Search

Interactions between galaxies leave distinguishable imprints in the form of tidal features which hold important clues about their mass assembly. Unfortunately, these structures are difficult to detect because they are low surface brightness features so deep observations are needed. Upcoming surveys promise several orders of magnitude increase in depth and sky coverage, for which automated methods for tidal feature detection will become mandatory. We test the ability of a convolutional neural network to reproduce human visual classifications for tidal detections. We use as training $\sim$6000 simulated images classified by professional astronomers. The mock Hyper Suprime Cam Subaru (HSC) images include variations with redshift, projection angle and surface brightness ($\mu_{lim}$ =26-35 mag arcsec$^{-2}$). We obtain satisfactory results with accuracy, precision and recall values of Acc=0.84, P=0.72 and R=0.85, respectively, for the test sample. While the accuracy and precision values are roughly constant for all surface brightness, the recall (completeness) is significantly affected by image depth. The recovery rate shows strong dependence on the type of tidal features: we recover all the images showing shell features and 87% of the tidal streams; these fractions are below 75% for mergers, tidal tails and bridges. When applied to real HSC images, the performance of the model worsens significantly. We speculate that this is due to the lack of realism of the simulations and take it as a warning on applying deep learning models to different data domains without prior testing on the actual data., Comment: 13 pages, 10 figures, accepted for publication in MNRAS

We study systematics associated with estimating simple stellar population (SSP) parameters -- age, metallicity [M/H], $\alpha$-enhancement [$\alpha$/Fe] and IMF shape -- and associated $M_*/L$ gradients, of elliptical slow rotators (E-SRs), fast rotators (E-FRs) and S0s from stacked spectra of galaxies in the MaNGA survey. These systematics arise from (i) how one normalizes the spectra when stacking; (ii) having to subtract emission before estimating absorption line strengths; (iii) the decision to fit the whole spectrum or just a few absorption lines; (iv) SSP model differences (e.g. isochrones, enrichment, IMF). The MILES+Padova SSP models, fit to the H$_\beta$, $\langle$Fe$\rangle$, TiO$_{\rm 2SDSS}$ and [MgFe] Lick indices in the stacks, indicate that out to the half-light radius $R_e$: (a) ages are younger and [$\alpha$/Fe] values are lower in the central regions but the opposite is true of [M/H]; (b) the IMF is more bottom-heavy in the center, but is close to Kroupa beyond about $R_e/2$; (c) this makes $M_*/L$ about $2\times$ larger in the central regions than beyond $R_e/2$. While the models of Conroy et al. (2018) return similar [M/H] and [$\alpha$/Fe] profiles, the age and (hence) $M_*/L$ profiles can differ significantly even for solar abundances and a Kroupa IMF; different responses to non-solar abundances and IMF parametrization further compound these differences. There are clear (model independent) differences between E-SRs, E-FRs and S0s: younger ages and less enhanced [$\alpha$/Fe] values suggest that E-FRs and S0s are not SSPs, but relaxing this assumption is unlikely to change their inferred $M_*/L$ gradients significantly., Comment: 22 pages, 23 figures, accepted for publication in MNRAS

We compare the two largest galaxy morphology catalogues, which separate early and late type galaxies at intermediate redshift. The two catalogues were built by applying supervised deep learning (convolutional neural networks, CNNs) to the Dark Energy Survey data down to a magnitude limit of $\sim$21 mag. The methodologies used for the construction of the catalogues include differences such as the cutout sizes, the labels used for training, and the input to the CNN - monochromatic images versus $gri$-band normalized images. In addition, one catalogue is trained using bright galaxies observed with DES ($i<18$), while the other is trained with bright galaxies ($r<17.5$) and `emulated' galaxies up to $r$-band magnitude $22.5$. Despite the different approaches, the agreement between the two catalogues is excellent up to $i<19$, demonstrating that CNN predictions are reliable for samples at least one magnitude fainter than the training sample limit. It also shows that morphological classifications based on monochromatic images are comparable to those based on $gri$-band images, at least in the bright regime. At fainter magnitudes, $i>19$, the overall agreement is good ($\sim$95\%), but is mostly driven by the large spiral fraction in the two catalogues. In contrast, the agreement within the elliptical population is not as good, especially at faint magnitudes. By studying the mismatched cases we are able to identify lenticular galaxies (at least up to $i<19$), which are difficult to distinguish using standard classification approaches. The synergy of both catalogues provides an unique opportunity to select a population of unusual galaxies., Comment: 17 pages, 14 figures (1 appendix for galaxy examples including 3 figures)

