Lawrence Berkeley National Laboratory Recent Work

Title The Fifteenth Data Release of the Sloan Digital Sky Surveys: First Release of MaNGA-derived Quantities, Data Visualization Tools, and Stellar Library

Permalink https://escholarship.org/uc/item/9fp0m8s1

Journal Astrophysical Journal, Supplement Series, 240(2)

ISSN 0067-0049

Authors Aguado, DS Ahumada, R Almeida, A et al.

Publication Date 2019-02-01

DOI 10.3847/1538-4365/aaf651

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California Draft version December 12, 2018 Preprint typeset using LATEX style emulateapj v. 12/16/11

THE FIFTEENTH DATA RELEASE OF THE SLOAN DIGITAL SKY SURVEYS: FIRST RELEASE OF MANGA DERIVED QUANTITIES, DATA VISUALIZATION TOOLS AND STELLAR LIBRARY

D. S. Aguado1, Romina Ahumada2, Andres´ Almeida3, Scott F. Anderson4, Brett H. Andrews5, Borja Anguiano6, Erik Aquino Ort´ız7, Alfonso Aragon-Salamanca´ 8, Maria Argudo-Fernandez´ 9,10, Marie Aubert11, Vladimir Avila-Reese7, Carles Badenes5, Sandro Barboza Rembold12,13, Kat Barger14, Jorge Barrera-Ballesteros15, Dominic Bates16, Julian Bautista17, Rachael L. Beaton18, Timothy C. Beers19, Francesco Belfiore20, Mariangela Bernardi21, Matthew Bershady22, Florian Beutler17, Jonathan Bird23, Dmitry Bizyaev24,25, Guillermo A. Blanc18, Michael R. Blanton26, Michael Blomqvist27, Adam S. Bolton28, Med´ eric´ Boquien9, Jura Borissova29,30, Jo Bovy31,32, William Nielsen Brandt33,34,35, Jonathan Brinkmann24, Joel R. Brownstein36, Kevin Bundy20, Adam Burgasser37, Nell Byler4, Mariana Cano Diaz7, Michele Cappellari38, Ricardo Carrera39, Bernardo Cervantes Sodi40, Yanping Chen41, Brian Cherinka15, Peter Doohyun Choi42, Haeun Chung43, Damien Coffey44, Julia M. Comerford45, Johan Comparat44, Kevin Covey46, Gabriele da Silva Ilha12,13, Luiz da Costa13,47, Yu Sophia Dai48, Guillermo Damke3,50, Jeremy Darling45, Roger Davies38, Kyle Dawson36, Victoria de Sainte Agathe51, Alice Deconto Machado12,13, Agnese Del Moro44, Nathan De Lee23, Aleksandar M. Diamond-Stanic52, Helena Dom´ınguez Sanchez´ 21, John Donor14, Niv Drory53, Helion´ du Mas des Bourboux36, Chris Duckworth16, Tom Dwelly44, Garrett Ebelke6, Eric Emsellem54,55, Stephanie Escoffier11, Jose´ G. Fernandez-Trincado´ 56,2,57, Diane Feuillet58, Johanna-Laina Fischer21, Scott W. Fleming59, Amelia Fraser-McKelvie8, Gordon Freischlad24, Peter M. Frinchaboy14, Hai Fu60, Llu´ıs Galbany5, Rafael Garcia-Dias1,61, D. A. Garc´ıa-Hernandez´ 1,61, Luis Alberto Garma Oehmichen7, Marcio Antonio Geimba Maia13,47,Hector´ Gil-Mar´ın62,63, Kathleen Grabowski24, Meng Gu64, Hong Guo65, Jaewon Ha42, Emily Harrington66,67, Sten Hasselquist68, Christian R. Hayes6, Fred Hearty33, Hector Hernandez Toledo7, Harry Hicks69, David W. Hogg26, Kelly Holley-Bockelmann23, Jon A. Holtzman68, Bau-Ching Hsieh70, Jason A. S. Hunt32, Ho Seong Hwang43,Hector´ J. Ibarra-Medel7, Camilo Eduardo Jimenez Angel1,61, Jennifer Johnson71, Amy Jones72, Henrik Jonsson¨ 73, Karen Kinemuchi24,68, Juna Kollmeier18, Coleman Krawczyk17, Kathryn Kreckel58, Sandor Kruk38, Ivan Lacerna74,30, Ting-Wen Lan75, Richard R. Lane76,30, David R. Law59, Young-Bae Lee42, Cheng Li77, Jianhui Lian17, Lihwai Lin (林俐暉)70, Yen-Ting Lin70, Chris Lintott38, Dan Long24, Penelope´ Longa-Pena˜ 9, J. Ted Mackereth78, Axel de la Macorra7, Steven R. Majewski6, Olena Malanushenko24, Arturo Manchado1,61,79, Claudia Maraston17, Vivek Mariappan33, Mariarosa Marinelli80, Rui Marques-Chaves1,61, Thomas Masseron1,61, Karen L. Masters (何凱論)67,17,81, Richard M. McDermid82, Nicolas´ Medina Pena˜ 29, Sofia Meneses-Goytia17, Andrea Merloni44, Michael Merrifield8, Szabolcs Meszaros83,84, Dante Minniti85,30,86, Rebecca Minsley52, Demitri Muna87, Adam D. Myers88, Preethi Nair72, Janaina Correa do Nascimento12,13, Jeffrey A. Newman5, Christian Nitschelm9, Matthew D Olmstead89, Audrey Oravetz24, Daniel Oravetz24, Rene´ A. Ortega Minakata7, Zach Pace22, Nelson Padilla76, Pedro A. Palicio1,61, Kaike Pan24, Hsi-An Pan70, Taniya Parikh17, James Parker III24, Sebastien Peirani90, Samantha Penny17, Will J. Percival91,92,17, Ismael Perez-Fournon1,61, Thomas Peterken8, Marc Pinsonneault71, Abhishek Prakash93, Jordan Raddick15, Anand Raichoor94, Rogemar A. Riffel12,13, Rogerio´ Riffel95,13, Hans-Walter Rix58, Annie C. Robin57, Alexandre Roman-Lopes49, Benjamin Rose19, Ashley J. Ross71, Graziano Rossi42, Kate Rowlands15, Kate H. R. Rubin96, Sebastian´ F. Sanchez´ 7, Jose´ R. Sanchez-Gallego´ 4, Conor Sayres4, Adam Schaefer22, Ricardo P. Schiavon78, Jaderson S. Schimoia12,13, Edward Schlafly97, David Schlegel97, Donald Schneider33,34, Mathias Schultheis98, Hee-Jong Seo99, Shoaib J. Shamsi67, Zhengyi Shao65, Shiyin Shen65, Shravan Shetty22, Gregory Simonian71, Rebecca Smethurst8, Jennifer Sobeck4, Barbara J. Souter15, Ashley Spindler100, David V. Stark75, Keivan G. Stassun23, Matthias Steinmetz101, Thaisa Storchi-Bergmann95,13, Guy S. Stringfellow45, Genaro Suarez´ 7, Jing Sun14, Manuchehr Taghizadeh-Popp15,102, Michael S. Talbot36, Jamie Tayar71, Aniruddha R. Thakar15, Daniel Thomas17, Patricia Tissera85, Rita Tojeiro16, Nicholas W. Troup6, Eduardo Unda-Sanzana9, Octavio Valenzuela7, Mariana Vargas-Magana˜ 103, Jose Antonio Vazquez Mata7, David Wake104, Benjamin Alan Weaver28, Anne-Marie Weijmans16, Kyle B. Westfall20, Vivienne Wild16, John Wilson6, Emily Woods52, Renbin Yan105, Meng Yang16, Olga Zamora1,61, Gail Zasowski36, Kai Zhang105, Zheng Zheng48, Zheng Zheng36, Guangtun Zhu15,106, Joel C. Zinn71, Hu Zou48 Draft version December 12, 2018 arXiv:1812.02759v2 [astro-ph.IM] 10 Dec 2018 Abstract Twenty years have passed since first light for the Sloan Digital Sky Survey (SDSS). Here, we release data taken by the fourth phase of SDSS (SDSS-IV) across its first three years of operation (July 2014–July 2017). This is the third data release for SDSS-IV, and the fifteenth from SDSS (Data Release Fifteen; DR15). New data come from MaNGA – we release 4824 datacubes, as well as the first stellar spectra in the MaNGA Stellar Library (MaStar), the first set of survey-supported analysis products (e.g. stellar and gas kinematics, emission line, and other maps) from the MaNGA Data Analysis Pipeline (DAP), and a new data visualisation and access tool we call “Marvin”. The next data release, DR16, will include new data from both APOGEE-2 and eBOSS; those surveys release no new data here, but we document updates and corrections to their data processing pipelines. The release is cumulative; it also includes the most recent reductions and calibrations of all data taken by SDSS since first light. In this paper we describe the location and format of the data and tools and cite technical references describing how it was obtained and processed. The SDSS website (www.sdss.org) 2 SDSS Collaboration

has also been updated, providing links to data downloads, tutorials and examples of data use. While SDSS-IV will continue to collect astronomical data until 2020, and will be followed by SDSS-V (2020– 2025), we end this paper by describing plans to ensure the sustainability of the SDSS data archive for many years beyond the collection of data. Subject headings: Atlases — Catalogs — Surveys

