Presentazione di PowerPoint

1
Observations of Globular Clusters (of relevance
for the MODEST collaboration)
Giampaolo Piotto Dipartimento di
Astronomia Universita di Padova
Collaborators Jay Anderson, Luigi R. Bedin,
Santi Cassisi, Francesca
De Angeli, Ivan R. King,
Yazan Momany, Marco Montalto, Alejandra
Recio Blanco, Sandro Villanova

2
Recent Instrumental Advances
New instruments for both imaging and spectroscopy
have strongly affected the research topics in
globular cluster astronomy. We have also started
to take advantage of the 20-25 year baseline of
images on solid state digital imagers and,
overall, of more than 10 year baseline of HST
imaging for for high accuracy proper motions!
3
High Precision Astrometry on WFPC2/ACS HST Images
(Anderson and King 2002, 2003)
Just the random error remains 0.02 pxl on the
WFPC2 (0.01 pxl on the ACS) which
corresponds to 1 mas (PC) on a single imagewith N
images N 1 mas /sqrt(N) (in the PC case)
4
Astrometry (1) Identify cluster members for
deep surveys
Hunting the bottom of the Main Sequence down
to the hydrogen burning limit (HBL)
NGC6121M4
(Bedin, Anderson, King, Piotto 2001, ApJL, 560,
L75)
5
Luminosity-Radius Relation (LRR) The
models cannot fit the main sequence at
intermediate and high metallicities
low M/H
intermediate M/H
NGC 6397
King, Anderson, Cool, Piotto, (1999)
M4
(Bedin et al. 2001)
6

Mass functions in different radial
bins Observational constraint on mass
segregation. Set constraints on the
cluster dynamical model.
NGC 6121 M4
Bedin et al. 2004, in prep
7
Ongoing projects
Cluster Camera Fe/H NGC6397
WFPC2 -2.2 NGC6121 WFPC2/ACS
-1.2 NGC104 ACS -0.7 NGC6791
ACS 0.4 NGC5139 ACS
-1.6/-0.5
8
Example 47 Tucanae CMD spanning more than 17
magnitudes, from the RGB tip down to
Mv15, close to the HBL

GO9444
GO9648
Bedin, Piotto, King, Anderson, in prep.
9
Present Day Mass Function
Initial Mass Function
Ongoing Projects (in coll. with D.
Heggie) HST NGC 2808 NGC 5024 VLT NGC
6981 Archive A lot of data in both HST and VLT
archive.
10
ABSOLUTE MOTIONS
Astrometry (2) Measure proper motions
11
of M4
(U,V,W)LSR ( 53- 3, -202-20, 0-
4)Km/s
(P, Q, Z)LSR ( 54- 3, 16-20, 0-
4)Km/s
12
Once corrected m l cos b and mb for the Sun
peculiar motion we can get
Bedin, Piotto, King, Anderson 2003, AJ, 126, 247
13
Astrometry (3) GEOMETRICAL DISTANCES OF GLOBULAR
CLUSTERS This is our major project, at the
moment Globular cluster age measurement error is
dominated by uncertainty on distance, which is
at least 10 gt 0.2 mag distance modulus,
which translates in a gt25 error in age!!!
14
Direct measurements of distances are several
years away (GAIA, SIM,) and we have to rely on
standard candles, whose luminosiy is still
poorly known, and sometimes strongly dependent
on other parameters as metallicity (e.g. RR
Lyrae). We need reliable measures of
distances for as many GGCs as possible, covering
a wide range of metallicities in order to measure
accurate ages
15
Our method is very simple (in principle
) we compare the dispersion of the internal
proper motions (an angular quantity) with the
dispersion of the radial velocity (a linear
quantity) it is not a new idea, but now
16
INTERNAL DYNAMICS
(Bedin et al. 2003)
17
and thanks to instruments like the
high resolution multifiber spectrograph FLAMES_at_VL
T We get thousands of radial velocities per
night!!!
18
The main source of error is the sampling error
1/sqrt(2N). For a typical sample of 3,000 stars
this implies an error of 1.3 in the distance.
The distance scale obtained will not be only
sound, but its uncertainty will no longer
contribute to the uncertainty on the age
estimates.
19
NGC 2808
20
M4
FLAMES_at_VLT/ESO ESO-071.D-0205(A) ESO-072.D-0742(A)
plus several HST GO, (last GO-10146)
21
(Formula from Pryor Meylan 1993)
Error budget is very important!
Harris
2003 2.2 kpc Diff 20 closer!!! Better
agreement with Peterson, Rees Cudworth et al.
d1.72/-0.14 kpc
This is a preliminary calculation!!!
22
The sources of systematic errors are
- estimates of the observational
errors PLUS - mass segregation
- rotation MODEL !!!
- anisotropies
23