The magnetization of Zn1-xCoxO (0.0055 < x < 0.073) nanoparticles has been measured as a function of temperature T (1.7 K < T , 10 K) and for magnetic field up to 65 kOe using a SQUID magnetometer. Samples were synthesized by three different growth methods: microwave-assisted hydrothermal, combustion reaction and sol-gel. For all studied samples, the magnetic properties derive from the antiferromagnetic (AF) spin clustering due to the Co2+ nearest neighbors. At T >= 6 K, the magnetization of the Co2+ ions has a Brillouin-type behavior, but below 6 K, it shows a notable deviation. We have shown that the observed deviation may be derived from single-ion anisotropy (SIA) with uniaxial symmetry. Results of fits show that the axial-SIA parameter D (typically D = 4.4 K) is slightly larger that the bulk value D = 3.97 K. No significant change of D has been observed as a function of the Co concentration or the growth process. For each sample, the SIA fit gave also the effective concentration (x) corresponding to the technical saturation value of the magnetization. Comparison of the concentration dependence of x with predictions based on cluster models shows an enhancement of the AF spin clustering independent of the growth method. This is ascribed to a clamped non-random distribution of the cobalt ions in the nanoparticles. The approach of the local concentration (xL) has been used to quantify the observed deviation from randomicity. Assuming a ZnO core/ Zn1-xCoxO shell nanoparticle, the thickness of the shell has been determined from the ratio xL/x.

Gradients in the stellar populations (SP) of galaxies -- e.g., in age, metallicity, stellar Initial Mass Function (IMF) -- can result in gradients in the stellar mass to light ratio, $M_*/L$. Such gradients imply that the distribution of the stellar mass and light are different. For old SPs, e.g., in early-type galaxies at $z\sim 0$, the $M_*/L$ gradients are weak if driven by variations in age and metallicity, but significantly larger if driven by the IMF. A gradient which has larger $M_*/L$ in the center increases the estimated total stellar mass ($M_*$) and reduces the scale which contains half this mass ($R_{e,*}$), compared to when the gradient is ignored. For the IMF gradients inferred from fitting MILES simple SP models to the H$_\beta$, $\langle$Fe$\rangle$, [MgFe] and TiO$_{\rm 2SDSS}$ absorption lines measured in spatially resolved spectra of early-type galaxies in the MaNGA survey, the fractional change in $R_{e,*}$ can be significantly larger than that in $M_*$, especially when the light is more centrally concentrated. The $R_{e,*}-M_*$ correlation which results is offset by 0.3 dex to smaller sizes compared to when these gradients are ignored. Comparisons with `quiescent' galaxies at higher-$z$ must account for evolution in SP gradients (especially age and IMF) and the light profile before drawing conclusions about how $R_{e,*}$ and $M_*$ evolve. The implied merging between higher-$z$ and the present is less contrived if $R_{e,*}/R_e$ at $z\sim 0$ is closer to our IMF-driven gradient calibration than to unity., Comment: 16 pages, 15 figures, accepted for publication in MNRAS

We present the MaNGA PyMorph photometric Value Added Catalogue (MPP-VAC-DR17) and the MaNGA Deep Learning Morphological VAC (MDLM-VAC-DR17) for the final data release of the MaNGA survey, which is part of the SDSS Data Release 17 (DR17). The MPP-VAC-DR17 provides photometric parameters from S\`ersic and S\`ersic+Exponential fits to the 2D surface brightness profiles of the MaNGA DR17 galaxy sample in the $g$, $r$, and $i$ bands (e.g. total fluxes, half light radii, bulge-disk fractions, ellipticities, position angles, etc.). The MDLM-VAC-DR17 provides Deep Learning-based morphological classifications for the same galaxies. The MDLM-VAC-DR17 includes a number of morphological properties: e.g., a T-Type, a finer separation between elliptical and S0, as well as the identification of edge-on and barred galaxies. While the MPP-VAC-DR17 simply extends the MaNGA PyMorph photometric VAC published in the SDSS Data Release 15 (MPP-VAC-DR15) to now include galaxies which were added to make the final DR17, the MDLM-VAC-DR17 implements some changes and improvements compared to the previous release (MDLM-VAC-DR15): namely, the low-end of the T-Types are better recovered in this new version. The catalogue also includes a separation between Early- or Late-type Galaxies (ETG, LTG), which classifies the two populations in a complementary way to the T-Type, especially at the intermediate types (-1 < T-Type < 2), where the T-Type values show a large scatter. In addition, $k-$fold based uncertainties on the classifications are also provided. To ensure robustness and reliability, we have also visually inspected all the images. We describe the content of the catalogues and show some interesting ways in which they can be combined., Comment: Accepted for publication in MNRAS. Catalogues will be released with SDSS Data Release 17

