Bouwens RJ, Illingworth GD, Thompson RI, Blakeslee JP, Dickinson ME, Broadhurst TJ, Eisenstein DJ, Fan X, Franx M, Meurer G, et al. Star Formation at z ~ 6: The Hubble Ultra Deep Parallel Fields. The Astrophysical Journal Letters. 2004;606 :L25-L28. Publisher's VersionAbstract

We report on the i-dropouts detected in two exceptionally deep AdvancedCamera for Surveys fields (B435, V606,i775, and z850 with 10σ limits of 28.8,29.0, 28.5, and 27.8, respectively) taken in parallel with the UltraDeep Field Near-Infrared Camera and Multi-Object Spectrometerobservations. Using an i-z>1.4 cut, we find 30 i-dropouts over 21arcmin2 down to z850,AB=28.1, or 1.4 i-dropoutsarcmin-2, with significant field-to-field variation (asexpected from cosmic variance). This extends i-dropout searches some~0.9 mag further down the luminosity function than was possible in theGreat Observatories Origins Deep Survey (GOODS) fields, yielding a ~7times increase in surface density. An estimate of the size evolution forUV-bright objects is obtained by comparing the composite radial fluxprofile of the bright i-dropouts (z850,AB<27.2) withscaled versions of the Hubble Deep Field-North and -South U-dropouts.The best fit is found with a(1+z)-1.57+0.50-0.53 scaling in size(for fixed luminosity), extending lower redshift (1-2.8 innumber and (1+z)0.1 in luminosity, suggesting a rest-framecontinuum UV luminosity density at z~6 that is just0.38+0.09-0.07 times that at z~3.8. Our inclusionof the size evolution makes the present estimate lower than previous z~6estimates.Based on observations made with the NASA/ESA Hubble Space Telescope,which is operated by the Association of Universities for Research inAstronomy, Inc., under NASA contract NAS5-26555. These observations areassociated with program 9803.

Liebert J, Bergeron P, Eisenstein D, Harris HC, Kleinman SJ, Nitta A, Krzesinski J. A Helium White Dwarf of Extremely Low Mass. The Astrophysical Journal Letters. 2004;606 :L147-L149. Publisher's VersionAbstract

We analyze the spectrum of an unusually low mass white dwarf, SDSSJ123410.37-022802.9 (0335-264-52000), found in our recent, white dwarfcatalog from the first data release (DR1) of the Sloan Digital SkySurvey (SDSS). Two independent, model atmosphere fits result in anaccurate determination of atmospheric and stellar parameters. The morehands-on analysis yields Teff=17,470+/-750 K andlogg=6.38+/-0.05. We argue that the object cannot be a main-sequence Astar, a horizontal-branch star, or a subdwarf B star. Instead, it isinterpreted as a very low mass white dwarf with a core composed ofhelium, with mass ~0.18-0.19 Msolar, similar to thatpublished previously for the unusual companion to the millisecond pulsarJ1012+5307. The star probably remains in a binary, perhaps even with anundiscovered or dead pulsar companion. However, the companion might be amore ordinary star, provided Roche lobe mass transfer began shortlyafter the now-visible component left the main sequence. A second SDSSlow-mass white dwarf candidate is also analyzed, but the spectrum forthis fainter object is of poorer quality. The sample appears to includeadditional, similar candidates, worthy of more accurate observation.Correct identification of any additional white dwarfs of extremely lowmass requires careful observation and interpretation.

Kleinman SJ, Harris HC, Eisenstein DJ, Liebert J, Nitta A, Krzesiński J, Munn JA, Dahn CC, Hawley SL, Pier JR, et al. A Catalog of Spectroscopically Identified White Dwarf Stars in the First Data Release of the Sloan Digital Sky Survey. The Astrophysical Journal. 2004;607 :426-444. Publisher's VersionAbstract

We present the full spectroscopic white dwarf and hot subdwarf samplefrom the Sloan Digital Sky Survey (SDSS) first data release, DR1. Wefind 2551 white dwarf stars of various types, 240 hot subdwarf stars,and an additional 144 objects we have identified as uncertain whitedwarf stars. Of the white dwarf stars, 1888 are nonmagnetic DA types and171 are nonmagnetic DBs. The remaining (492) objects consist of alldifferent types of white dwarf stars: DO, DQ, DC, DH, DZ, hybrid starssuch as DAB, etc., and those with nondegenerate companions. We fit theDA and DB spectra with a grid of models to determine the Teffand logg for each object. For all objects, we provide coordinates,proper motions, SDSS photometric magnitudes, and enough information toretrieve the spectrum/image from the SDSS public database. This catalognearly doubles the known sample of spectroscopically identified whitedwarf stars. In the DR1 imaged area of the sky, we increase the knownsample of white dwarf stars by a factor of 8.5. We also comment onseveral particularly interesting objects in this sample.

