Figure 2.
Half-light radius versus
stellar mass for galaxies
with photometric
redshifts 2 half-light
radius versus stellar
mass for galaxies with
photometric redshifts
2 < z < 4. Red symbols
indicate objects best
fit with de Vaucouleurs
profiles and blue symbols
objects best fit by
exponential disks. Objects
from the CANDELS survey
by HST are shown as faint
dots. Stars indicate the
host galaxies of starbursts
detected by Herschel.
The dotted cyan line
indicates the resolution of
HST, and the dot-dashed
black line shows that of
GeMS/GSAOI. Note that
the estimates from the
Gemini data tend towards
smaller sizes at a given
stellar mass than those
from HST.
As a pilot program, we observed three of
these fields (including ES1C; Figure 1, previ-
ous page) with GeMS/MCAO for between 30
and 90 minutes each (Lacy et al., 2018). The
resulting images have limiting magnitudes
of 24-24.6 and resolutions (Full-Width at
Half-Maximum of the PSF) of 0.07-0.16 arc-
second, depending on the integration time
and observing conditions.
What Can We Learn From
Higher Resolution?
Galaxy sizes (at a fixed stellar mass) are seen
to grow rapidly with cosmic time from z ~ 2
to z ~ 1. The mechanism for this is currently
the subject of debate, though mergers are
likely to play an important role. The smaller
Gemini PSF allows more accurate estimates
of the sizes of the most compact galaxies
(those with scale sizes < 1 kpc) than previ-
ously possible. At z < 2, we obtained results
on galaxy sizes as a function of stellar mass
that are similar to those from studies with
the HST (e.g., the CANDELS survey, van der
Wel et al., 2014). At z < 2, however, we see
evidence of a higher fraction of compact
star-forming galaxies (Figure 2). Although
this needs to be confirmed by obtaining
higher signal-to-noise profiles from deeper
observations, this could imply an even more
extreme size evolution in the galaxy popula-
tion than currently assumed.
In the GeMS fields there are several sources
detected in the far-infrared HerMES survey
(which used the Herschel Space Observatory).
At the redshifts we are seeing them (z ~ 1-3),
they correspond to ultraluminous infrared
galaxies (ULIRGs). In the local Universe, the
far-infrared emission from ULIRGs is typically
powered by starbursts and AGN. To obtain
redshift estimates, and to disentangle the
contribution of these two power sources, we
used multiwavelength data from surveys in
the optical and infrared, as well as new ra-
dio continuum data from the Australia Tele-
scope Compact Array and Very Large Array.
The ULIRGs we identify consist of a combi-
nation of pure starburst galaxies and com-
posite AGN/starburst objects. We find that
the ULIRGS with strong AGN tend to reside
in hosts with smaller scale sizes than purely
star-forming galaxies of similar infrared lu-
minosity.
Like their local counterparts, the
ULIRGs in this study seem to show
signs of recent merger activity, such
as highly disturbed morphologies.
We also find a candidate triple AGN
system (Figure 3), which consists of
three AGN with photometric red-
shifts of z = 1.4 (spectroscopic red-
shifts are required to confirm the
triple AGN system): one is a radio-
loud AGN, suggesting the presence
of radio jets and lobes; one is a Type-
2 AGN, showing both narrow and
broadened optical spectral emis-
sion lines; and one is a Type-1 AGN,
showing narrow emission lines only
(though still wider than emission
lines in normal galaxies). Both the
10
GeminiFocus
January 2019 / 2018 Year in Review