We present morphological classifications of $\sim$27 million galaxies from the Dark Energy Survey (DES) Data Release 1 (DR1) using a supervised deep learning algorithm. The classification scheme separates: (a) early-type galaxies (ETGs) from late-types (LTGs); and (b) face-on galaxies from edge-on. Our Convolutional Neural Networks (CNNs) are trained on a small subset of DES objects with previously known classifications. These typically have $\mathrm{m}_r \lesssim 17.7~\mathrm{mag}$; we model fainter objects to $\mathrm{m}_r < 21.5$ mag by simulating what the brighter objects with well determined classifications would look like if they were at higher redshifts. The CNNs reach 97\% accuracy to $\mathrm{m}_r<21.5$ on their training sets, suggesting that they are able to recover features more accurately than the human eye. We then used the trained CNNs to classify the vast majority of the other DES images. The final catalog comprises five independent CNN predictions for each classification scheme, helping to determine if the CNN predictions are robust or not. We obtain secure classifications for $\sim$ 87\% and 73\% of the catalog for the ETG vs. LTG and edge-on vs. face-on models, respectively. Combining the two classifications (a) and (b) helps to increase the purity of the ETG sample and to identify edge-on lenticular galaxies (as ETGs with high ellipticity). Where a comparison is possible, our classifications correlate very well with S\'ersic index (\textit{n}), ellipticity ($\epsilon$) and spectral type, even for the fainter galaxies. This is the largest multi-band catalog of automated galaxy morphologies to date., Comment: Accepted for publication in MNRAS (2021 February 22); 17 pages, 16 figures

This is the third paper of a series where we study the stellar population gradients (SP; ages, metallicities, $\alpha$-element abundance ratios and stellar initial mass functions) of early type galaxies (ETGs) at $z\le 0.08$ from the MaNGA-DR15 survey. In this work we focus on the S0 population and quantify how the SP varies across the population as well as with galactocentric distance. We do this by measuring Lick indices and comparing them to stellar population synthesis models. This requires spectra with high signal-to-noise which we achieve by stacking in bins of luminosity (L$_r$) and central velocity dispersion ($\sigma_0$). We find that: 1) There is a bimodality in the S0 population: S0s more massive than $3\times 10^{10}M_\odot$ show stronger velocity dispersion and age gradients (age and $\sigma_r$ decrease outwards) but little or no metallicity gradient, while the less massive ones present relatively flat age and velocity dispersion profiles, but a significant metallicity gradient (i.e. [M/H] decreases outwards). Above $2\times10^{11}M_\odot$ the number of S0s drops sharply. These two mass scales are also where global scaling relations of ETGs change slope. 2) S0s have steeper velocity dispersion profiles than fast rotating elliptical galaxies (E-FRs) of the same luminosity and velocity dispersion. The kinematic profiles and stellar population gradients of E-FRs are both more similar to those of slow rotating ellipticals (E-SRs) than to S0s, suggesting that E-FRs are not simply S0s viewed face-on. 3) At fixed $\sigma_0$, more luminous S0s and E-FRs are younger, more metal rich and less $\alpha$-enhanced. Evidently for these galaxies, the usual statement that 'massive galaxies are older' is not true if $\sigma_0$ is held fixed., Comment: Accepted for publication in MNRAS. 15 pages, 20 figures

Early-type galaxies -- slow and fast rotating ellipticals (E-SRs and E-FRs) and S0s/lenticulars -- define a Fundamental Plane (FP) in the space of half-light radius $R_e$, enclosed surface brightness $I_e$ and velocity dispersion $\sigma_e$. Since $I_e$ and $\sigma_e$ are distance-independent measurements, the thickness of the FP is often expressed in terms of the accuracy with which $I_e$ and $\sigma_e$ can be used to estimate sizes $R_e$. We show that: 1) The thickness of the FP depends strongly on morphology. If the sample only includes E-SRs, then the observed scatter in $R_e$ is $\sim 16\%$, of which only $\sim 9\%$ is intrinsic. Removing galaxies with $M_*<10^{11}M_\odot$ further reduces the observed scatter to $\sim 13\%$ ($\sim 4\%$ intrinsic). The observed scatter increases to the $\sim 25\%$ usually quoted in the literature if E-FRs and S0s are added. If the FP is defined using the eigenvectors of the covariance matrix of the observables, then the E-SRs again define an exceptionally thin FP, with intrinsic scatter of only $5\%$ orthogonal to the plane. 2) The structure within the FP is most easily understood as arising from the fact that $I_e$ and $\sigma_e$ are nearly independent, whereas the $R_e-I_e$ and $R_e-\sigma_e$ correlations are nearly equal and opposite. 3) If the coefficients of the FP differ from those associated with the virial theorem the plane is said to be `tilted'. If we multiply $I_e$ by the global stellar mass-to-light ratio $M_*/L$ and we account for non-homology across the population by using S\'ersic photometry, then the resulting stellar mass FP is less tilted. Accounting self-consistently for $M_*/L$ gradients will change the tilt. The tilt we currently see suggests that the efficiency of turning baryons into stars increases and/or the dark matter fraction decreases as stellar surface brightness increases., Comment: 13 pages, 9 figures, 3 tables, accepted for publication in MNRAS

