The Data Release of the Sloan Digital Sky Survey-II Supernova Survey
Masao Sako, Bruce Bassett, Andrew C. Becker, Peter J. Brown, Heather Campbell, Rachel Cane, David Cinabro, Chris B. D'Andrea, Kyle S. Dawson, Fritz DeJongh, Darren L. Depoy, Ben Dilday, Mamoru Doi, Alexei V. Filippenko, John A. Fischer, Ryan J. Foley, Joshua A. Frieman, Lluis Galbany, Peter M. Garnavich, Ariel Goobar, Ravi R. Gupta, Gary J. Hill, Brian T. Hayden, Renee Hlozek, Jon A. Holtzman, Ulrich Hopp, Saurabh W. Jha, Richard Kessler, Wolfram Kollatschny, Giorgos Leloudas, John Marriner, Jennifer L. Marshall, Ramon Miquel, Tomoki Morokuma, Jennifer Mosher, Robert C. Nichol, Jakob Nordin, Matthew D. Olmstead, Linda Ostman, Jose L. Prieto, Michael Richmond, Roger W. Romani, Jesper Sollerman, Max Stritzinger, Donald P. Schneider, Mathew Smith, J. Craig Wheeler, Naoki Yasuda, Chen Zheng
aa r X i v : . [ a s t r o - ph . C O ] J a n Submitted to ApJS
Preprint typeset using L A TEX style emulateapj v. 5/2/11
THE DATA RELEASE OF THE SLOAN DIGITAL SKY SURVEY-II SUPERNOVA SURVEY
Masao Sako , Bruce Bassett , Andrew C. Becker , Peter J. Brown , Heather Campbell , Rachel Cane ,David Cinabro , Chris B. D’Andrea , Kyle S. Dawson , Fritz DeJongh , Darren L. Depoy , Ben Dilday ,Mamoru Doi , Alexei V. Filippenko , John A. Fischer , Ryan J. Foley , Joshua A. Frieman ,Lluis Galbany , Peter M. Garnavich , Ariel Goobar , Ravi R. Gupta , Gary J. Hill ,Brian T. Hayden , Ren´ee Hlozek , Jon A. Holtzman , Ulrich Hopp , Saurabh W. Jha ,Richard Kessler , Wolfram Kollatschny , Giorgos Leloudas , John Marriner , Jennifer L. Marshall ,Ramon Miquel , Tomoki Morokuma , Jennifer Mosher , Robert C. Nichol , Jakob Nordin ,Matthew D. Olmstead , Linda ¨Ostman , Jose L. Prieto , Michael Richmond , Roger W. Romani ,Jesper Sollerman , Max Stritzinger , Donald P. Schneider , Mathew Smith , J. Craig Wheeler ,Naoki Yasuda , and Chen Zheng Submitted to ApJS
ABSTRACTThis paper describes the data release of the Sloan Digital Sky Survey-II (
SDSS-II ) Supernova Sur-vey conducted between 2005 and 2007. Light curves, spectra, classifications, and ancillary data arepresented for 10,258 variable and transient sources discovered through repeat ugriz imaging of
SDSS
Stripe 82, a 300 deg area along the celestial equator. This data release is comprised of all transientsources brighter than r ≃ . SDSS-III BOSS spectrographs. Photometric classifications are pro-vided for the candidates with good multi-color light curves that were not observed spectroscopically.From these observations, 4607 transients are either spectroscopically confirmed, or likely to be, su-pernovae, making this the largest sample of supernova candidates ever compiled. We present a newmethod for SN host-galaxy identification and derive host-galaxy properties including stellar masses,star-formation rates, and the average stellar population ages from our SDSS multi-band photometry.We derive
SALT2 distance moduli for a total of 1443
SN I a with spectroscopic redshifts as well as pho-tometric redshifts for a further 677 purely-photometric
SN I a candidates. Using the spectroscopicallyconfirmed subset of the three-year
SDSS-II SN I a sample and assuming a flat
ΛCDM cosmology, wedetermine Ω M = 0 . ± .
093 (statistical error only) and detect a non-zero cosmological constant at5.7 σ . Subject headings: cosmology: observations — supernovae: general — surveys Department of Physics and Astronomy, University of Penn-sylvania, 209 South 33rd Street, Philadelphia, PA 19104, USA African Institute for Mathematical Sciences, Muizenberg,7945, Cape Town, South Africa South African Astronomical Observatory, Cape Town,South Africa Department of Mathematics and Applied Mathematics,University of Cape Town, Cape Town, South Africa Department of Astronomy, University of Washington, Box351580, Seattle, WA 98195, USA Department of Physics & Astronomy, Texas A&M Univer-sity, College Station, TX 77843, USA Institute of Cosmology and Gravitation, Dennis SciamaBuilding, Burnaby Road, University of Portsmouth,Portsmouth, PO1 3FX, UK Institute of Astronomy, University of Cambridge, MadingleyRoad, Cambridge CB3 0HA, UK Wayne State University, Department of Physics and Astron-omy, Detroit, MI 48202, USA Department of Physics and Astronomy, University of Utah,Salt Lake City, UT 84112, USA Center for Particle Astrophysics, Fermi National Accelera-tor Laboratory, P.O. Box 500, Batavia, IL 60510, USA Las Cumbres Observatory Global Telescope Network, 6740Cortona Dr. Suite 102, Goleta, CA 93117, USA Institute of Astronomy, Graduate School of Science, TheUniversity of Tokyo, 2-21-1 Osawa, Mitaka, Tokyo 181-0015,Japan Kavli Institute for the Physics and Mathematics of theUniverse (Kavli IPMU, WPI), Todai Institutes for AdvancedStudy, the University of Tokyo, Kashiwa 277-8583, Japan Department of Astronomy, University of California, Berke- ley, CA 94720-3411, USA Astronomy Department, University of Illinois at Urbana-Champaign, 1002 West Green Street, Urbana, IL 61801 USA Department of Physics, University of Illinois Urbana-Champaign, 1110 W. Green Street, Urbana, IL 61801 USA Department of Astronomy and Astrophysics, The Univer-sity of Chicago, 5640 South Ellis Avenue, Chicago, IL 60637,USA Kavli Institute for Cosmological Physics, The University ofChicago, 5640 South Ellis Avenue, Chicago, IL 60637, USA Institut de F´ısica d’Altes Energies, E-08193 Bellaterra(Barcelona), Spain CENTRA Centro Multidisciplinar de Astrof´ısica, InstitutoSuperior T´ecnico, Av. Rovisco Pais 1, 1049-001 Lisbon, Portu-gal Department of Physics, University of Notre Dame, 225Nieuwland Science Hall, Notre Dame, IN 46556, USA Oskar Klein Centre, Stockholm University, SE-106 91Stockholm, Sweden Department of Physics, Stockholm University, SE-106 91Stockholm, Sweden Argonne National Laboratory, 9700 South Cass Avenue,Lemont, IL 60439, USA McDonald Observatory, University of Texas at Austin,Austin, TX 7871, USA Lawrence Berkeley National Laboratory, 1 Cyclotron RoadMS 50B-4206, Berkeley, CA 94720, USA Department of Astronomy, Princeton University, Prince-ton, New Jersey 08544, USA Department of Astronomy, MSC 4500, New Mexico StateUniversity, P.O. Box 30001,Las Cruces, NM 88003, USA
Sako et al. INTRODUCTION
In response to the astounding discovery of the late-time acceleration of the expansion rate of the Uni-verse (Riess et al. 1998; Perlmutter et al. 1999), a num-ber of large-scale supernova ( SN ) surveys were launched.These experiments included programs to observe lowredshift SN such as the Nearby Supernova Factory(Aldering et al. 2002), the Carnegie Supernova Project(Hamuy et al. 2006), and the Center for Astrophysics SN Program (Hicken et al. 2009). At higher redshift, newsurveys included
ESSENCE (Miknaitis et al. 2007), theSupernova Legacy Survey (
SNLS ; Astier et al. 2006), anddedicated
HST observations by Riess et al. (2007). At in-termediate redshifts, the Sloan Digital Sky Survey (
SDSS ;York et al. 2000) bridged the gap between the local anddistant SN searches by providing repeat observations ofa 300 deg stripe of sky at the equator (known as Stripe82) and discovered thousands of Type Ia SN ( SN I a) overthe redshift range 0 . < z < . SDSS SN
Survey. This search was adedicated multi-band, magnitude-limited survey, whichprovided accurate multi-color photometry for tens ofthousands of transient objects, all with a well-determineddetection efficiency. The data have lead to precise mea-surements of the SN rate as a function of redshift, en-vironment, and SN type (Dilday et al. 2008, 2010a,b;Smith et al. 2012; Taylor et al. 2014), and have leadto important new constraints on cosmology with de-tailed studies of systematic uncertainties (Kessler et al.2009a; Sollerman et al. 2009; Lampeitl et al. 2010a;Betoule et al. 2014). The large survey volume andhigh cadence have enabled early discoveries of rareevents (Phillips et al. 2007; McClelland et al. 2010; Universitaets-Sternwarte Munich, Scheiner Str 1, D 81679Munich, Germany MPI f. Extraterrestrische Physik, Giessenbachstrasse, D85741 Garching, Germany Department of Physics and Astronomy, Rutgers, the StateUniversity of New Jersey, 136 Frelinghuysen Road, Piscataway,NJ 08854, USA Institut f¨ur Astrophysik, Universit¨at G¨ottingen, Friedrich-Hund Platz 1, 37077 G¨ottingen, Germany Dark Cosmology Centre, Niels Bohr Institute, Universityof Copenhagen, Juliane Maries Vej 30, 2100 Copenhagen,Denmark Instituci´o Catalana de Recerca i Estudis Avan¸cats, E-08010Barcelona, Spain Space Sciences Lab, University of California Berkeley, 7Gauss Way, Berkeley, CA 94720, USA Department of Astrophysical Sciences, Princeton Univer-sity, Princeton, NJ 08540, USA School of Physics and Astronomy, Rochester Institute ofTechnology, Rochester, New York 14623, USA Department of Physics, Stanford University, Palo Alto, CA94305, USA Department of Astronomy, Stockholm University, SE-10691 Stockholm, Sweden Department of Physics and Astronomy, Aarhus University,Denmark Department of Astronomy and Astrophysics, The Pennsyl-vania State University, University Park, PA 16802, USA Institute for Gravitation and the Cosmos, The Pennsylva-nia State University, University Park, PA 16802, USA Department of Physics, University of the Western Cape,Cape Town, 7535, South Africa Department of Astronomy, University of Texas at Austin,Austin, TX 78712, USA McCully et al. 2013), as well as detailed statistical stud-ies of normal events (Hayden et al. 2010a,b).The extensive, well-calibrated
SDSS galaxy cataloghas also helped revolutionize the study of
SN I a andthe dependence on their host-galaxy properties. Forexample, Lampeitl et al. (2010b) and Johansson et al.(2013) showed a clear correlation between SN Hubbleresiduals and the stellar mass of the host. The ori-gin of this correlation remains unclear, but Gupta et al.(2011) found evidence for the correlation being due tothe age of the stellar population (cf. Johansson et al.2013), while D’Andrea et al. (2011) found the correla-tion was likely related to the gas-phase metallicity us-ing a sub-sample of star-forming
SDSS host galaxies.Hayden et al. (2013) have used the fundamental metal-licity relation (Mannucci et al. 2010) to further reducethe Hubble residuals, suggesting again that metallic-ity is the underlying physical parameter responsible forthe correlation. Galbany et al. (2012), however, did notdetect an obvious correlation between Hubble residu-als and distance to the SN from the center of the hostgalaxy, as might be expected due to metallicity gradi-ents, but the are not as sensitive as the more directmetallicity measurements presentend in D’Andrea et al.(2011). Galbany et al. (2012) also found that extinctionand SN I a color decrease with increasing distance fromthe center of the host, and that the average SN light curveshape differs significantly in elliptical and spiral galax-ies as seen in many previous studies (Hamuy et al. 1996;Gallagher et al. 2005; Sullivan et al. 2006). Xavier et al.(2013) found that SN I a properties in rich galaxy clus-ters are, on average, different from those in passive fieldgalaxies, possibly due to differences in age of the stellarpopulations. Finally, Smith et al. (2014) studied the ef-fects of weak gravitational lensing on the
SDSS-II SN I adistance measurements.The SN spectra presented in this data release are acollection of data from 11 different telescopes and in-cludes some spectra taken to determine galaxy prop-erties long after the SN had faded. We did not at-tempt a detailed spectroscopic analysis of the full sam-ple beyond transient classification and redshift measure-ment, but subsets of the data were previously published(Zheng et al. 2008; Konishi et al. 2011a; ¨Ostman et al.2011) and analyzed to quantitatively measure spec-tral features (Konishi et al. 2011b; Nordin et al. 2011a,b;Foley et al. 2012).Since spectra were not obtained for all discovered tran-sients (as is true for all SN surveys), Sako et al. (2011)analyzed the light curves of the full sample of variableobjects and identified ∼ SN I acandidates with quantitative estimates for the classi-fication efficiency and sample purity. In the absenceof a SN spectrum, the identification and placement of SN I a on a Hubble diagram is greatly aided by a knowl-edge of the host-galaxy redshift. Many host galaxyspectroscopic redshifts were measured by the
SDSS-I and
SDSS-II surveys, but the
SDSS-III (Eisenstein et al.2011) Baryon Oscillation Spectroscopic Survey (
BOSS ;Dawson et al. 2013) ancillary program (Olmstead et al.2013) provided redshifts for most of the observable SN host galaxies. Hlozek et al. (2012) and Campbell et al.(2013) presented Hubble diagrams using photometric SN DSS-II SN Data Release 3classification and host redshifts, and demonstrated thatstatistically competitive cosmological constraints can beobtained with limited spectroscopic follow up of ac-tive SN candidates. The SDSS work on photometricidentification represents an important example analy-sis for ongoing and future large surveys, such as Pan-
STARRS (Scolnic et al. 2013a; Rest et al. 2013),
DES (Bernstein et al. 2012) and
LSST (Tyson 2002), wherefull spectroscopic follow up of all active SN candidateswill be impractical.This paper presents a catalog of 10,258 SDSS sourcesthat were identified as part of the
SDSS SN search. Theimages and object catalogs provided herein were pro-duced by the standard
SDSS survey pipeline as presentedin
SDSS
Data Release 7 (Abazajian et al. 2009). Ourtransient catalog is presented as a machine readable ta-ble in the on-line version of this paper, and the format ofthe catalog is described in Table 1. Detailed descriptionsof general properties ( § § SN I a light curve fits ( §
7) for selected sources, and hostgalaxy identifications ( §
8) are given along with trun-cated tables of catalog data. The photometric data isdescribed in §
5. Many sources have associated opticalspectra, which are described and cataloged in § SDSS-II SUPERNOVA SURVEY
The
SDSS-II SN data were obtained during three-month campaigns in the Fall of 2005, 2006, and 2007as part of the extension of the original
SDSS . A smallamount of engineering data were collected in 2004(Sako et al. 2005), but are not included in this paper,since the cadence and survey duration were not ade-quate for detailed light curve studies. The
SDSS telescope(Gunn et al. 2006) and imaging camera (Gunn et al.1998) produce photometric measurements in each of the ugriz
SDSS filters (Fukugita et al. 1996) spanning thewavelength range of 350 to 1000 nm. The most usefulfilters for observing
SDSS SN , however, are g , r , and i because the SN are difficult to detect in u and z exceptat low redshifts ( z . . SN I a) due to the relativelypoor throughput of those filters.The
SDSS SN survey is a “rolling search”, where a por-tion of the sky is repeatedly scanned to discover new SN and to measure the light curves of the ones previouslydiscovered. The survey observed Stripe 82, which is 2.5 ◦ wide in Declination between Right Ascension of 20 h and04 h . The camera is operated in drift scan mode withall filters being observed nearly simultaneously with afixed exposure time of 55 seconds each. Full coverage ofStripe82 was obtained in two nights (with offset camerapositions), but the average cadence was approximatelyfour nights because of inclement weather and interfer-ence from moonlight. The coverage and cadence of thesurvey is shown in Figure 1. The repeated scans wereused by Annis et al. (2011) to produce and analyze deepcoadded images. The survey is sensitive to SN I a be-yond a redshift of 0.4, but beyond a redshift of 0.2 thecompleteness, and the ability to obtain high-quality pho-tometry, deteriorates.The
SDSS camera images were processed by the
SDSS imaging software (Stoughton et al. 2002) and SN were identified via a frame subtraction technique(Alard & Lupton 1998). Objects detected after framesubtraction in two or more filters were placed in a m b e r o f s ca n s
82N 82S −50 0 50 m ea n ca d e n ce RA (deg) m b e r o f s ca n s
82N 82S −50 0 50 m ea n ca d e n ce RA (deg) m b e r o f s ca n s
82N 82S −50 0 50 m ea n ca d e n ce RA (deg)
Figure 1.
