SDSS DR7 Research Articles

GPz: non-stationary sparse Gaussian processes for heteroscedastic uncertainty estimation in photometric redshifts

The next generation of cosmology experiments will be required to use photometric redshifts rather than spectroscopic redshifts. Obtaining accurate and well-characterized photometric redshift distributions is therefore critical for Euclid, the Large Synoptic Survey Telescope and the Square Kilometre Array. However, determining accurate variance predictions alongside single point estimates is crucial, as they can be used to optimize the sample of galaxies for the specific experiment (e.g. weak lensing, baryon acoustic oscillations, supernovae), trading off between completeness and reliability in the galaxy sample. The various sources of uncertainty in measurements of the photometry and redshifts put a lower bound on the accuracy that any model can hope to achieve. The intrinsic uncertainty associated with estimates is often non-uniform and input-dependent, commonly known in statistics as heteroscedastic noise. However, existing approaches are susceptible to outliers and do not take into account variance induced by non-uniform data density and in most cases require manual tuning of many parameters. In this paper, we present a Bayesian machine learning approach that jointly optimizes the model with respect to both the predictive mean and variance we refer to as Gaussian processes for photometric redshifts (GPz). The predictive variance of the model takes into account both the variance due to data density and photometric noise. Using the SDSS DR12 data, we show that our approach substantially outperforms other machine learning methods for photo-z estimation and their associated variance, such as TPZ and ANNz2. We provide a Matlab and Python implementations that are available to download at https://github.com/OxfordML/GPz .

Covering factors of the dusty obscurers in radio-loud and radio-quiet quasars

We compare covering factors of circumnuclear dusty obscurers in radio-loud and radio-quiet quasars. The radio-loud quasars are represented by a sample of FR II quasars obtained by cross-matching a catalog of the FR II radio sources selected by van Velzen et al. with the SDSS DR7 catalog of quasars. Covering factors of FR II quasars are compared with covering factors of the radio-quiet quasars matched with them in redshift, black hole mass, and Eddington-ratio. We found that covering factors, proxied by the infrared-to-bolometric luminosity ratio, are on average slightly smaller in FR II quasars than in radio-quiet quasars, however, this difference is statistically significant only for the highest Eddington ratios. For both samples, no statistically significant dependence of a median covering factor on Eddington ratio, black hole mass, nor redshift can be claimed.

The fraction of BL Lacertae objects in groups of galaxies

Context: BL Lac objects are a rare class of active galactic nuclei that typically show featureless optical spectra that make it difficult to estimate the redshift. A novel method for estimating the redshift of BL Lac objects has recently been proposed which assumes that these objects typically reside in groups of galaxies. Aims: The aim of this work is to estimate the fraction of BL Lac objects that reside in groups of galaxies. Methods: We use a sample of groups (M12 catalogue) selected by applying a friends-of-friends algorithm in the SDSS DR12. Galaxies in the M12 sample were cross-correlated with the sample of blazars in the BZCAT. Results: We found that 121 galaxies in the M12 catalogue are blazars in the BZCAT sample, all but one are BL Lac objects, and a large fraction are classified as BL Lac-galaxy dominated. Analysing the fraction of galaxies in groups as a function of redshift we have estimated a correction factor that takes into account the typical incompleteness of the catalogues based on the friends-of-friends algorithm. Once this factor was applied to the sample of BL Lac objects with a counterpart in the M12 catalogue, we found that the percentage of BL Lac objects in groups is $\gtrsim 67 \pm 8\%$. Conclusions: The high rate of BL Lac objects in groups found in this work strongly supports a recent method that has successfully estimated the redshift of BL Lac objects with featureless spectra.