1. INTRODUCTION The Sloan Digital Sky Survey (SDSS; York et al. 2000) data releases began with the Early Data Release, or [email protected] 1 Instituto de Astrof´ısica de Canarias, E-38205 La Laguna, 32 Dunlap Institute for Astronomy and Astrophysics, Univer- Tenerife, Spain sity of Toronto, 50 St. George Street, Toronto, Ontario M5S 2 Departamento de Astronomia, Casilla 160-C, Universidad 3H4, Canada de Concepcion, Concepcion, Chile 33 Department of Astronomy and Astrophysics, Eberly Col- 3 Instituto de Investigaci´on Multidisciplinario en Ciencia lege of Science, The Pennsylvania State University, 525 Davey y Tecnolog´ıa, Universidad de La Serena, Benavente 980, La Laboratory, University Park, PA 16802, USA Serena, Chile 34 Institute for Gravitation and the Cosmos, The Pennsylva- 4 Department of Astronomy, Box 351580, University of nia State University, University Park, PA 16802, USA Washington, Seattle, WA 98195, USA 35 Department of Physics, The Pennsylvania State University, 5 PITT PACC, Department of Physics and Astronomy, University Park, PA 16802, USA University of Pittsburgh, Pittsburgh, PA 15260, USA 36 Department of Physics and Astronomy, University of Utah, 6 Department of Astronomy, University of Virginia, 530 115 S. 1400 E., Salt Lake City, UT 84112, USA McCormick Road, Charlottesville, VA 22904-4325, USA 37 Center for Astrophysics and Space Science, University of 7 Instituto de Astronom´ıa, Universidad Nacional Aut´onoma California San Diego, La Jolla, CA 92093, USA de M´exico,A.P. 70-264, 04510, M´exico,D.F., M´exico 38 Sub-department of Astrophysics, Department of Physics, 8 School of Physics & Astronomy, University of Nottingham, University of Oxford, Denys Wilkinson Building, Keble Road, Nottingham, NG7 2RD, UK Oxford OX1 3RH, UK 9 Centro de Astronom´ıa(CITEVA), Universidad de Antofa- 39 Astronomical Observatory of Padova, National Institute of gasta, Avenida Angamos 601 Antofagasta, Chile Astrophysics, Vicolo Osservatorio 5 - 35122 - Padova, Italy 10 Chinese Academy of Sciences South America Center for 40 Instituto de Radioastronom´ıa y Astrof´ısica, Universidad Astronomy, China-Chile Joint Center for Astronomy, Camino Nacional Aut´onomade M´exico, Campus Morelia, A.P. 3-72, El Observatorio 1515, Las Condes, Santiago, Chile C.P. 58089 Michoac´an,M´exico 11 Aix Marseille Univ, CNRS/IN2P3, CPPM, Marseille, 41 New York University Abu Dhabi, P. O BOX 129188, Abu France Dhabi, UAE 12 Departamento de F´ısica, CCNE, Universidade Federal de 42 Department of Astronomy and Space Science, Sejong Santa Maria, 97105-900, Santa Maria, RS, Brazil University, Seoul 143-747, Republic of Korea 13 Laborat´orio Interinstitucional de e-Astronomia, 77 Rua 43 Korea Institute for Advanced Study, 85 Hoegiro, General Jos´eCristino, Rio de Janeiro, 20921-400, Brazil Dongdaemun-gu, Seoul 02455, Republic of Korea 14 Department of Physics and Astronomy, Texas Christian 44 Max-Planck-Institut f¨urextraterrestrische Physik, Gießen- University, Fort Worth, TX 76129, USA bachstr. 1, D-85748 Garching, Germany 15 Department of Physics and Astronomy, Johns Hopkins 45 Center for Astrophysics and Space Astronomy, Department University, 3400 N. Charles St., Baltimore, MD 21218, USA of Astrophysical and Planetary Sciences, University of Colorado, 16 School of Physics and Astronomy, University of St An- 389 UCB, Boulder, CO 80309-0389, USA drews, North Haugh, St Andrews, KY16 9SS, UK 46 Department of Physics and Astronomy, Western Washing- 17 Institute of Cosmology & Gravitation, University of ton University, 516 High Street, Bellingham, WA 98225, USA Portsmouth, Dennis Sciama Building, Portsmouth, PO1 3FX, 47 Observat´orio Nacional, R. Gal. Jose Cristino 77, Rio de UK Janeiro, RJ 20921-400, Brazil 18 The Observatories of the Carnegie Institution for Science, 48 National Astronomical Observatories, Chinese Academy of 813 Santa Barbara St., Pasadena, CA 91101, USA Sciences, 20A Datun Road, Chaoyang District, Beijing 100012, 19 Department of Physics and JINA Center for the Evolution China of the Elements, University of Notre Dame, Notre Dame, IN 49 Departamento de F´ısica,Facultad de Ciencias, Universidad 46556, USA de La Serena, Cisternas 1200, La Serena, Chile 20 University of California Observatories, University of Cali- 50 AURA Observatory in Chile, Cisternas 1500, La Serena, fornia, Santa Cruz, CA 95064, USA Chile 21 Department of Physics and Astronomy, University of 51 LPNHE, CNRS/IN2P3, Universit´ePierre et Marie Curie Pennsylvania, Philadelphia, PA 19104, USA Paris 6, Universit´eDenis Diderot Paris, 4 place Jussieu, 75252 22 Department of Astronomy, University of Wisconsin- Paris CEDEX, France Madison, 475 N. Charter St., Madison, WI 53726, USA 52 Department of Physics and Astronomy, Bates College, 44 23 Vanderbilt University, Department of Physics & Astron- Campus Avenue, Lewiston, ME 04240, USA omy, 6301 Stevenson Center Ln., Nashville, TN 37235, USA 53 McDonald Observatory, The University of Texas at Austin, 24 Apache Point Observatory, P.O. Box 59, Sunspot, NM 1 University Station, Austin, TX 78712, USA 88349, USA 54 European Southern Observatory, Karl-Schwarzschild-Str. 25 Sternberg Astronomical Institute, Moscow State Univer- 2, 85748 Garching, Germany sity, Universitetskij pr. 13, 119991 Moscow, Russia 55 Univ Lyon, Univ Lyon1, Ens de Lyon, CNRS, Centre 26 Center for Cosmology and Particle Physics, Department of de Recherche Astrophysique de Lyon UMR5574, F-69230 Physics, New York University, 726 Broadway, Room 1005, New Saint-Genis-Laval France York, NY 10003, USA 56 Instituto de Astronom´ıay Ciencias Planetarias, Universi- 27 Aix Marseille Univ, CNRS, LAM, Laboratoire dad de Atacama, Copayapu 485, Copiap´o,Chile d’Astrophysique de Marseille, Marseille, France 57 Institut UTINAM, CNRS UMR6213, Univ. Bourgogne 28 National Optical Astronomy Observatory, 950 North Franche-Comt´e, OSU THETA Franche-Comt´e-Bourgogne, Cherry Avenue, Tucson, AZ 85719, USA Observatoire de Besan¸con, BP 1615, 25010 Besan¸con Cedex, 29 Instituto de F´ısica y Astronom´ıa, Universidad de Val- France para´ıso, Av. Gran Breta˜na1111, Playa Ancha, Casilla 5030, 58 Max-Planck-Institut f¨ur Astronomie, K¨onigstuhl 17, D- Chile 69117 Heidelberg, Germany 30 Instituto Milenio de Astrof´ısica, Av. Vicu˜naMackenna 59 Space Telescope Science Institute, 3700 San Martin Drive, 4860, Macul, Santiago, Chile Baltimore, MD 21218, USA 31 Department of Astronomy and Astrophysics, University of 60 Department of Physics & Astronomy, University of Iowa, Toronto, 50 St. George Street, Toronto, ON, M5S 3H4, Canada Iowa City, IA 52245, USA Data Release 15 3