Number of scans versus right ascension (shown in de-grees) of the
SDSS SN equatorial stripe (Stripe 82) is shown alongwith the mean cadence for each year (2005-2007) of the survey. Thecoverage in right ascension increased slightly as the template imagecoverage increased while the mean cadence was approximately fourdays for all three observing seasons. database of detections. These detected objects werescanned visually and were designated candidates if theywere not obvious artifacts. Spectroscopic measurementswere made for promising candidates depending on theavailability and capabilities of telescopes. The candidateselection and spectroscopic identification have been de-scribed by Sako et al. (2008). In three observing seasons,the
SDSS-II SN
Survey discovered 10,258 new variableobjects and spectroscopically identified 500
SN I a and 81core-collapse SN ( CC SN ). SN CANDIDATE CATALOG
Table 1 describes the format of the
SDSS-II SN cat-alog, which includes information on the 10,258 sources Sako et al.
Number of epochs with S/N>5 N u m be r o f S N C and i da t e s All CandidatesType>104
Figure 2.
The distribution of number of epochs observed per SNis shown. An epoch consists of one night of observation in all 5SDSS filters without any requirement that there was a detectablesignal in any of the filters. There are typically about 20 epochs inan observing season, but a small fraction of SN lie in the overlapregion and are observed with twice the cadence or up to 40 timesper season. detected on two or more nights. The full catalog is madeavailable online; a small portion is reproduced as an ex-ample in Table 2.General photometric properties include the J2000coordinates of the SN candidate, the number ofepochs detected by the search pipeline ( Nsearchepoch )and final photometry pipeline above
S/N > NepochSNR5 ), and r -band magnitude ( Peakrmag ) and
MJD ( MJDatPeakrmag ) of the brightest measurement.We show the distribution of
NepochSNR5 for all candi-dates in Figure 2 as an indication of the general qualityof the light curves.We provide the heliocentric redshift ( zspecHelio ) anduncertainty ( zspecerrHelio ) when spectroscopic mea-surements are available. The source of the redshift isfrom the host galaxy spectrum or, if the host galaxy red-shift is not known, from the SN spectrum. More de-tails on the spectra are given in §
6. The number ofspectra available as part of this Data Release are givenas nSNspec (the number of SN spectra) and nGALspec (the number of host galaxy spectra) in the catalog. Thegalaxy spectra include cases where the galaxy spectrumis obtained from the SN spectroscopic observation butwith an aperture chosen to enhance the galaxy light andcases where a spectrum was taken when the SN was nolonger visible for the purpose of measuring the galaxyredshift and possibly other galaxy properties. Galaxyspectra that were taken with the SDSS spectrograph(Smee et al. 2013) are not included in these totals, but objIDHost gives
SDSS DR
SDSS database. Spectra as part of the
SDSS-III BOSS pro-gram are also not included in these totals. They arediscussed in Campbell et al. (2013) and Olmstead et al.(2013), but their redshifts are listed under zspecHelio .Finally, we provide the
CMB -frame redshifts and uncer-tainties in zCMB and zerrCMB , respectively.Some sources (most of the spectroscopically identified SN ) were assigned a standard name by the IAU ; the nameis listed for those sources that have been assigned one.The peak r -band magnitude observed is plotted versus Redshift P ea k O b s e r v ed M agn i t ude (r- band ) Not SNIaSNIa
Figure 3.
The peak r -band magnitude observed is shown as afunction of redshift. Black points shown are for all candidatesclassified as SN I a. All other SN candidates are shown in blue. redshift in Figure 3.The candidates are classified according to their lightcurves and spectra (when available), and the results ofthe classification are shown in Table 2. Visual scanningremoved most of the artifacts, so almost all of the objectsin the catalog are variable astronomical sources, some ofwhich are only visible for a limited period of time (for ex-ample, supernovae). The multi-night requirement elimi-nates rapidly moving objects, which are primarily main-belt asteroids. A summary of the number of objects ineach classification is shown in Table 3. The classification“Unknown” means that the light curve was too sparseand/or noisy to make a useful classification, “Variable”means that the source was observed in more than one ob-serving season, and “AGN” means that an optical spec-trum was identified as having features associated with anactive galaxy, primarily broad hydrogen emission lines.The other categories separate the source light curves into3 SN types: Type II, Type Ibc (either Ib or Ic), andType Ia. A prefix “p” indicates a purely-photometrictype where the redshift is unknown and that the identi-fication has been made with the photometric data only.A prefix “z” indicates that a redshift is measured fromits candidate host galaxy and the classification uses thatredshift as a prior. The SN classifications without a prefixare made based on a spectrum (including a few non- SDSS spectra). The Type Ib and Ic spectra identifications areshown separately. The “
SN I a?” classification is basedon a spectrum that suggests a
SN I a but is inconclusive.The details and estimated accuracy of the classificationscheme are given in the next section.Some of the SN candidates in the catalog have associ-ated notes. Notes indicate SN where the typing spectrumwas obtained by other groups (and is not included in the SDSS data release) and indicate SN candidates that mayhave peculiar features. The bulk of the spectroscopicallyidentified SN I a are consistent with normal
SN I a features,but a few were identified as having some combination ofpeculiar spectral and light curve features. We did notsearch for these peculiar features in a systematic way,but we have noted the likely peculiar features that werefound. Some
SN I a have poor fits to the
SN I a light curvemodel or unlikely parameters for normal
SN I a, but wehave not noted these, preferring to just present the fit pa-rameters. Table 4 describes the codes that may appearDSS-II SN Data Release 5in the notes column (item 136) of Table 1. PHOTOMETRIC CLASSIFICATION
This section describes our method for photometric clas-sification of the SN candidates. First, we reject likelynon- SN events as those showing variability over two ormore seasons. The exact nature of these sources is notknown, but the majority are most likely variable starsand active galactic nuclei. A total of 3225 are identifiedas “Variable” in Table 2.All remaining candidates showed variability duringonly a single season and are therefore viable SN can-didates. Their light curves were then analyzed withthe Photometric SN IDentification (
PSNID ) software(Sako et al. 2011), first developed for spectroscopic tar-geting and subsequently extended to identify and an-alyze photometric
SN I a samples. In short, the soft-ware compares the observed photometry against a grid of
SN I a light curve models and core-collapse SN ( CC SN )templates, and computes the Bayesian probabilities ofwhether the candidate belongs to a Type Ia, Ib/c, orII SN . The technique is similar to that developed byPoznanski et al. (2007), except that we subclassify the CC SN into Type Ib/c and II . Extensive tests and tun-ing were performed using the large (but still limited)sample of spectroscopic confirmations from SDSS-II andsimulations as described in Sako et al. (2011). Thelight curve templates used in the analysis presented hereare the same as those from Sako et al. (2011).
PSNID and the templates are now part of the
SNANA package(Kessler et al. 2009b).The Bayesian probabilities are useful because they rep-resent the relative likelihood of SN types, whereas thebest-fit minimum reduced χ ( χ r ), or more precisely thefit probability P fit , provides an absolute measure of thelikelihood. The combination of the Bayesian probability( P Ia ) and the goodness-of-fit ( P fit ) provides reliable clas-sification of SN I a candidates. The expected level of con-tamination and efficiency can be estimated from eitherlarge datasets or simulations. Sako et al. (2011) usedthis method to identify
SN I a candidates from
SDSS-II .The
SN I a classification purity and efficiency were esti-mated to be 91% and 94%, respectively. The one majordrawback of this techinique, however, was the generalunreliability of classifying
CC SN .To make further improvements, we developed an ex-tension to
PSNID that uses the Bayesian classificationdescribed above as an initial filter, but subsequently re-fines the classification using a kd-tree nearest-neighbor( NN ) technique. We call this method PSNID/NN , and itis based on the fact that different SN types populate a dis-tinct region in extinction, light-curve shape, and redshiftparameter space when fit to an SN I a model. This is il-lustrated in Figure 4.
SN I b/c are generally redder (large A V ) and they fade more rapidly (large ∆ m ( B )) com-pared to SN I a. SN II , on the other hand, have broad, flatlight curves (small ∆ m ( B )). As described below, thismethod makes substantial improvements to both SN I aand
CC SN classification.In this method, every SN in the data sample is com-pared against a training set and the most likely type isdetermined from the statistics of its neighbors in a multi-dimensional parameter space. Ideally, the training set isa large, uniform, and unbiased sample of spectroscopi- cally confirmed SN , but such training sets do not existat the low-flux limit of the SDSS-II SN sample. Our cur-rent implementation, therefore, uses simulated SN from SNANA . The simulation is based on well-measured
CC SN template light curves, which are used to simulate eventsof different magnitudes and redshifts. However, the un-derlying library is small (only 42
CC SN template lightcurves), and adequacy of this sample size has yet tobe rigorously verified. We simulated 10 seasons worthof SN candidates using a mix of SN I a, SN I b/c, and
SN II identical to that used in the SN Classification Chal-lenge (Kessler et al. 2010b,c). For each SN candidate inthe data sample, we calculate Cartesian distances in 3-dimensional parameter space ( A V , ∆ m ( B ), z ) to eachsimulated SN (labeled i ) using the following formula: d = c z ( z SN − z i ) + c ∆ m (∆ m , SN − ∆ m ,i ) + c A V ( A V, SN − A V,i ) , (1)where c z , c ∆ m , and c A V are coefficients determinedand optimized using simulations for both the data andtraining sets. The classification probabilities are deter-mined by counting the numbers of SN I a, SN I b/c, and
SN II in the training set that are within a certain dis-tance d max . Since this distance is degenerate with theoverall normalization of the other three coefficients, weset d max = 1 .
0. The optimized set of coefficients are c z = 160, c ∆ m = 60, and c A V = 10 assuming d max = 1.For each SN candidate in the data sample we countthe number of simulated SN from each type N type within d SN < d max . The nearest-neighbor probabilities P NN , type are then determined using, P NN , type = N type N Ia + N Ibc + N II . (2)The final classification is performed using the Bayesian,nearest-neighbor, and fit probabilities. For a candidateto be a photometric SN I a candidate, we require, • P Ia > P Ibc and P Ia > P II • P NN , Ia > P NN , Ibc and P NN , Ia > P NN , II • P fit ≥ .
01 for
SN I a model • Detections at − ≤ T rest ≤ +5 days and +5
Regions occupied by SN Ia (black), SN Ibc (red) and SN II (blue) in ∆ m ( B ) – A V space in different redshift slices for asimulated SDSS-II SN Survey. The panels are z < . . < z < . . < z < . . < z < . z > . z (bottom right). available. In these cases, the candidates are classified asz SN I a, z
SN I bc, or z
SN II in Table 2. Otherwise, we use aflat redshift prior and the candidates are denoted p
SN I a,p
SN I bc, or p
SN II .All candidates that do not meet any of the criteriaabove are declared “unknown”. The statistics of the SN candidate classification are shown in Table 3. Simulationresults are shown in Figure 5 where we compare classifi-cation performance between the Bayesian-only methodand with the nearest-neighbor probabilities. For theBayesian-only method, the SN I a classification figure-of-merit (defined as the product of the efficiency and purity)has a very broad maximum when we require P Ia > P Ia > P Ia > .