We estimate ages, metallicities, $\alpha$-element abundance ratios and stellar initial mass functions of elliptical (E) and S0 galaxies from the MaNGA-DR15 survey. We stack spectra and use a variety of single stellar population synthesis models to interpret the absorption line strengths in these spectra. We quantify how these properties vary across the population, as well as with galactocentric distance. This paper is the first of a series and is based on a sample of pure elliptical galaxies at z $\le$ 0.08. We show that the properties of the inner regions of Es with the largest luminosity (L$_r$) and central velocity dispersion ($\sigma_0$) are consistent with those associated with the commonly used Salpeter IMF, whereas a Kroupa-like IMF is a better description at $\sim$ 0.8R/Re (assuming [Ti/Fe] variations are limited). For these galaxies the stellar mass-to-light ratio decreases at most by a factor of 2 from the central regions to Re. In contrast, for lower L$_r$ and $\sigma_0$ galaxies, the IMF is shallower and M$_{*}$/L$_r$ in the central regions is similar to the outskirts. Although a factor of 2 is smaller than previous reports based on a handful of galaxies, it is still large enough to matter for dynamical mass estimates. Accounting self-consistently for these gradients when estimating both M$_{*}$ and M$_{dyn}$ brings the two into good agreement: gradients reduce M$_{dyn}$ by $\sim$ 0.2 dex while only slightly increasing the M$_{*}$ inferred using a Kroupa IMF. This is a different resolution of the M$_{*}$-M$_{dyn}$ discrepancy than has been followed in the recent literature where M$_{*}$ of massive galaxies is increased by adopting a Salpeter IMF while leaving Mdyn unchanged. A companion paper discusses how stellar population differences are even more pronounced if one separates slow from fast rotators., Comment: 22 pages, 29 figures, accepted for publication in MNRAS

We present estimates of stellar population (SP) gradients from stacked spectra of slow (SR) and fast (FR) rotator elliptical galaxies from the MaNGA-DR15 survey. We find that: 1) FRs are $\sim 5$ Gyrs younger, more metal rich, less $\alpha$-enhanced and smaller than SRs of the same luminosity $L_r$ and central velocity dispersion $\sigma_0$. This explains why when one combines SRs and FRs, objects which are small for their $L_r$ and $\sigma_0$ tend to be younger. Their SP gradients are also different. 2) Ignoring the FR/SR dichotomy leads one to conclude that compact galaxies are older than their larger counterparts of the same mass, even though almost the opposite is true for FRs and SRs individually. 3) SRs with $\sigma_0\le 250$ km s$^{-1}$ are remarkably homogeneous within $\sim R_e$: they are old, $\alpha$-enhanced and only slightly super-solar in metallicity. These SRs show no gradients in age and $M_*/L_r$, negative gradients in metallicity, and slightly positive gradients in [$\alpha$/Fe] (the latter are model dependent). SRs with $\sigma_0\ge 250$ km $s^{-1}$ are slightly younger and more metal rich, contradicting previous work suggesting that age increases with $\sigma_0$. They also show larger $M_*/L_r$ gradients. 4) Self-consistently accounting for $M_*/L$ gradients yields $M_{\rm dyn}\approx M_*$ because gradients reduce $M_{\rm dyn}$ by $\sim 0.2$ dex while only slightly increasing the $M_*$ inferred using a Kroupa (not Salpeter) IMF. 5) The FR population all but disappears above $M_*\ge 3\times 10^{11}M_\odot$; this is the same scale at which the size-mass correlation and other scaling relations change. Our results support the finding that this is an important mass scale which correlates with the environment and above which mergers matter., Comment: 22 pages, 28 figures. Accepted for publication in MNRAS

We analyze the optical morphologies of galaxies in the IllustrisTNG simulation at $z\sim0$ with a Convolutional Neural Network trained on visual morphologies in the Sloan Digital Sky Survey. We generate mock SDSS images of a mass complete sample of $\sim12,000$ galaxies in the simulation using the radiative transfer code SKIRT and include PSF and noise to match the SDSS r-band properties. The images are then processed through the exact same neural network used to estimate SDSS morphologies to classify simulated galaxies in four morphological classes (E, S0/a, Sab, Scd). The CNN model finds that $\sim95\%$ of the simulated galaxies fall in one the four main classes with high confidence. The mass-size relations of the simulated galaxies divided by morphological type also reproduce well the slope and the normalization of observed relations which confirms the realism of optical morphologies in the TNG suite. However, the Stellar Mass Functions decomposed into different morphologies still show significant discrepancies with observations both at the low and high mass end. We find that the high mass end of the SMF is dominated in TNG by massive disk galaxies while early-type galaxies dominate in the observations according to the CNN classifications. The present work highlights the importance of detailed comparisons between observations and simulations in comparable conditions., Comment: submitted to MNRAS, comments welcome