EDR, in June 2001 (Stoughton et al. 2002) and have SDSS-II (DR6–DR7; Frieman et al. 2008; Yanny et al. been heavily used by astronomers and the broader pub- 2009), SDSS-III (DR8–DR12; Eisenstein et al. 2011), and lic since that time (Raddick et al. 2014a,b). Here we SDSS-IV (DR13–DR15; Blanton et al. 2017). Kollmeier present the fifteenth public data release from SDSS, or et al. (2017) describe the plans for SDSS-V, to start in DR15, made publicly available on 10th December 2018. mid-2020. SDSS has been marked by four phases so far, with The data releases contain information about SDSS op- plans for a fifth. Details are available in the papers de- tical broad band imaging, optical spectroscopy, and in- scribing SDSS-I (EDR, DR1–DR5; York et al. 2000), frared spectroscopy. Currently, SDSS-IV conducts opti- cal and infrared spectroscopy (using two dedicated spec- trographs; Smee et al. 2013; Wilson et al. 2018) at the 61 Departamento de Astrof´ısica, Universidad de La Laguna 2.5-meter Sloan Foundation Telescope at Apache Point (ULL), E-38206 La Laguna, Tenerife, Spain Observatory (APO; Gunn et al. 2006) and infrared spec- 62 Sorbonne Universit´es,Institut Lagrange de Paris (ILP), 98 bis Boulevard Arago, 75014 Paris, France troscopy at the du Pont Telescope at Las Campanas Ob- 63 Laboratoire de Physique Nucl´eaireet de Hautes Energies, servatory (LCO; Bowen & Vaughan 1973). Universit´ePierre et Marie Curie, 4 Place Jussieu, 75005 Paris, SDSS-IV began observations in July 2014, and consists France 64 Harvard-Smithsonian Center for Astrophysics, 60 Garden of three programs: St., Cambridge, MA 02138, USA 65 Shanghai Astronomical Observatory, Chinese Academy of 1. The extended Baryon Oscillation Spectroscopic Science, 80 Nandan Road, Shanghai 200030, China Survey (eBOSS; Dawson et al. 2016) is surveying 66 Department of Physics, Bryn Mawr College, Bryn Mawr, galaxies and quasars at redshifts z ∼ 0.6–3.5 for PA 19010, USA 67 Department of Physics and Astronomy, Haverford College, large scale structure. It includes two major sub- 370 Lancaster Avenue, Haverford, PA 19041, USA programs: 68 Department of Astronomy, New Mexico State University, Box 30001, MSC 4500, Las Cruces NM 88003, USA SPectroscopic IDentification of ERosita 69 School of Maths and Physics, University of Portsmouth, • Portsmouth, PO1 3FX, UK Sources (SPIDERS; Dwelly et al. 2017) 70 Academia Sinica Institute of Astronomy and Astrophysics, investigates the nature of X-ray emitting P.O. Box 23-141, Taipei 10617, Taiwan sources, including active galactic nuclei and 71 Department of Astronomy, Ohio State University, 140 W. 18th Ave., Columbus, OH 43210, USA galaxy clusters. 72 Department of Physics and Astronomy, University of Time Domain Spectroscopic Survey (TDSS; Alabama, Tuscaloosa, AL 35487-0324, USA • 73 Lund Observatory, Department of Astronomy and Theoret- Morganson et al. 2015) is exploring the phys- ical Physics, Lund University, Box 43, SE-22100 Lund, Sweden ical nature of time-variable sources through 74 Instituto de Astronom´ıa, Universidad Cat´olicadel Norte, spectroscopy. Av. Angamos 0610, Antofagasta, Chile 75 Kavli Institute for the Physics and Mathematics of the Universe, Todai Institutes for Advanced Study, the University 2. Mapping Nearby Galaxies at APO (MaNGA; of Tokyo, Kashiwa, Japan 277- 8583 Bundy et al. 2015) uses integral field spectroscopy 76 Instituto de Astrof´ısica,Pontificia Universidad Cat´olicade Chile, Av. Vicuna Mackenna 4860, 782-0436 Macul, Santiago, Chile 92 Perimeter Institute for Theoretical Physics, 31 Caroline St. 77 Tsinghua Center for Astrophysics & Department of North, Waterloo, ON N2L 2Y5, Canada Physics, Tsinghua University, Beijing 100084, China 93 Infrared Processing and Analysis Center, California Insti- 78 Astrophysics Research Institute, Liverpool John Moores tute of Technology, MC 100-22, 1200 E California Boulevard, University, IC2, Liverpool Science Park, 146 Brownlow Hill, Pasadena, CA 91125, USA Liverpool L3 5RF, UK 94 Institute of Physics, Laboratory of Astrophysics, Ecole 79 Consejo Superior de Investigaciones Cient´ıficas, Spain. Polytechnique F´ed´eralede Lausanne (EPFL), Observatoire de [email protected] Sauverny, 1290 Versoix, Switzerland 80 Department of Physics, Virginia Commonwealth Univer- 95 Instituto de F´ısica,Universidade Federal do Rio Grande do sity, Richmond, VA 23220-4116, USA Sul Av. Bento Gon¸calves 9500, CEP 91501-970, Porto Alegre, 81 SDSS-IV Spokesperson (Corresponding Author, spokesper- RS, Brazil. [email protected]) 96 Department of Astronomy, San Diego State University, San 82 Diego, CA 92182, USA Department of Physics and Astronomy, Macquarie Univer- 97 sity, Sydney NSW 2109, Australia Lawrence Berkeley National Laboratory, 1 Cyclotron Road, 83 Berkeley, CA 94720, USA ELTE Gothard Astrophysical Observatory, H-9704 Szom- 98 bathely, Szent Imre herceg st. 112, Hungary Laboratoire Lagrange, Universit´eCˆoted’Azur, Observa- 84 Premium Postdoctoral Fellow of the Hungarian Academy toire de la Cˆoted’Azur, CNRS, Blvd de l’Observatoire, F-06304 Nice, France of Sciences 99 85 Departamento de F´ısica, Facultad de Ciencias Exactas, Department of Physics and Astronomy, Ohio University, Clippinger Labs, Athens, OH 45701, USA Universidad Andres Bello, Av. Fernandez Concha 700, Las 100 Condes, Santiago, Chile Department of Physical Sciences, The Open University, 86 Milton Keynes, MK7 6AA, UK Vatican Observatory, V00120 Vatican City State, Italy 101 87 Center for Cosmology and AstroParticle Physics, The Ohio Leibniz-Institut f¨urAstrophysik Potsdam (AIP), An der Sternwarte 16, D-14482 Potsdam, Germany State University, 191 W. Woodruff Ave., Columbus, OH 43210, 102 USA Institute for Data Intensive Engineering and Science, 88 Department of Physics and Astronomy, University of Johns Hopkins University, 3400 N. Charles St, Baltimore, MD 21218, USA Wyoming, Laramie, WY 82071, USA 103 89 Kings College, 133 North River St, Wilkes-Barre, PA 18711 Instituto de F´ısica, Universidad Nacional Aut´onomade M´exico,Apdo. Postal 20-364, M´exico USA 104 90 Institut d‘Astropysique de Paris, UMR 7095, CNRS - Department of Physics, University of North Carolina Asheville, One University Heights, Asheville, NC 28804, USA UPMC, 98bis bd Arago, 75014 Paris, France 105 91 Department of Physics and Astronomy, University of Department of Physics and Astronomy, University of Kentucky, 505 Rose St., Lexington, KY, 40506-0055, USA Waterloo, 200 University Ave W, Waterloo, ON N2L 3G1, 106 Canada Hubble Fellow 4 SDSS Collaboration

(IFS) to study a representative sample of ∼10,000 by the Data Analysis Pipeline (see section 4.1.2). nearby galaxies. These products are available for all data cubes in DR15, with the exception of the cubes generated 3. APOGEE-2 (the second phase of the APO Galac- by some ancillary programs (i.e., Coma, IC342 and tic Evolution Experiment or APOGEE; Majewski M31) if they do not have redshifts (e.g. sky fields). et al. 2017) performs a large-scale and systematic investigation of the entire Milky Way Galaxy with 3. Alongside the new MaNGA data, and data prod- near-infrared, high-resolution, and multiplexed in- ucts, DR15 also marks the launch of Marvin: a new strumentation. tool to visualize and analyze MaNGA datacubes and maps (see section 4.2). SDSS-IV has had two previous data releases (DR13 and DR14; Albareti et al. 2017; Abolfathi et al. 2018 4. DR15 is the first public data release for the respectively), containing the first two years of eBOSS, MaNGA Stellar Library MaStar (see section 4.3), MaNGA, and APOGEE-2 data and new calibrations of containing 3326 optical stellar spectra. the SDSS imaging data set. DR15 contains new reductions and new data from 5. In addition to updates to two previously released MaNGA. This release includes the first three years of Value Added Catalogs (VACs), DR15 also includes MaNGA data plus a new suite of derived data products six new VACs contributed by the MaNGA team based on the MaNGA data cubes, a new data access tool (see Table 2). This brings the total number of for MaNGA known as Marvin, and data from a large an- VACs in the SDSS public data releases to 40. cillary programme aimed at improving the stellar library available for MaNGA (MaStar; the MaNGA Stellar Li- 6. Finally, DR15 includes a re-release of all previ- brary). ous versions of SDSS data releases. This includes The full scope of the data release is described in Sec- the most recent data releases for APOGEE-2 and tion 2, and information on data distribution is given in eBOSS (described in Abolfathi et al. 2018, DR14), Section 3. Each of the sub-surveys is described in its and the most recent release of the SDSS imag- own section, with MaNGA in Section 4 and APOGEE-2 ing data (described in Albareti et al. 2017, DR13). and eBOSS (including SPIDERS and TDSS) in Section Data of previous SDSS surveys are also included: 5.1 and 5.2, respectively. We discuss future plans for the Legacy Spectra were finalized in DR8 (Aihara SDSS-IV and beyond in Section 6. Readers wanting a et al. 2011), and the SEGUE-1 and SEGUE-2 spec- glossary of terms and acronyms used in SDSS can find tra in DR9 (Ahn et al. 2012). The MARVELS one at https://www.sdss.org/dr15/help/glossary/. spectra were last re-reduced for DR12 (Alam et al. 2015). 2. SCOPE OF DATA RELEASE 15

As with all previous SDSS public data releases, DR15 is 3. DATA DISTRIBUTION cumulative and includes all data products that have been publicly released in earlier SDSS data releases. All previ- The DR15 data can be accessed through a variety of ous releases are archived online to facilitate science repli- mechanisms, depending on the type of data file and the cation; however we recommend new users always make needs of the user. All data access methods are described use of the latest DR (even when using older data) to en- on the SDSS Web site (https://www.sdss.org/dr15/ sure they are using the most recent reduction routines. data_access), and we also provide tutorials and exam- The scope of DR15 is shown in Table 1, and its compo- ples for accessing and working with SDSS data products nents can be summarized as follows. at (https://www.sdss.org/dr15/tutorials. We de- scribe our four main data access mechanisms below. 1. MaNGA integral-field spectroscopic data from 285 All raw and processed imaging and spectroscopic data plates, including 119 plates observed between 26 can be accessed through the Science Archive Server September 2016 (MJD 57658) and 29 June 2017 (SAS, (https://data.sdss.org/sas/dr15). This site (MJD 57934) that are newly released data in DR15. includes intermediate data products and VACs. The This data set is identical to the internally released SAS is a file-based system, from which data can be MaNGA Product Launch-7 (MPL-7), and contains directly downloaded by browsing, or in bulk mode us- the same set of galaxies but processed with a differ- ing rsync, wget or Globus Online. Bulk download- ent version of the reduction pipeline as the earlier ing methods are outlined at https://www.sdss.org/ internally released MPL-6. DR15 contains 4824 dr15/data_access/bulk. All data files available on reconstructed 3D data cubes, of which 4688 are the SAS have a data model (https://data.sdss.org/ target galaxies (the remainder are ancillary targets datamodel), which provides a detailed overview of the which include galaxies, parts of galaxies, and some content of each data file. deep sky fields). This dataset includes 67 repeat Processed optical and infrared spectra, as well as pro- observations, so that the total number of unique cessed imaging, can also be accessed on the SAS through galaxies in this data release is 4621. Most of these the Science Archive Webapp (SAW), an interactive web galaxies are part of the MaNGA main sample, but application (webapp; http://dr15.sdss.org links to ancillary target galaxies are also included in this the DR15 version). In DR15, the SAW is serving MaS- count (see Table 4 for a summary of these). tar spectra for the first time. Through this webapp, users can display individual spectra and overlay model fits in- 2. In addition to the MaNGA data cubes, DR15 also cluded on the SAS. There is a search option available to releases for the first time data products generated select spectra based on e.g., plate number, coordinates, Data Release 15 5

TABLE 1 Reduced SDSS-IV spectroscopic data in DR15

Target Category # DR13 # DR13+14 # DR13+14+15 eBOSS LRG samples 32968 138777 138777 ELG Pilot Survey 14459 35094 35094 Main QSO Sample 33928 188277 188277 Variability Selected QSOs 22756 87270 87270 Other QSO samples 24840 43502 43502 TDSS Targets 17927 57675 57675 SPIDERS Targets 3133 16394 16394 Standard Stars/White Dwarfs 53584 63880 63880 APOGEE-2 All Stars 164562 263444 263444 NMSU 1-meter stars 894 1018 1018 Telluric stars 17293 27127 27127 APOGEE-N Commissioning stars 11917 12194 12194 MaNGA Cubes 1390 2812 4824 MaNGA main galaxy sample: PRIMARY v1 2 600 1278 2126 SECONDARY v1 2 473 947 1665 COLOR-ENHANCED v1 2 216 447 710 MaStar (MaNGA Stellar Library) 0 0 3326 Other MaNGA ancillary targets1 31 121 324 1 Many MaNGA ancillary targets were also observed as part of the main galaxy sample, and are counted twice in this table; some ancillary targets are not galaxies.