99) with the Bayesian-only method. The full sum-mary of efficiencies and purities of classification of all SN types with flat- z and spec- z priors is listed in Table 5. PHOTOMETRY
Light curves are constructed using the Scene ModelingPhotometry software (
SMP ; Holtzman et al. 2008).
SMP assumes that the pixel data can be described by the sumof a point source that is fixed in space but varying inmagnitude with time, a galaxy background that is con-stant in time but has an arbitrary spatial distribution,and a sky background that is constant over a wider areabut varies in brightness at each observation. The galaxybackground is parameterized as an arbitrary amplitudeon a 15 ×
15 grid of pixels of size 0 . ′′ . The fitting pro-cess accounts for the variations in point spread function( PSF ) to model the distribution of light for each nightof observation. The SN magnitudes and SDSS referenceDSS-II SN Data Release 7 . . . e ff i c i e n c y , pu r it y , F o M P Ia (Bayesian)Bayesian only (flat z) 0.2 0.4 0.6 0.8 1 . . . e ff i c i e n c y , pu r it y , F o M P Ia (Bayesian + NNType)Bayesian+NN (flat z)0.2 0.4 0.6 0.8 1 . . . e ff i c i e n c y , pu r it y , F o M P Ia (Bayesian)Bayesian only (spec z) 0.2 0.4 0.6 0.8 1 . . . e ff i c i e n c y , pu r it y , F o M P Ia (Bayesian + NNType)Bayesian+NN (spec z) Figure 5.
The SN Ia photometric classification efficiency (black), purity (red), and figure of merit (product of the efficiency and purity;blue) as a function of the P Ia probability cut for simulated SDSS-II SN data. The top panels show results from Bayesian-only (left) andwith the nearest-neighbor extension ( PSNID/NN ) for a flat redshift prior. The bottom panels show the same for a spectroscopic redshiftprior. We required log( P fit ) > − .
0. Note that the purity using the Bayesian-only method is never above ∼ stars on the same image are measured simultaneouslyusing the same PSF so the SN magnitudes are measuredrelative to a calibrated SDSS star catalog.A complete set of light curve photometric data for all10,258 SN candidates is given on the SDSS
Data Releaseweb page (SDSS 2013). The format of the data is de-scribed on the web page and is the same as the previouslyreleased first-year data sample (Holtzman et al. 2008).The magnitudes quoted in these data files, and elsewherein this paper, are the
SDSS standard inverse hyperbolicsine magnitudes defined by Lupton et al. (1999). Mag-nitudes are given in the
SDSS native system and differfrom the AB system by an additive constant given in § AB -corrected and are expressed in µ J. The magnitudes andfluxes are reported in a way that is consistent with thefirst-year data sample except that the calibration of
SDSS native magnitudes to µ J has changed as described belowin § SMP fitting procedure.
Photometric Uncertainties
A substantial effort has been made to ensure accurateestimates of the uncertainties in the
SDSS light curve fluxmeasurements. An important feature of
SMP is that itworks on the original images ( i.e. , without resamplingpixels) so that a simple propagation of pixel-by-pixel un-certainties from photo-electron statistics offers a robustestimate of the photometric uncertainty. The galaxy model is remapped for each image, but the galaxy iswell measured in the reference images and the error isalmost always subdominant. In addition to the pixelstatistical uncertainty,
SMP computes a “frame error”that accounts for zero point uncertainty, galaxy modeluncertainty, and systematic sky background uncertainty.The error model was tested by Holtzman et al. (2008)using pre-explosion epochs (known zero flux), artificialsupernovae (computer generated), and real stars. Theconclusion was that the error model provides a good de-scription of the observed photometric errors.After running the
SMP code, we re-examined the pho-tometric errors by examining the light curve residualsrelative to the
SALT2 (Guy et al. 2010) model. Wealso investigated the distribution of residuals using pre-explosion epochs, where the residuals do not depend onthe
SN I a model. For these data the largest errors arisefrom statistical uncertainties and possible errors in mod-eling the galaxy background light. We also examinedthe distribution of residuals relative to the
SALT2 lightcurve model when there was a significant signal (morethan 2 σ above the sky background). In this latter case,uncertainties in the light curve model and zeropoints con-tribute to the width of the distribution of residuals. Forthese tests we used spectroscopically confirmed SN I a ex-cluding peculiar types and further limited the sample tothose SN whose SALT2 fit parameters indicated normalstretch | x | < c < .
2. The g -band distributions of the normalized residuals (residualdivided by the uncertainty) are shown in left-hand pan-els of Figure 6. A normal Gaussian distribution (not a Sako et al.fit) is shown for comparison. While both distributionsare quite close to the expected normal Gaussian, thepre-explosion epoch distribution (upper left) is slightlywider than the curve and the distribution with signifi-cant signal ( σ >
2) is narrower. The normalized residualdistribution for the pre-explosion epochs could be largerif the photometry underestimates the error in modelingthe galaxy background. When there is significant signal,the distribution of normalized residuals could be smallerbecause of an overestimate of the zero-pointing error orthe light curve model uncertainty, which is included inthe estimated errors. Since the zero-point errors are atleast partially correlated between epochs, the fit param-eters (especially the SN color parameter) can absorb partof the zero-point error, and therefore decrease the widthof the distribution of residuals. While the measurementerrors are considerably larger in u and z bands, the dis-tribution of the normalized residuals are similar for theother SDSS filters, indicating that the error estimates areapproximately correct.Based on these distributions, we adjusted the errorsaccording to the prescription σ ′ = q σ + c f (3)The constant c f was adjusted to result in an rms ofunity for the pre-explosion epoch distributions. Thesesmall adjustments are within the errors quoted byHoltzman et al. (2008). The values used for the erroradjustments for all five filters are shown in Table 6. Theresulting g -band distributions of normalized residuals areshown on the right-hand side of Figure 6. Our choice ofthe form in Equation (3) also slightly reduces the widthof the distribution of residuals with σ >
2. We did notattempt additional modifications to the errors to bringthe σ > u -band, where it is common to have many pointsmeasured with large errors. The overall SALT2 lightcurvefit mean confidence level (derived from the χ /dof) is in-creased from 0.28 to 0.57 as a result of this change.We also observe a small, but statistically significantoffset in the mean residual of the pre-explosion epochs.The largest offset was found for r -band where the off-set was 0 . σ , where σ is the width of the normalizeddistribution. We did not correct this offset because wewere uncertain whether subtracting a constant flux fromall epochs would be an appropriate correction. We diddetermine, however, that adding a constant flux offset toour data had a negligible effect on the SALT2 light curvefit probability.
Star catalog calibration
The star catalog calibration is discussed in detail byBetoule et al. (2013), where the
SDSS stellar photome-try calibration is described in detail and the
SDSS pho-tometry is compared with the Supernova Legacy Survey(
SNLS ) photometry. The starting point for the
SDSS SN calibration is a preliminary version of the Ivezi´c et al.(2007) star catalog that was used for
SMP photometry inHoltzman et al. (2008). This catalog uses the stellar lo- cus to calibrate the stellar colors but relies on photometryfrom the
SDSS
Photometric Telescope ( PT ) to establishthe relative zeropoint for r -band. As explained in detailby Betoule et al. (2013), there is a significant flat-fieldingerror in the PT photometry, leading to a photometry thatwas biased as a function of declination. We determinedcorrections to the Ivezi´c et al. (2007) star catalog using SDSS
Data Release 8 (Aihara et al. 2011), whose cali-bration is based on the method of Padmanabhan et al.(2008). This method, the so-called “Ubercal” method,re-determines the nightly zeropoints based only on theinternal consistency of the 2.5 m telescope observations.Our adjustments to the stellar photometry were typicallywithin a range of 2%, but corrections of up to 5% weremade in the u -band. The corrections improved the agree-ment with the SNLS photometry. Instead of recomput-ing the SN magnitudes relative to the new star catalog,we simply applied the corrections to the SN magnitudesfound using the Ivezi´c et al. (2007) catalog.Neither the star catalog of Ivezi´c et al. (2007) (basedon the stellar locus) nor SDSS
Data Release 8 attemptsto improve the absolute calibration of
SDSS photom-etry. The photometry is tied to an absolute scaleby BD+17 ◦ SDSS filter response curves (Doi et al. 2010) and the
HST standard spectra (Bohlin 2007) given in the
HST
CAL-SPEC database (CALSPEC 2006). When the syntheticphotometry of these standards is compared to the
SDSSPT photometry, we obtain an absolute calibration, whichis expressed as “ AB Offsets” from the nominal
SDSS cal-ibration (see Oke & Gunn 1983 for a description of the AB magnitude system). The differences between our cur-rent results and those of Holtzman et al. (2008) are thatwe have: 1) used the recently published SDSS filter re-sponse curves, 2) used more recent
HST spectra, and 3)re-derived the PT to 2.5 m telescope photometric trans-formation, including corrections for the recently discov-ered non-uniformity of the PT flat field. Details of AB system calibration may be found in Betoule et al. (2013).Table 7 lists the AB offsets to be applied to the SDSS SN data. We use the average of three solar analogs (P041C,P177D, and P330E) because these stars are similar incolor to the stars used to determine the (assumed) linearcolor transformation between the PT and 2.5 m telescope.The uncertainty is calculated from the dispersion of theresults for the solar analogs. The value determined forBD+17 ◦ AB offsetspresented here and Table 1 of Holtzman et al. (2008) isthe u -band offset with ∆ AB ∼ .
03, which differs pri-marily because of the different filter response curve for u -band, as discussed in detail by Doi et al. (2010).It is important to note that the SN light curve pho-tometry is given in the SDSS natural system – the samesystem that is used for all the
SDSS data releases. The AB offsets must be added to the SN light curve magni-tudes in order to place them on a calibrated AB system. u -band uncertainties There has been some concern in the literature aboutthe accuracy of the u -band photometry. The observa-tions reported by Jha et al. (2006), for example, used aDSS-II SN Data Release 9 -4 -2 0 2 4 Normalized Residual F r equen cy / . Early pointsBefore adjustmentMean=0.07Rms=1.15 -4 -2 0 2 4
Normalized Residual F r equen cy / . Early pointsAfter adjustmentMean=0.04Rms=1.00 -4 -2 0 2 4
Normalized Residual F r equen cy / . Significant pointsBefore adjustmentMean=0.03Rms=0.93 -4 -2 0 2 4
Normalized Residual F r equen cy / . Significant pointsAfter adjustmentMean=0.02Rms=0.86
Figure 6.
The normalized residuals for SN Ia light curve fits to the g -band data before (left-hand panels) and after (right-hand panels)the adjustment described in the text. The top panels use data prior to the SN explosion (“Early points”), which is therefore independentof the light curve model. The bottom panels show the residuals for points where the detected flux is two or more standard deviations abovebackground (“Significant points”). diverse set of telescopes and cameras and were not sup-ported by a large, uniform survey like SDSS . For thesereasons, one might question whether there are substan-tial errors in the u -band calibration. For example, inthe SNLS u band are de-weighted. The quality ofthe SDSS u -band data benefits greatly from an extensive,accurate star catalog of SDSS
Stripe 82. For example,Figure 7 shows the variations in stellar magnitudes inthe Ivezi´c et al. (2007) catalog, showing a repeatabilityof 0.03 mag over most of the magnitude range. Thesesecondary stars, which are the
SMP photometric refer-ences, are measured several times during photometricconditions so that the calibration error is typically 0.01to 0.02 magnitude per star. The
SMP normally uses atleast three calibration stars in u -band so that the typi-cal zero point error (which is included in the SMP frameerror) is comparable to the overall u -band scale error of0.0089 (Betoule et al. 2013).A check of SDSS SN photometry is described inMosher et al. (2012), who compared
SDSS and CarnegieSupernova Project (Contreras et al. 2010) measurementson a subset of
SN I a observed by both surveys. Forthe 32 u -band observations, they find agreement of0 . ± .
014 mag, and comparable agreement in theother bands. SPECTRA
SDSS SN spectra were obtained with the Hobbey-Eberly Telescope (
HET ), the Apache Point Observatory3.5m Telescope (
APO ), the Subaru Telescope, the 2.4-m
14 16 18 20
Average u-band magnitude R m s m agn i t ude d i ff e r en c e Figure 7.
The rms photometric scatter of repeated measurementsof SDSS Stripe 82 stars in u -band. Hiltner Telescope at the Michigan-Dartmouth-MIT Ob-servatory (
MDM ), the European Southern Observatory(
ESO ) New Technology Telescope (
NTT ), the Nordic Op-tical Telescope (
NOT ), the Southern African Large Tele-scope (
SALT ), the William Herschel Telescope (
WHT ),the Telescopio Nazionale Galileo (
TNG ), the Keck I Tele-scope, and the Magellan Telescope. Table 8 providesdetails of the instrumental configurations used at eachtelescope. These observations resulted in confirmationof 500
SN I a, 19
SN I b/c, and 62
SN II . A total of 1360unique spectra are part of this data release. In manycases, we provide extractions of the SN and host galaxy0 Sako et al. -20 0 20 40 60 SN Phase (days in rest-frame) N u m be r o f S N / da ys Figure 8.