We present STEEL a STatistical sEmi-Empirical modeL designed to probe the distribution of satellite galaxies in groups and clusters. Our fast statistical methodology relies on tracing the abundances of central and satellite haloes via their mass functions at all cosmic epochs with virtually no limitation on cosmic volume and mass resolution. From mean halo accretion histories and subhalo mass functions the satellite mass function is progressively built in time via abundance matching techniques constrained by number densities of centrals in the local Universe. By enforcing dynamical merging timescales as predicted by high-resolution N-body simulations, we obtain satellite distributions as a function of stellar mass and halo mass consistent with current data. We show that stellar stripping, star formation, and quenching play all a secondary role in setting the number densities of massive satellites above $M_*\gtrsim 3\times 10^{10}\, M_{\odot}$. We further show that observed star formation rates used in our empirical model over predict low-mass satellites below $M_*\lesssim 3\times 10^{10}\, M_{\odot}$, whereas, star formation rates derived from a continuity equation approach yield the correct abundances similar to previous results for centrals., Comment: 21 pages, 17 Figures. MNRAS, in press

We describe the SDSS-IV MaNGA PyMorph Photometric (MPP-VAC) and MaNGA Deep Learning Morphology (MDLM-VAC) Value Added Catalogs. The MPP-VAC provides photometric parameters from S\'ersic and S\'ersic+Exponential fits to the 2D surface brightness profiles of the MaNGA DR15 galaxy sample. Compared to previous PyMorph analyses of SDSS imaging, our analysis of the MaNGA DR15 incorporates three improvements: the most recent SDSS images; modified criteria for determining bulge-to-disk decompositions; and the fits in MPP-VAC have been eye-balled, and re-fit if necessary, for additional reliability. A companion catalog, the MDLM-VAC, provides Deep Learning-based morphological classifications for the same galaxies. The MDLM-VAC includes a number of morphological properties (e.g., a TType, and a finer separation between elliptical and S0 galaxies). Combining the MPP- and MDLM-VACs allows to show that the MDLM morphological classifications are more reliable than previous work. It also shows that single-S\'ersic fits to late- and early-type galaxies are likely to return S\'ersic indices of $n \le 2$ and $\ge 4$, respectively, and this correlation between $n$ and morphology extends to the bulge component as well. While the former is well-known, the latter contradicts some recent work suggesting little correlation between $n$-bulge and morphology. Combining both VACs with MaNGA's spatially resolved spectroscopy allows us to study how the stellar angular momentum depends on morphological type. We find correlations between stellar kinematics, photometric properties, and morphological type even though the spectroscopic data played no role in the construction of the MPP- and MDLM-VACs., Comment: 22 pages, 30 figures and 4 tables. Accepted for publication in MNRAS. The MPP-VAC and MDLM-VAC are available online at https://www.sdss.org/dr15/data_access/value-added-catalogs/manga-pymorph-dr15-photometric-catalog and https://www.sdss.org/dr15/data_access/value-added-catalogs/manga-morphology-deep-learning-dr15-catalog

Deep Learning (DL) algorithms for morphological classification of galaxies have proven very successful, mimicking (or even improving) visual classifications. However, these algorithms rely on large training samples of labelled galaxies (typically thousands of them). A key question for using DL classifications in future Big Data surveys is how much of the knowledge acquired from an existing survey can be exported to a new dataset, i.e. if the features learned by the machines are meaningful for different data. We test the performance of DL models, trained with Sloan Digital Sky Survey (SDSS) data, on Dark Energy survey (DES) using images for a sample of $\sim$5000 galaxies with a similar redshift distribution to SDSS. Applying the models directly to DES data provides a reasonable global accuracy ($\sim$ 90%), but small completeness and purity values. A fast domain adaptation step, consisting in a further training with a small DES sample of galaxies ($\sim$500-300), is enough for obtaining an accuracy > 95% and a significant improvement in the completeness and purity values. This demonstrates that, once trained with a particular dataset, machines can quickly adapt to new instrument characteristics (e.g., PSF, seeing, depth), reducing by almost one order of magnitude the necessary training sample for morphological classification. Redshift evolution effects or significant depth differences are not taken into account in this study., Comment: Accepted for publication in MNRAS

We use machine learning to identify in color images of high-redshift galaxies an astrophysical phenomenon predicted by cosmological simulations. This phenomenon, called the blue nugget (BN) phase, is the compact star-forming phase in the central regions of many growing galaxies that follows an earlier phase of gas compaction and is followed by a central quenching phase. We train a Convolutional Neural Network (CNN) with mock "observed" images of simulated galaxies at three phases of evolution: pre-BN, BN and post-BN, and demonstrate that the CNN successfully retrieves the three phases in other simulated galaxies. We show that BNs are identified by the CNN within a time window of $\sim0.15$ Hubble times. When the trained CNN is applied to observed galaxies from the CANDELS survey at $z=1-3$, it successfully identifies galaxies at the three phases. We find that the observed BNs are preferentially found in galaxies at a characteristic stellar mass range, $10^{9.2-10.3} M_\odot$ at all redshifts. This is consistent with the characteristic galaxy mass for BNs as detected in the simulations, and is meaningful because it is revealed in the observations when the direct information concerning the total galaxy luminosity has been eliminated from the training set. This technique can be applied to the classification of other astrophysical phenomena for improved comparison of theory and observations in the era of large imaging surveys and cosmological simulations., Comment: Accepted for publication in ApJ