THE LUMINOSITY DEPENDENCE OF QUASAR UV CONTINUUM SLOPE: DUST EXTINCTION SCENARIO

ABSTRACT We investigate the UV continuum slope α of a large quasar sample from SDSS DR7. By using specific continuum windows, we build two samples at lower ( ) and higher ( ) redshifts, which correspond to the continuum slopes at longer (near-UV) and shorter (far-UV) rest wavelength ranges, respectively. Overall, the average continuum slopes are −0.36 and −0.51 for and with similar dispersions . For both samples, we confirm the luminosity dependence of the continuum slope, i.e., fainter quasars have redder spectra. We further find that both and have a common upper limit ( ), which is almost independent of the quasar luminosity . This finding implies that the intrinsic quasar continuum (or the bluest quasar), at any luminosity, obeys the standard thin-disk model. We propose that the other quasars with redder α are caused by the reddening from the dust locally. With this assumption, we employ the dust extinction scenario to model the observed relation. We find that a typical value of (depending on the types of extinction curve) of the quasar local dust is enough to explain the discrepancy of α between the observation ( ) and the standard accretion disk model prediction ( ).

ASYMMETRY BETWEEN GALAXIES WITH CLOCKWISE HANDEDNESS AND COUNTERCLOCKWISE HANDEDNESS

ABSTRACT While it is clear that spiral galaxies can have different handedness, galaxies with clockwise patterns are assumed to be symmetric to galaxies with counterclockwise patterns in all of their other characteristics. Here, we use data from SDSS DR7 to show that photometric data can distinguish between clockwise and counterclockwise galaxies. Pattern recognition algorithms trained and tested using the photometric data of a clean, manually crafted data set of 13,440 spiral galaxies with can predict the handedness of a spiral galaxy in ∼64% of the cases, which is significantly higher than the mere chance accuracy of 50% ( ). Experiments with a different data set of 10,281 automatically classified galaxies showed similar results of ∼65% classification accuracy, suggesting that the observed asymmetry is also consistent in data sets annotated in a fully automatic process, without human intervention. That shows that the photometric data collected by SDSS is sensitive to the handedness of the galaxy. Analysis of the number of galaxies classified as clockwise and counterclockwise by crowdsourcing shows that manual classification between spiral and elliptical galaxies can be affected by the handedness of the galaxy, and therefore the galaxy morphology analyzed by citizen science campaigns might be biased by the galaxy handedness. The code and data used in the experiment are publicly available, and the experiment can be easily replicated.

Probing large scale homogeneity and periodicity in the LRG distribution using Shannon entropy

We quantify the degree of inhomogeneity in the Luminous Red Galaxy (LRG) distribution from the SDSS DR7 as a function of length scales by measuring the Shannon entropy in independent and regular cubic voxels of increasing grid sizes. We also analyze the data by carrying out measurements in overlapping spheres and find that it suppresses inhomogeneities by a factor of 5 to 10 on different length scales. Despite the differences observed in the degree of inhomogeneity both the methods show a decrease in inhomogeneity with increasing length scales which eventually settle down to a plateau at $\sim 150 \, h^{-1} \rm {Mpc}$. Considering the minuscule values of inhomogeneity at the plateaus and their expected variations we conclude that the LRG distribution becomes homogeneous at $150 \, h^{-1} \rm {Mpc}$ and beyond. We also use the Kullback-Leibler divergence as an alternative measure of inhomogeneity which reaffirms our findings. We show that the method presented here can effectively capture the inhomogeneity in a truly inhomogeneous distribution at all length scales. We analyze a set of Monte Carlo simulations with certain periodicity in their spatial distributions and find periodic variations in their inhomogeneity which helps us to identify the underlying regularities present in such distributions and quantify the scale of their periodicity. We do not find any underlying regularities in the LRG distribution within the length scales probed.

