• astrophysics
• Galaxies
• HST
• photometry
• physics
• stellar physics
• stellar populations

This work is aimed to the study of the stellar populations belonging to the Dwarf Spiral Galaxy
M33, through the reduction and the analysis of space-based photometric data and the interpre-
tation of the results through theoretical models.
Thanks to the current space-based and ground based telescopes it is possible to observe
Local Group galaxies, resolving their single components down to low luminosities. Moreover the
current computational capabilities allow to obtain photometric data with high accuracy also in
crowded fields like those in galaxy disks. This allows to study the stellar populations, trying to
recover their evolutionary history. The study of stellar populations also allows to obtain several
distance indicators from which galactic distances are derived and thus the structure of the local
universe can be inquired. These distance indicators are needed to calibrate secondary distance
indicator that are used to infer the cosmic scale distances and through them the Hubble constant.
M33 is a Spiral Galaxy, structured in a disk, where the bulk of its stellar and gas populations
reside, a central compact region called nucleus and an outer, less dense, star distribution called
halo. M33, belonging to Local Group of galaxies, is an ideal candidate thanks to its relatively
low distance (about 880 kpc) and to an intermediate line of sight inclination (about 56 degrees ), which makes
possible to recognise its radial disk structure. Thanks to this properties, this galaxy has been
subjected to several studies form various authors that had observed fields located in many part
of the disk, from the most internal till the outskirts of M33, but there is still uncertainties on the
distance modulus, whose estimation range from 24.50 mag to 24.92 mag. The Colour-Magnitude
Diagrams (CMD) of the various fields show a complex stellar population, as expected for a galaxy,
and chemical composition radial gradient have been determined, but the Age-Metallicitybump 2
relation still today non understood and we are far to reconstruct its evolutionary history.
For my thesis work I reduced new photometric datasets, obtained with the ACS/WFC and
the WFC3/IR instruments on board of the Hubble Space Telescope (HST), relative to two fields,
located at about 10 arcominutes and 14 arcominutes from M33 center, in the optical F606W
and F814W bands for the external field (Parallel) and the near-infrared F160W band for the
internal field (Disk2). In order to obtain the CMD for the internal field, also archival HST
data have been reduced (in the optical bands F475W, F606W and F814W), that overlap to
Disk2 pointing. The adopted reduction data is the following: firs I applied the correction for
the geometrical distortions to the HST frames, which have been previously pre-reduced through
their HST instrumental calibration pipelines. Then I performed the sources research and Point
Spread Function (PSF) photometric procedures with the DAOPHOT/ALLFRAME softwares,
which are optimized for crowded fields. In more details, when the candidate sources list is
obtained. These candidate stars must be checked by view on the frames, in order to choose the
most isolated stars. These will be used to built the PSF model, i.e. the model of bi-dimensional
distribution of photons on the detector for each source. The semi-analytic profile obtained for
each of the selected stars is averaged in the PSF model, that is then used to perform a non linear
lest-square fit of all the sources in each frames, to better measure the magnitude of faint stars in
crowded fields and to reject non stellar sources. The whole procedures is quite interactive and
it requires many iterations in order to obtain a sufficiently accurate photometry. The obtained
stellar catalogues contain about > 750,000 for Disk2 field and > 550,000 for the parallel field.
From these data I obtained the CMD of various stellar populations.
Then I compared these CMD with the theoretical prediction of two different sets of model in
order to establish the main characteristics. Before proceeding with the data interpretation, the
main differences among the two set of theoretical models, the first calculated by the Pisa Stellar
Group and the second by the Teramo Stellar group (BaSTI), have been described, in order to
roughly evaluate the current uncertainties on the results due the still present indetermination
on the adopted physics inputs. The CMDs of the two fields result very similar, showing a
luminosity extended Main Sequence (phase of central burning of H into He) which is compatible
with very recent star formation episodes ( 30 Myr ago), a Red Giant Branch (RGB) (phase
of shell burning of H into He) extended up to the central He ignition and a very populated
Red Clump (phase of central burning of He into C/O), features being highlighted also by other
authors. Nevertheless among the observed features little attention have received an over density
of stars along the RGB, which could be attributed to two different evolutionary phases RGB,
or to the Asymptotic Giant Branch (AGB) clump. In order to recognize this observed feature,
I compared with the observed data with the theoretical positions of RGB bump and of AGB
clump, derived form isochrones, adopting two values of reddening that represent the inferior and
superior limit of reddening, which is not precisely known. Furthermore I also compared the
theoretical magnitude differences between RGB bump and RC and between AGB clump and RC
with the observed magnitude differences between the peaks values of the corresponding features,
in order to avoid the dependence on the adopting reddening value. From the comparisons with
the theoretical predictions, the star over density results compatible with the AGB clump, which
is thus populated with stars of intermediate ages (2 ÷ 5 Gyr), that have finished the central He
burning and that are currently performing the double shell burning (H into He and He into C/O).
The analysis also show that stars with very low metallicity ([Fe/H]<-1.5) should be a minority of
the total population, given that the corresponding RGB bump is not visible. The RC luminosity
results in accordance with the theoretical predictions, in agreement with results of other authors.
There are indication that the AGB bump luminosity can be used as distance indicator,
because it has only a slightly age dependence and metallicity. Thus I used the luminosity of the
discovered AGB bump as distance indicator, obtaining a true distance modulus in agreement with
recent independent estimations for the galaxy. Also the RGB tip luminosity is widely used as a
standard candle, because low mass stars ignite central He burning in similar structure condition
and thus at similar luminosity. I also used the TRGB method determining a distance modulus
in accordance with the AGB distance modulus and other estimations obtained with the same
method.