In this work we explore the possibility of applying machine learning methods designed for one-dimensional problems to the task of galaxy image classification. The algorithms used for image classification typically rely on multiple costly steps, such
as the Point Spread Function (PSF) deconvolution and the training and application of complex Convolutional Neural Networks (CNN) of thousands or even millions of parameters. In our approach, we extract features from the galaxy images by analysing the elliptical isophotes in their light distribution and collect the information in a sequence. The sequences obtained with this method present definite features allowing a direct distinction between galaxy types, as opposed to smooth Sersic profiles. Then, we train and classify the sequences with machine learning algorithms, designed through the platform Modulos AutoML, and study how they optimize the classification task. As a demonstration of this method, we use the second public release of the Dark Energy Survey (DES DR2). We show that by applying it to this sample we are able to successfully distinguish between early-type and late-type galaxies, for images with signal-to-noise ratio greater then 300. This yields an accuracy of $86%$ for the early-type galaxies and $93%$ for the late-type galaxies, which is on par with most contemporary automated image classification approaches. Our novel method allows for galaxy images to be accurately classified and is faster than other approaches. Data dimensionality reduction also implies a significant lowering in computational cost. In the perspective of future data sets obtained with e.g. Euclid and the Vera Rubin Observatory (VRO), this work represents a path towards using a well-tested and widely used platform from industry in efficiently tackling galaxy classification problems at the peta-byte scale.
To fully understand cosmic black hole growth we need to constrain the population of heavily obscured active galactic nuclei (AGN) at the peak of cosmic black hole growth ($zsim$1-3). Sources with obscuring column densities higher than $mathrm{10^{24}
}$ atoms $mathrm{cm^{-2}}$, called Compton-thick (CT) AGN, can be identified by excess X-ray emission at $sim$20-30 keV, called the Compton hump. We apply the recently developed Spectral Curvature (SC) method to high-redshift AGN (2<z<5) detected with Chandra. This method parametrizes the characteristic Compton hump feature cosmologically redshifted into the X-ray band at observed energies <10 keV. We find good agreement in CT AGN found using the SC method and bright sources fit using their full spectrum with X-ray spectroscopy. In the Chandra deep field south, we measure a CT fraction of $mathrm{17^{+19}_{-11}%}$ (3/17) for sources with observed luminosity $mathrm{>5times 10^{43}}$ erg $mathrm{s^{-1}}$. In the Cosmological evolution survey (COSMOS), we find an observed CT fraction of $mathrm{15^{+4}_{-3}%}$ (40/272) or $mathrm{32pm11 %}$ when corrected for the survey sensitivity. When comparing to low redshift AGN with similar X-ray luminosities, our results imply the CT AGN fraction is consistent with having no redshift evolution. Finally, we provide SC equations that can be used to find high-redshift CT AGN (z>1) for current (XMM-Newton) and future (eROSITA and ATHENA) X-ray missions.
We report on the detection of a remarkable new fast high-energy transient found in the Chandra Deep Field-South, robustly associated with a faint ($m_{rm R}=27.5$ mag, $z_{rm ph}$$sim$2.2) host in the CANDELS survey. The X-ray event is comprised of 1
15$^{+12}_{-11}$ net 0.3-7.0 keV counts, with a light curve characterised by a $approx$100 s rise time, a peak 0.3-10 keV flux of $approx$5$times$10$^{-12}$ erg s$^{-1}$ cm$^{-2}$, and a power-law decay time slope of $-1.53pm0.27$. The average spectral slope is $Gamma=1.43^{+0.23}_{-0.13}$, with no clear spectral variations. The hbox{X-ray} and multi-wavelength properties effectively rule out the vast majority of previously observed high-energy transients. A few theoretical possibilities remain: an orphan X-ray afterglow from an off-axis short-duration Gamma-ray Burst (GRB) with weak optical emission; a low-luminosity GRB at high redshift with no prompt emission below $sim$20 keV rest-frame; or a highly beamed Tidal Disruption Event (TDE) involving an intermediate-mass black hole and a white dwarf with little variability. However, none of the above scenarios can completely explain all observed properties. Although large uncertainties exist, the implied rate of such events is comparable to those of orphan and low-luminosity GRBs as well as rare TDEs, implying the discovery of an untapped regime for a known transient class, or a new type of variable phenomena whose nature remains to be determined.
Recent work suggests blue ellipticals form in mergers and migrate quickly from the blue cloud of star-forming galaxies to the red sequence of passively evolving galaxies, perhaps as a result of black hole feedback. Such rapid reddening of stellar pop
ulations implies that large gas reservoirs in the pre-merger star-forming pair must be depleted on short time scales. Here we present pilot observations of atomic hydrogen gas in four blue early-type galaxies that reveal increasing spatial offsets between the gas reservoirs and the stellar components of the galaxies, with advancing post-starburst age. Emission line spectra show associated nuclear activity in two of the merged galaxies, and in one case radio lobes aligned with the displaced gas reservoir. These early results suggest that a kinetic process (possibly feedback from black hole activity) is driving the quick truncation of star formation in these systems, rather than a simple exhaustion of gas supply.