Multifrequency studies of galaxies and groups

To understand the role of the environment in galaxy formation, evolution, and present-day properties, it is essential to study the multi-frequency behavior of different galaxy populations under various environmental conditions. We crossmatch the SDSS DR10 group catalog with GAMA Data Release 2 and Wide-field Survey Explorer (WISE) data to construct a catalog of 1651 groups and 11436 galaxies containing photometric information in 15 different wavebands ranging from ultraviolet (0.152 {\mu}m) to mid-infrared (22 {\mu}m). We perform the spectral energy distribution (SED) fitting of galaxies using the MAGPHYS code and estimate the rest frame luminosities and stellar masses. We use the 1/Vmax method to estimate the galaxy stellar mass and luminosity functions, and the luminosity density field of galaxies to define the large scale environment of galaxies. The stellar mass functions of both central and satellite galaxies in groups are different in low and high density large scale environments. Satellite galaxies in high density environments have a steeper low mass end slope compared to low density environments, independently of the galaxy morphology. Central galaxies in low density environments have a steeper low mass end slope but the difference disappears for fixed galaxy morphology. The characteristic stellar mass of satellite galaxies is higher in high density environments and the difference exists only for galaxies with elliptical morphologies. Galaxy formation in groups is more efficient in high density large scale environments. Groups in high density environments have higher abundances of satellite galaxies, irrespective of the satellite galaxy morphology. The elliptical satellite galaxies are generally more massive in high density environments. The stellar masses of spiral satellite galaxies show no dependence on the large scale environment.

Photometric redshifts for the SDSS Data Release 12

We present the methodology and data behind the photometric redshift database of the Sloan Digital Sky Survey Data Release 12 (SDSS DR12). We adopt a hybrid technique, empirically estimating the redshift via local regression on a spectroscopic training set, then fitting a spectrum template to obtain K-corrections and absolute magnitudes. The SDSS spectroscopic catalog was augmented with data from other, publicly available spectroscopic surveys to mitigate target selection effects. The training set is comprised of $1,976,978$ galaxies, and extends up to redshift $z\approx 0.8$, with a useful coverage of up to $z\approx 0.6$. We provide photometric redshifts and realistic error estimates for the $208,474,076$ galaxies of the SDSS primary photometric catalog. We achieve an average bias of $\overline{\Delta z_{\mathrm{norm}}} = 5.84 \times 10^{-5}$, a standard deviation of $\sigma \left(\Delta z_{\mathrm{norm}}\right)=0.0205$, and a $3\sigma$ outlier rate of $P_o=4.11\%$ when cross-validating on our training set. The published redshift error estimates and photometric error classes enable the selection of galaxies with high quality photometric redshifts. We also provide a supplementary error map that allows additional, sophisticated filtering of the data.

Stellar mass functions: methods, systematics and results for the local Universe

We present a comprehensive method for determining stellar mass functions, and apply it to samples in the local Universe. We combine the classical 1/Vmax approach with STY, a parametric maximum likelihood method and SWML, a non-parametric maximum likelihood technique. In the parametric approach, we are assuming that the stellar mass function can be modelled by either a single or a double Schechter function and we use a likelihood ratio test to determine which model provides a better fit to the data. We discuss how the stellar mass completeness as a function of z biases the three estimators and how it can affect, especially the low mass end of the stellar mass function. We apply our method to SDSS DR7 data in the redshift range from 0.02 to 0.06. We find that the entire galaxy sample is best described by a double Schechter function with the following parameters: $\log (M^{*}/M_\odot) = 10.79 \pm 0.01$, $\log (\Phi^{*}_1/\mathrm{h^3\ Mpc^{-3}}) = -3.31 \pm 0.20$, $\alpha_1 = -1.69 \pm 0.10$, $\log (\Phi^{*}_2/\mathrm{h^3\ Mpc^{-3}}) = -2.01 \pm 0.28$ and $\alpha_2 = -0.79 \pm 0.04$. We also use morphological classifications from Galaxy Zoo and halo mass, overdensity, central/satellite, colour and sSFR measurements to split the galaxy sample into over 130 subsamples. We determine and present the stellar mass functions and the best fit Schechter function parameters for each of these subsamples.

Dependence of the clustering properties of galaxies on galaxy age

From the Main galaxy data of the Sloan Digital Sky Survey Data Release 10 (SDSS DR10), we construct two volume-limited samples with the luminosity −20.5 ≤Mr ≤−18.5 and −22.5 ≤Mr ≤−20.5, respectively, and explore the dependence of the clustering properties of galaxies on galaxy age by cluster analysis. Statistical analyses in two volume-limited Main galaxy samples can reach the same conclusions: young galaxies preferentially form isolated galaxies, close pairs and small groups at all scales, whereas old galaxies preferentially inhabit dense groups and clusters.