TABLE 2 New or Updated Value Added Catalogs

Description Section Reference(s) Mini data release, 31 July 2018 eBOSS DR14 QSO LSS catalogs §5.2 Ata et al. (2018) eBOSS DR14 LRG LSS catalogs §5.2 Bautista et al. (2018) Optical Emission Line Properties and Black Hole Mass §5.3 Coffey et al. (2018) Estimates for SPIDERS DR14 Quasars Open Cluster Chemical Abundance and Mapping Catalog §5.1.2 Donor et al. (2018) DR15, 10 December 2018 GEMA-VAC: Galaxy Environment for MaNGA VAC §4.5.4 M. Argudo-Fernandez et al. (in prep). MaNGA Spectroscopic Redshifts §4.5.5 Talbot et al. (2018) MaNGA Pipe3D: Spatially resolved and integrated §4.5.1 S´anchez et al. (2016, 2018) properties of DR15 MaNGA galaxiesa MaNGA Firefly Stellar Populations a §4.5.1 Goddard et al. (2017a); Wilkinson et al. (2017); Parikh et al. (2018) MaNGA PyMorph DR15 photometric catalog §4.5.2 Fischer et al. (2018) MaNGA Morphology Deep Learning DR15 catalog §4.5.2 Dom´ınguezS´anchez et al. (2018) HI-MaNGA Data Release 1 §4.5.3 Masters et al. (2018) MaNGA Morphologies from Galaxy Zoo §4.5.2 Willett et al. (2013); Hart et al. (2016) aupdate to DR14 VAC redshift or ancillary observing program. Searches can be the CAS. The CAS can be accessed through the Sky- saved for future references as permalinks. Spectra can Server webapp (https://skyserver.sdss.org), which be directly downloaded from the SAS through the we- provides Explore tools as well as the option of browser- bapp, and links are included to the SkyServer Explore based queries in synchronous mode. CASJobs (https: page for each object. The user can select SDSS data re- //skyserver.sdss.org/casjobs) is suitable for more leases back to DR8 (the SAW was originally developed extensive queries, which are executed in asynchronous during SDSS-III so serves data from that phase of SDSS or batch mode, and offers users personal storage space onwards only), but is encouraged to always use the most for query results (Li & Thakar 2008). The CAS is in- recent data release at https://data.sdss.org/home. tegrated with SciServer (https://www.sciserver.org), MaNGA datacubes and maps are not available in the which offers several data-driven science services, includ- SAW, but can be visualized and analyzed through Mar- ing SciServer Compute: a system that allows users to vin. Marvin, described in detail in § 4.2, provides links run Jupyter notebooks in Docker containers, directly ac- to the SAS for downloading data files, as well as the Sky- cessing the SDSS catalogs. Server Explore page. All the data reduction software that is used by the The Catalog Archive Server (CAS; Thakar 2008; various SDSS-IV teams to reduce and process their data Thakar et al. 2008) stores the catalogs of DR15: these in- (including links to the Marvin Repository on Github) clude photometric, spectroscopic and derived properties. is publicly available at https://www.sdss.org/dr15/ Some of the VACs also have catalogs that are stored on software/products. 6 SDSS Collaboration

4. MANGA similar to those released in DR13 and DR14 (and iden- The MaNGA survey uses a custom built instrument tical to the internal collaboration release MPL-7), and (Drory et al. 2015) which feeds fibers from a suite of consist of multi-extension FITS files giving the flux, in- hexagonal bundles into the BOSS spectrograph (Smee verse variance, mask, and other information for each ob- et al. 2013). Over its planned five years of operations ject. The metadata from all of our observations is sum- MaNGA aims to get data for ∼10,000 nearby galaxies marized in a FITS binary table, “drpall-v2 4 3.fits”, (Law et al. 2015; Yan et al. 2016b,a; and see Wake et al. detailing the coordinates, targeting information, redshift, data quality, etc. The version of the MaNGA DRP used (2017) for details on the sample selection). 1 DR15 consists of MaNGA observations taken dur- for DR15 (v2 4 3 ) incorporates some significant changes ing the first three years of the survey (up to Summer compared to the DR14 version of the pipeline (v2 1 2). 2017) and nearly doubles the sample size of fully-reduced These changes include: galaxy data products previously released in DR14 (Abol- • The MaNGA DRP has been extended to produced fathi et al. 2018). These data products include raw data, one-dimensional reduced spectra for each of the intermediate reductions such as flux-calibrated spectra MaStar targets observed during bright time; de- from individual exposures, and final data cubes and tails of these modifications are described in 4.3. row-stacked spectra (RSS) produced using the MaNGA § Data Reduction Pipeline (DRP; Law et al. 2016). DR15 • DR15 introduces some significant changes in the includes DRP data products for 4,824 MaNGA cubes overall flux calibration relative to DR13/DR14 distributed amongst 285 plates, corresponding to 4,621 (and relative to the description by Yan et al. unique galaxies plus 67 repeat observations and 118 spe- 2016b). Foremost among these is the use of BOSZ cial targets from the ancillary programs (see §4.4). Un- stellar spectral models (Bohlin et al. 2017) instead like in previous data releases, data cubes and summary of the original Kurucz templates built in 2003 to RSS files are no longer produced for the twelve 7-fiber derive the spectrophotometric calibration based on mini-bundles on each plate that target bright stars and contemporal observations of standard stars with are used to derive the spectrophotometric calibration the MaNGA 7-fiber mini-bundles. Since the BOSZ vector for each exposure (see Yan et al. 2016b); these templates picked by the pipeline are bluer by 0.03 observations will from here on instead be included in the mag in SDSS u − r color than the original Kurucz MaStar stellar spectral library (see §4.3). 2003 templates, this change slightly increases the In addition, for the first time DR15 includes the re- overall flux blue-ward of 4000 Ain˚ the MaNGA data lease of derived spectroscopic products (e.g., stellar kine- cubes. Test observations of hot white dwarfs com- matics, emission-line diagnostic maps, etc.) from the pared to ideal blackbody models generally show MaNGA Data Analysis Pipeline (K. Westfall et al., in better performance using the new BOSZ calibra- prep; F. Belfiore et al., in prep), see §4.1.2. tion (as illustrated in Figure 3). Additionally, the We provide the sky footprint of MaNGA galaxies re- throughput loss vector applied to the observational leased in DR15 in Figure 1; while the projected final sur- data is now smoother at many wavelengths; high- vey footprint is shown overlaid on the footprint of other frequency basis spline fits are still used in telluric relevant surveys and for two different expectations for regions, but the spline has a much lower frequency weather at the telescope in Figure 2. outside the telluric regions to avoid introducing ar- tifacts due to slight template mismatches. This 4.1. MaNGA Data and Data Products significantly reduces the amount of artificially high 4.1.1. The Data Reduction Pipeline frequency low level (∼ few percent) variations seen The MaNGA Data Reduction Pipeline (DRP) is in the resulting spectra from earlier versions. The the IDL-based software suite that produces final flux- list of telluric regions is also updated. calibrated data cubes from the raw dispersed fiber spec- • Many aspects of the spectral line-spread func- tra obtained at APO. The DRP is described in detail tion (LSF) estimation in the DRP have changed by Law et al. (2016) and consists of two stages. The in DR15 in order to improve the level of agree- ‘2d’ DRP processes individual exposures, applying bias ment with independent estimates (observations of and overscan corrections, extracting the one-dimensional bright stars and galaxies previously observed at fiber spectra, sky-subtracting and flux-calibrating the higher spectral resolution, observations of the so- spectra, and combining information from the four indi- lar spectrum, etc.). These changes include the vidual cameras in the BOSS spectrographs into a single use of a Gaussian comb method to propagate set of row-stacked spectra (mgCFrame files) on a com- LSF estimates through the wavelength rectifica- mon wavelength grid. The ‘3d’ DRP uses astrometric tion step, computation of both pre-pixelized and information to combine the mgCFrame fiber spectra from post-pixelized LSF estimates2, improved interpola- individual exposures into a composite data cube on a tion over masked regions, and a modified arc lamp regularized 0.500 grid, along with information about the inverse variance, spaxel mask, instrumental resolution, 1 https://svn.sdss.org/public/repo/manga/mangadrp/tags/ and other key parameters. The mgCFrame per-exposure v2_4_3 files are produced on both linear and logarithmic wave- 2 i.e., whether the best-fit Gaussian model of the lines is de- termined by evaluation at the pixel midpoints (post-pixellized) or length grids directly from the raw detector pixel sam- integrated over the pixel boundaries (pre-pixelized). The two tech- pling, and used to construct the corresponding logarith- niques can differ at the 10% level for marginally undersampled mic and linearly-sampled data cubes. lines, and the appropriate value to use in later analyses depends The DRP data products release in DR15 are largely on the fitting algorithm. Data Release 15 7