Tegmark M, Strauss MA, Blanton MR, Abazajian K, Dodelson S, Sandvik H, Wang X, Weinberg DH, Zehavi I, Bahcall NA, et al. Cosmological parameters from SDSS and WMAP. Physical Review D. 2004;69 :103501. Publisher's VersionAbstract

We measure cosmological parameters using the three-dimensional powerspectrum P(k) from over 200 000 galaxies in the Sloan Digital Sky Survey(SDSS) in combination with Wilkinson Microwave Anisotropy Probe (WMAP)and other data. Our results are consistent with a “vanilla”flat adiabatic cold dark matter model with a cosmological constantwithout tilt (ns=1), running tilt, tensor modes, or massiveneutrinos. Adding SDSS information more than halves the WMAP-only errorbars on some parameters, tightening 1σ constraints on the Hubbleparameter from h≈0.74+0.18-0.07 toh≈0.70+0.04-0.03, on the matter density fromΩm≈0.25±0.10 toΩm≈0.30±0.04 (1σ) and on neutrino

Pope AC, Matsubara T, Szalay AS, Blanton MR, Eisenstein DJ, Gray J, Jain B, Bahcall NA, Brinkmann J, Budavari T, et al. Cosmological Parameters from Eigenmode Analysis of Sloan Digital Sky Survey Galaxy Redshifts. The Astrophysical Journal. 2004;607 :655-660. Publisher's VersionAbstract

We present estimates of cosmological parameters from the application ofthe Karhunen-Loève transform to the analysis of thethree-dimensional power spectrum of density fluctuations using SloanDigital Sky Survey galaxy redshifts. We use Ωmh andfbbm to describe theshape of the power spectrum, σL8g for the(linearly extrapolated) normalization, and β to parameterize lineartheory redshift-space distortions. On scales k<~0.16hMpc-1, our maximum likelihood values areΩmh=0.264+/-0.043, fb=0.286+/-0.065,σL8g=0.966+/-0.048, and β=0.45+/-0.12.When we take a prior on Ωb from the Wilkinson MicrowaveAnisotropy Probe (WMAP), we find Ωmh=0.207+/-0.030,which is in excellent agreement with WMAP and the Two-Degree Field. Thisindicates that we have reasonably measured the gross shape of the powerspectrum, but we have difficulty breaking the degeneracy betweenΩmh and fb, because the baryon oscillationsare not resolved in the current spectroscopic survey window function.

Abazajian K, Adelman-McCarthy JK, Agüeros MA, Allam SS, Anderson K, Anderson SF, Annis J, Bahcall NA, Baldry IK, Bastian S, et al. The Second Data Release of the Sloan Digital Sky Survey. The Astronomical Journal. 2004;128 :502-512. Publisher's VersionAbstract

The Sloan Digital Sky Survey (SDSS) has validated and made publiclyavailable its Second Data Release. This data release consists of 3324deg2 of five-band (ugriz) imaging data with photometry forover 88 million unique objects, 367,360 spectra of galaxies, quasars,stars, and calibrating blank sky patches selected over 2627deg2 of this area, and tables of measured parameters from

Hall PB, Snedden SA, Niederste-Ostholt M, Eisenstein DJ, Strauss MA, York DG, Schneider DP. A Quasar without Broad Lyα Emission. The Astronomical Journal. 2004;128 :534-543. Publisher's VersionAbstract