• Model predictions
• O Data

We fit the observed radial velocities and
internal proper motions with a superposition of
orbits constructed with an axisymmetric dynamical
model (Schwarzschild models). The orbit library
is generated using the code developped by
Gebhardt et al. (2000). F. De Angeli PhD thesis
Preliminary results for 47Tuc
24
Ongoing work on proper motions example
HST observations completed last month
GO9899, PI Piotto
25
Ongoing work on radial velocities example
NGC2808 2000 stars observed (FLAMES_at_VLT, PI
Piotto)
In addition NGC6121 (2600)
NGC6397 (1700) NGC6752
(1500)
26
Geometrical distance project priority list NGC
6121 Least model dependent! NGC 2808 NGC 6752 NGC
6397 NGC 5139 NGC 104 plus 7 other clusters
with at least two epoch HST observations
27
Why should all this be of interest for MODEST?

• From the various proper motion projects we get
• Accurate proper motions AND radial velocities for
  up to a few thousand stars, from the cluster
  center to many core radii from the center
• Mass functions, in a few cases down to 0.1 solar
  masses
• Mass segregation
• For a selected number of clusters, accurate
  distances and ages
• In some cases, absolute motion

28
Accurate Reddening and Metallicity measurement
with GIRAFFE/UVESFLAMES_at_VLT
Cluster Giraffe UVES
Ongoing ESO program (PI Gratton) Targets Metalli
cities with 0.03dex uncertainties (UVES
data) Reddening with 0.005 mag. uncertainty
(GIRAFFE data
Coupled with the geometric distance project we
should be able to measure GC ages with a few 100
Myr uncertainties
29

• 74 GC cores observed with the WFPC2 in the F439W
  and F555W band all clusters with (m-M)Blt18
• Data reduced with DAOPHOT and ALLFRAME
• Data calibrated to both HST Flight and standard
  Johnson B, V systems following Dolphin (2000)
• Completeness available for all the CMD branches
  (7100 experiments with more than 5 million
  artificial stars)
• Photometric data and completeness are
  available at http//dipastro.pd.astro.it/globulars
  
• The database has proven to be a mine of
  information

Piotto et al. (2002), AA, 391, 945
30
Relative Ages of Galactic Globular Clusters
Within each single bin, GCs are coeval, with an
age dispersion less than 1Gyr (smaller for the
most metal poor clusters). Clusters with
Fe/Hlt-1.5 appears 1.5-2 Gyr younger, but this
second results is totally model dependent.
31
Omega Centaurithe population puzzle goes deeper
Astrometry (4) Omega Centauri. Accurate
astrometry implies accurate photometry!