The distribution in time when SN Ia spectra were ob-served relative to peak brightness in B -band. spectra separately. The majority of the SN spectra suf-fer contamination from the host galaxy, and we did notattempt to remove that contamination. Contaminationof the galaxy spectrum by SN light may also be an issuein some of the galaxy spectra.Most SN spectra were taken when the SN candidateswere near peak brightness. The distribution of observa-tion times relative to peak brightness is shown in Fig-ure 8. Of the 889 SN candidates with measured spectra,177 have two or more spectra, and 16 have five or morespectra.The spectra were all observed using long slit spectro-graphs, but they were observed under a variety of condi-tions with the procedures determined by the individualobservers. Some spectra were observed at the parallac-tic angle while other spectra were observed with the slitaligned to pass through both the SN and the host, ornearest, galaxy. The different slit sizes and observingconditions result in slit losses that are not well character-ized for most of the spectra. The spectra were processedby the observers, or their collaborators, using proceduresdeveloped for each particular telescope.The spectra are calibrated to standard star observa-tions, but with the exception of the Keck spectra, thequality of the calibration is not verified. Telluric lines aregenerally removed, but residual absorption features orsky lines may be present. We provide uncertainties for allthe spectra, but the uncertainties are generally limited tostatistical errors. Because of the non-uniformities in thesample, and uncontrolled systematic errors, we cannotmake a general statement about the accuracy of all thespectra. Some subsamples of spectra have been subjectedto detailed analyzes ( ¨Ostman et al. 2011; Konishi et al.2011a,b; Foley et al. 2012) and more detailed informa-tion on corrections and systematic errors can be foundin these references.The SN spectral classification and redshift determi-nation methods are described in Zheng et al. (2008).Briefly, the spectra were compared to template spec-tra and the best matching template spectrum was de-termined. Each spectrum was classified as “None” (nopreferred match, usually because the spectrum was toonoisy), “Galaxy” (spectrum of a normal galaxy withno evidence for a SN ), “AGN” (spectrum of an activegalaxy) or a SN type: “Ia” (Type Ia), “Ia?” (possi- ble Type Ia), “Ia-pec” (peculiar Type Ia), “Ib” (TypeIb), “Ic” (Type Ic), or “II” (Type II). The redshifts aregenerally determined by cross-correlation with templatespectra, but for some of the galaxy redshifts observed in2008 were determined by measuring line centroids. Allredshifts are presented in the heliocentric frame.The list of spectra is displayed in Table 9. Each obser-vation is uniquely specified by the SN candidate ID andspectrum ID . The observing telescope is listed and theclassification of the spectrum described above is listed inthe column labeled “Evaluation”. Separate redshifts aregiven for the galaxy and SN spectra, when available. Themean value of the SN I a redshifts are offset from the hostgalaxy by 0 . ± . SN I aredshift). The offset probably arises from variations inthe SN template spectra that were used to determine the SN redshifts. A similar offset (0 . SN spectrum, the uncertaintyfloor is set to δz = 0 . δz =0 . SDSS and
BOSS spectrographs have uncertainties set by their respectivepipelines as quoted in their catalogs. SN IA SAMPLE AND
SALT2
ANALYSIS
We provide results from light curve fits as a referenceto serve as a check for those who wish to make theirown fits using different methods or selection criteria andfor those less critical applications that can use our lightcurve fits directly. Using the
SNANA version 10.31b pack-age (Kessler et al. 2009b) implementation of the
SALT2SN I a light curve model (Guy et al. 2010), we determinelight curve parameters for two kinds of fits. The first usesfixed spectroscopic redshifts (either from the SN spec-trum or the host galaxy), and fits four parameters: timeof peak brightness ( t ), color ( c ), the shape (stretch) pa-rameter ( x ), and the luminosity scale ( x ). The secondfit ignores spectroscopic redshift (when known) and in-cludes the redshift as a fifth fitted parameter as describedin Kessler et al. (2010a). For comparison, we have alsoused the MLCS A V play similar roles to the SALT2 parameters x and c , respectively.To ensure reasonable fits, we applied selection criteriaas summarized in Table 10. Note that SN I a fits are maderegardless of the SN type classification. The SNANA in-put files for these fits are available on the data releaseweb pages (SDSS 2013).We also placed some requirements on the photomet-ric measurements that were used in the fit. We excludeepochs where
SMP was judged to be unreliable (a photo-metric flag of 1024 or larger) and epochs earlier than15 days or later than 45 days (in the rest frame). Inaddition, 151 epochs in 105 different SN were designatedoutliers based on the inspection of the light curve fitsand were not used to determine the light curve parame-ters. These outlier epochs are included in both the ASCII and
SNANA data releases, and a list of these epochs isincluded in the
SNANA release.DSS-II SN Data Release 11 ✲(cid:0) ✲✁ ✵ ✁ (cid:0)❙✂✄☎✁ ✆✶✲✝✞✵✲✵✞✟✵✞✵✵✞✟✝✞✵✝✞✟✁✞✵▼✠✡☛☞✌☞❉ ③✍✎✏✑ ✒✓✔✕✖✍✎✏✑✗ ✒✘✔✖✍✎✏✑ ✒✙✙✔✖✚ ✥ ✛ ✜✢✜✣✣✤ ✛ ✜✢✦✧✧✤ ①★ ✩ ✜✢✜✧✪✪ ①★✷✛ ✜✢✜✜✫✬ ①★✸➧ ✜✢✜✜✧✤ ➧ ✜✢✜✜✧✣ ➧ ✜✢✜✜✪✜ ➧ ✜✢✜✜✭✫✲✵✞(cid:0) ✲✵✞✁ ✵✞✵ ✵✞✁ ✵✞(cid:0) ✵✞✮ ✵✞✯❙✂✄☎✁ ✰✲✁✲✝✵✝✁▼✠✡☛☞✌☞✱ ❱ ③✍✎✏✑ ✒✓✔✳✖✍✎✏✑✗ ✒✘✓✖✍✎✏✑ ✒✙✙✳✖ ❆✴ ✹ ✜✢✦✤✦✤ ✩ ✦✢✧✦✧✪ ❝➧ ✜✢✜✜✫✪ ➧ ✜✢✜✫✦✪
Figure 9.
Comparisons of the MLCS2k2 and SALT2 light-curvefit parameters for SN Ia from the spectroscopic (SNIa, SNIa?) orphotometric (with host redshift; zSNIa) samples. The top panelshows the light curve shape parameters, MLCS2k2 ∆ versus SALT2 x
1, for the 1259 SN Ia where these parameters were well measured.The solid purple curve is a cubic polynomial fit to the data, withcoefficients as displayed, restricted to points with − < x < − < ∆ < A V versus SALT2 c , for 1265 SN Ia. The solid purple line is a linearregression fit using the Bayesian Gaussian mixture model of Kelly(2007), via the IDL routine linmix err.pro , over the restricteddata range − . < c < . − < A V < Some representative 4-parameter fit results are shownin Table 11 (
SALT2
MLCS
SALT2
SALT2 and
MLCS
SALT2 c is comparedwith MLCS A V and SALT2 x is compared with MLCS
SALT2 and
MLCS
MLCS x = − A V and c is tighter, with just a handful of outliers.There is also a clear color zeropoint offset between the Redshift -0.6-0.4-0.200.20.40.60.81 SA L T c Host redshiftSN redshift
Redshift -8-6-4-202468 Host redshiftSN redshift SA L T x Figure 10.
SALT2 color (top) and x as a function of redshift forthe spectroscopic SN Ia (black) and z host -Ia (red) samples. The z host -Ia sample is noticeably redder (more positive values of c ) thanthe spec-Ia sample, but the x distributions are indistinguishable. fitters, c ≈ − . A V = 0.We have also compared the SALT2 parameters x and c in Figure 10 for the 4-parameter fit sample, showing sep-arately the sample where the redshift is obtained fromthe SN spectrum as opposed to the host galaxy spec-trum. We expect the sample with SN spectra to be bi-ased because of the spectroscopic target selection. Fig-ure 10 shows no evidence of a bias in x but a cleardifference in c , which is consistent with the findings ofCampbell et al. (2013) where the weighted mean SALT2 colors of the spectroscopically-confirmed
SN I a whereslightly bluer than for the whole sample (including manyphotometrically–classified
SN I a). This effect is presum-ably because reddened Type Ia SN were less likely to beselected for spectroscopy. Distance moduli
We have used the results of our 4-parameter
SALT2 fits to compute the distance modulus to the
SDSS-II SN ,excluding those events where the fit parameter uncer-tainty was large ( δt > δx > SALT2 mu(Marriner et al. 2011), which is also part of
SNANA , tocompute the
SALT2 α and β parameters and computedthe distance modulus according to the relationship, µ = − . x − M + αx − βc, (4)2 Sako et al.where µ is the distance modulus, and M = − . α = 0 . ± .
010 and β = 3 . ± .
13. Only the spectroscopically confirmed
SN I a were used to determine these parameters and theintrinsic scatter was assumed to be entirely due to varia-tions in peak B -band magnitude with no color variations(Marriner et al. 2011). Including the photometric SN I asample, we get α = 0 . ± .
009 and β = 2 . ± . SN I a reported in Kessler et al. (2009a) and elsewhere(Lampeitl et al. 2010a). The differences arise from thefollowing changes: • Re-calibration and updated AB offsets(Betoule et al. 2013) • Fitting ugriz instead of gri • For
MLCS • Updated (Guy et al. 2010)
SALT2 model (see Sec-tion § SN previously published in Kessler et al.(2009a), the difference in µ versus redshift is shown inFig. 11 for MLCS
SALT2 . For
MLCS u band has an importanteffect. For SALT2 the difference increases with redshift.
SALT2 versions
The results of the
SALT2 fits depend on the versionof the code used, the spectral templates, and the colorlaw. Our fits use the
SALT2 model as implemented in
SNANA version 10.31b and the spectral templates andcolor law reported in Guy et al. (2010, G10). Most ofthe prior work with the
SDSS sample used the earlierversions of the spectral templates and color law givenin Guy et al. (2007, G07) with the notable exceptionof Campbell et al. (2013), which used G10. For the
SDSS data, the largest differences in the fitted param-eters arises from the difference in the color law betweenG07 and G10. The
SDSS-II - SNLS joint light curve anal-ysis paper on cosmology (Betoule et al. 2014) releases anew version of the
SALT2 model that is based on addingthe full
SDSS-II spectroscopically confirmed SN sampleto the SALT2 training set.Figure 12 shows the different versions of the color lawand the range of wavelengths sampled for each photomet-ric band assuming an
SDSS redshift range of 0 < z < . SN fits for the SALT2 color parameter ( c ), where eachpoint is a particular SN with both fits using the spectraltemplates from G10 but different color laws.Although there is some scatter, the relationship be-tween the two fits can be described approximately by aline. δc = 0 . c + 0 .
00 (5)
Redshift -0.15-0.1-0.0500.050.10.15 ! µ ( T h i s w o r k – K e ss l e r ) (a) Redshift -0.2-0.15-0.1-0.0500.050.1 ! µ ( T h i s w o r k – K e ss l e r ) (b) Figure 11.
The differences in distance modulus between theresults in this paper and the results published in Kessler et al.(2009a) for (a)
MLCS
Wavelength (A) -20246 A tt enua t i on ( m ag ) u g r i zGuy 10Guy 07 Figure 12.
The figure displays the color laws from Guy et al.(2007, dotted line) and Guy et al. (2010, solid line). The horizontallines ( ugriz ) indicate the range for the mean wavelength responseof each filter, respectively, over the redshift range of 0 . < z < . i -band and, at higher redshift,in g -band. We conclude that the G07 color law results in a value ofthe c parameter that is 20% higher than G10 on average.The effects of the differences in the spectral templatesand changes to the
SNANA code are much smaller.
Comparison of
SDSS u-band with model
DSS-II SN Data Release 13 -0.2 0 0.2 0.4 0.6 0.8 c (Guy10) -0.100.10.2 c ( G u y ) - c ( G u y ) Figure 13.
A comparison of derived SALT2 c using color lawsfrom Guy et al. (2007) and Guy et al. (2010) for the spectroscop-ically confirmed SN Ia sample. straight line fit, where the errorsalong the horizontal axis are ignored and all data is given equalweight. The result is δ c = 0 . ∗ c + 0 . To address concerns about ultraviolet measurements,we compared our u -band data with the predictions ofthe SALT2 and
MLCS gri banddata and comparing the measured u -band flux with thatpredicted by the model. The results are shown in Figure14 for the G07 model, the G10 model, and MLCS u -band flux compared toour data at early times with the exception of the earliestpoint for the G10 model. Both the G07 and G10 modelslie above MLCS . ± .
008 magnitudes higher thanour data, G07 is 0 . ± .
009 and mlcs2k2 is 0 . ± . SDSS dataconcerning the differences between
MLCS
SALT2 .The
SDSS light curve fits are relatively insensitive to thisdifference because of the poor instrumental sensitivity inthe u -band; it is more important for the high redshiftdata where an accurate rest-frame u -band measurementis necessary to obtain an accurate measurement of thecolor. Hubble Diagram and Cosmological Constraints
We compare the redshift determination from the 5-parameter
SALT2 fit and spectroscopic redshift in Fig-ure 15. Good agreement between the two redshifts isseen although the photometric error is often large. Av-eraging SN in bins of redshift reveals a net redshiftbias of the photometric redshifts relative to those mea-sured spectroscopically. The bias has been seen previ-ously (Kessler et al. 2010a; Campbell et al. 2013), andwas shown to agree well with the bias observed with sim-ulated SN I a light curves.Figure 16 shows the Hubble diagram for the SN thatmeet our fit selection criteria and have spectroscopic red-shifts ( δz < . SN that have been typed with spectra and the bottompanel (b) shows the 827 SN where the redshift is deter-mined from the host galaxy. We present these plots toshow the full sample and have not attempted to makeselections or to optimize the determination of cosmolog-ical parameters. The obvious outlier at z = 0 .
043 is -20 -10 0 10 20 30 40
Time (days in SN rest frame) -0.8-0.6-0.4-0.200.20.40.60.8 R e s i dua l fl u x ( fl u x / m ode l - ) Guy10Guy07MLCS2k2 -20 -10 0 10 20 30 40
Time (days in SN rest frame) -1-0.500.51 R e s i dua l fl u x ( fl u x / m ode l - ) Individual pointsWeighted Average
Figure 14.
The average u band residual for the SALT2 (G07and G10) models and the MLCS2k2 model are shown (top). Themeasured u -band flux is compared with the prediction from the fitusing g , r , and i band data only. The points are a weighted averageof the residuals shown in 3-day intervals measured in the SN Ia restframe. The bottom panel shows the same weighted average for theG10 model, but also shows the individual points that comprise theaverage. z l c z l c − z s p ec z spec bias rms Figure 15.