The environmental dependence of the stellar mass fundamental plane of early-type galaxies

Aims. We investigate the environmental dependence of the stellar mass fundamental plane (FP$_*$) using the early-type galaxy sample from the Sloan Digital Sky Survey Data Release 7 (SDSS DR7). Methods. The FP$_*$ is calculated by replacing the luminosity in the fundamental plane (FP) with stellar mass. Based on the SDSS group catalog, we characterize the galaxy environment according to the mass of the host dark matter halo and the position in the halo. In halos with the same mass bin, the color distributions of central and satellite galaxies are different. Therefore, we calculate FP$_*$ coefficients of galaxies in different environments and compare them with those of the FP to study the contribution of the stellar population. Results. We find that coefficient $a$ of the FP$_*$ is systematically larger than that of the FP, but coefficient $b$ of the FP$_*$ is similar to the FP. Moreover, the environmental dependence of the FP$_*$ is similar to that of the FP. For central galaxies, FP$_*$ coefficients are significantly dependent on the halo mass. For satellite galaxies, the correlation between FP$_*$ coefficients and the halo mass is weak. Conclusions. We conclude that the tilt of the FP is not primarily driven by the stellar population.

Shell-like structures in our cosmic neighbourhood

Signatures of the processes in the early Universe are imprinted in the cosmic web. Some of them may define shell-like structures characterised by typical scales. We search for shell-like structures in the distribution of nearby rich clusters of galaxies drawn from the SDSS DR8. We calculate the distance distributions between rich clusters of galaxies, and groups and clusters of various richness, look for the maxima in the distance distributions, and select candidates of shell-like structures. We analyse the space distribution of groups and clusters forming shell walls. We find six possible candidates of shell-like structures, in which galaxy clusters have maxima in the distance distribution to other galaxy groups and clusters at the distance of about 120 Mpc/h. The rich galaxy cluster A1795, the central cluster of the Bootes supercluster, has the highest maximum in the distance distribution of other groups and clusters around them at the distance of about 120 Mpc/h among our rich cluster sample, and another maximum at the distance of about 240 Mpc/h. The structures of galaxy systems causing the maxima at 120 Mpc/h form an almost complete shell of galaxy groups, clusters and superclusters. The richest systems in the nearby universe, the Sloan Great Wall, the Corona Borealis supercluster and the Ursa Major supercluster are among them. The probability that we obtain maxima like this from random distributions is lower than 0.001. Our results confirm that shell-like structures can be found in the distribution of nearby galaxies and their systems. The radii of the possible shells are larger than expected for a BAO shell (approximately 109 Mpc/h versus approximately 120 Mpc/h), and they are determined by very rich galaxy clusters and superclusters with high density contrast while BAO shells are barely seen in the galaxy distribution. We discuss possible consequences of these differences.

Local starburst galaxies and their descendants

Despite strong interest in the starburst (hereafter SB) phenomenon, the concept remains ill-defined. We use a strict definition of SB to examine the statistical properties of local SB and post-starburst (hereafter PB) galaxies. We also seek relationships to active galaxies. Potential SB galaxies are selected from the SDSS DR7 and their stellar content is analysed. We apply an age dependent dust attenuation correction and derive star formation rates (SFR), ages and masses of the young and old populations. The photometric masses nicely agree with dynamical masses derived from the H-alpha emission line width. To select SB galaxies, we use the birthrate parameter b=SFR/<SFR>, requiring b>=3. The PB sample is selected from the citerion EW(Hdelta_abs)>=6 A. Only 1% of star-forming galaxies are found to be SB galaxies. They contribute 3-6% to the stellar production and are therefore unimportant for the local star formation activity. The median SB age is 70 Myr, roughly independent of mass. The b-parameter strongly depends on burst age. Values close to b=60 are found at ages ~10 Myr, while almost no SBs are found at ages >1 Gyr. The median baryonic burst mass fraction of sub-L* galaxies is 5%, decreasing slowly with mass. The median mass fraction of the recent burst in the PB sample is 5-10%. The age-mass distribution of the progenitors of the PBs is bimodal with a break at log(M)~10.6 above which the ages are doubled. The SB and PB luminosity functions (hereafter LFs) follow each other closely until M_r~-21, when AGNs begin to dominate. The PB LF continues to follow the AGN LF while SB loose significance. This suggests that the number of luminous SBs is underestimated by about one dex at high luminosities, due to large amounts of dust and/or AGN blending. It also indicates that the SB phase preceded the AGN phase. We also discuss the conditions for global gas outflow caused by stellar feedback.