Within a galaxy the stellar mass-to-light ratio $\Upsilon_*$ is not constant. Spatially resolved kinematics of nearby early-type galaxies suggest that allowing for a variable initial mass function (IMF) returns significantly larger $\Upsilon_*$ gradients than if the IMF is held fixed. If $\Upsilon_*$ is greater in the central regions, then ignoring the IMF-driven gradient can overestimate $M_*^{\rm dyn}$ by as much as a factor of two for the most massive galaxies, though stellar population estimates $M_*^{\rm SP}$ are also affected. Large $\Upsilon_*$-gradients have four main consequences: First, $M_*^{\rm dyn}$ cannot be estimated independently of stellar population synthesis models. Second, if there is a lower limit to $\Upsilon_*$ and gradients are unknown, then requiring $M_*^{\rm dyn}=M_*^{\rm SP}$ constrains them. Third, if gradients are stronger in more massive galaxies, then $M_*^{\rm dyn}$ and $M_*^{\rm SP}$ can be brought into agreement, not by shifting $M_*^{\rm SP}$ upwards by invoking constant bottom-heavy IMFs, as advocated by a number of recent studies, but by revising $M_*^{\rm dyn}$ estimates in the literature downwards. Fourth, accounting for $\Upsilon_*$ gradients changes the high-mass slope of the stellar mass function $\phi(M_*^{\rm dyn})$, and reduces the associated stellar mass density. These conclusions potentially impact estimates of the need for feedback and adiabatic contraction, so our results highlight the importance of measuring $\Upsilon_*$ gradients in larger samples., Comment: 13 pages, 7 figures, MNRAS in press

We present a morphological catalogue for $\sim$ 670,000 galaxies in the Sloan Digital Sky Survey in two flavours: T-Type, related to the Hubble sequence, and Galaxy Zoo 2 (GZ2 hereafter) classification scheme. By combining accurate existing visual classification catalogues with machine learning, we provide the largest and most accurate morphological catalogue up to date. The classifications are obtained with Deep Learning algorithms using Convolutional Neural Networks (CNNs). We use two visual classification catalogues, GZ2 and Nair & Abraham (2010), for training CNNs with colour images in order to obtain T-Types and a series of GZ2 type questions (disk/features, edge-on galaxies, bar signature, bulge prominence, roundness and mergers). We also provide an additional probability enabling a separation between pure elliptical (E) from S0, where the T-Type model is not so efficient. For the T-Type, our results show smaller offset and scatter than previous models trained with support vector machines. For the GZ2 type questions, our models have large accuracy (> 97\%), precision and recall values (> 90\%) when applied to a test sample with the same characteristics as the one used for training. The catalogue is publicly released with the paper., Comment: 18 pages, 21 figures; Accepted for publication in MNRAS

Spatially resolved kinematics of nearby galaxies has shown that the ratio of dynamical- to stellar population-based estimates of the mass of a galaxy ($M_*^{\rm JAM}/M_*$) correlates with $\sigma_e$, if $M_*$ is estimated using the same IMF for all galaxies and the stellar M/L ratio within each galaxy is constant. This correlation may indicate that, in fact, the IMF is more dwarf-rich for galaxies with large $\sigma$. We use this correlation to estimate a dynamical or IMF-corrected stellar mass, $M_*^{\rm \alpha_{JAM}}$, from $M_{*}$ and $\sigma_e$ for a sample of $6 \times 10^5$ SDSS galaxies for which spatially resolved kinematics is not available. We also compute the `virial' mass estimate $k(n,R)\,R_e\,\sigma_R^2/G$, where $n$ is the Sersic index, in the SDSS and ATLAS$^{\rm 3D}$ samples. We show that an $n$-dependent correction must be applied to the $k(n,R)$ values provided by Prugniel & Simien (1997). Our analysis also shows that the shape of the velocity dispersion profile in the ATLAS$^{\rm 3D}$ sample varies weakly with $n$: $(\sigma_R/\sigma_e) = (R/R_e)^{-\gamma(n)}$. The resulting stellar mass functions, based on $M_*^{\rm \alpha_{JAM}}$ and the recalibrated virial mass, are in good agreement. If the $M_*^{\rm \alpha_{JAM}}/M_* - \sigma_e$ correlation is indeed due to the IMF, and stellar M/L gradients can be ignored, then our $\phi(M_*^{\rm \alpha_{JAM}})$ is an estimate of the stellar mass function in which $\sigma_e$-dependent variations in the IMF across the population have been accounted for. Using a Fundamental Plane based observational proxy for $\sigma_e$ produces comparable results. By demonstrating that cheaper proxies are sufficiently accurate, our analysis should enable a more reliable census of the mass in stars for large galaxy samples, at a fraction of the cost. Our results are provided in tabular form., Comment: 17 pages, 19 figures, 4 tables. Accepted for publication by MNRAS. Tables 1, C1 and C2 are provided as ancillary files