Fig. 1.— The sky distribution (Mollewiede equatorial projection for Dec > −20◦) of MaNGA plates released in DR15 (purple). This is overlaid on a plot of all possible MaNGA plates (in grey). MaNGA targets are selected from a sample with SDSS-I photometry and redshifts; hence this footprint corresponds to the Data Release 7 imaging data (Abazajian et al. 2009). Each plate contains 17 MaNGA targets, and around 30% of all possible plates will be observed in the full 6-year survey. The most likely final footprint is indicated in Figure 2

Fig. 2.— The sky distribution (in a rectangular projection for clarity) of the MaNGA projected final footprint overlaid with information about other surveys. Because MaNGA targets are selected from a sample with SDSS-I photometry and redshifts, the selection of all possible plates (grey) corresponds to the Data Release 7 imaging data (Abazajian et al. 2009). Each plate contains 17 MaNGA targets, and around 30% of all possible plates will be observed in the full 6-year survey; this plot indicates the likely final footprint for (a) typical weather condition (Tier 1) and (b) good weather conditions (Tier 2). Completed plates noted on this plot show all observed plates at the time this was created; which is approximately one year of observing more than is being released in DR15. Where those plates are not filled in they have HI follow-up from the HI-MaNGA program (Masters et al. 2018; some, but not all of these data are released as a VAC in DR15 - see §4.5.3). For the most up-to-date version of this plot see https://www.sdss.org/surveys/manga/forecast/ reference line list to improve LSF estimation and scribing the spatial covariance introduced in the wavelength calibration in the far blue by rejecting data cubes by the cube building algorithm. This poor-quality lines. The DRP data products contain information is provided in the form of sparse cor- additional extensions to describe this new informa- relation matrices at the characteristic wavelengths tion, including a 3D cube describing the effective of the SDSS griz filters, and can be interpolated to LSF at each spaxel within the MaNGA data cube estimate the correlation matrix at any other wave- as a function of wavelength; this combines the in- length in the MaNGA data cubes. Note that the formation known about the LSF in each individual DR14 paper incorrectly stated that those data in- fiber spectrum to describe the net effect of stack- cluded these extensions. They did not (the team- ing spectra with slightly different resolution. The internal MPL-5 which is the most similar MPL to LSF changes and assessment against various ob- DR14 did, but DR14 itself did not), so this is the servational calibrators will be described in greater first release of these extensions. detail by D. Law et al. (in prep). • The DRPall summary file for DR15 contains ten • The DRP data cubes now contain extensions de- additional columns with respect to DR14. These 8 SDSS Collaboration

lems, large numbers of dead fibers, etc.). About 1% of the data cubes are flagged as significantly problematic (i.e., have the CRITICAL quality bit set) and should be treated with extreme caution. Additionally, there is a 3d mask extension to each data cube that contains spaxel-by-spaxel information about problematic regions within the cube. This mask identifies issues such as dead fibers (which can cause local glitches and holes within the cube), foreground stars that should be masked by analysis packages such as the DAP, etc. Although the vast majority of cosmic rays and other transient features are detected by the DRP and flagged (either for removal or masking), lower-intensity glitches (e.g, where the edge of a cosmic ray track intersects with a bright emission line) can sometimes be missed and propagate into the final datacubes where they show up as unmasked hot pixels. Future improvements to the DRP may further address this issue, but caution is thus always advised when searching for isolated emission features in the data Fig. 3.— This figure illustrates the flux calibration difference be- cubes. tween DR14 and DR15 MaNGA data reductions. The upper panel For information on downloading MaNGA data in DR15 shows the spectra for an Oke standard, HZ 21, a T=100,000 K please see 3; new for DR15 is the Marvin interface to star (Oke & Shipman 1971; Reynolds et al. 2003), as given by the § CALSPEC database (black) and by MaNGA in DR14 (red) and MaNGA data (see §4.2 below). DR15 (blue) averaging over 9 exposures taken on plate 7444. The small difference at the blue wavelengths can be seen more obvi- 4.1.2. The Data Analysis Pipeline ously in the bottom panel where we divide these three spectra by a T=100,000 K blackbody spectrum normalized at 6000-6100A. The MaNGA data-analysis pipeline (DAP) is the Ignoring the absorption lines, this provides a test to our flux cal- SDSS-IV software package that analyzes the data pro- ibration. Using the BOSZ templates in DR15, the resulting con- duced by the MaNGA data-reduction pipeline (DRP). tinuum of this white dwarf agrees much better with the blackbody The DAP currently focuses on “model-independent” spectrum below 5000A, significantly improved compared to DR14, which uses the version of the Kurucz models (Kurucz 1979; Kurucz properties; i.e., those relatively basic spectral properties & Avrett 1981) which were produced in 2003, and to CALSPEC. that require minimal assumptions to derive. For DR15, (One can also compare this with Figure 9 of Yan et al. (2016b)). these products include stellar and ionized-gas kinematics, columns include an estimate of the targeting red- nebular emission-line fluxes and equivalent widths, and shift z that is used as the starting guess by the spectral indices for numerous absorption features, such DAP when analyzing the MaNGA data cubes. z is as the Lick indices (Worthey & Ottaviani 1997; Trager generally identical to the NASA-Sloan Atlas (NSA) et al. 1998) and D4000 (Bruzual A. 1983). Examples (Blanton et al. 2011) catalog redshift for the ma- of the DAP provided measurements and model fits are jority of MaNGA galaxies, but the origin of the shown in Figure 4, discussed through the rest of this Sec- redshift can vary for galaxies in the ∼ 25 MaNGA tion. ancillary programs. Additional columns include a An overview of the DAP is provided by K. Westfall et variety of estimates of the volume weights for the al., (in prep). There, we describe the general workflow of MaNGA primary and secondary galaxy samples. the pipeline, explain the detailed algorithm used for each of its primary products, provide high-level assessments • Additional under-the-hood modifications to the of its performance, and describe the delivered data prod- DRP have been made for DR15 that provide minor ucts in detail. In-depth assessments of the stellar kine- bug fixes and performance improvements. These matics, ionized-gas kinematics, emission-line fluxes, and include modifications to the reference pixel flat- emission-line equivalent widths are provided by K. West- fields for certain MJDs, updates to the reference fall et al., (in prep), and F. Belfiore et al., (in prep). All bias and bad pixel masks, better rejection of sat- survey-provided properties are currently derived from the urated pixels, updates to the algorithms governing datacubes sampled in constant steps of the logarithm of weighting of the wavelength rectification algorithm wavelength (i.e., the LOGCUBE files). However, the core near ultra-bright emission lines, etc. A detailed functions are developed to consider each spectrum largely change-log can be found in the DRP online repos- independently. itory.3 The DAP allows for a number of different options when analyzing the data, which we refer to as the analy- When working with the MaNGA data, note that there sis mode or DAPTYPE. In DR15, the DAPTYPE joins the are several quality-control features that should be used keywords identifying the type of spatial binning ap- to ensure the best scientific quality output. First, each plied (e.g., Voronoi binned to S/N& 10, VOR10), the MaNGA data cube has a FITS header keyword DRP3QUAL parametric form of the line-of-sight velocity distribution that describes the overall quality of the cube (identify- (LOSVD) used for the stellar kinematics (a Gaussian ing issues such as focus problems, flux calibration prob- function, GAU), and the template set used to model the stellar continuum (a hierarchically clustered distillation 3 https://svn.sdss.org/public/repo/manga/mangadrp/tags/ of the MILES stellar library, MILESHC). For DR15, two v2_4_3/RELEASE_NOTES DAPTYPEs have been made available, VOR10-GAU-MILESHC Data Release 15 9

15 V [km/s] (a) 200 5 10 (c) NaD

100 4 ] 5 c e

s 3 c (a) 0 r 0 a

[ [O III]

Mg I 5 2

100 Ca H&K 10 (b) 1 ] l

200 0 15 e x

2 a p H Flux [erg/s/cm ] 6 0.4 (b) 15 s 1 / H Å 0 / 2 10 1 0.3 [N II] (c) m [N II] [S II] c / ] s

5 / c g e

r 0.2 s e c

(a) 7 r 0 1 7 a 1 [

0 0 5 1 0.1 1 [

y t

10 (b) i

s 0.0 n 15 e D

x u 15 D4000 l 1.0 (c) F 2.1 10 0.8 H (c) [O III] 2.0 Mg I ] 5 c

e 0.6 1.9 s

c (a) r 0 a [ 1.8 0.4 5 [O II] 1.7 10 (b) 0.2 1.6 15 0.0

10 0 10 1.5 4000 4500 5000 5500 6000 6500 7000 [arcsec] [Å]