Matter perturbations in scaling cosmology

A suitable nonlinear interaction between dark matter with an energy density $\rho_{M}$ and dark energy with an energy density $\rho_{X}$ is known to give rise to a non-canonical scaling $\rho_{M} \propto \rho_{X}a^{-\xi}$ where $\xi$ is a parameter which generally deviates from $\xi =3$. Here we present a covariant generalization of this class of models and investigate the corresponding perturbation dynamics. The resulting matter power spectrum for the special case of a time-varying Lambda model is compared with data from the SDSS DR9 catalogue. We find a best-fit value of $\xi = 3.25$ which corresponds to a decay of dark matter into the cosmological term. Our results are compatible with the $\Lambda$CDM model at the 2$\sigma$ confidence level.

Baryon acoustic oscillations from the SDSS DR10 galaxies angular correlation function

The 2-point angular correlation function $w(\ensuremath{\theta})$ (2PACF), where $\ensuremath{\theta}$ is the angular separation between pairs of galaxies, provides the transversal baryon acoustic oscillation (BAO) signal almost model independently. In this paper we use 409 337 luminous red galaxies in the redshift range $z=[0.440,0.555]$ obtained from the tenth data release of the Sloan Digital Sky Survey (SDSS DR10) to estimate ${\ensuremath{\theta}}_{\mathrm{BAO}}(z)$ from the 2PACF at six redshift shells. Since noise and systematics can hide the BAO signature in the $w\text{\ensuremath{-}}\ensuremath{\theta}$ plane, we also discuss some criteria to localize the acoustic bump. We identify two sources of model dependence in the analysis, namely, the value of the acoustic scale from cosmic microwave background (CMB) measurements and the correction in the ${\ensuremath{\theta}}_{\mathrm{BAO}}(z)$ position due to projection effects. Constraints on the dark energy equation-of-state parameter $\mathrm{w}(z)$ from the ${\ensuremath{\theta}}_{\mathrm{BAO}}(z)$ diagram are derived, as well as from a joint analysis with current CMB measurements. We find that the standard $\mathrm{\ensuremath{\Lambda}}\mathrm{CDM}$ model as well as some of its extensions are in good agreement with these ${\ensuremath{\theta}}_{\mathrm{BAO}}(z)$ measurements.

Comparisons of the galaxy age, stellar velocity dispersion and K-band luminosity distributions between grouped galaxies and isolated ones

In two volume-limited Main galaxy samples of the Sloan Digital Sky Survey Data Release 10 (SDSS DR10), we compare the age, stellar velocity dispersion and K-band luminosity distributions of grouped galaxies with those of isolated galaxies, to explore the environmental dependence of these properties of galaxies. It is found that grouped galaxies have preferentially larger stellar velocity dispersions and are preferentially older than isolated galaxies. We also note apparent difference of K-band luminosity distribution at both extremes of density in the luminous volume-limited Main galaxy sample: grouped galaxies are preferentially more luminous than isolated galaxies, while this difference in the faint volume-limited Main galaxy sample is very small.