Comparison of spectroscopic and light curve photo-metric redshifts for the confirmed SN Ia sample. the under-luminous SN SN for determina-tion of cosmological parameters were presented previ-ously (Campbell et al. 2013). Redshift D i s t an c e M odu l u s (a) Redshift D i s t an c e M odu l u s (b) Figure 16.
Hubble diagram of the spectroscopic SN Ia (top) and z host -Ia (bottom). The large scatter in the z host -Ia sample espe-cially at low redshift ( z < .
2) is most likely due to contaminationfrom CC SN.
The Hubble diagrams shown in Figure 16 are not cor-rected for biases due to selection effects. Since the
SDSSSN survey is a magnitude limited survey a bias towardsbrighter SN is expected, particularly at the higher red-shifts. Correction for bias was a particularly impor-tant effect in the analysis of Campbell et al. (2013) andBetoule et al. (2014), who used photometrically identi-fied SN in addition to the spectroscopically confirmedsample. Figure 17 shows the bias expected from a sim-ulation of the SDSS SN survey for two sample detectionthresholds: requiring at least one light curve point tobe observed in each of 3 filters above background by5 σ ( SNRMAX
3) and 10 σ . The expected bias for a 5 σ threshold, which is typical for SDSS-II , is small but stillsignificant for a precise determination of cosmological pa-rameters. These two different bias corrections illustratethat the correction is important and that it depends onthe selection criteria for each particular analysis. The SN -0.08-0.06-0.04-0.0200.020.040.060.08 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 redshift µ f i t - µ s i m SNRMAX3 > > Figure 17.
The bias in distance modulus as a function of redshiftfor the SDSS sample for two different example selection criteria. detection efficiency is discussed in more detail elsewhere(Dilday et al. 2010a).We present a brief cosmological analysis of our fullthree-year spectroscopically–confirmed
SN I a sample inTable 3 and with
SALT2 fit parameters in the range ofnormal
SN I a ( − . < c < . − . < x < . SN I a. As-suming a ΛCDM cosmology, we simultaneously fit Ω M and Ω Λ using the sncosmo mcmc module within SNANA ,and show their joint constraints in Figure 18. In thisanalysis, we have corrected for the expected selection bi-ases (including Malmquist bias) using the 5 σ threshholdcurve (Figure 17), and have marginalised over H and thepeak absolute magnitude of SN I a, but only show statis-tical errors in Figure 18. Acceleration (Ω Λ > Ω M /
2) isdetected at a confidence of 3.1 σ . If we further assume aflat geometry, then we determine Ω M = 0 . ± .
093 and
Omega Λ > σ confidence (statisticalerror only). In Figure 19, we show the residuals of thedistance moduli with respect to this best fit cosmology,including varying Ω M by ± σ from this best fit. Over-all, our cosmological constraints are not as competitiveas higher redshift samples of SN I a because of the lim-ited redshift range of our
SDSS-II SN sample. Therefore,we refer the reader to Betoule et al. (2014) for a moreextensive analysis of the full
SDSS-II spectroscopically–confirmed SN sample combined with other SN datasets(low redshift samples, SNLS , HST ) and other cosmologi-cal measurements. HOST GALAXIES
A wealth of data on the SN host galaxies is availablefrom the SDSS
Data Release 8 ( DR
8; Aihara et al. 2011).In Section 8.1 we describe the host-galaxy identificationmethod used in this paper, which we suggest for futureanalyses. In Section 8.2 we describe the host-galaxyproperties computed from
SDSS data and presented inTable 1, and explain differences with values reportedin previous analyses (Lampeitl et al. 2010b; Smith et al.2012; Gupta et al. 2011).DSS-II SN Data Release 15 Ω M Ω Λ Figure 18.
The 68% and 95% contours (statistical errors only)for the joint fit to Ω M and Ω Λ for the full three-year SDSS-IIspectroscopic sample. The dashed line represents Ω Λ = Ω M / ΛCDM cosmology, we determine Ω M = 0 . ± . Redshift -1-0.500.51 H ubb l e R e s i dua l s DataOmega_M=0.49Omega_M=0.15
Figure 19.
Residuals of Hubble diagram for the full three-yearSDSS-II spectroscopic sample relative to a best-fit cosmology as-suming a flat ΛCDM cosmology with Ω M = 0 . M two standard deviations higherand lower are also shown. Host Galaxy Identification
We use a more sophisticated methodology for selectingthe correct SN host galaxy than trivially selecting thenearest galaxy with the smallest angular separation tothe supernova. We instead use a technique that accountsfor the probability based on the local surface brightnesssimilar to that used in Sullivan et al. (2006).We begin by searching DR ′′ radius of each SN candidate position and considerall the objects as possible host galaxy candidates. Wecharacterize each host galaxy by an elliptical shape. Wechose the elliptical approximation, because the model-independent isophotal parameters were determined to beless reliable and were therefore not included in DR r -band. The sec-ond moments are given in DR Q and U , from which one can compute the el-lipticity and orientation of the ellipse. The major axis ofthe ellipse is set equal to the Petrosian half-light radius( SDSS parameter
PetroR50 ) in the r -band; this radiusencompasses 50% of the observed galaxy light. We found this parameter to be a more robust representation of thegalaxy size than the deVRad and expRad profile fit radii,which too often had values that indicated a failure of theprofile fit.For each potential host galaxy, we calculate the ellip-tical light radius in the direction of the SN and call thisthe directional light radius ( DLR ). Next, we compute theratio of the SN -host separation to the DLR and denotethis normalized distance as d DLR . We then order thenearby host galaxy candidates by increasing d DLR anddesignate the first-ranked object as the host galaxy. Forparticular objects where this fails (the mechanism for de-termining this is described later), due to values of
Q, U or PetroR50 that are missing or poorly measured, weselect the next nearest object in d DLR as the host. In ad-dition, we impose a cut on the maximum allowed d DLR for a nearby object to be a host. This cut is chosen tomaximize the fraction of correct host matches while min-imizing the fraction of incorrect ones as explained below.If there is no host galaxy candidate meeting these crite-ria, we consider the candidate to be hostless.Determining an appropriate d DLR cutoff requires thatwe first estimate the efficiency of our matching algorithm.We estimate our efficiency by selecting a sample of pos-itively identified host galaxies based on the agreementbetween the SN redshift and the redshift of the hostgalaxy from SDSS DR SN of all types via visual inspection of images.We then consider the 172 host galaxies that have red-shifts in DR
8. The distribution of differences in the SN redshift and host galaxy redshift for this sample is shownin Figure 20. The prominent peak at zero and lack of out-liers is proof that these SN are correctly matched withthe host galaxy. The small offset between the host galaxyredshift and the redshift obtained from the SN spectrumwas discussed in §
6. Of the 172 host galaxies, 150 have aredshift agreement of ± .
01 or better, and we designatethis sample of SN -host galaxy pairs as the “truth sam-ple”. We plot the distribution as a function of d DLR nor-malized to the data, as the dashed blue curve in Fig. 21(top panel).
SN redshift − host redshift (DR8) N u m b e r Figure 20.
The distribution of the difference in redshift (SN spec-trum redshift minus host galaxy spectrum redshift) for the sampleof spectroscopically-confirmed SNe whose hosts have redshifts inDR8.
Figure 21.
Top : The distribution in d DLR is shown for the truthsample (dashed blue line), the contamination of false host galaxymatches (dotted red line) and the sum of the two distributions(solid blackline). The full sample is shown as the open circleswith Poisson error bars and should be compared to the solid line.
Bottom : The efficiency and purity of the host galaxy selection isshown as a function of d DLR and the matching criterion at d DLR =4 indicated.
The efficiency for the identification of the full
SDSS sample needs to include the SN which are hostless. Us-ing the sample of spectroscopic SN I a with z < . SDSS-II SN survey is esti-mated to be 100% efficient (Dilday et al. 2010a) for spec-troscopic measurement, we estimated the rate of hostless SN under the assumption that this low- z host sample isrepresentative of the true SN I a host distribution. We ob-tained
SDSS ugriz model magnitudes and errors for thelow- z host sample from the DR and used them and the measured redshifts tocompute the best-fit model spectral energy distributions( SED ) using the code kcorrect v4 2 (Blanton & Roweis2007). The spectra were shifted to redshift bins of 0.05up to z = 0 .
45, and we computed the expected appar-ent magnitudes of the hosts at those redshifts. We thenweighted these magnitudes in the various z -bins by theredshift distribution of the entire spectroscopic SN I asample to mimic the observed r -band distribution forthe whole redshift range. We identified those hosts that fell outside the DR r -band magnitude limit of 22.2 ashostless. From this analysis, we predict a hostless rateof 12% for the SDSS sample. Normalizing the truth dis-tribution to 88% and taking the cumulative sum gives usan estimate of the efficiency of our matching method asa function of d DLR , which is shown as the blue curve inFig. 21 (bottom panel).Unfortunately, we do not have spectroscopic redshiftsfor all candidates nor all potential host galaxies, so wecan not rely on agreement between redshifts for the pu-rity of the sample. In order to estimate the rate ofmisidentification, we chose a set of 10,000 random co-ordinates in the SN survey footprint and applied ourmatching algorithm using the DR d DLR of SN candidates with unrelated galaxies. We realize thatin reality, SNe will occur galaxies rather than randomlyon the sky but a more sophisticated background esti-mate involving random galaxies and an assumed d DLR distribution is left for future work. The top panel of Fig.21 summarizes the situation: the distribution in d DLR isshown for the truth galaxies (dashed blue line), the ex-pected distribution of background galaxies is shown asthe dotted red line, and the solid black line is the sum ofthe two. The data sample is shown as the open circles.While the data is similar to expectations, it is notablymore peaked at low values of d DLR than the truth sam-ple would lead us to expect. The difference in the distri-butions is partly due to the fact that the truth sample(being constructed from the sample of spectroscopically-confirmed SNe) is biased against SNe that occur near thecore of their host galaxy where a spectroscopic confirma-tion is very difficult or impossible. We therefore expectthat many more SNe will reside at low d DLR than thetruth sample predicts. In addition, there may be diffi-culty in determining accurate galaxy shape parametersfor the fainter galaxies that comprise our full sample.Normalizing the host distribution for the random pointsand taking the cumulative sum yields the contaminationrate as a function of d DLR . In the bottom panel we plotthe estimated sample purity (1 − contamination) as thered curve on the bottom panel of Figure 21.We choose d DLR = 4 as our matching criterion in or-der to obtain high purity (97%) while still obtaining agood efficiency (80%). For that criterion we find that16% of our SN candidates are hostless. We expect the ob-served rate of hostless SN to be higher than the predictedrate because of the inefficiency of our d DLR < d DLR = 4. While the measured rate of hostless galaxiesagrees fairly well with expectations, we suspect that ourefficiency is underestimated because of the difference in d DLR distributions between the truth sample and the fullsample and also the corrections made by visual scanning,which are described below.There are many ways in which a host galaxy can bemisidentified. If the true host is not found (which canhappen when it is too faint or near a bright star or satel-lite track), it will not be selected. Even having a match-ing SN redshift and host galaxy redshift does not guar-antee a correct match in the presence of galaxy groups,clusters, or mergers. For nearby ( z . .
05) candidates,the SN can be offset by more than 30 ′′ from the cen-DSS-II SN Data Release 17ter of the host galaxy or the SDSS galaxy reconstructionmay erroneously detect multiple objects in a large, ex-tended galaxy. More distant candidates suffer from ahigher density of plausible host galaxies, which also tendto be fainter and more point-like.We attempted to mitigate these issues by examininga subset of all candidates and manually correcting anyobvious mistakes made by the host-matching algorithm.In total, only 116 host galaxies were corrected and thedetails regarding their selection are given below. Firstwe examined the images of several hundred of the lowestredshift candidates, since there are relatively few of themand they can exhibit some of the issues with host match-ing listed in the previous paragraph. In addition, of the3000 host galaxies targeted by
BOSS (Campbell et al.2013; Olmstead et al. 2013) we found that for ≈
350 can-didates either the DR BOSS target coordinatesby more than 1 . ′′ . We visually inspected these ≈ BOSS targets, the d DLR algorithm choice was changed for 116host galaxies. The majority of these (69) had no hostidentified because of our selection criterion of d DLR < d DLR . However, we did not assign a host based onvisual inspection if there was no corresponding objectin the DR d DLR disagreed withthe visual scanning result. Most of these were causedby improperly deblended galaxies or regions where therewere multiple candidates and visual pattern recognitionproved superior; poor estimate of the galaxy size param-eters was likely a factor in these corrections. There are7 candidates that were changed to be hostless becausethe d DLR < Host Galaxy Properties
Much can be learned about SN through the proper-ties of their host galaxies. We can derive several suchproperties by fitting host galaxy photometry to galaxyspectral energy distribution ( SED ) models. We beginby retrieving the
SDSS ugriz model magnitudes (whichyield the most accurate galaxy colors) and their errorsfrom DR SN host galaxy sample. For all SN host galaxies with a spectroscopic redshift from eitherthe host or the SN we use the redshifts, host magnitudes,and magnitude errors in conjunction with stellar popula-tion synthesis (SPS) codes to estimate physical proper-ties of our hosts such as stellar mass, star-formation rate,and average age. In this work, we obtain these prop-erties using two different methods used by Gupta et al.(2011) and Smith et al. (2012), respectively. The for-mer method utilizes SED models from the code FlexibleStellar Population Synthesis (
FSPS ; Conroy et al. 2009;Conroy & Gunn 2010) while the latter utilizes
SED mod-els from the code
P´EGASE.2 (Fioc & Rocca-Volmerange1997; Le Borgne et al. 2004). The current results, how-ever, are not identical to the previously published resultsin that
SDSS DR
SDSS co-add catalog (Annis et al. 2011) were used previously, and Gupta et al. (2011) augmented
SDSS photometry with UV and near-IR data. While theco-add catalog is certainly deeper, it is more prone toproblems like artifacts or galactic substructures being de-tected as objects. The previous works cited above usedrelatively small SN samples and identified host galaxiesby visual inspection, while in this paper we must relyon our automated algorithm which would fail on suchproblematic cases in the co-add catalogs.In Table 12, we display the host properties calculatedusing FSPS for a few SN candidates. Table 1 containsthe full sample and also calculations using P´EGASE.2 of the analogous quantities using the same photometricdata. The galaxy stellar mass (log( M ), where M is ex-pressed in units of M ⊙ ) is shown in Table 12 with itsuncertainty while the same information is presented as arange (log( M lo ) and log( M hi ) in Table 1). All the calcu-lated parameters are presented in the same way: Table12 shows the uncertainties while Table 1 gives the range.Table 12 also shows the logarithm of the specific star-forming rate log(sSFR), where s SFR is the mass of starsformed in M ⊙ per year per galaxy stellar mass) averagedover the most recent 250 Myr. The mass-weighted aver-age age of the galaxy is also given in units of Gyr. Wegive analogous quantities for P´EGASE.2 in Table 1 exceptthat we give the logarithm of the star-forming rate ( i.e. ,not normalized to the galaxy stellar mass) and age is theage of the best fitting template (in Gyr) since these arethe natural, more fundamental outputs of the
P´EGASE.2 code.Figure 22 shows the distribution of galaxies as a func-tion of logarithm of galaxy stellar mass versus the loga-rithm of star-forming rate for
P´EGASE.2 (top) and
FSPS (bottom). The two distributions are similar overall butthere are significant differences as well. In the analy-sis of Smith et al. (2012), galaxies are split into groupsbased on their s
SFR : highly star-forming galaxies havelog(sSFR) ≥ − .