The z=3.02 quasar SDSS J095253.83+011421.9 exhibits broad metal-lineemission (C IV FWHM~=9000 km s-1), but broad Lyαemission is not present. Instead, only a narrow Lyα line isobserved (FWHM~=1140 km s-1). The large CIV/Lyα ratioin the broad-line region (BLR) emission from this object can be matchedmost closely by a BLR dominated by gas at very high densities(1015 cm-3), which suppresses the Lyαemission, and illuminated by an incident power law extending to ~200μm, which yields increased emission from purely collisionally excitedcoolant lines (such as C IV, N V, and O VI) but not from recombinationlines like Lyα. However, the strong C III emission predicted bythis model is not observed, and the observed broad C III] emission mustcome from a lower density BLR component and should be accompanied bybroad Lyα emission, which is not observed. The least unlikelyexplanation for this spectrum seems to be that any intrinsic broadLyα emission is removed by smooth N V absorption in the red wingof the Lyα emission line and by smooth Lyα absorption in theblue wing of the Lyα emission line. This postulated smoothabsorption would be in addition to the strong, associated, narrowabsorption seen in numerous ions. Smooth absorption in Lyα, N V,and O VI, but not in C IV, would be unusual, but not impossible,although it is suspicious that the postulated absorption must almostexactly cancel the postulated intrinsic broad emission. We conclude thatthe spectrum of SDSS J0952+0114 appears unique (among ~=3600 SDSSspectra of quasars at z>2.12) because of some combination of unusualparameters, and we discuss possible observations to determine thecombination of circumstances responsible for the spectrum.

Ivezić Ž, Lupton RH, Schlegel D, Boroski B, Adelman-McCarthy J, Yanny B, Kent S, Stoughton C, Finkbeiner D, Padmanabhan N, et al. SDSS data management and photometric quality assessment. Astronomische Nachrichten. 2004;325 :583-589. Publisher's VersionAbstract

We summarize the Sloan Digital Sky Survey data acquisition andprocessing steps, and describe runQA, a pipeline designed for automateddata quality assessment. In particular, we show how the position of thestellar locus in color-color diagrams can be used to estimate theaccuracy of photometric zeropoint calibration to better than 0.01 mag in0.03 deg2 patches. Using this method, we estimate thattypical photometric zeropoint calibration errors for SDSS imaging dataare not larger than ˜0.01 mag in the g, r, and i bands, 0.02 magin the z band, and 0.03 mag in the u band (root-mean-scatter forzeropoint offsets).

Finkbeiner DP, Padmanabhan N, Schlegel DJ, Carr MA, Gunn JE, Rockosi CM, Sekiguchi M, Lupton RH, Knapp GR, Ivezić Ž, et al. Sloan Digital Sky Survey Imaging of Low Galactic Latitude Fields: Technical Summary and Data Release. The Astronomical Journal. 2004;128 :2577-2592. Publisher's VersionAbstract

The Sloan Digital Sky Survey (SDSS) mosaic camera and telescope haveobtained five-band optical-wavelength imaging near the Galactic planeoutside of the nominal survey boundaries. These additional data wereobtained during commissioning and subsequent testing of the SDSSobserving system, and they provide unique wide-area imaging data inregions of high obscuration and star formation, including numerous youngstellar objects, Herbig-Haro objects, and young star clusters. Becausethese data are outside the survey regions in the Galactic caps, they arenot part of the standard SDSS data releases. This paper presents imagingdata for 832 square degrees of sky (including repeats), in thestar-forming regions of Orion, Taurus, and Cygnus. About 470deg2 are now released to the public, with the remainder tofollow at the time of SDSS Data Release 4. The public data in Orioninclude the star-forming region NGC 2068/NGC 2071/HH 24 and a large partof Barnard's loop.

Smolčić V, Ivezić Ž, Knapp GR, Lupton RH, Pavlovski K, Ilijić S, Schlegel D, Smith AJ, McGehee PM, Silvestri NM, et al. A Second Stellar Color Locus: a Bridge from White Dwarfs to M stars. The Astrophysical Journal Letters. 2004;615 :L141-L144. Publisher's VersionAbstract

We report the discovery of a locus of binary stars in the Sloan DigitalSky Survey (SDSS) g-r versus u-g color-color diagram that connects thecolors of white dwarfs and M dwarfs. While its contrast with respect tothe main stellar locus is only ~1:2300, this previously unrecognizedfeature includes 863 stars from the SDSS Data Release 1 (DR1). Theposition and shape of the feature are in good agreement with predictionsof a simple binary star model that consists of a white dwarf and an Mdwarf, with the components' luminosity ratio controlling the positionalong this binary system locus. SDSS DR1 spectra for 47 of these objectsstrongly support this model. The absolute magnitude-color distributioninferred for the white dwarf component is in good agreement with themodels of Bergeron et al.