32
The problem of the double MS and of the multiple
SGBs and TO
Bedin, Piotto et al. 2004, ApJL, 605, L125
33
(No Transcript)
34
While the multiple TO could be understood in
terms of a metallicity (and age) spread, the
double main sequence represents a real puzzle.
35
Is it a structure in the background?
Bedin, Piotto, Anderson et al. 2004, ApJL, 605,
L125
Leon, Meylan, Combes 2000
36
Bedin et al. (2004) have proposed an alternative
explanation for the Omega Centauri double main
sequence It represents a population of
super-helium rich stars (Ygt0.30), which might
be originated by material polluted by
intermediate mass (1.5-3 solar masses) AGB
star ejecta. This would be consistent with
1) The increase of s-process elements with
metallicity found by Smith et al. (2000)
2) The anomalously hot horizontal branch 3)
The lack of correlation between period shift and
metallicity among RRLyr stars (Gratton et
al. 1986) ESO DDT project (PI Piotto) approved
for 15hr at FLAMES_at_VLT in order to verify this
hypothesis 3 HST extra orbits allocated on DDT
(GO10101, PI King)
37
17 blue main sequence 17 red main sequence 33
upper SGB 32 middle SGB 23 lower SG
FLAMES GIRAFFE Observatios in June2004
38
First results the double main sequence
RedMS Rad. Vel. 235-11km/s Fe/H-1.56 BlueMS
Rad. Vel. 232-6km/s Fe/H-1.27 It is more
metal rich!
17x12204 hours i.t.
Piotto et al., ApJL, in preparation
39
Other chemical elements
Blue Main Sequence C/Fe0.0 N/Fe1.0 Ba/Fe
0.7
Red Main Sequence C/Fe0.0 N/Fe1.0 Ba/Fe
0.4
The blue main sequence stars are richer in Ba
(s-process element), but NOT carbon rich. This
is the second important result. The fact that
there is no significant radial velocity
difference and no significant difference in
proper motion make the background object
explanation even more unlike. The only other
possibility is indeed a strong He overabundance
40
An overabundance of helium (Y0.40) indeed can
reproduce the observed blue main sequence.
The fact that the more metal rich, and possibly
helium richer stars are not carbon rich seems
to exclude that the cloud has been contaminated
by AGB ejecta. According to Thielemann et al.
(1996) SNe from 8-12 solar mass stars should
produce a huge amount of helium. Material
polluted by these SNe could in principle
originate stars with the chemical content of
the blueMS stars in Omega Centauri.
41
Future Plans

• Observations
• 1) Reduce the new ACS/HST images (foreseen for
  June 2005)
• to follow the two MSs in Omega Cen down to the
  hydrogen
• burning limit Use the first epoch of the same
  field for accurate
• proper motions of the stars in the two MSs
• 2) With improved ACS photometry search for main
  sequence splits
• and/or broadening in other globular clusters.
• Theory (of interest for MODEST!)
• Investigate the fraction of material ejected by
  SNe from 8-12
• solar mass stars that can be retained within the
  cluster
• (see also proposal at the end of the talk).

42
NEXT STEP FOR ASTROMETRY GROUND-BASED
ASTROMETRY Example WFI_at_2.2m ESO 12 mas/frame
A post doc (Ramakant Singh Yadav) full time
dedicated in Padova
43
IN JUST 2.8yr
NGC 6121-M4-WFI_at_2.2mESO
44
Blue Stragglers from the snapshot catalog

• Blue stragglers (BS) are present in all of our 74
  CMDs
• Almost 3000 BSs have been extracted from 62 GCs
• The location of BSs in the CMD depends on
  metallicity
• The brightest BSs have always a mass less than
  1.6 solar masses
• In all GCs, BSs are significantly more
  concentrated than other cluster stars.

45
Piotto, De Angeli et al. (2004, ApJL, 605, L125)
Ns represents the density of stars in a
cluster. (i.e. the observed number of stars has
been divided by the fraction of the cluster light
sampled by our WPFC2 images, and then divided by
the total cluster light). There is a
significant correlation between the BSS
frequency and the total cluster luminosity
(mass) and a very mild anticorrelation with the
central collision rate.
46

• Here, we plot the
• estimated total number
• of stars, obtained from
• the observed counts,
• divided by the fraction
• of the cluster light
• sampled by our images
• Note that
• The total number
• of HB stars scales
• linearly with Mv,
• or the total mass,
• as expected.
• 2) The number of BS
• is largely independent
• from the total mass

