroscopic determination LTE,
ionization equilibrium, no dependence on eq.
width, no dependence on exc. potential Sun plus
2 well-studied giants in literature for zero
point correction HD113226 (?Vir), HD27371
(Hyades giant)

9
Parameters determination
Comparison of the stellar parameters with
theoretical isochrones (Girardi et al. 2000)
using a modified version of the Bayesian
estimation of Jorgenson and Lindegren (2005)
algorithm. The total probability distribution
function of belonging to M,t space is computed
from the Mv, Teff, Fe/H probabilities.

10
Sanity Checks
Several sanity checks were made (B-V)-(B-V)o,
Gravity, Radius

Teff-(B-V) transformation can be improved,
known to be not suitable for cool stars
lt(B-V)-(B-V)ogt -0.009 ? 0.03
11
Sanity Checks

Spectro - log(g) gt evolutionary log(g) known
non-LTE, strong dependence on ? ?? 0.07 km/s
--gt ?log(g) -0.2 Effect on Fe/H negligible
(FeI)
Red points lunar occultation or long baseline
interferometry. Agreement to better than 6
12
Results
? 0.12 for Mgt1.2M and Fe/H0

13
Age-Metallicity Relationship
The cumulative Fe/H/age bin (?t) is computed by
weighting the measured Fe/H with the
probability of belonging to ?t
The youngest stars (age lt 1 gyrs, 17.1 stars)
are compatible with a single value of Fe/H
(sigma0.09).
Spread become significant at higher ages at 4
Gyrs much larger than measurement errors. AMR
is flat up to the largest age only -0.2 -0.3
at 12 Gyrs The results of the young bin are
different from what found in main sequence
surveys Giants (and subgiants) seems on the
other hand very good tracers of the young
population

14

STARS Hosting Planets
4 Stars (3 in South, 1 in the North) have been
shown to host planets
(Setiawan et
al. 04, 05, Döllinger 06)
?
?
?

?
GIANTS hosting planets do not seem to be
preferentially metal rich!