Eisenstein D, White M. Theoretical uncertainty in baryon oscillations. Physical Review D. 2004;70 :103523. Publisher's VersionAbstract

We discuss the systematic uncertainties in the recovery of dark energyproperties from the use of baryon acoustic oscillations as a standardruler. We demonstrate that while unknown relativistic components in theuniverse prior to recombination would alter the sound speed, theinferences for dark energy from low-redshift surveys are unchanged solong as the microwave background anisotropies can measure the redshiftof matter-radiation equality, which they can do to sufficient accuracy.The mismeasurement of the radiation and matter densities themselves (asopposed to their ratio) would manifest as an incorrect prediction forthe Hubble constant at low-redshift. In addition, these anomalies doproduce subtle but detectable features in the microwave anisotropies.

Bouwens RJ, Thompson RI, Illingworth GD, Franx M, van Dokkum PG, Fan X, Dickinson ME, Eisenstein DJ, Rieke MJ. Galaxies at z~7-8: z850-Dropouts in the Hubble Ultra Deep Field. The Astrophysical Journal Letters. 2004;616 :L79-L82. Publisher's VersionAbstract

We have detected likely z~7-8 galaxies in the144''×144'' Near-Infrared Camera andMulti-Object Spectrometer (NICMOS) observations of the Hubble Ultra DeepField. Objects are required to be >=3 σ detections in bothNICMOS bands, J110 and H160. The selectioncriteria for this sample are(z850-J110)AB>0.8,(z850-J110)AB>0.66(J110-H160)AB+0.8,(J110-H160)AB<1.2 and no detectionat less than 8500 Å. The five selected sources have totalmagnitudes H160,AB~27. Four of the five sources are quiteblue compared to typical lower redshift dropout galaxies and areclustered within a 1 arcmin2 region. Because all five sourcesare near the limit of the NICMOS data, we have carefully evaluated theirreality. Each of the candidates is visible in different splits of thedata and a median stack. We analyzed several noise images and estimatethe number of spurious sources to be 1+/-1. A search using anindependent reduction of this same data set clearly revealed three ofthe five candidates and weakly detected a fourth candidate, suggestingthat the contamination could be higher. For comparison with predictionsfrom lower redshift samples, we take a conservative approach and adoptfour z~7-8 galaxies as our sample. With the same detection criteria onsimulated data sets, assuming no evolution from z~3.8, we predict 10sources at z~7-8, or 14 if we use a more realistic (1+z)-1size scaling. We estimate that the rest-frame continuum UV (~1800Å) luminosity density at z~7.5 (integrated down to0.3L*z=3) is just0.20+0.12-0.08 times that found at z~3.8 (or0.20+0.23-0.12 times this quantity includingcosmic variance). Effectively this sets an upper limit on the luminositydensity down to 0.3L*z=3 and is consistent withsignificant evolution at the bright end of the luminosity function fromz~7.5 to 3.8. Even with the lower UV luminosity density at z~7.5, itappears that galaxies could still play an important role in reionizationat these redshifts, although definitive measurements remain to be made.Based on observations made with the NASA/ESA Hubble Space Telescope,which is operated by the Association of Universities for Research inAstronomy, Inc., under NASA contract NAS 5-26555.

Csabai I, Budavári T, Connolly AJ, Szalay AS, Győry Z, Benítez N, Annis J, Brinkmann J, Eisenstein D, Fukugita M, et al. The Application of Photometric Redshifts to the SDSS Early Data Release. The Astronomical Journal. 2003;125 :580-592. Publisher's VersionAbstract

The Early Data Release (EDR) from the Sloan Digital Sky Survey providesone of the largest multicolor photometric catalogs currently availableto the astronomical community. In this paper we present the firstapplication of photometric redshifts to the ~6 million extended sourcesin these data (with 1.8 million sources having r'<21). Utilizing arange of photometric redshift techniques, from empirical to template andhybrid techniques, we investigate the statistical and systematicuncertainties present in the redshift estimates for the EDR data. Forr'<21, we find that the redshift estimates provide realistic redshifthistograms with an rms uncertainty in the photometric redshift relationof 0.035 at r'<18 and rising to 0.1 at r'<21. We conclude bydescribing how these photometric redshifts and derived quantities, suchas spectral type, rest-frame colors, and absolute magnitudes, are storedin the SDSS database. We provide sample queries for searching onphotometric redshifts and list the current caveats and issues thatshould be understood before using these photometric redshifts instatistical analyses of the SDSS galaxies.