Fig. 4.— Example data provide by the MaNGA data-analysis pipeline (DAP) for MaNGA observation 8138-12704, MaNGA ID 1- 339041, following the hybrid binning approach (DAPTYPE is HYB10-GAU-MILESHC). The left columns shows maps, or images, of some of the DAP derived quantities. Namely, from top to bottom, the stellar velocity field, Hα flux, and D4000 spectral index, where the measured value is indicated by the colorbar to the right of each map panel. The effective beam size for the MaNGA observations (FWHM∼200. 5) is shown by the gray circle in the bottom left of each map panel. Three spaxels are highlighted and labeled as (a), (b), and (c), according to their spectra plotted in the right column. Each spectrum panel shows the observed MaNGA spectrum (black), stellar-continuum-only model (blue), and best-fitting (stars+emission lines) model (red); the residuals between the data (black) and the model (red) are shown in gray. Note that the red and blue lines are identical except for regions with nebular emission. A few salient emission and absorption features are marked in each panel. Inset panels provide a more detailed view of the quality of the fitted models in the regions highlighted with gray boxes. The spectrum panels only show the spectral regions fit by the DAP, which is limited by the MILES spectral range for DR15. and HYB10-GAU-MILESHC. That is, only the binning ap- the second mode (HYB10-GAU-MILESHC) only performs proach differs between the two available DAPTYPEs, pri- the stellar kinematics on the binned spectra; the subse- marily distinguishing whether or not the main analysis quent emission-line and spectral-index measurements are steps are performed on binned spectra or individual spax- all performed on the individual spaxels. This “hybrid” els. The stellar LOSVD is always assumed to be Gaus- binning approach is likely the approach that most users sian, and the 42 templates resulting from a hierarchical will want to use in their analysis. The main exception clustering analysis of the MILES stellar library (S´anchez- to this is if any subsequent analyses depend on, e.g., the Bl´azquezet al. 2006; Falc´on-Barrosoet al. 2011) are al- availability of emission-line models for the binned spec- ways used for the continuum templates; details regarding tra, as is the case for the FIREFLY VAC (Wilkinson et al. the latter is discussed in K. Westfall et al., (in prep). 2017, see §4.5.1). The example data shown in Figure 4 In the first mode (VOR10-GAU-MILESHC), the spaxels is for observation 8138-12704 following from the hybrid are binned using the Voronoi-binning scheme from Cap- binning approach. Close inspection of the stellar velocity pellari & Copin (2003) to a minimum g-band S/N of 10 field will show that outermost regions have been binned per spectral pixel. The first mode then performs all sub- together, all showing the same stellar velocity measure- sequent analysis on those binned spectra. Alternatively, ment. However, the Hα flux and D4000 maps have mea- 10 SDSS Collaboration surements for each spaxel. least in the sense of not including any measure- The DAP is executed for all observations obtained ments we know to be dubious. However, the DAP by the MaNGA survey; however, some observations, makes use of an UNRELIABLE flag that is intended primarily those obtained for our ancillary science pro- to be more of a warning that users should con- grams, do not have all the required parameters currently sider how the measurements affect their science as needed as input by the DAP. Additionally, a few obser- opposed to an outright rejection of the value. The vations (<0.3%) trip corner failure modes of the DAP UNRELIABLE flag is put to limited use in DR15, only leading to errors in the construction of its main output flagging measurements that hinge on bandpass in- files. These issues mean that not all LOGCUBE files pro- tegrals (emission-line moments and non-parametric vided by the DRP have associated DAP products. For equivalent widths, and spectral indices) where any those observations that are successfully analyzed (4718 pixels are masked within the bandpass. However, in total), the DAP provides two main output files for this bit may become more extensively used in fu- each DAPTYPE, the MAPS file and the model LOGCUBE ture releases as we continue to vet the results of the file. Examples of how to access and plot the data in analysis. A more extensive discussion of the mask these files are provided in a set of tutorials on the data- bits and their usage is provided by K. Westfall et release website at https://www.sdss.org/dr15/manga/ al., (in prep). manga-tutorials/dap-tutorial/. The MAPS file contains all of the derived properties or- 2. To keep the format of the output files consistent ganized as a series of maps, or images, that have the with the DRP LOGCUBE files, the binned spectra same on-sky projection as a single wavelength channel and binned-spectra measurements are repeated for in the analyzed DRP LOGCUBE file. The images in the each spaxel within a given bin. This means that, left panels of Figure 4 are example maps taken from the e.g., the stellar velocity dispersion measured for a DAP MAPS file for observation 8138-12704. The maps are given binned spectrum is provided in the output organized in a series of extensions grouped by the mea- DAP map at the location of each spaxel in that bin. surement they provide. Some extensions contain a single Of course, when analyzing the output, one should image with all of the relevant data, whereas other ex- most often only be concerned with the unique mea- tensions have multiple images, one for, e.g., each of the surements for each observation. To this end, we measured emission lines. For example, the STELLAR VEL provide an extension in the MAPS file that provides extension has a single image with the measured single- a “bin ID” for each spaxel. Spaxels excluded from component stellar velocity measured for each spatial bin any analysis (as in the buffer region during the dat- (like that shown in Figure 4), while the SPECINDEX exten- acube construction) are given a bin ID of -1. This sion has 46 images, organized similarly to the wavelength allows the user to select all the unique measure- channels in the DRP datacubes (the D4000 map shown ments by finding the locations of all unique bin in Figure 4 is in the 44th channel of the SPECINDEX ex- ID values, ignoring anything with a bin ID of -1. tension). Tutorials for selecting the unique measurements in The DAP-output model LOGCUBE file provides both the the DAP output maps are provided via the data- S/N-binned spectra and the best-fitting model spectra. release website at https://www.sdss.org/dr15/ From these files, users can plot the best-fitting model manga/manga-tutorials/dap-tutorial/. spectra against the data, as demonstrated in Figure 4, as an assessment of the success of the DAP. This is par- 3. Corrections are provided for a few quantities in ticularly useful when a result of the fit, e.g. the Hα flux, the MAPS file that have not been applied to the data seems questionable. Indeed, K. Westfall et al., in prep, in the output files. The stellar velocity dispersion note regimes where the DAP has not been appropriately and ionized-gas velocity dispersions are provided tailored to provide a successful fit; this is particularly as measured from the core pPXF software (Cappel- true for spectra with very broad emission lines, such as lari & Emsellem 2004; Cappellari 2017) used by the broad-line regions of AGN. Users are encouraged to the DAP. This means that any instrumental effects make sure they are well aware of these limitations in present during the fitting process are also present in the context of their science goals. Finally, in combina- the output data. For both the stellar and ionized- tion with the DRP LOGCUBE file, users can use the model gas dispersions, we have estimated the instrumen- LOGCUBE data to construct emission-line-only or stellar- tal corrections for each measurement and provided continuum-only data cubes by subtracting the relevant the result in extensions in the MAPS file. These model data. corrections should be applied when using the data Although we have endeavored to make the output data for science. For the velocity dispersion measure- user-friendly, there are a few usage quirks of which users ments, our purpose in not applying the corrections should be aware: ourselves is to allow the user freedom in how they deal with measurements of the dispersion that are 1. As with all SDSS data products, users are strongly below our measurement of the instrumental reso- encouraged to understand and use the provided lution. Such issues can be pernicious at low ve- quality flags, for these data provided as masks. locity dispersion and the treatment of these data The mask bits provide important information as can have significant effects on, e.g., the construc- to whether or not users should trust the provided tion of a radially averaged velocity dispersion pro- measurements in their particular use case. The file (see K. Westfall et al., in prep. who discuss conservative approach of ignoring any measure- this at length, and also Penny et al. 2016 who dis- ment where the mask bit is nonzero is safe, at cuss this issue for dwarf galaxies). Corrections are Data Release 15 11

also provided (but not applied) for the spectral in- • perform powerful queries on metadata. dices to convert the measurement to zero velocity dispersion at the spectral resolution of the MILES • abstract the MaNGA datamodel and write code stellar templates (Beifiori et al. 2011) used during which is agnostic to where the data actually lives. the stellar-continuum fit. Additional detail regard- • make better visualization and scientific decisions by ing these corrections is provided in K. Westfall et mitigating common mistakes when accessing these al., (in prep.) and tutorials demonstrating how to type of data. apply them to the data are provided via the data- release website. Marvin has two main components: a webapp and a Python package of tools, both using an underlying 4. In the hybrid binning scheme, the stellar kinemat- Marvin API (or Application Programming Interface). ics are performed on the binned spectra, but the The webapp, Marvin Web5, provides an easily accessible emission-line fits are performed on the individual interface for searching the MaNGA dataset and visual spaxels. When comparing the model to the data, exploration of individual MaNGA galaxies. The Marvin the user must compare the emission-line modeling suite of Python tools, Marvin Tools, provides seamless results to the DRP LOGCUBE spectra, not the binned programmatic access to the MaNGA data for more in- spectra provided in the DAP model LOGCUBE file, depth scientific analysis and inclusion in your science unless the “binned” spectrum is actually from a workflow. Marvin contains a multi-modal data access single spaxel. Tutorials for how to overplot the system that provides remote access to MaNGA files or correct stellar-continuum and emission-line models sub-data contained within, download MaNGA files to are provided via the data-release website. work with on the users local machine, and seamlessly Finally, similar to the DRPall file provided by the transition between the two with a negligible change in MaNGA DRP, the DAP constructs a summary table syntax. called the DAPall file. This summary file collates use- Existing 3d data cube visualizers in astronomy, as well ful data from the output DAP files, as well as providing as in other scientific disciplines, often come as standalone some global quantities drawn from basic assessments of desktop applications designed to visualize and interact the output maps, that may be useful for sample selection. with individual files local to a client machine. However, For example, the DAPall file provides the luminosity- these tools are highly specific, limited to exploring files weighted stellar velocity dispersion and integrated star- one at a time, and still require manually downloading formation rate within 1 Re. The sophistication of these all data locally. While Marvin is a tool for 3d cube vi- measurements are limited in some cases. For example, sualization, its focus is on streamlined data access from the star-formation rate provided is simply based on the local or remote sources, with a clear separation of com- measured Hα luminosity and does not account for in- ponents into browser-based visualization and program- ternal attenuation or sources of Hα emission that are matic data tools, rather than on providing yet another unrelated to star formation, as such we caution users to desktop-based cube viewer. Marvin’s design allows for make use of this for science only after understanding the users to rapidly explore and access the data in a manner implications of this caveat. Development and refinement of their choosing, whilst still providing enough flexibility of DAPall output will continue based on internal and to, if desired, plug the data into existing cube viewers community input. Additional discussion of how these available in the astronomy community. properties are derived is provided by K. Westfall et al., The components of Marvin are described in more detail (in prep). in the Marvin paper (B. Cherinka et al. in prep.) as well as in the Marvin documentation6, which also contains 4.2. Marvin Access to MaNGA tutorials and example Jupyter notebooks. In addition, Marvin (Cherinka et al. 2018)4 is a new tool designed we briefly introduce them below. for streamlined access to the MaNGA data, optimized 4.2.1. Marvin Web for overcoming the challenges of searching, accessing, and visualizing the complexity of the MaNGA dataset. The Marvin Web provides quick visual access to the Whereas previous generations of SDSS took spectra only set of MaNGA galaxies. It provides a dynamic, inter- of the centers of galaxies, MaNGA takes many spectra active, point-and-click view of individual galaxies to ex- of each galaxy, in a hexagonal grid across the face of plore the output from the MaNGA DRP and DAP (§4.1.1 each (IFU bundle), which are combined into a final data §4.1.2 respectively), along with galaxy information from cube. This means that for each object there is not a the NSA catalog (Blanton et al. 2011) 7. single spectrum, but in fact a suite of complex results in We show a screen-shot of the View-Spectra page of one or more data cubes. The motivation of Marvin arises Marvin Web in Figure 5. By clicking anywhere within from the additional complexity of MaNGA data, namely, the galaxy IFU bundle on the SDSS three-color image, the spatial interconnectivity of its spectra. or any Data Analysis 2D Map, the user can explore the Marvin allows the user to: spectrum at that location for quick inspection. The visu- alized spectrum is interactive as well, allowing panning • access reduced MaNGA datacubes local, remotely, and zooming. or via a web interface. 5 https://dr15.sdss.org/marvin • access and visualize data analysis products. 6 https://sdss-marvin.readthedocs.io/en/stable/ 7 https://www.sdss.org/dr15/manga/ 4 https://www.sdss.org/dr15/manga/marvin/ manga-target-selection/nsa/ 12 SDSS Collaboration