We constrain the total accreted mass density in supermassive black holes at z>6, inferred via the upper limit derived from the integrated X-ray emission from a sample of photometrically selected galaxy candidates. Studying galaxies obtained from the
deepest Hubble Space Telescope images combined with the Chandra 4 Msec observations of the Chandra Deep Field South, we achieve the most restrictive constraints on total black hole growth in the early Universe. We estimate an accreted mass density <1000Mo Mpc^-3 at z~6, significantly lower than the previous predictions from some existing models of early black hole growth and earlier prior observations. These results place interesting constraints on early black growth and mass assembly by accretion and imply one or more of the following: (1) only a fraction of the luminous galaxies at this epoch contain active black holes; (2) most black hole growth at early epochs happens in dusty and/or less massive - as yet undetected - host galaxies; (3) there is a significant fraction of low-z interlopers in the galaxy sample; (4) early black hole growth is radiatively inefficient, heavily obscured and/or is due to black hole mergers as opposed to accretion or (5) the bulk of the black hole growth occurs at late times. All of these possibilities have important implications for our understanding of high redshift seed formation models.
Using multiwavelength surveys of active galactic nuclei across a wide range of bolometric luminosities (10^{43}<L_{bol}(erg/s<5x10^{46}) and redshifts (0<z<3), we find a strong, redshift-independent correlation between the AGN luminosity and the frac
tion of host galaxies undergoing a major merger. That is, only the most luminous AGN phases are connected to major mergers, while less luminous AGN appear to be driven by secular processes. Combining this trend with AGN luminosity functions to assess the overall cosmic growth of black holes, we find that ~50% by mass is associated with major mergers, while only 10% of AGN by number, the most luminous, are connected to these violent events. Our results suggest that to reach the highest AGN luminosities -where the most massive black holes accreted the bulk of their mass - a major merger appears to be required. The luminosity dependence of the fraction of AGN triggered by major mergers can successfully explain why the observed scatter in the M-sigma relation for elliptical galaxies is significantly lower than in spirals. The lack of a significant redshift dependence of the L_{bol}-f_{merger} relation suggests that downsizing, i.e., the general decline in AGN and star formation activity with decreasing redshift, is driven by a decline in the frequency of major mergers combined with a decrease in the availability of gas at lower redshifts.
We present a study of local post-starburst galaxies (PSGs) using the photometric and spectroscopic observations from the Sloan Digital Sky Survey (SDSS) and the results from the Galaxy Zoo project. We find that the majority of our local PSG populatio
n have neither early- nor late- type morphologies but occupy a well-defined space within the colour-stellar mass diagram, most notably, the low-mass end of the green valley below the transition mass thought to be the mass division between low-mass star-forming galaxies and high-mass passively-evolving bulge-dominated galaxies. Our analysis suggests that it is likely that a local PSG will quickly transform into red, low-mass early-type galaxies as the stellar morphologies of the green PSGs largely resemble that of the early-type galaxies within the same mass range. We propose that the current population of PSGs represents a population of galaxies which is rapidly transitioning between the star-forming and the passively-evolving phases. Subsequently, these PSGs will contribute towards the build-up of the low-mass end of the red sequence once the current population of young stars fade and stars are no longer being formed. These results are consistent with the idea of downsizing where the build-up of smaller galaxies occurs at later epochs.
We use textit{GALEX} (Galaxy Evolution Explorer) near-UV (NUV) photometry of a sample of early-type galaxies selected in textit{SDSS} (Sloan Digital Sky Survey) to study the UV color-magnitude relation (CMR). $NUV-r$ color is an excellent tracer of e
ven small amounts ($sim 1$% mass fraction) of recent ($la 1$ Gyr) star formation and so the $NUV-r$ CMR allows us to study the effect of environment on the recent star formation history. We analyze a volume-limited sample of 839 visually-inspected early-type galaxies in the redshift range $0.05 < z < 0.10$ brighter than $M_{r}$ of -21.5 with any possible emission-line or radio-selected AGN removed to avoid contamination. We find that contamination by AGN candidates and late-type interlopers highly bias any study of recent star formation in early-type galaxies and that, after removing those, our lower limit to the fraction of massive early-type galaxies showing signs of recent star formation is roughly $30 pm 3%$ This suggests that residual star formation is common even amongst the present day early-type galaxy population. We find that the fraction of UV-bright early-type galaxies is 25% higher in low-density environments. However, the density effect is clear only in the lowest density bin. The blue galaxy fraction for the subsample of the brightest early-type galaxies however shows a very strong density dependence, in the sense that the blue galaxy fraction is lower in a higher density region.
We have studied ~2100 early-type galaxies in the SDSS DR3 which have been detected by the GALEX Medium Imaging Survey (MIS), in the redshift range 0 < z < 0.11. Combining GALEX UV photometry with corollary optical data from the SDSS, we find that, at
a 95 percent confidence level, at least ~30 percent of galaxies in this sample have UV to optical colours consistent with some recent star formation within the last Gyr. In particular, galaxies with a NUV - r colour less than 5.5 are very likely to have experienced such recent star formation, taking into account the possibility of a contribution to NUV flux from the UV upturn phenomenon. We find quantitative agreement between the observations and the predictions of a semi-analytical LCDM hierarchical merger model and deduce that early-type galaxies in the redshift range 0 < z < 0.11 have ~1 to 3 percent of their stellar mass in stars less than 1 Gyr old. The average age of this recently formed population is ~300 to 500 Myrs. We also find that monolithically evolving galaxies, where recent star formation can be driven solely by recycled gas from stellar mass loss, cannot exhibit the blue colours (NUV - r < 5.5) seen in a significant fraction (~30 percent) of our observed sample.