Eisenstein DJ, Hogg DW, Fukugita M, Nakamura O, Bernardi M, Finkbeiner DP, Schlegel DJ, Brinkmann J, Connolly AJ, Csabai I, et al. Average Spectra of Massive Galaxies in the Sloan Digital Sky Survey. The Astrophysical Journal. 2003;585 :694-713. Publisher's VersionAbstract

We combine Sloan Digital Sky Survey spectra of 22,000 luminous, red,bulge-dominated galaxies to get high signal-to-noise ratio averagespectra in the rest-frame optical and ultraviolet (2600-7000 Å).The average spectra of these massive, quiescent galaxies are early typewith weak emission lines and with absorption lines indicating anapparent excess of α-elements over solar abundance ratios. We makeaverage spectra of subsamples selected by luminosity, environment, andredshift. The average spectra are remarkable in their similarity. Whatvariations do exist in the average spectra as a function of luminosityand environment are found to form a nearly one-parameter family inspectrum space. We present a high signal-to-noise ratio spectrum of thevariation. We measure the properties of the variation with a modifiedversion of the Lick index system and compare to model spectra fromstellar population syntheses. The variation may be a combination of ageand chemical abundance differences, but the conservative conclusion isthat the quality of the data considerably exceeds the current state ofthe models.

Hogg DW, Blanton MR, Eisenstein DJ, Gunn JE, Schlegel DJ, Zehavi I, Bahcall NA, Brinkmann J, Csabai I, Schneider DP, et al. The Overdensities of Galaxy Environments as a Function of Luminosity and Color. The Astrophysical Journal Letters. 2003;585 :L5-L9. Publisher's VersionAbstract

We study the mean environments of galaxies in the Sloan Digital SkySurvey (SDSS) as a function of rest-frame luminosity and color.Overdensities in galaxy number are estimated in 8 and 1 h-1Mpc spheres centered on 115,000 galaxies taken from the SDSSspectroscopic sample. We find that, at constant color, overdensity isindependent of luminosity for galaxies with the blue colors of spirals.This suggests that at fixed star formation history, spiral-galaxy massis a very weak function of environment. Overdensity does depend onluminosity for galaxies with the red colors of early types; bothlow-luminosity and high-luminosity red galaxies are found to be inhighly overdense regions.

Bernardi M, Sheth RK, Annis J, Burles S, Eisenstein DJ, Finkbeiner DP, Hogg DW, Lupton RH, Schlegel DJ, SubbaRao M, et al. Early-Type Galaxies in the Sloan Digital Sky Survey. I. The Sample. The Astronomical Journal. 2003;125 :1817-1848. Publisher's VersionAbstract

A sample of nearly 9000 early-type galaxies, in the redshift range0.01<=z<=0.3, was selected from the Sloan Digital Sky Survey(SDSS) using morphological and spectral criteria. This paper describeshow the sample was selected, presents examples of images andseeing-corrected fits to the observed surface brightness profiles,describes our method for estimating K-corrections, and shows that theSDSS spectra are of sufficiently high quality to measure velocitydispersions accurately. It also provides catalogs of the measuredphotometric and spectroscopic parameters. In related papers, these dataare used to study how early-type galaxy observables, includingluminosity, effective radius, surface brightness, color, and velocitydispersion, are correlated with one another.

Bernardi M, Sheth RK, Annis J, Burles S, Eisenstein DJ, Finkbeiner DP, Hogg DW, Lupton RH, Schlegel DJ, SubbaRao M, et al. Early-type Galaxies in the Sloan Digital Sky Survey. II. Correlations between Observables. The Astronomical Journal. 2003;125 :1849-1865. Publisher's VersionAbstract

