Transcript Stars with T eff

Current uncertainties
in
Stellar Evolution Models
Santi Cassisi
INAF - Astronomical Observatory of Teramo, Italy
The “ingredients”
An evolutionary code
• Numerics
• Boundary conditions
•1D versus 3D
Equation of State
Radiative opacity
Conductive opacity
Nuclear reaction rates
Neutrino energy losses
Physical inputs
•
•
•
•
•
Mixing treatment
• Overshooting
• Superadiabatic convection
• Non-canonical processes
Microscopic mechanism
• Atomic diffusion
• Radiative levitation
Additional mechanism
• Mass loss
• Rotation
• Magnetic field
Some pieces of evidence…
Very Low Mass stars:
King et al. (1998)
Zoccali et al. (2000)
✓ Surface boundary conditions
✓ Equation of State
✓ Opacity
Richer et al. (2008)
Segransan et al. (2003)
Eclipsing binary: an important benchmark
1/2
The case of V69 in the Galactic GC 47Tuc
(Thompson et al. 2010)
BaSTI
Victoria
≈1.5Gyr
Dartmouth
When the differences (He content,
heavy elements distribution, diffusion
efficiency, etc…) are taken properly
into account, the difference can be
reduced to about 0.8Gyr,…
The Age – Luminosity calibration: the clock
A comparison among the various stellar model libraries suggests that an
uncertainty of about 1Gyr (i.e. ≈10%) do exist at the older ages…
Eclipsing binary: an important benchmark
2/2
The case of V20 in the Galactic Open Cluster NGC6791
(Grundahl et al. 2008)
(m-M)V=13.46 ± 0.10
E(B-V)=0.15 ± 0.02
Victoria-Regina (t=8.5Gyr)
Photometry by Stetson et al. (2003)
Kalirai et al.(2007)
Red Giant Branch Stars
The location and slope are
dependent on the metallicity…;
strongly
The RGB Tip brightness is one of the most
important “primary” distance indicators;
47 Tuc
HST Snapshot
Piotto et al. (1999)
RGB star counts are quite important:
•to check the inner chemical stratification;
•being RGB stars among the brightest and
cooler objects, their number (+ AGB
stars)controls the integrated properties in
the NIR bands;
•the RGB/AGB number ratio provides hints
on the Star Formation History of complex
stellar populations (Greggio 2002);
Accurate RGB modeling is mandatory for interpreting data of unresolved stellar
systems using population synthesis tools as well as for estimating the properties of
resolved systems by means of isochrone fitting techniques
The state-of-art of RGB models:
the luminosity function
Theoretical predictions about RGB star
counts appear a quite robust result
RGB bump
M13: Sandquist et al. (2010)
What is present situation about the level of agreement between
between theory and observations concerning the RGB bump
brightness?
The RGB bump brightness
To overcome problems related to still-present indetermination on GC distance
modulus and reddening, it is a common procedure to compare theory with
observations by using the ΔV(Bump-HB) parameter
Monelli et al. (2010)
Does it exist a real problem in RGB
stellar models or is there a problem in
the data analysis?
The brightness of the Red Giant Branch Tip
RGB tip
The I-Cousin band TRGB magnitude
is one of the most important
primary distance indicators:
• age independent for t>2-3Gyrs;
• metallicity independent for [M/H]<−0.9
The TRGB brightness is a strong
function of the He core mass at
the He-burning ignition
TRGB: He core mass – luminosity
≈ 0.03M
Salaris, Cassisi & Weiss (2001)
These differences are – often but not always…- those expected when
considering the different physical inputs adopted in the model computations
TRGB: He core mass & luminosity
• last generations of stellar models agree – almost all – within ≈ 0.003M
• a fraction of the difference in McHe is due to the various initial He
contents – but in the case of the Padua models…
• the difference in Mbol(TRGB) is of the order of 0.15 mag when excluding
the Padua models…
The TRGB brightness as Standard Candle:
theoretical calibrations
ω Cen – Bellazzini et al. (2001)
The I-band theoretical calibrations appear sistematically brighter by about 0.15 mag
The TRGB brightness: theory versus observations (an update)
Updated RGB models are now in
agreement with empirical data at
the level of better than 0.5σ
In the near-IR bands, the same calibration
is in fine agreement with empirical
constraints (but in the J-band…)
The reliability of this comparison would be largely improved by:
• increasing the GC sample…;
• reducing the still-existing uncertainties in the color-Teff transformations
The Horizontal Branch
The brightness:
• a primary standard candle
• Star counts  The R parameter
The ZAHB luminosity is mainly fixed by
the mass size of the He core@TRGB
Any physical inputs affecting the value of
McHe, has a strong impact on the ZAHB
luminosity
The color distribution:
• the 2° parameter problem
The color location along the HB DOES depend on
the mass loss efficiency along the RGB
Peculiar “patterns”:
• rotation
• surface chemical abundances
The ZAHB brightness: an update
De Santis & Cassisi (1999)
• The difference among the most recent models is about 0.15 mag
• All models but the Dotter’s ones, predict the same dependence on [M/H]
Mass loss along the RGB: the impact on the HB
The impact of mass loss phenomenon on the evolutionary properties of RGB stars
is (…not always!...) negligible, but…it is very important for the Horizontal Branch
High mass-loss efficiency
Dorman, Rood & O’Connell (1993)
low mass-loss efficiency
The integrated magnitudes & colors of stellar systems
can be largely affected by the HB morphology (see Conroy’s talk…)
Mass loss on the RGB: not good news
“Investigations of the impact of RGB mass loss upon the HB morphology have
mostly relied on the Reimers’s (1975) formula, and it is widely used as a LAW”
(Catelan 2005)
But…