Additional pages Marvin Web provides are: Stellar spectral libraries are an essential tool for many fields in astronomy. They are especially useful for mod- • a Query page, for searching the MaNGA dataset eling spectra of external galaxies, including fitting for through an SQL-like interface. redshift and stellar kinematics, fitting the continuum to isolate emission lines, and calculating stellar population • a Plate page, containing all MaNGA galaxies ob- served on a given SDSS plate. models (e.g. Leitherer et al. 1999; Bruzual & Charlot 2003; Maraston 2005; Vazdekis et al. 2010; Conroy & • and an Image Roulette page, for randomly sam- Gunn 2010; Conroy 2013) for deriving age, metallicity, pling images of MaNGA galaxies. and stellar mass of the stellar populations from inte- grated light spectra. They are also useful for Galactic Tutorials for navigating Marvin Web can be astronomy and stellar astronomy. Although theoretical found at https://www.sdss.org/dr15/manga/ spectral libraries have been substantially improved over manga-tutorials/marvin-tutorial/marvin-web/. the years, they are still not realistic enough for certain Marvin Web is designed as a gateway to entry into real stellar types (e.g. very cold stars and Carbon stars) due MaNGA data, providing commonly-desired functionality to the incomplete line list and difficult-to-model physical all in one location, as well as code snippets to help tran- effects, such as convection, microturbulence, and devi- sition users into a more programmatic environment using ations from plane-parallel geometry and local thermo- the Marvin Tools. dynamic equilibrium (non-LTE). Therefore, empirical li- braries are still needed for many applications and for cal- 4.2.2. The Marvin Tools ibrating the theoretical models, provided one is able to Marvin Tools provides a programmatic interaction assign robust stellar parameters to the empirical spectra. with the MaNGA data, enabling rigorous and repeatable At the beginning of the MaNGA survey, there were science-grade analyses. Marvin Tools come in the form no empirical stellar libraries available covering the whole of a Python package that provides convenience classes MaNGA wavelength range with a spectral resolution that and functions that simplify the processes of searching, is equal to or higher than that of MaNGA. Current state- accessing, downloading, and interacting with MaNGA of-the-art empirical stellar libraries also have some other data, selecting a sample, running user-defined analysis shortcomings. Some libraries have issues with flux cali- code, and producing publication quality figures. bration or telluric subtraction. Furthermore, all existing The Marvin Tools are a pip-installable product, pack- libraries have limited stellar parameter space coverage, aged under sdss-marvin, with full installation instruc- lacking sufficient sampling in especially cool dwarfs, car- tions at the Marvin documentation8, and source code on bon stars, metal-poor stars, and very hot stars. They Github9. also do not sufficiently sample the [α/Fe] vs. [Fe/H] space Overall, Marvin Tools allow for easier access to the (see Maraston & Str¨omb¨ack 2011) for a discussion of all data without knowing much about the data model, by these problems). These issues prompted us to take ad- seamlessly connecting all the MaNGA data products, vantage of a parallel observing opportunity in SDSS-IV eliminating the need to micromanage a multitude of files. for assembling an empirical stellar spectral library that The user can do all their analysis from one interface. samples a wider stellar parameter space with a larger number of stars than any previous library, and matches 4.2.3. Queries in Marvin MaNGA’s wavelength coverage and spectral resolution. Both Marvin Web and Tools provide interfaces for Included in this data release is the first version of the searching the MaNGA dataset through a Structured MaNGA Stellar Library (MaStar). These observations Query Language (SQL)-like interface, either via a web- are performed by piggybacking on the APOGEE-2N ob- form or a Python class. The Marvin Query system uses servations during bright time. MaNGA fiber bundles are a simplified SQL syntax that focuses only on a filter con- plugged along with APOGEE fibers on these APOGEE- dition using boolean logic operators, and a list of param- led plates to observe selected stars. As a result, the MaS- eters to return. This eliminates the need to learn the tar stellar spectra are observed using exactly the same full SQL language and the detailed MaNGA database instrument as MaNGA galaxies so they provide an ideal layout. With this query system, users can make queries set of templates for modeling stellar continuum and stel- across the entire MaNGA sample using traditional global lar populations in MaNGA galaxies. galaxy properties (functionality to perform intra-galaxy The program has so far observed several thousands of queries using individual spaxel measurements is planned stars, each with several epochs of observation. The ver- for a future release). Tutorials for querying with Marvin sion we are releasing in DR15 includes 8646 good quality spectra for 3321 unique stars, which cover a wide range can be found for the web10 and for the tools11. in stellar parameter space. The details of the target se- 4.3. MaStar: A Large and Comprehensive Stellar lection, data reduction, flux calibration, and stellar pa- Spectral Library rameter distribution are described by Yan et al. (2018). Here we provide a brief summary.

8 https://sdss-marvin.readthedocs.io/en/stable/ 4.3.1. Target Selection installation.html A good target selection is essential for achieving a 9 https://github.com/sdss/marvin 10 https://www.sdss.org/dr15/manga/manga-tutorials/ wide sampling of stellar parameter space. We aim to marvin-tutorial/marvin-web/ cover the stellar parameter space as completely as possi- 11 https://sdss-marvin.readthedocs.io/en/stable/query. ble and sample it roughly evenly. We base our selection html primarily on existing stellar parameter catalogs, includ- Data Release 15 13