A magnitude-limited sample of nearly 9000 early-type galaxies, in theredshift range 0.01<=z<=0.3, was selected from the Sloan DigitalSky Survey using morphological and spectral criteria. The sample wasused to study how early-type galaxy observables, including luminosity L,effective radius Ro, surface brightness Io, color,and velocity dispersion σ, are correlated with one another.Measurement biases are understood with mock catalogs that reproduce allof the observed scaling relations and their dependences on fittingtechnique. At any given redshift, the intrinsic distribution ofluminosities, sizes, and velocity dispersions in our sample are allapproximately Gaussian. A maximum likelihood analysis shows thatσ~L0.25+/-0.012,Ro~L0.63+/-0.025, andRo~I-0.75+/-0.02 in the r* band. In addition, themass-to-light ratio within the effective radius scales asMo/L~L0.14+/-0.02 orMo/L~M0.22+/-0.05o, and galaxies withlarger effective masses have smaller effective densities:Δo~M-0.52+/-0.03o. Theserelations are approximately the same in the g*, i*, and z* bands.Relative to the population at the median redshift in the sample,galaxies at lower and higher redshifts have evolved only little, withmore evolution in the bluer bands. The luminosity function is consistentwith weak passive luminosity evolution and a formation time of about 9Gyr ago.

Bernardi M, Sheth RK, Annis J, Burles S, Eisenstein DJ, Finkbeiner DP, Hogg DW, Lupton RH, Schlegel DJ, SubbaRao M, et al. Early-Type Galaxies in the Sloan Digital Sky Survey. III. The Fundamental Plane. The Astronomical Journal. 2003;125 :1866-1881. Publisher's VersionAbstract

A magnitude-limited sample of nearly 9000 early-type galaxies in theredshift range 0.01<=z<=0.3 was selected from the Sloan DigitalSky Survey (SDSS) using morphological and spectral criteria. Thefundamental plane relation in this sample isRo1.49+/-0.05I-0.75+/-0.01oin the r* band. It is approximately the same in the g*, i*, and z*bands. Relative to the population at the median redshift in the sample,galaxies at lower and higher redshifts have evolved only a little. Ifthe fundamental plane is used to quantify this evolution, then theapparent magnitude limit can masquerade as evolution; once thisselection effect has been accounted for, the evolution is consistentwith that of a passively evolving population that formed the bulk of itsstars about 9 Gyr ago. One of the principal advantages of the SDSSsample over previous samples is that the galaxies in it lie inenvironments ranging from isolation in the field to the dense cores ofclusters. The fundamental plane shows that galaxies in dense regions areslightly different from galaxies in less dense regions.

Eisenstein DJ. Deprojecting Densities from Angular Cross-Correlations. The Astrophysical Journal. 2003;586 :718-722. Publisher's VersionAbstract

We present a model-independent, spherically symmetric density estimatorto be used in the cross-correlation of imaging catalogs with objects ofknown redshift. The estimator is a simple modification of the usualprojected density estimator, with weightings that produce a sphericalaperture rather than a cylindrical one.

Blanton MR, Brinkmann J, Csabai I, Doi M, Eisenstein D, Fukugita M, Gunn JE, Hogg DW, Schlegel DJ. Estimating Fixed-Frame Galaxy Magnitudes in the Sloan Digital Sky Survey. The Astronomical Journal. 2003;125 :2348-2360. Publisher's VersionAbstract

Broadband measurements of flux for galaxies at different redshiftsmeasure different regions of the rest-frame galaxy spectrum. Certainastronomical questions, such as the evolution of the luminosity functionof galaxies, require transforming these inherently redshift-dependentmagnitudes into redshift-independent quantities. To prepare to addressthese astronomical questions, investigated in detail in subsequentpapers, we fit spectral energy distributions (SEDs) to broadbandphotometric observations in the context of the optical observations ofthe Sloan Digital Sky Survey (SDSS). Linear combinations of fourspectral templates can reproduce the five SDSS magnitudes of allgalaxies to the precision of the photometry. Expressed in theappropriate coordinate system, the locus of the coefficients multiplyingthe templates is planar and, in fact, nearly linear. The resultingreconstructed SEDs can be used to recover fixed-frame magnitudes over arange of redshifts. This process yields consistent results in the sensethat, within each sample, the intrinsic colors of similar type galaxiesare nearly constant with redshift. We compare our results with simplerinterpolation methods and galaxy spectrophotometry from the SDSS. Thesoftware that generates these results is publicly available and easilyadapted to handle a wide range of galaxy observations.Based on observations obtained with the Sloan Digital Sky Survey.