5, moderately star-forming galaxieshave − . < log(sSFR) < − .
5, and passive galaxieshave log(sSFR) ≤ − .
0. Galaxies classified as passiveby
P´EGASE.2 were assigned random log(SFR) values be-tween − − ∼ − . P´EGASE.2 templates between star-forming and completely passive galaxies. The
FSPS cal-culations do not provide such a clear distinction betweenpassive and star-forming, so we somewhat arbitrarily de-fine passive galaxies as those with log(SFR) < − .
0. Forboth analyses we see that, with a few exceptions, themost massive galaxies are classified as passive comparedto the less massive galaxies which are classified as star-forming.Figures 23 (top) compares the stellar mass calculatedwith
P´EGASE.2 and
FSPS . Galaxies are split accordingto the s
SFR scheme described above, with red circles in-dicating passive, green triangles indicating moderatelystar-forming, and blue diamonds indicating highly star-forming. The mass estimates show good agreement, withthe stellar mass estimated from
FSPS being marginallyhigher than that estimated from the
P´EGASE.2 tem-plates. Figure 23 (bottom) compares the
SFR estimatedby both methods. We find that 68% (24%) of galax-8 Sako et al.
Figure 22.
The distribution stellar mass and star-formation ratefor the SN candidate host galaxies with a spectroscopic redshiftfor the P´EGASE.2 analysis (Smith et al. 2012, top panel) and theFSPS analysis (Gupta et al. 2011, bottom panel). Lines of con-stant specific star formation rate separate the regions of high andmoderate star formation (top panel) and the separation betweenstar-forming and passive galaxies is shown with dashed lines in eachpanel. For P´EGASE.2, galaxies with a log(SFR) < − − − ies are found to be star-forming (passive), respectively,by both analyses, and 6% are found to be passive by P´EGASE.2 and star-forming by
FSPS and 2% vice versa.In general, the
SFR show good agreement between thetwo methods, with a larger scatter than that observedfor the mass estimates. For galaxies classified as star-forming, the
SFR estimated by
P´EGASE.2 are systemat-ically higher than those estimated by
FSPS . The differ-ences in derived galaxy properties are likely due to thedifferences in the available
SED templates and how they
Figure 23.
The comparison between P´EGASE.2 and FSPSgalaxy stellar masses and SFRs for the the SN candidate host galax-ies with a spectroscopic redshift. As in Figure 22 for P´EGASE.2,galaxies with log(SFR) < − − − are parametrized in FSPS compared with
P´EGASE.2 . SUMMARY
This paper represents the final Data Release of the
SDSS-II SN
Survey of 10,258 candidates. A new methodof classification based on the light curve data has beenpresented and applied to the candidates. Reference lightcurve fits are provided for
SALT2 and
MLCS SN observations with their hostgalaxies was presented, including a quantitative estimateof efficiency and false-positive association. Host galaxyproperties were computed from the photometric data us-ing two computer programs: P´EGASE.2 and
FSPS . Atable listing the 1360 spectra that were obtained in con-DSS-II SN Data Release 19junction with the
SDSS SN search was presented. A webpage reference to the complete light curve data and re-duced spectra was given. A complete set of photometricdata for all the
SDSS SN candidates has been presentedand is released on the
SDSS SN data release web page.All the spectra taken in conjunction with the
SDSS-IISN survey are also released. In addition, we have pro-vided light curve fits and host galaxy identifications andestimated host galaxy parameters.0 Sako et al.Funding for the
SDSS and
SDSS-II has been providedby the Alfred P. Sloan Foundation, the Participating In-stitutions, the National Science Foundation, the U.S.Department of Energy, the National Aeronautics andSpace Administration, the Japanese Monbukagakusho,the Max Planck Society, and the Higher EducationFunding Council for England. The
SDSS
Web Site is .The
SDSS is managed by the Astrophysical ResearchConsortium for the Participating Institutions. The Par-ticipating Institutions are the American Museum of Nat-ural History, Astrophysical Institute Potsdam, Univer-sity of Basel, Cambridge University, Case Western Re-serve University, University of Chicago, Drexel Univer-sity, Fermilab, the Institute for Advanced Study, theJapan Participation Group, Johns Hopkins University,the Joint Institute for Nuclear Astrophysics, the KavliInstitute for Particle Astrophysics and Cosmology, theKorean Scientist Group, the Chinese Academy of Sci-ences (
LAMOST ), Los Alamos National Laboratory, theMax-Planck-Institute for Astronomy (
MPIA ), the Max-Planck-Institute for Astrophysics (
MPA ), New MexicoState University, Ohio State University, University ofPittsburgh, University of Portsmouth, Princeton Uni-versity, the United States Naval Observatory, and theUniversity of Washington.The Hobby-Eberly Telescope (
HET ) is a joint projectof the University of Texas at Austin, the Pennsyl-vania State University, Stanford University, Ludwig-Maximillians-Universit¨at M¨unchen, and Georg-August-Universit¨at G¨ottingen. The
HET is named in honor ofits principal benefactors, William P. Hobby and RobertE. Eberly. The Marcario Low-Resolution Spectrographis named for Mike Marcario of High Lonesome Optics,who fabricated several optics for the instrument butdied before its completion; it is a joint project of theHobby-Eberly Telescope partnership and the Institutode Astronom´ıa de la Universidad Nacional Aut´onoma deM´exico. The Apache Point Observatory 3.5-meter tele-scope is owned and operated by the Astrophysical Re-search Consortium. We thank the observatory director,Suzanne Hawley, and site manager, Bruce Gillespie, fortheir support of this project. The Subaru Telescope isoperated by the National Astronomical Observatory ofJapan. The William Herschel Telescope is operated bythe Isaac Newton Group, and the Nordic Optical Tele-scope is operated jointly by Denmark, Finland, Iceland,Norway, and Sweden, both on the island of La Palmain the Spanish Observatorio del Roque de los Mucha-chos of the Instituto de Astrofisica de Canarias. Ob-servations at the
ESO
New Technology Telescope at LaSilla Observatory were made under programme ID s 77.A-0437, 78.A-0325, and 79.A-0715. Kitt Peak NationalObservatory, National Optical Astronomy Observatory,is operated by the Association of Universities for Re-search in Astronomy, Inc. ( AURA ) under cooperativeagreement with the National Science Foundation. The
WIYN
Observatory is a joint facility of the Universityof Wisconsin-Madison, Indiana University, Yale Univer-sity, and the National Optical Astronomy Observatories.The W.M. Keck Observatory is operated as a scientificpartnership among the California Institute of Technol-ogy, the University of California, and the National Aero-nautics and Space Administration. The Observatory was made possible by the generous financial support ofthe W.M. Keck Foundation. The South African LargeTelescope of the South African Astronomical Observa-tory is operated by a partnership between the NationalResearch Foundation of South Africa, Nicolaus Coperni-cus Astronomical Center of the Polish Academy of Sci-ences, the Hobby-Eberly Telescope Board, Rutgers Uni-versity, Georg-August-Universit¨at G¨ottingen, Universityof Wisconsin-Madison, University of Canterbury, Uni-versity of North Carolina-Chapel Hill, Dartmough Col-lege, Carnegie Mellon University, and the United King-dom
SALT consortium. The Telescopio Nazionale Galileo(
TNG ) is operated by the Fundaci´on Galileo Galilei of theItalian
INAF (Istituo Nazionale di Astrofisica) on the is-land of La Palma in the Spanish Observatorio del Roquede los Muchachos of the Instituto de Astrof´ısica de Ca-narias.A. V. Filippenko has received generous financial as-sistance from the Christopher R. Redlich Fund, the
TABASGO
Foundation, and
NSF grant
AST -1211916.Supernova research at Rutgers University is supported inpart by
NSF CAREER award
AST -0847157 to S. W. Jha.G. Leloudas is supported by the Swedish Research Coun-cil through grant No. 623-2011-7117. M. D. Stritzingergratefully acknowledges generous support provided bythe Danish Agency for Science and Technology and In-novation realized through a Sapere Aude Level 2 grant.DSS-II SN Data Release 21 The meaning of the photometric flags is detailed inHoltzman et al. (2008). http://skyservice.pha.jhu.edu/casjobs/ REFERENCESAbazajian, K. N., Adelman-McCarthy, J. K., Ag¨ueros, M. A., etal. 2009, ApJS, 182, 543Alard, C., & Lupton, R. H. 1998, ApJ, 503, 325Aldering, G., Adam, G., Antilogus, P., et al. 2002, Proc. SPIE,4836, 61Annis, J., Soares-Santos, M., Strauss, M. A., et al. 2011,arXiv:1111.6619Aihara, H., Allende Prieto, C., An, D., et al. 2011, ApJS, 193, 29Astier, P., Guy, J., Regnault, N., et al. 2006, A&A, 447, 31Bernstein, J. P., Kessler, R., Kuhlmann, S., et al. 2012, ApJ, 753,152Betoule, M., Marriner, J., Regnault, N., et al. 2013, A&A, 552,A124Betoule, M., Kessler, R., et al. 2014, (in prep.)Blanton, M. R., & Roweis, S. 2007, AJ, 133, 734Bohlin, R. C. 2007, The Future of Photometric,Spectrophotometric and Polarimetric Standardization, 364, 315
HST
CALSPEC database
Campbell, H., D’Andrea, C. B., Nichol, R. C., et al. 2013, ApJ,763, 88Conley, A., Guy, J., Sullivan, M., et al. 2011, ApJS, 192, 1Conroy, C., & Gunn, J. E. 2010, ApJ, 712, 833Conroy, C., Gunn, J. E., & White, M. 2009, ApJ, 699, 486Contreras, C., Hamuy, M., Phillips, M. M., et al. 2010, AJ, 139,519D’Andrea, C. B., et al. 2011, ApJ, 743, 172Dawson, K. S., Schlegel, D. J., Ahn, C. P., et al. 2013, AJ, 145, 10Dilday, B., Kessler, R., Frieman, J. A., et al. 2008, ApJ, 682, 262Dilday, B., Smith, M., Bassett, B., et al. 2010, ApJ, 713, 1026Dilday, B., Bassett, B., Becker, A., et al. 2010, ApJ, 715, 1021Doi, M., Tanaka, M., Fukugita, M., et al. 2010, AJ, 139, 1628Eisenstein, D. J., Weinberg, D. H., Agol, E., et al. 2011, AJ, 142,72Fioc, M., & Rocca-Volmerange, B. 1997, A&A, 326, 950Foley, R. J., Filippenko, A. V., Kessler, R., et al. 2012, AJ, 143,113Foley, R. J., Challis, P. J., Chornock, R., et al. 2013, ApJ, 767, 57Frieman, J. A., Bassett, B., Becker, A., et al. 2008, AJ, 135, 338Fukugita, M., Ichikawa, T., Gunn, J. E., et al. 1996, AJ, 111, 1748Galbany, L., Miquel, R., ¨Ostman, L., et al. 2012, ApJ, 755, 125Gallagher, J. S., Garnavich, P. M., Berlind, P., et al. 2005, ApJ,634, 210Gunn, J. E., Carr, M., Rockosi, C., et al. 1998, AJ, 116, 3040Gunn, J. E., Siegmund, W. A., Mannery, E. J., et al. 2006, AJ,131, 2332Gupta, R. R., D’Andrea, C. B., Sako, M., et al. 2011, ApJ, 740, 92Guy, J., Astier, P., Baumont, S., et al. 2007, A&A, 466, 11Guy, J., Sullivan, M., Conley, A., et al. 2010, A&A, 523, A7Hamuy, M., Phillips, M. M., Suntzeff, N. B., et al. 1996, AJ, 112,2398Hamuy, M., Folatelli, G., Morrell, N. I., et al. 2006, PASP, 118, 2Hayden, B. T., Garnavich, P. M., Kessler, R., et al. 2010, ApJ,712, 350Hayden, B. T., Garnavich, P. M., Kasen, D., et al. 2010, ApJ,722, 1691Hayden, B. T., Gupta, R. R., Garnavich, P. M., et al. 2013, ApJ,764, 191Hicken, M., Challis, P., Jha, S., et al. 2009, ApJ, 700, 331Hill, G. J., Nicklas, H. E., MacQueen, P. J., et al. 1998,Proc. SPIE, 3355, 375Hlozek, R., Kunz, M., Bassett, B., et al. 2012, ApJ, 752, 79Holtzman, J. A., Marriner, J., Kessler, R., et al. 2008, AJ, 136,2306Ivezi´c, ˇZ., Smith, J. A., Miknaitis, G., et al. 2007, AJ, 134, 973Jha, S., Kirshner, R. P., Challis, P., et al. 2006, AJ, 131, 527 Jha, S., Riess, A. G., & Kirshner, R. P. 2007, ApJ, 659, 122Johansson, J., Thomas, D., Pforr, J., et al. 2013, MNRAS, 435,1680Kashikawa, N., Inata, M., Iye, M., et al. 2000, Proc. SPIE, 4008,104Kelly, B. C. 2007, ApJ, 665, 1489Kessler, R., Becker, A. C., Cinabro, D., et al. 2009, ApJS, 185, 32Kessler, R., Bernstein, J. P., Cinabro, D., et al. 2009, PASP, 121,1028Kessler, R., Cinabro, D., Bassett, B., et al. 2010, ApJ, 717, 40Kessler, R., Conley, A., Jha, S., & Kuhlmann, S. 2010,arXiv:1001.5210Kessler, R., Bassett, B., Belov, P., et al. 2010, PASP, 122, 1415Kessler, R., Guy, J., Marriner, J., et al. 2013, ApJ, 764, 48Konishi, K., Yasuda, N., Tokita, K., et al. 2011, arXiv:1101.1565Konishi, K., Frieman, J. A., Goobar, A., et al. 2011,arXiv:1103.2497Lampeitl, H., Nichol, R. C., Seo, H.