A number of recent estimates of the total luminosities of galaxies in the SDSS are significantly larger than those reported by the SDSS pipeline. This is because of a combination of three effects: one is simply a matter of defining the scale out to which one integrates the fit when defining the total luminosity, and amounts on average to < 0.1 mags even for the most luminous galaxies. The other two are less trivial and tend to be larger; they are due to differences in how the background sky is estimated and what model is fit to the surface brightness profile. We show that PyMorph sky estimates are fainter than those of the SDSS DR7 or DR9 pipelines, but are in excellent agreement with the estimates of Blanton et al. (2011). Using the SDSS sky biases luminosities by more than a few tenths of a magnitude for objects with half-light radii > 7 arcseconds. In the SDSS main galaxy sample these are typically luminous galaxies, so they are not necessarily nearby. This bias becomes worse when allowing the model more freedom to fit the surface brightness profile. When PyMorph sky values are used, then two component Sersic-Exponential fits to E+S0s return more light than single component deVaucouleurs fits (up to ~0.2 mag), but less light than single Sersic fits (0.1 mag). Finally, we show that PyMorph fits of Meert et al. (2015) to DR7 data remain valid for DR9 images. Our findings show that, especially at large luminosities, these PyMorph estimates should be preferred to the SDSS pipeline values., Comment: 12 pages, 17 figures, accepted for publication in MNRAS

The Sloan Digital Sky Survey pipeline photometry underestimates the brightnesses of the most luminous galaxies. This is mainly because (i) the SDSS overestimates the sky background and (ii) single or two-component Sersic-based models better fit the surface brightness profile of galaxies, especially at high luminosities, than does the de Vaucouleurs model used by the SDSS pipeline. We use the PyMorph photometric reductions to isolate effect (ii) and show that it is the same in the full sample as in small group environments, and for satellites in the most massive clusters as well. None of these are expected to be significantly affected by intracluster light (ICL). We only see an additional effect for centrals in the most massive halos, but we argue that even this is not dominated by ICL. Hence, for the vast majority of galaxies, the differences between PyMorph and SDSS pipeline photometry cannot be ascribed to the semantics of whether or not one includes the ICL when describing the stellar mass of massive galaxies. Rather, they likely reflect differences in star formation or assembly histories. Failure to account for the SDSS underestimate has significantly biased most previous estimates of the SDSS luminosity and stellar mass functions, and therefore Halo Model estimates of the z ~ 0.1 relation between the mass of a halo and that of the galaxy at its center. We also show that when one studies correlations, at fixed group mass, with a quantity which was not used to define the groups, then selection effects appear. We show why such effects arise, and should not be mistaken for physical effects., Comment: 15 pages, 17 figures, accepted for publication in MNRAS. The PyMorph luminosities and stellar masses are available at https://www.physics.upenn.edu/~ameert/SDSS_PhotDec/

[abridged] We quantify the evolution of the stellar mass functions of star-forming and quiescent galaxies as a function of morphology from $z\sim 3$ to the present. Our sample consists of ~50,000 galaxies in the CANDELS fields ($\sim880$ $arcmin^2$), which we divide into four main morphological types, i.e. pure bulge dominated systems, pure spiral disk dominated, intermediate 2-component bulge+disk systems and irregular disturbed galaxies. Our main results are: Star-formation: At $z\sim 2$, 80\% of the stellar mass density of star-forming galaxies is in irregular systems. However, by $z\sim 0.5$, irregular objects only dominate at stellar masses below $10^9M\odot$. A majority of the star-forming irregulars present at $z\sim 2$ undergo a gradual transformation from disturbed to normal spiral disk morphologies by $z\sim 1$ without significant interruption to their star-formation. Rejuvenation after a quenching event does not seem to be common except perhaps for the most massive objects. Quenching: We confirm that galaxies reaching a stellar mass of $M_*\sim10^{10.8}M_\odot$ ($M^*$) tend to quench. Also, quenching implies the presence of a bulge: the abundance of massive red disks is negligible at all redshifts over 2~dex in stellar mass. However the dominant quenching mechanism evolves. At $z>2$, the SMF of quiescent galaxies above $M^*$ is dominated by compact spheroids. Quenching at this early epoch destroys the disk and produces a compact remnant unless the star-forming progenitors at even higher redshifts are significantly more dense. At $1

We quantify the systematic effects on the stellar mass function which arise from assumptions about the stellar population, as well as how one fits the light profiles of the most luminous galaxies at z ~ 0.1. When comparing results from the literature, we are careful to separate out these effects. Our analysis shows that while systematics in the estimated comoving number density which arise from different treatments of the stellar population remain of order < 0.5 dex, systematics in photometry are now about 0.1 dex, despite recent claims in the literature. Compared to these more recent analyses, previous work based on Sloan Digital Sky Survey (SDSS) pipeline photometry leads to underestimates of rho_*(> M_*) by factors of 3-10 in the mass range 10^11 - 10^11.6 M_Sun, but up to a factor of 100 at higher stellar masses. This impacts studies which match massive galaxies to dark matter halos. Although systematics which arise from different treatments of the stellar population remain of order < 0.5 dex, our finding that systematics in photometry now amount to only about 0.1 dex in the stellar mass density is a significant improvement with respect to a decade ago. Our results highlight the importance of using the same stellar population and photometric models whenever low and high redshift samples are compared., Comment: 18 pages, 17 figures, accepted for publication in MNRAS. The PyMorph luminosities and stellar masses are available at https://www.physics.upenn.edu/~ameert/SDSS_PhotDec/

