Open Positions
PhD Thesis: Disks or Spheroids? The shapes of high-redshift galaxies
Background:
In the present-day universe,
comparable fractions of stars are in spheroids and in disks
(about 58+-7% of
stars are in spheroids and 42+-7% are in disks, Benson et al 2007).
Most stars
are believed to start their lives in stellar disks (born from gas
disks); if a
major merger occurs, those stellar disks get ‘scrambled’
and become spheroids.
Measuring the fraction of stars in disk and in spheroids at any epoch,
tells us
therefore something about which stars have lived ‘dynamically
quiescent’ lives
since their birth, and which have been subject to mergers. The 3D
shapes of
galaxies are not directly measurable, only their projected 2D shapes.
If one
can select an ensemble of galaxies that have similar intrinsic
properties (e.g.
stellar mass, star-formation rate, size), but is seen from random
viewing
angles, then the 3D shape can be inferred from the distribution of
projected
shapes under reasonable assumptions, such as approximate axisymmetry
(e.g.
Padilla and Strauss, 2008).
Project goal:
The immediate goal of the project
is to understand the shapes (disks or spheroids) of galaxies at
redshift ~2,
the epoch at which most stars formed. This is interesting and
important,
because we have no direct information whether stars formed in disks at
that
epoch and, since the merger rates were much higher at earlier epochs
than now,
we don’t know how long stellar disk survive (until the next major
merger).The
successful shapeanalysis can therefore provide key input in
understanding
galaxy evolution.
The basic approach would be to identify different samples of z~2 (1.5<z<2.5) galaxies that are viewed at random viewing angles, analyse the distribution of their projected shapes (i.e. observed flattening) and infer something about their disk- or spheroid-likeness. The different samples of galaxies differ e.g. in stellar mass (do the most massive galaxies tend to be spheroidal, as they are today? e.g. van der Wel et al 2009) or in star-formation rate (do starforming galaxies look disk-like even at z~2?).
Project outline and practical steps:
Cosmologically distant galaxies
are small (0.5”), therefore such an analysis has to be based on
HST imaging. At
wavelengths
λrest < 4000A the
light
from galaxies is dominated by very young stars, which may have a quite
different geometry than the overall stellar mass distribution; at z~2
this
means that observations have to made at (I+z)*4000A = 1.2μm, or
longer. Only
now, does HST have a wide-field near-IR camera. This camera will carry
out a
900 orbit imaging survey (PI Faber; starting Summer 2010), in which the
advisors are co-investigators (with a stated science interest within
the collaboration
in the above problem).
This will be the data basis of
the thesis project; but an initial analysis can be made (and software
can be
written/optimized), already now on existing public data with
HST’s WFC3 camera.
Broadly speaking the steps are:
• Define a sample of z~2 galaxies
in the HST survey area; this requires ‘photometric
redshifts’ from others and a
way to pick the sample that is ‘viewing angle independent’
sample selection
(this requires selection at long wavelength with IRAC), to be done by
you.
• Set up ‘machinery’ (presumably
the code GALFIT) that measures shapes of galaxies.(code exists but
parameters
have to be adjusted to the problem at hand).
• Get ellipticity distributions
for different sub-samples (do extensive testing/modeling) to assure
that they
are real/correct.
• Compare resulting 3D shape
distributions with predictions from so-called semianalytic models (in
collaborations with theorists at MPIA and elsewhere).
• Answer questions posed
initially…possible thesis papers are:
‘Were the most massive galaxies
round at high redshift?’
‘Did stars also form mostly in
disks at high redshift?’
‘Is the radial light profile (de
Vaucouleur or exponential) correlated with the shape of the
galaxy?’
Practical consideration:
This project requires:
familiarizing with many concepts of redshift galaxy research; using
existing
analysis software; writing your own modeling software (not very
extensive),
i.e. becoming proficient in either C, IDL, Python etc.. This will be
your
project, but it is embedded in a larger collaboration. This provides
opportunity
for collaborative visits to the US and to attend collaboration
meetings. This
thesis can lead to an important result, but requires hard-work and
self-motivation.
Availability: pending, please inquire
Please send a brief application to rix@mpia.de
last updated: March 2010