various prescriptions do exist

they predict quite different
mass loss efficiency
These formulae are not able to reproduce the mass-loss rates
measured by Origlia et al. (2002, 2007)…
HB stars show a number of peculiarities
Discontinuities in the abundance ratios
Diffusive processes (atomic diffusion + radiative levitation) are
really at work in HB stars! What about stellar models…?
Stellar model predictions
(Michaud et al. 2007,2008)
Various masses to cover range of Teff
along the HB;
In color from 15 to 30 Myr after ZAHB;
Same turbulence model for all masses;
Data for M15 by Behr (2003)
• Stars with Teff < 11000 K have same
metals as giants of cluster;
• Stars with Teff > 11000 K have X100
overabundance;
• Overabundances explained… but normal
ones suggest something else also present
since overbundances by X5 expected;
Discontinuity in the rotation rates
Globular clusters
(Behr 00 + 03, Recio-Blanco et al. 02 + 04)
Field Stars
Some embarrassments:
•they rotate…
•some of them rotate fast…
•it seems to exist a discontinuity…
[Fe/H] and rotation along the HB:
an observational link
•
Data from Behr et al. (1999,
2003) - black: M3;
red: M13;
green: M15; blue: M68;
brown:
NGC 288;
•
Stars with Teff < 11000 K have
[Fe/H] as RGB stars;
•
Stars with Teff > 11000 K have
[Fe/H] values from 10 to 100 times
larger;
•
Stars with anomalies have slow
rotation; note star with arrow;
Any clues from stellar models?
Meridional circulation in HB stars
Trend with the Teff of the limiting rotational velocity for He settling in
presence of meridional circulation for two cases:
• Ciculation enters the He convection zone (dots);
• Circulation does not enter into the convection zone (triangles);
Dotted curve: ~max observed Vsin i;
Circulation wipes out
about 11000 K;
anomalies
below
Dark gray region: no anomalies observed;
White region: anomalies observed;
Quievy et al. (2009)
but NO clues on why
HB stars rotate…
The Asymptotic Giant Branch
The AGB evolutionary phase is very important for many reasons as:
• Neutron Capture Nucleosynthesis
• Population tracers
• Integrated properties of resolved & unresolved stellar populations
Marigo et al. (2008)
But… it is in the age regime when AGB stars dominate the SED, where
different population synthesis models give - quite - different results
SED for SSP
In some cases - as in Maraston (2005) - the AGB contribution to both the
bolometric and near-IR light of a stellar population, is much larger (a
factor of 2 or more…) than in other models…
see the talks by Bruzual and Conroy
AGB stellar models: why a THORNY problem?
pulsations
Mass
loss
mixing
opacity
burning(s)
Nucleosynthesis
Brightness
Effective temperature scale  colors
Evolutionary lifetime
Initial – Final mass relation
The TDU efficiency: an unsettled issue
Problem:
How to treat the mixing during the TDU?
Solution(s):
•Bare Schwarzschild criterion
•Envelope overshoot
•Time dependent mixing
Free (!) parameter(s)
•Diffusive process
The mixing efficiency during the TDU has important effects on:
• the rate of surface C-enhancement;
• the effective temperature scale and colors;
• the mass loss efficiency and, in turn, the TP stage lifetime;
• the amount of s-elements @ the stellar surface ;
The 2th problem: opacity for C-enhanced mixtures
Long time ago, Scalo & Ulrich (1975) showed that: TiO and H2O are the most
important molecules in the oxygen-rich regime (C/O<1), while carbon-bearing
molecules (C2, CN, C2H2 and C3) dominate the opacity for C/O>1
A crucial issue!
Fundamental further steps ahead have
been NOW made (Lederer & Aringer 2008,
Marigo & Aringer 2009, Weiss & Ferguson 2009)
What is the impact on the AGB stars effective temperature scale?
The importance of an appropriate treatment of C-rich mixture opacity
Direct effect:
• huge decrease of the effective temperature
Marigo & Girardi (2007)
Indirect effect:
• strong increase of the mass loss efficiency…
Fully evolutionary AGB models: is there a general consensus?
Weiss & Ferguson (2009) versus Karakas (2003) and Wassiliadis & Wood (1993)
No overshooting
Overshooting
+ WF09
A comparison among independent “fully AGB models” shows that:
• relevant differences exist both in the TP lifetimes and TPs number;
• sometime the differences have no explanation (as between K93 and WV93…);
• significant differences do exist also for the He core mass predictions…;
Conclusions