47

Evolutionary pathway to produce Blue Stragglers
in GCs
A more massive main sequence star exchanges into
a binary containing two main sequence stars. The
primary evolves off the main sequence and fills
the Roche lobe. The secondary gains mass and
becomes a blue straggler. Blue stragglers will
form earlier in binaries containing more
massive stars, i.e. in high collision rate
clusters. Given the finite lifetime of a blue
straggler, the blue straggler population (from
primordial binaries) in the most crowded clusters
today could be lower than in very sparse clusters.
Davies, Piotto, De Angeli 2004, MNRAS, 349, 129
48

Production of Blue Stragglers in GCs
Davies, Piotto, De Angeli 2004
49
Blue Straggler Luminosity Function
On the basis of our model, we expect to find
predominantly BS produced by collisions in
clusters with Mvlt-8.8. These BS are expected to
be brighter (Bailyn and Pinsonneault 1995) This
prediction seems to be confirmed by the
observed luminosity function.
50
We have extended our investigation to open
clusters
NEW!!!! The trends continues into the mass
regimes of (relatively) old open
clusters (agegt0.5Gyr). (The high noise for open
clusters is mainly due to the small number of red
clump stars.)
GCs
Open clusters
51
BSS in Open Clusters If we include the
total cluster sample, the anticorrelation with
the total magnitude (mass) is even more
evident (extending the trend already observed for
GCs). Apparently, there is also a correletion
with the cluster age, with older clusters having
more BSS
Log(age)
Total Absolute Magnitude
52

Extended horizontal branches 22 out of the 74
clusters of the snapshot database show a blue
tail which extends to Tegt20.000K, entering into
the EHB regime. A number of these have been
identified as EHB clusters within the snapshot
project. In practice, 25-30 of the clusters of
our sample have a blue tail extended to
Te20.000K or more. EHB are not so rare, after
all!
WHY?
53

Horizontal Branch Extension
For each cluster we fitted a model to obtain the
temperature of its hottest stars, as an index
of the HB extension. Then we started
by exploring simple pairwise correlations.
54
Metallicity the first parameter
Clearly there is a correlation between the HB
extension and metallicity. The metallicity is
the first parameter, afterall. There is also a
large dispersion. Indeed, The metallicity
explains only 32 of the total
variance. Basically, this is the second
parameter problem.
55
New important correlations the total absolute
magnitude
The total absolute magnitude accounts for 19 of
the total variance.
Note the if we remove the most metal rich
clusters (for which the metallicity effect
dominates), the correlation between the HB
extension and the total absolute magnitude
(mass) is much stronger.
56
No correlation with the central density or
other relevant cluster parameters
57
Multicomponent Analysis
PCA analysis confirms that the HB
extension correlates with Fe/H and Mv (i.e.
total mass)

58
Why the dependence on the total cluster mass? A
possible explanation could be related to what we
have found in Omega Centauri
self pollution!
IF a significant fraction of the material lost
by intermediate mass AGB stars and/or SNe can be
retained by the cluster and contaminate the
medium from which less massive stars are still
forming,we would end with low mass stars richer
in helium. Stars richer in helium would become
bluer HB stars. In this scenario more massive
clusters would be able to retain material from
the AGB/SNe ejecta than less massive ones, and
therefore would end with more extended HBs as
observed!
DAntona et al. (2002)
59
A proposal for MODEST collaboration

• The new results in Omega Centauri and on the
  extension of the
• dependence of the horizontal branch in globular
  clusters on the
• cluster total mass rise a number of questions.
• Can the ejecta from SNe generated by 8-12 solar
  mass stars
• be retained inside a globular cluster?
• 2) Can the ejecta from intermediate mass AGB
  stars be retained
• inside a globular cluster?
• 3) Which is the fraction of retained ejecta as a
  function of the
• cluster mass, mass function, etc.?
• 4) Which is the resulting chemical contamination?