Supermassive black hole -- host galaxy relations are key to the computation of the expected gravitational wave background (GWB) in the pulsar timing array (PTA) frequency band. It has been recently pointed out that standard relations adopted in GWB computations are in fact biased-high. We show that when this selection bias is taken into account, the expected GWB in the PTA band is a factor of about three smaller than previously estimated. Compared to other scaling relations recently published in the literature, the median amplitude of the signal at $f=1$yr$^{-1}$ drops from $1.3\times10^{-15}$ to $4\times10^{-16}$. Although this solves any potential tension between theoretical predictions and recent PTA limits without invoking other dynamical effects (such as stalling, eccentricity or strong coupling with the galactic environment), it also makes the GWB detection more challenging., Comment: 6 pages 4 figures, submitted to MNRAS letters

We describe the luminosity function, based on Sersic fits to the light profiles, of CMASS galaxies at z ~ 0.55. Compared to previous estimates, our Sersic-based reductions imply more luminous, massive galaxies, consistent with the effects of Sersic- rather than Petrosian or de Vaucouleur-based photometry on the Sloan Digital Sky Survey (SDSS) main galaxy sample at z ~ 0.1. This implies a significant revision of the high mass end of the correlation between stellar and halo mass. Inferences about the evolution of the luminosity and stellar mass functions depend strongly on the assumed, and uncertain, k+e corrections. In turn, these depend on the assumed age of the population. Applying k+e corrections taken from fitting the models of Maraston et al. (2009) to the colors of both SDSS and CMASS galaxies, the evolution of the luminosity and stellar mass functions appears impressively passive, provided that the fits are required to return old ages. However, when matched in comoving number- or luminosity-density, the SDSS galaxies are less strongly clustered compared to their counterparts in CMASS. This rules out the passive evolution scenario, and, indeed, any minor merger scenarios which preserve the rank ordering in stellar mass of the population. Potential incompletenesses in the CMASS sample would further enhance this mismatch. Our analysis highlights the virtue of combining clustering measurements with number counts., Comment: Accepted for publication in MNRAS, 15 pages, 12 figures

We present a catalog of visual like H-band morphologies of $\sim50.000$ galaxies ($H_{f160w}<24.5$) in the 5 CANDELS fields (GOODS-N, GOODS-S, UDS, EGS and COSMOS). Morphologies are estimated with Convolutional Neural Networks (ConvNets). The median redshift of the sample is $\sim1.25$. The algorithm is trained on GOODS-S for which visual classifications are publicly available and then applied to the other 4 fields. Following the CANDELS main morphology classification scheme, our model retrieves the probabilities for each galaxy of having a spheroid, a disk, presenting an irregularity, being compact or point source and being unclassifiable. ConvNets are able to predict the fractions of votes given a galaxy image with zero bias and $\sim10\%$ scatter. The fraction of miss-classifications is less than $1\%$. Our classification scheme represents a major improvement with respect to CAS (Concentration-Asymmetry-Smoothness)-based methods, which hit a $20-30\%$ contamination limit at high z. The catalog is released with the present paper via the $\href{http://rainbowx.fis.ucm.es/Rainbow_navigator_public}{Rainbow\,database}$, Comment: Accepted for publication in ApjS. Figure 10 summarizes the excellent agreement between our classification and a pure visual one. Table 3 shows the content of the catalogs. The catalogs are available from the Rainbow database (http://rainbowx.fis.ucm.es/Rainbow_navigator_public) based on the selections from the CANDELS team and cross-matched with 3D-HST v4.1 catalogs

In this paper we consider a multivariate model-based approach to measure the dynamic evolution of tail risk interdependence among US banks, financial services and insurance sectors. To deeply investigate the risk contribution of insurers we consider separately life and non-life companies. To achieve this goal we apply the multivariate student-t Markov Switching model and the Multiple-CoVaR (CoES) risk measures introduced in Bernardi et. al. (2013b) to account for both the known stylised characteristics of the data and the contemporaneous joint distress events affecting financial sectors. Our empirical investigation finds that banks appear to be the major source of risk for all the remaining sectors, followed by the financial services and the insurance sectors, showing that insurance sector significantly contributes as well to the overall risk. Moreover, we find that the role of each sector in contributing to other sectors distress evolves over time accordingly to the current predominant financial condition, implying different interconnection strength., Comment: arXiv admin note: text overlap with arXiv:1312.6407