James Angthopo
Galaxy formation and evolution is an important field in Astrophysics with many
unanswered questions. The two important questions we focus on are what physical
processes are associated with galaxy evolution and how quickly these processes
age the galaxies. To answer these questions, we make use of large surveys and
simulations to study the galaxy properties; finding a distinct bimodal
distribution which separates
young and old galaxies. Between this bimodality, we locate a sparse region,
thought of as the transition region. Understanding the properties of galaxies
located in this transition region is essential, as they enable us to understand what physical
processes are important in galaxy evolution as well as how quickly these
processes will age the galaxy. However, a great hurdle associated with this is finding
an accurate way of defining these regions, as depending on the
parameter space used, we can suffer from different forms of degeneracies, hence
mixing populations from separate regions. Therefore in this poster
we present a novel method of defining these three regions, which limits the
mixing of these populations; more so than previous methods. By using this
cleaner definition, we also study the properties of galaxy population in this
transition region, where we find both agreements, in general, and disagreements,
when considering more detailed analysis, with previous methods.
We also use this definition to
compare observation with state-of-art simulations to try and find what
the simulations reproduces correctly and what the simulation fails to reproduce,
hence providing a path upon which the simulations can be improved.
This presentation examines a novel method of exploring galaxy
evolution, which is less dependent on modelling.
Galaxies are primarily made up of stars, gas and dust. Dust particles
are tiny grains with a wide range of chemical composition. (A) shows
how galaxies look as they evolve from young, Star-Forming (SF),
to old, Quiescent (Q), galaxies. SF galaxies have high star formation rates
while Q galaxies have little/no ongoing star formation.
(B) shows the effect of dust. We can see both ageing and dust
reddens the galaxy. The reason for this is shown in (C); where we see that
dust absorbs/scatters light predominantly at bluer wavelengths.
Furthermore, the wavelength dependence of dust scattering/absorption
causes greater dust sensitivity when measurements are made over
larger wavelength range. (D) shows the spectra of two different
galaxies. The top spectra is of a young galaxy, while the bottom spectra
could either be of a dusty young, SF, galaxy or an old, Q, galaxy.
SLIDE 2
(A) shows the typical spectral range of a broadband colour, consisting
of the flux ratio between the g and r bands. Colour is measured
over a wide wavelength range, thus making it sensitive to dust.
(B) shows the distribution of galaxies from the SDSS on colour vs
velocity dispersion plane (the latter is a proxy for gravitational
potential/stellar mass). The blue, green and white lines show the
Blue Cloud (BC), Green Valley (GV) and Red Sequence (RS),
respectively, where young, transitioning, and old galaxies reside
respectively. Without any dust correction, the overlap of different
regions is evident. Therefore dust modelling is required to
mitigate this degeneracy.
(C) shows the spectral range used for calculating the 4000A break
index, Dn(4000). Both g-r and Dn(4000) can be considered as age
proxies. However, owing to this index being measured over a shorter
wavelength interval, it is resilient to dust.
(D) shows the novel method of studying the galaxy evolution regions,
i.e. using Dn(4000) vs velocity dispersion rather than colour. This
method provides, for the same SDSS galaxies, a clean separation
between BC, GV and RS, without the need to apply dust correction
(thus avoiding inherent systematics).
SLIDE 3
We compare the observational data, SDSS (A), with state-of-art
simulation, EAGLE (B). We analyse the galaxy distribution on the
Dn(4000) vs stellar mass plane, where we ensure the distribution
of SDSS and EAGLE to be identical in stellar mass. The plot shows the
density of galaxies, where purple (orange) are the most (least) dense
regions. In both (A) and (B) the cyan, green and white data points
show BC, GV and RS obtained for SDSS. EAGLE produces a shallower
RS than SDSS, as it has more galaxies at high Dn(4000) in lower
mass galaxies. Furthermore, it also produces older young galaxies as
its BC is at greater Dn(4000) than SDSS galaxies. Similarly, EAGLE GV
and RS galaxies also have higher Dn(4000) values.
SLIDE 4
Here we compare the fraction of different types of GV galaxies for
SDSS and EAGLE. (A) shows the division of GV into lower, lGV, middle,
mGV and upper, uGV, terciles. (B) and (C) shows fractions of Q
(red), SF (blue), and active galactic nuclei (green), AGN, galaxies for
SDSS and EAGLE respectively. (B) shows that AGN contribute
significantly to the population of GV galaxies, specially at high
stellar mass, which the simulation is able to reproduce in
(C). EAGLE also reproduces the general increase (decrease)
in Q (SF) fraction with the increase in stellar mass.
However, for fixed mass, EAGLE underproduces (overproduces) SF (Q)
galaxies in GV, providing a path upon which the simulations
can be improved.