-J., et al. 2010, MNRAS, 401,2331Lampeitl, H., Smith, M., Nichol, R. C., et al. 2010, ApJ, 722, 566Le Borgne, D., Rocca-Volmerange, B., Prugniel, P., et al. 2004,A&A, 425, 881Lupton, R. H., Gunn, J. E., & Szalay, A. S. 1999, AJ, 118, 1406Mannucci, F., Cresci, G., Maiolino, R., Marconi, A., & Gnerucci,A. 2010, MNRAS, 408, 2115McClelland, C. M., Garnavich, P. M., Galbany, L., et al. 2010,ApJ, 720, 704McCully, C., Jha, S. W., Foley, R. J., et al. 2013, arXiv:1309.4457Marriner, J., Bernstein, J. P., Kessler, R., et al. 2011, ApJ, 740, 72Miknaitis, G., Pignata, G., Rest, A., et al. 2007, ApJ, 666, 674Mosher, J., Sako, M., Corlies, L., et al. 2012, AJ, 144, 17Nordin, J., ¨Ostman, L., Goobar, A., et al. 2011, A&A, 526, A119Nordin, J., ¨Ostman, L., Goobar, A., et al. 2011, ApJ, 734, 42Oke, J. B., & Gunn, J. E. 1983, ApJ, 266, 713Oke, J. B., Cohen, J. G., Carr, M., et al. 1995, PASP, 107, 375Olmstead, M. D., Brown, P. J., Sako, M., et al. 2013,arXiv:1308.6818¨Ostman, L., Nordin, J., Goobar, A., et al. 2011, A&A, 526, A28Padmanabhan, N., Schlegel, D. J., Finkbeiner, D. P., et al. 2008,ApJ, 674, 1217Perlmutter, S., Aldering, G., Goldhaber, G., et al. 1999, ApJ,517, 565Phillips, M. M., Li, W., Frieman, J. A., et al. 2007, PASP, 119,360Poznanski, D., Maoz, D., & Gal-Yam, A. 2007, AJ, 134, 1285Rest, A., Scolnic, D., Foley, R. J., et al. 2013, arXiv:1310.3828Riess, A. G., Filippenko, A. V., Challis, P., et al. 1998, AJ, 116,1009Riess, A. G., Strolger, L.-G., Casertano, S., et al. 2007, ApJ, 659,98Sako, M., Romani, R., Frieman, J., et al. 2005, 22nd TexasSymposium on Relativistic Astrophysics, 415Sako, M., Bassett, B., Becker, A., et al. 2008, AJ, 135, 348Sako, M., Bassett, B., Connolly, B., et al. 2011, ApJ, 738, 162Scolnic, D., Rest, A., Riess, A., et al. 2013, arXiv:1310.3824Scolnic, D. M., Riess, A. G., Foley, R. J., et al. 2013,arXiv:1306.4050SDSS SN Web Page( http://sdssdp62.fnal.gov/sdsssn/DataRelease/index.html )Smee, S. A., Gunn, J. E., Uomoto, A., et al. 2013, AJ, 146, 32Smith, M., Nichol, R. C., Dilday, B., et al. 2012, ApJ, 755, 61Smith, M., Bacon, D. J., Nichol, R. C., et al. 2014, ApJ, 780, 24Sollerman, J., M¨ortsell, E., Davis, T. M., et al. 2009, ApJ, 703,1374Stoughton, C., Lupton, R. H., Bernardi, M., et al. 2002, AJ, 123,485Sullivan, M., Le Borgne, D., Pritchet, C. J., et al. 2006, ApJ, 648,868Taylor, M., et al. 2014 (in prep.)Tyson, J. A. 2002, Proc. SPIE, 4836, 10Xavier, H. S., Gupta, R. R., Sako, M., et al. 2013, MNRAS, 434,1443York, D. G., Adelman, J., Anderson, J. E., Jr., et al. 2000, AJ,120, 1579Zheng, C., Romani, R. W., Sako, M., et al. 2008, AJ, 135, 1766
Table 1
SDSS SN Catalog a Item Format Symbol Description (units)1 I5 CID
SDSS
Candidate Identification Number2 F12.6 RA SN Right ascension (J2000, degrees)3 F11.6 DEC SN Declination (J2000, degrees)4 I5 Nsearchepoch Number of search detection epochs5 A13 IAUName Name assigned by the International Astronomical Union6 A11 Classification Candidate PSNID type (see Table 3)7 F6.1 Peakrmag Measured peak asinh magnitude ( r -band)8 F10.1 MJDatPeakrmag Modified Julian Date (MJD) of observed peak brightness ( r -band)9 I5 NepochSNR5 Number of epochs with S/N >
510 I5 nSNspec Number of SN spectra11 I5 nGALspec Number of host galaxy spectra12 F10.6 zspecHelio Heliocentric redshift13 F10.6 zspecerrHelio Heliocentric redshift uncertainty14 F10.6 zCMB CMB-frame redshift15 F10.6 zerrCMB CMB-frame redshift uncertainty
SALT2
SALT2 x (normalization) parameter17 E10.2 x0errSALT2zspec SALT2 x (normalization) parameter uncertainty18 F6.2 x1SALT2zspec SALT2 x (shape) parameter19 F6.2 x1errSALT2zspec SALT2 x (shape) parameter uncertainty20 F6.2 cSALT2zspec SALT2 c (color) parameter21 F6.2 cerrSALT2zspec SALT2 c (color) uncertainty22 F10.2 PeakMJDSALT2zspec SALT2
MJD at peak in B -band23 F7.2 PeakMJDderrSALT2zspec SALT2
MJD at peak in B -band uncertainty24 F7.2 muSALT2zspec SALT2mu distance modulus25 F6.2 muerrSALT2zspec SALT2mu distance modulus uncertainty26 F8.3 fitprobSALT2zspec SALT2 fit chi-squared probability27 F8.2 chi2SALT2zspec
SALT2 fit chi-squared28 I5 ndofSALT2zspec
SALT2 number of light curve points used
MLCS
MLCS
MLCS
MLCS V -band extinction ( A V )32 F6.2 averrMLCS2k2zspec MLCS V -band extinction ( A V ) uncertainty33 F10.2 PeakMJDMLCS2k2zspec MLCS B -band34 F7.2 PeakMJDerrMLCS2k2zspec MLCS B -band uncertainty35 F7.2 muMLCS2k2zspec MLCS
MLCS
MLCS
MLCS
MLCS
SN I a Bayesian probability ( z spec prior)41 E10.2 logprobIaPSNIDzspec SN I a log( P fit ) ( z spec prior)42 I5 lcqualityIaPSNIDzspec SN I a light curve quality ( z spec prior)43 F7.3 PIbcPSNIDzspec SN I b/c Bayesian probability ( z spec prior)44 E10.2 logprobIbcPSNIDzspec SN I b/c log( P fit ) ( z spec prior)45 I5 lcqualityIbcPSNIDzspec SN I b/c light curve quality ( z spec prior)46 F7.3 PIIPSNIDzspec SN II
Bayesian probability ( z spec prior)47 E10.2 logprobIIPSNIDzspec SN II log( P fit ) ( z spec prior)48 I5 lcqualityIIPSNIDzspec SN II light curve quality ( z spec prior)49 I5 NnnPSNIDzspec Number of nearest neighbors ( z spec prior)50 F7.3 PnnIaPSNIDzspec SN I a nearest-neighbor probability ( z spec prior)51 F7.3 PnnIbcPSNIDzspec SN I b/c nearest-neighbor probability ( z spec prior)52 F7.3 PnnIIPSNIDzspec SN II nearest-neighbor probability ( z spec prior)53 F8.4 zPSNIDzspec PSNID redshift ( z spec prior)54 F8.4 zerrPSNIDzspec PSNID redshift uncertainty ( z spec prior)55 F6.2 dm15PSNIDzspec PSNID ∆ m ( B ) ( z spec prior)56 F6.2 dm15errPSNIDzspec PSNID ∆ m ( B ) uncertainty ( z spec prior)57 F6.2 avPSNIDzspec PSNID A V ( z spec prior)58 F6.2 averrPSNIDzspec PSNID A V uncertainty ( z spec prior)59 F10.2 PeakMJDPSNIDzspec PSNID T max ( z spec prior)60 F7.2 PeakMJDerrPSNIDzspec PSNID T max uncertainty ( z spec prior)61 I5 SNIbctypePSNIDzspec Best-fit SN I b/c template ( z spec prior)62 I5 SNIItypePSNIDzspec Best-fit SN II template ( z spec prior)SALT2 5-parameter fits (ignoring spectroscopic redshift information)63 E10.2 x0SALT2flat SALT2 x (normalization) parameter (flat- z prior)64 E10.2 x0errSALT2flat SALT2 x (normalization) parameter uncertainty (flat- z prior)65 F6.2 x1SALT2flat SALT2 x (shape) parameter (flat- z prior) DSS-II SN Data Release 23
Table 1 — Continued
Item Format Symbol Description (units)66 F6.2 x1errSALT2flat
SALT2 x (shape) parameter uncertainty (flat- z prior)67 F6.2 cSALT2flat SALT2 c (color) parameter(flat- z prior)68 F6.2 cerrSALT2flat SALT2 c (color) parameter uncertainty (flat- z prior)69 F10.2 PeakMJDSALT2flat SALT2 T max (flat- z prior)70 F7.2 PeakMJDerrSALT2flat SALT2 T max uncertainty (flat- z prior)71 F7.2 zphotSALT2flat SALT2 fitted redshift (heliocentric frame)72 F6.2 zphoterrSALT2flat
SALT2 fitted redshift uncertainty (heliocentric frame)73 F8.3 fitprobSALT2flat
SALT2 fit chi-squared probability74 F8.2 chi2SALT2flat
SALT2 fit chi-squared75 I5 ndofSALT2flat
SALT2 number of light curve points usedPSNID parameters ignoring spectroscopic redshift information76 F8.3 PIaPSNIDflat
SN I a Bayesian probability (flat- z prior)77 F8.2 logprobIaPSNIDflat SN I a log( P fit ) (flat- z prior)78 I5 lcqualityIaPSNIDflat SN I a light curve quality (flat- z prior)79 F7.3 PIbcPSNIDflat SN I b/c Bayesian probability (flat- z prior)80 E10.2 logprobIbcPSNIDflat SN I b/c log( P fit ) (flat- z prior)81 I5 lcqualityIbcPSNIDflat SN I b/c light curve quality (flat- z prior)82 F7.3 PIIPSNIDflat SN II
Bayesian probability (flat- z prior)83 E10.2 logprobIIPSNIDflat SN II log( P fit ) (flat- z prior)84 I5 lcqualityIIPSNIDflat SN II light curve quality (flat- z prior)85 I5 NnnPSNIDflat Number of nearest neighbors (flat- z prior)86 F7.3 PnnIaPSNIDflat SN I a nearest-neighbor probability (flat- z prior)87 F7.3 PnnIbcPSNIDflat SN I b/c nearest-neighbor probability (flat- z prior)88 F7.3 PnnIIPSNIDflat SN II nearest-neighbor probability (flat- z prior)89 F8.4 zPSNIDflat PSNID redshift (flat- z prior)90 F8.4 zerrPSNIDflat PSNID redshift uncertainty (flat- z prior)91 F6.2 dm15PSNIDflat PSNID ∆ m ( B ) (flat- z prior)92 F6.2 dm15errPSNIDflat PSNID ∆ m ( B ) uncertainty (flat- z prior)93 F6.2 avPSNIDflat PSNID A V (flat- z prior)94 F6.2 averrPSNIDflat PSNID A V uncertainty (flat- z prior)95 F10.2 PeakMJDPSNIDflat PSNID T max (flat- z prior)96 F7.2 PeakMJDerrPSNIDflat PSNID T max uncertainty (flat- z prior)97 I5 SNIbctypePSNIDflat Best-fit SN I b/c template (flat- z prior)98 I5 SNIItypePSNIDflat Best-fit SN II template (flat- z prior)Host galaxy information99 I21 objIDHost Host galaxy object ID in SDSS
DR8 Database100 F13.6 RAhost Right ascension of galaxy host (degrees)101 F11.6 DEChost Declination of galaxy host (degrees)102 F6.2 separationhost Distance from SN to host (arc-sec)103 F6.2 DLRhost Normalized distance from SN to host ( d DLR )104 F7.2 zphothost Host photometric redshift (KF algorithm)105 F6.2 zphoterrhost zphothost uncertainty106 F7.2 zphotRFhost Host photometric redshift (RF algorithm)107 F6.2 zphotRFerrhost zphotRFhost uncertainty108 F8.3 dereduhost Host galaxy u -band magnitude (dereddened)109 F7.3 erruhost Host galaxy u -band magnitude uncertainty110 F8.3 deredghost Host galaxy g -band magnitude (dereddened)111 F7.3 errghost Host galaxy g -band magnitude uncertainty112 F8.3 deredrhost Host galaxy r -band magnitude (dereddened)113 F7.3 errrhost Host galaxy r -band magnitude uncertainty114 F8.3 deredihost Host galaxy i -band magnitude (dereddened)115 F7.3 errihost Host galaxy i -band magnitude uncertainty116 F8.3 deredzhost Host galaxy z -band magnitude (dereddened)117 F7.3 errzhots Host galaxy z -band magnitude (dereddened)Galaxy Parameters Calculated with FPPS118 F7.2 logMassFSPS FSPS log( M ), M =Galaxy Mass (M in units of M ⊙ )119 F7.2 logMassloFSPS FSPS
Lower limit of uncertainty in log( M )120 F7.2 logMasshiFSPS FSPS
Upper limit of uncertainty in log ( M )121 F8.2 logSSFRFSPS FSPS log( sSF R ) sSF R =Galaxy Specific Star Forming Rate ( SF R in M ⊙ /yr)122 F8.2 logSSFRloFSPS FSPS
Lower limit of uncertainty in log( sSF R )123 F8.2 logSSFRhiFSPS
FSPS
Upper limit of uncertainty in log( sSF R )124 F7.2 ageFSPS
FSPS galaxy age (Gyr)125 F7.2 ageloFSPS
FSPS
Lower limit of uncertainty in age126 F7.2 agehiFSPS
FSPS
Upper limit of uncertainty in age127 F8.2 minredchi2FSPS Reduced chi-squared of best
FSPS template fitGalaxy Parameters Calculated with
P´EGASE.2
128 F8.2 logMassPEGASE
P´EGASE.2 log( M ), M=Galaxy Mass (M in units of M ⊙ )129 F8.2 logMassloPEGASE P´EGASE.2
Lower limit of uncertainty in log( M )130 F8.2 logMasshiPEGASE P´EGASE.2
Upper limit of uncertainty in log(
SF R ) Table 1 — Continued
Item Format Symbol Description (units)131 F9.2 logSFRPEGASE
P´EGASE.2 log(
SF R ) SFR=Galaxy star forming rate ( M ⊙ /yr)132 F9.2 logSFRloPEGASE P´EGASE.2
Lower limit of uncertainty in log(
SF R )133 F9.2 logSFRhiPEGASE
P´EGASE.2
Upper limit of uncertainty in log(
SF R )134 F8.2 agePEGASE
P´EGASE.2 galaxy age (Gyr)135 F8.2 minchi2PEGASE Reduced chi-squared of best
P´EGASE.2 fit136 I3 notes See list of notes in Table 4 a The full table is published in its entirety in the electronic edition of The Astrophysical Journal. Only the column names and tableformat is shown here.