A SEARCH FOR HYPERLUMINOUS X-RAY SOURCES IN THE XMM-NEWTON SOURCE CATALOG

ABSTRACT We present a new method to identify luminous off-nuclear X-ray sources in the outskirts of galaxies from large public redshift surveys, distinguishing them from foreground and background interlopers. Using the 3XMM-DR5 catalog of X-ray sources and the SDSS DR12 spectroscopic sample of galaxies, with the help of this off-nuclear cross-matching technique, we selected 98 sources with inferred X-ray luminosities in the range 1041 &lt; L X &lt; 1044 erg s−1, compatible with hyperluminous X-ray objects (HLX). To validate the method, we verify that it allowed us to recover known HLX candidates such as ESO 243–49 HLX–1 and M82 X–1. From a statistical study, we conservatively estimate that up to 71 ± 11 of these sources may be foreground- or background sources, statistically leaving at least 16 that are likely to be HLXs, thus providing support for the existence of the HLX population. We identify two good HLX candidates and using other publicly available data sets, in particular the VLA FIRST in radio, UKIRT Infrared Deep Sky Survey in the near-infrared, GALEX in the ultraviolet and Canada–France–Hawaii Telescope Megacam archive in the optical, we present evidence that these objects are unlikely to be foreground or background X-ray objects of conventional types, e.g., active galactic nuclei, BL Lac objects, Galactic X-ray binaries, or nearby stars. However, additional dedicated X-ray and optical observations are needed to confirm their association with the assumed host galaxies and thus secure their HLX classification.

THE SWIFT AGN AND CLUSTER SURVEY. II. CLUSTER CONFIRMATION WITH SDSS DATA

We study 203 (of 442) Swift AGN and Cluster Survey extended X-ray sources located in the SDSS DR8 footprint to search for galaxy over-densities in three dimensional space using SDSS galaxy photometric redshifts and positions near the Swift cluster candidates. We find 104 Swift clusters with a >3sigma galaxy over-density. The remaining targets are potentially located at higher redshifts and require deeper optical follow-up observations for confirmation as galaxy clusters. We present a series of cluster properties including the redshift, BCG magnitude, BCG-to-X-ray center offset, optical richness, and X-ray luminosity. We also detect red sequences in ~85% of the 104 confirmed clusters. The X-ray luminosity and optical richness for the SDSS confirmed Swift clusters are correlated and follow previously established relations. The distribution of the separations between the X-ray centroids and the most likely BCG is also consistent with expectation. We compare the observed redshift distribution of the sample with a theoretical model, and find that our sample is complete for z <~ 0.3 and is still 80% complete up to z ~= 0.4, consistent with the SDSS survey depth. These analysis results suggest that our Swift cluster selection algorithm has yielded a statistically well-defined cluster sample for further studying cluster evolution and cosmology. We also match our SDSS confirmed Swift clusters to existing cluster catalogs, and find 42, 23 and 1 matches in optical, X-ray and SZ catalogs, respectively, so the majority of these clusters are new detections.

A CATALOG OF GALEX ULTRAVIOLET EMISSION FROM SPECTROSCOPICALLY CONFIRMED M DWARFS*

ABSTRACT We present a catalog of Galaxy Evolution Explorer Near-UV (NUV) and Far-UV (FUV) photometry for the Palomar/MSU and SDSS DR7 spectroscopic M dwarf catalogs. The catalog contains NUV measurements matched to 577 spectroscopically confirmed M dwarfs and FUV measurements matched to 150 spectroscopically confirmed M dwarfs. Using these data, we find that NUV and FUV luminosities strongly correlate with Hα emission, a typical indicator of magnetic activity in M dwarfs. We also examine the fraction of M dwarfs with varying degrees of strong line emission at NUV wavelengths. Our results indicate that the frequency of M dwarf NUV emission peaks at intermediate spectral types, with at least ∼30% of young M4–M5 dwarfs having some level of activity. For mid-type M dwarfs, we show that NUV emission decreases with distance from the Galactic plane, a proxy for stellar age. Our complete matched source catalog is available online.

Schechter Function Model for the QSO Luminosity Function from the SDSS DR7

A study of the optical luminosity function of Quasi Stellar Objects (QSOs) and its evolution with redshift is carried out using the data from the Sloan Digital Sky Survey Data Release Seven (SDSS DR7). It is shown that the observed QSO luminosity function is well fitted by a Schechter function model of the form , where is the break or characteristic luminosity with luminosity evolution characterized by a second order polynomial in red shift. The best fit parameters are determined by using the Levenberg-Marquardt method of nonlinear least square fit.