Table 2
SDSS-II SN Candidates a CID RA DEC n e b IAUName
Type
Peakrmag MJDatPeakrmag n n s b n g b z Helio δz Helio objIDHost
679 327.434978 0.657569 3 2005eh Unknown 21.8 53699.2 1 0 0 0.124957 0.000017 1237656238472888902680 327.555405 0.842584 21 · · ·
Variable 21.6 53685.1 1 0 0 · · · · · · · · ·
Unknown 21.8 53656.2 0 0 0 0.048551 0.000022 1237678617405227407685 337.823273 -0.882037 14 · · · pSNII 21.7 53656.2 10 0 0 · · · · · · · · ·
Unknown 21.4 53616.3 5 0 0 0.067866 0.000010 1237656906347708594689 345.314592 -0.866253 15 · · ·
Variable 21.3 53680.2 15 0 0 · · · · · · · · ·
Unknown 20.3 53616.2 9 0 0 0.130903 0.000021 1237663542608986381692 351.071097 -0.945665 18 · · ·
Variable 21.2 53663.2 15 0 0 0.197275 0.000030 1237656906351182046694 330.154633 -0.623472 22 · · ·
Unknown 19.7 53627.2 28 0 0 0.127493 0.000018 1237663542609183155695 352.963374 -0.963772 3 · · ·
Variable 22.7 53637.3 0 0 0 0.058267 0.000009 1237656906351968456696 354.180048 -1.020436 6 · · · psNIa 21.3 53623.3 5 0 0 · · · · · · · · ·
697 335.002430 -0.626145 8 · · ·
Unknown 21.6 53627.2 5 0 0 0.156675 0.000032 1237663542611280181698 335.302586 -0.554336 17 · · ·
Variable 21.6 53663.2 14 0 0 · · · · · · · · ·
Variable 21.7 53656.2 15 0 0 · · · · · · · · ·
AGN 21.5 53616.2 14 0 0 0.595712 0.000242 1237663542611542434 · · · a This table is a portion of the full SN catalog, which is published in its entirety as Table 1 in the electronic edition of The Astrophysical Journal. Selected columnsrelating to general properties of the entries are shown here for guidance regarding the form and content of these columns. b In the electronic edition ne , n ns , ng are called Nsearchepoch , NepochSNR5 , nSNspec and nGALspec , respectively. Table 3
Number of SN Candidates by type categoryType Type Code NumberUnknown 0 1584Variable 5 3225pSNII 101 1841pSNIbc 102 24pSNIa 103 677zSNII 104 411zSNIbc 105 62zSNIa 106 907AGN 110 906SLSN 114 3SNIb 111,115 a a a ,120 500Total 10,258 a The indicated types were confirmed with spectra obtained by observers who were unaffiliated with SDSS. DSS-II SN Data Release 25
Table 4
Explanation of SN Notes column (Item 136)Note Explanation1 SN typing based on spectra obtained by groups outside SDSS.The spectra used for typing are not included in the data release.2 Peculiar type Ia SN possibly similar to sn91bg3 Peculiar type Ia SN possibly similar to sn00cx4 Peculiar type Ia SN possibly similar to sn02ci5 Peculiar type Ia SN possibly similar to sn02cx
Table 5
PSNID/NN Typing Efficiency and PuritySN Type z -prior Efficiency PurityIa flat 97.5% 94.8% · · · z spec · · · z spec · · · z spec Table 6
Normalized residuals.Band Nominal Adjustment Corrected Corrected s < s > u g r i z Table 7
SDSS AB Offsets.Band AB Offset u − . g +0 . r +0 . i +0 . z +0 . Note . — All magnitudes in this paper are SDSS asinh magnitudes (Lupton et al. 1999) in the native system used by SDSS. The AB offsetsshould be added to the native magnitudes to obtain magnitudes calibrated to the AB system. Fluxes are expressed in mu J and have the ABoffsets already applied. The derivation of the AB offsets is described in the text and in more detail in Betoule et al. (2013).
Table 8
Instrument ConfigurationsTelescope Instrument Wavelength Range Resolution Reference or Link˚ A ˚ A HET LRS 4070 – 10700 20 Hill et al., 1998ARC DIS 3100 – 9800 8-9 Link a Subaru FOCAS 3650 – 6000 8 Kashikawa et al., 20004900 – 9000 12WHT ISIS 3900 – 8900 4.3 & 7.5 Link b MDM CCDS 3800 – 7300 15 Link c Keck LRIS 3200 – 9400 4.5 & 8.9 Oke et al., 1995TNG DOLORES 3800 – 7300 10 Link d NTT EMMI 3800 – 9200 17 Dekker et al., 1986NOT ALFOSC-FASU 3200 – 9100 21 Link e Magellan LDSS3 3800 – 9200 9.5 Link f SALT RSS 3800 – 8000 5.7 Burgh et al., 2003
Table 9
Spectroscopic Data a SDSS ID b Spec ID c Telescope Type(s) Observation Date Evaluation SN redshift Galaxy redshift701 2795 APO Gal 2008-09-02 Gal · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · a The full table is published in its entirety in the electronic edition of The Astrophysical Journal. A portion is shown here for guidanceregarding its form and content. b Internal SN candidate designation. c Internal spectrum identification number.
Table 10
Selection criteria for
SALT2 light curve fitsSNANA variable MLCS2k2
SALT2 (4-par)
SALT2 (5-par)redshift a · · · (0.45) · · · (0.7) · · · ( · · · )redshift err b · · · (0.011) · · · (0.011) · · · ( · · · )SNRMAX c · · · (3.0) · · · (3.0) · · · (3.0)Trestmin e · · · (10.0) · · · (10.0) · · · (0.0)Trestmax f · · · (0.0) · · · (0.0) · · · (10.0) a Maximum redshift selected. b Maximum redshift uncertainty for 4-parameter fits. c Maximum signal-to-noise ratio among epochs in one band. d The number of filters that must have at least one epoch meeting the SNRMAX requirement. e The earliest epoch measured in days in the rest frame relative to maximum light in B -band must occur before this time. f The latest epoch measured in days in the rest frame relative to maximum light in B -band must occur after this time. DSS-II SN Data Release 27
Table 11
SALT2 4-parameter Fit Results a CID Classification z b x δx x δx c b δc b t max b δt max b µ c δµ c P c χ dof c
703 zSNIa 0.298042 5.43e-05 3.47e-06 0.73 0.63 -0.01 0.05 53626.5 0.7 40.80 0.25 0.966 40.80 59735 zSNIa 0.190858 8.82e-05 1.23e-05 -2.66 0.58 0.01 0.09 53610.7 1.8 39.59 0.27 0.955 20.60 33739 SNIa 0.107638 4.05e-04 3.34e-05 -0.88 0.20 -0.00 0.04 53609.5 1.1 38.31 0.19 0.001 58.80 29744 SNIa 0.128251 2.74e-04 1.57e-05 1.37 0.37 0.06 0.03 53612.9 0.9 38.97 0.20 0.983 17.40 32762 SNIa 0.191381 1.29e-04 4.84e-06 1.09 0.29 -0.05 0.03 53625.2 0.3 40.04 0.20 0.802 46.90 56774 SNIa 0.093331 6.30e-04 2.63e-05 0.79 0.19 -0.05 0.03 53608.5 0.0 38.27 0.19 0.806 25.00 32779 zSNIa 0.238121 7.72e-05 3.68e-06 0.46 0.38 0.02 0.04 53626.9 0.4 40.30 0.21 0.991 42.80 67822 zSNIa 0.237556 6.82e-05 3.48e-06 -0.38 0.54 -0.09 0.04 53621.3 0.5 40.60 0.24 0.454 53.50 53841 zSNIa 0.299100 5.59e-05 3.88e-06 0.33 0.64 -0.14 0.05 53624.9 0.6 41.07 0.26 0.994 39.30 64859 zSNIa 0.278296 6.57e-05 3.33e-06 0.68 0.51 0.03 0.04 53624.2 0.7 40.49 0.23 0.710 69.70 77893 zSNIa 0.110133 8.20e-05 4.06e-06 -1.18 0.45 0.04 0.04 53620.2 0.5 39.87 0.21 0.006 69.90 43904 zSNIa 0.385316 3.79e-05 3.07e-06 1.13 2.42 -0.28 0.07 53620.6 4.0 42.06 0.43 0.992 40.00 64911 zSNIa 0.207264 4.97e-05 3.59e-06 -0.39 0.74 0.23 0.06 53621.7 0.8 40.00 0.26 0.800 45.10 54932 zSNIa 0.391335 3.13e-05 3.35e-06 3.39 1.38 0.01 0.07 53619.0 0.8 41.84 0.40 0.796 58.20 68986 zSNIa 0.280578 4.22e-05 2.74e-06 -0.23 1.03 0.01 0.06 53619.8 1.6 40.86 0.30 0.991 43.40 68 · · · a This table is a portion of the full SN catalog, which is published in its entirety as Table 1 in the electronic edition of The Astrophysical Journal.Selected columns relating to 4-parameter SALT2 light curve fits are shown here for guidance regarding the form and content of these columns.b In the electronic edition z , x , δx , x , δx , c , δc , t max , and δt max are called zspecHelio , x0SALT2zspec , x1SALT2zspec , x1errSALT2zspec , cSALT2zspec , cerrSALT2zspec , peakMJDSALT2zspec , peakMJDerrSALT2zspec , respectively.c In the electronic edition µ , δµ , P , χ , and dof are called muSALT2zspec , muerrSALT2zspec , fitprobSALT2zspec , chi2SALT2zspec , and ndofSALT2zspec ,respectively. Table 12
Derived Host Galaxy Parameters from FSPS and P´EGASE.2 a FSPS P´EGASE.2
CID objIDHost log( M ) b log(sSFR) b log(age) b χ r c log( M ) d log(SFR) d log(age) d χ
679 1237656238472888902 10 . +0 . − . − . +0 . − . . +1 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +2 . − . . +0 . − . − . +9 . − . . +0 . − . − . +0 . − . . +1 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +1 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +2 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +1 . − . . +0 . − . − . +9 . − . . +0 . − . − . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . − . +4 . − . . +2 . − . . +0 . − . − . +9 . − . . +0 . − . − . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +1 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +1 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +1 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . . +1 . − . . +0 . − . . +0 . − . · · · a This table is a portion of the full SN catalog, which is published in its entirety as Table 1 in the electronic edition of The Astrophysical Journal.Selected columns relating to host galaxy properties are shown here for guidance regarding the form and content of these columns.b In the electronic edition log ( M ) is called logMassFSPS and the upper limit is logMasshiFSPS and the lower limit is logMassloFSPS . log( sSF R ) iscalled logSSFRFSPS and the upper and lower limits are logSSFRhiFSPS and logSSFRloFSPS , respectively, and age is called ageFSPS with upper andlower limits agehiFSPS and ageloFSPS .c The reduced χ value of the fit. This column is called minredchi2FSPS in the electronic edition.d In the electronic edition log ( M ) is called logMassPEGASE and the upper limit is logMasshiPEGASE and the lower limit is logMassloPEGASE . log( SF R )is called logSFRPEGASE and the upper and lower limits are logSFRhiPEGASE and logSFRloPEGASE , respectively, and age is called agePEGASE .e The χ value of the fit. This column is called minchi2PEGASEminchi2PEGASE