Fig. 5.— Screenshot of the galaxy maps view of the Marvin Web for the MaNGA galaxy 12-193481 (Mrk 848). The SDSS three-color image of the galaxy is shown in the top left part of the figure. The upper right panel shows the spectrum of the spaxel at the position (37,37), which corresponds to the center of the bundle (shown by the red dot). The maps show: (lower left) stellar kinematics; (lower middle) Hα emission line flux; and (lower right) D4000 spectral index for this galaxy based on its unbinned spectral data cube from the MaNGA DAP (see §4.1.2). ing APOGEE-1 and -2 (Majewski et al. 2017), SEGUE These magnitude limits yield relatively few OB stars (Yanny et al. 2009), and LAMOST (Luo et al. 2015). and blue supergiants, as they have to be very distant Given the field plan of APOGEE-2, we select all the stars or very extincted to fall within this magnitude range. available in these catalogs in the planned APOGEE-2 Therefore, currently we are adjusting our exposure time footprint. For each star, we count its neighboring stars in certain fields to expand our parameter space distribu- in stellar parameter space (Teff , log g, [Fe/H]) and assign tion in the blue and luminous end. The first version of it a selection weight that is inversely proportional to its the library does not have many such stars but we will number of neighbors. The number of APOGEE-2 tar- improved on this for the final version which we expect to geting designs for each field is also taken into account. come out in the final SDSS-IV Data Release. We then draw our targets randomly in proportion to the normalized selection weight. This method flattens the 4.3.2. Observations stellar parameter space distribution and picks rare stars Observations for MaStar are obtained in a similar fash- in those fields where they are available. ion to the MaNGA observations except that they are con- In fields without stars with known stellar parame- ducted under bright time and without dithering. Since ters, we use spectral energy distribution (SED) fitting we are piggybacking on APOGEE-2, if APOGEE-2 vis- to search for hot and cool stars to patch the stellar pa- its a field multiple times, we would obtain multiple visits rameter distribution at the hot and cool ends. for the stars on that plate as well. Therefore, some stars The targets are required to have g or i-band magni- have many visits and some stars have only 1 visit. Each tude brighter than 17.5 in order to achieve a signal-to- visit of APOGEE-2 is typically 67 minutes long, which noise greater than 50 per pixel in 3 hours of integration, would allow us to take four 15-minutes exposures, un- although not all fields have the same integration time or less interrupted by weather or other reasons. Each plate the same number of visits. They are also required to be has 17 science targets and 12 standard stars, same as fainter than 12.7 magnitude in both g and i-band in order MaNGA. We take flat and arc frames before each visit. to stay below the saturation limit of the detector for 15 minute exposures. We later lowered the saturation limit 4.3.3. Data Reduction to 11.7 to include more luminous stars, with a slight off- The reduction of the MaStar data is handled by the set in fiber placement for stars with magnitudes between MaNGA Data Reduction Pipeline (DRP; see 4.1.1, Law 11.7 and 12.7. This slight offset does not affect our flux § et al. 2016). It has two stages. The first stage pro- calibration due to our unique calibration procedure. cesses the raw calibration frames and science frames 14 SDSS Collaboration to produce the sky-subtracted, flux-calibrated, camera- ture. In the current release, the spectra for these stars combined spectra for each fiber in each exposure. The are just flagged as having poor flux calibration (bit 5 of second stage differs between MaNGA galaxy data and MJDQUAL). MaStar stellar data. For MaStar stellar data, we eval- A significant fraction of the spectra also have unreli- uate the flux ratios among fibers in a bundle as a func- able radial velocity estimates. We used a rather limited tion of wavelength, and constrain the exact location of set of templates in our derivation of the radial velocities. the star relative to the fiber positions. This procedure Thus stars with very hot or very cool effective tempera- helps us derive the light loss due to the finite fiber aper- ture, and those with very high and low surface gravity, ture as a function of wavelength. The procedure takes are more likely to be affected by this issue. These can into account the profile of the PSF and the differential be identified by checking bit 6 of the MJDQUAL bitmask. atmosphere refraction. It is similar to how we handle All spectra are shifted to rest-frame according to the re- flux calibration in MaNGA data (Yan et al. 2016b). We ported heliocentric radial velocity, regardless of whether then correct the spectra for this aperture-induced light the measurement is robust or not. loss and arrive at the final flux-calibrated stellar spectra. In addition to these automated checks, we carried out a Comparison with photometry shows that our relative flux visual inspection campaign to ensure the quality of each calibration are accurate to 5% between g and r -band, spectrum. Using the Zooniverse Project Builder inter- and to 3% between r and i, and between i and z bands. face12, we started a private project for visually inspect- For each star, we combine the spectra from multiple ing the spectra. A total of 28 volunteers from within the exposures on the same night, and refer to these combined collaboration participated in the campaign, 10,797 visit spectra as “visit spectra”. We do not combine spectra spectra were inspected, each by at least 3 volunteers, from different nights together for the same star because to check for issues in flux calibration, sky subtraction, they can have different instrumental resolution vectors telluric subtraction, emission lines, etc. The results are and some stars could be variable stars. By summer 2017, input to the data reduction pipeline to assign the final we have obtained 17,309 visit-spectra for 6,042 unique quality flags. stars. Because not all visit-spectra are of high quality, as The primary set of spectra we are releasing contains we will discuss below, we selected only those with high only those spectra that are deemed to have sufficient quality and present this subset as the primary set to be quality to be useful. We have excluded from the pri- released. The primary set contains 8646 visit spectra for mary set those spectra with problematic sky subtraction 3321 unique stars. (bit 1 of MJDQUAL), low PSF covering fraction (bit 4), The final spectra are not corrected for foreground dust poor flux calibration (bit 5), or low S/N (bit 9), and extinction. Users should make these corrections before those identified as problematic by visual inspection (bit using them. 7). The primary set still contains spectra with unreliable heliocentric velocity measurement (bit 6), whose spectra 4.3.4. Quality Control would still be in the observed frame. It also contains A stellar library requires strict quality control. We spectra with strong emission lines (bit 8), some of which have a number of quality assessment carried out in the are intrinsic to the star. In addition, stars flagged to have pipeline to flag poor quality spectra. We identify cases high scattered light in the raw frame (bit 2) are also in- having low S/N, bad sky subtraction, high scattered cluded as they may not be affected significantly. Other light, low PSF-covering fraction, uncertain radial veloc- bits that are not mentioned above were never set in the ity measurement, and/or those with problematic flux cal- current data release. The users are strongly advised to ibration. Each spectrum we release has an associated check the quality flags when using the spectra. Detailed quality bitmask (MJDQUAL for each visit spectrum) giv- information about the quality flags can be found in Yan ing these quality information. We provide a summary of et al. (2018). these in Table 3 and describe them in more detail here. In addition to these basic quality checks, we are test- A large fraction of the observed stars have problematic ing the spectra by running them through a population flux calibration due to the fact that the standard stars synthesis code (Maraston 2005). This procedure, which on those plates have much less extinction than given by will be described in C. Maraston et al. (in prep)., allows the Schlegel et al. (1998, SFD) dust map. In the flux us to test the total effect of goodness of spectra plus as- calibration step of the MaNGA pipeline, we assume the signed stellar parameters and will be crucial for the joint standard stars are behind the Galactic dust, and we com- calculation of stellar parameters and stellar population pare the observed spectra of the standard stars with the models. Relevant to this description, this method allows dust-extincted theoretical models to derive the instru- us to spot bad or highly-extincted spectra. ment throughput curve. This assumption is valid for all galaxy plates which are at high Galactic latitude and 4.3.5. Stellar Parameter Distribution have relatively faint standard stars, placing them at a safe distance behind most of the dust. However, this as- Robust assignment of stellar parameters to the stars sumption fails for many fields targeted by MaStar, which are also critical for the stellar library. Our targets are are at low Galactic latitudes. Stars at low Galactic lati- selected from heterogeneous sources. Those selected tude are quite likely to be found in front of some fraction from APOGEE, SEGUE, and LAMOST have parame- of the dust in that direction. Thus, when applying the ters available from their respective catalogs. However, extinction given by SFD, we overly redden the theoret- they are measured with different methods and may not ical models and arrives at an incorrect flux calibration be consistent with each other. They also have different for these field. We have a solution to this problem which will be incorporated into the MaStar pipeline in the fu- 12 https://www.zooniverse.org/lab Data Release 15 15

TABLE 3 Qualitiy Control Bits MJDQUAL for MaStar. See §4.3.4 for full explanation.

Bit Description 1 Problematic sky subtraction 2 High scattered light in the raw frame 4 Low PSF covering fraction 5 Poor flux calibration 6 Unreliable radial velocity estimates 7 Flagged as unreliable from visual inspection 8 Strong emission lines 9 Low S/N boundaries applied in the determination of the param- eters. For our target selection purpose, we have made small constant corrections to the parameters to remove the overall systematic difference. These slightly adjusted parameters are included in the catalog we release. How- ever, the correction are done independent of detailed stellar types. As a result, the parameters from differ- ent catalogs can still have subtle stellar-type-dependent systematic differences. For individual stars, they could be used to determine the rough stellar type. But for the library as a whole, we caution against using these input parameters to compare the stars or to construct stellar population models with them. We are still in the process of determining stellar pa- rameters for all the stars in the MaStar library, in a way that is as homogeneous as possible. This is not an easy task, because for stars with different stellar types we need to rely on different spectral features and dif- ferent methodology. Although these are not yet avail- able in this version of the library, we present here the extinction-corrected Hertzsprung-Russell (HR) diagram Fig. 6.— Extinction-corrected HR diagram for MaStar targets, color-coded by our preliminary measurement of metallicity. The for our stars using photometry and parallax from Gaia g, r,and i bands are in the SDSS photometry system. The ab- DR2 (Gaia Collaboration et al. 2018; Evans et al. 2018) solute magnitudes are derived using parallax-based distance esti- and Gaia-parallax-based distance estimates from Bailer- mates from Bailer-Jones et al. (2018). Only stars for which we are able to get an approximate extinction correction are included here. Jones et al. (2018). This provides a rough idea of our This Figure is reproduced from Yan et al. (2018). stellar parameter coverage. Here we only plot stars that are either in directions with a total E(B−V ) less than 0.1 ties. But there is significant room for improvement. We mag or more than 300pc above or below the Milky Way need to cover more stars at the luminous end of the red mid-plane so that we can use the total amount of dust giant branch and at the blue end of the main sequence. measured by Schlegel et al. (1998) for the extinction cor- These will be added in future versions of the library. rection reliably. The photometry of our targets also come from various sources including PanSTARRS1 (Chambers 4.3.6. Data Access and usage information et al. 2016), APASS13, SDSS, Gaia DR1 (Gaia Collabo- The MaStar data for this release can be found ration et al. 2016), and Tycho-2 (Høg et al. 2000) for a in two main files: the “mastarall” and the few stars. For stars with PanSTARRS1 photometry, we “mastar-goodspec” files. Both files can be found on converted them to SDSS using the formula provided by the SAS (https://dr15.sdss.org/sas/dr15/manga/ Finkbeiner et al. (2016). For APASS, we assume they are spectro/mastar/v2_4_3/v1_0_2/, with datamodels in SDSS filters already. For all the other non-SDSS stars, at https://data.sdss.org/datamodel/files/MANGA_ we use Gaia DR2 photometry to derive the magnitudes SPECTRO_MASTAR/DRPVER/MPROCVER/. in SDSS gri bands according to the conversion given by The “mastarall” file contains four summary tables of Evans et al. (2018). In Figure 6, we show the r-band the basic information about the stars and of the infor- absolute magnitude (Mr) vs. g − i for these stars. The mation about their one or more visits. They include the color-coding are based on our preliminary measurement identifier (MaNGAID), astrometry, photometry, target- of metallicity using the ULySS pipeline (Koleva et al. ing bitmask which indicates source of photometry, input 2009, 2011) with MILES (S´anchez-Bl´azquez et al. 2006) stellar parameters, plate, IFU, modified Julian date of as the training set. the visit, derived heliocentric velocity, and spectral qual- From Figure 6, we can see that the current released ity information. MaNGAID is the identifier to identify subset of MaStar library already has a very good cover- unique stars, except in a few cases where the same star age across the HR diagram for a wide range of metallici- was assigned two different MaNGAIDs. These are docu- mented in detail online and in Yan et al. (2018). 13 https://www.aavso.org/apass The “mastar-goodspec” file contains the primary set 16 SDSS Collaboration of high quality visit spectra. In addition to identifica- 4.5. Value Added Catalogs tion information, we provide the wavelength, flux, in- As was the case previously in DR14, there are a verse variance, mask, spectral resolution vector, and the large number of Value Added Catalogs (VACs) linked to spectral quality bitmask. MaNGA data released in DR15. These either represent MaStar spectra can also be visualized in the SAW (see additional processing of DRP or DAP output, or follow- §3 for details). up programs or other data useful in combination with MaNGA data. We summarize new or updated VACs be- 4.4. Other Ancillary Programs low. We refer the reader to the DR13 paper (Albareti et al. 2017) for the most complete list of MaNGA ancillary 4.5.1. Spectral Modeling programs14. These approved programmes make use of In DR15 there are new releases for both the FIREFLY ∼5% of MaNGA bundles, and provide in Table 4 an up- (Goddard et al. 2017a) and Pipe3D (S´anchez et al. 2016) dated list of the number of bundles available in each doc- stellar population modeling and emission line analysis umented sample. VACs. Full details of both can be found in the DR14 There are three new ancillary programs for DR15. paper (and references therein), so we give only updates These provide observations in fields of IC 342, M31 specific to this DR15 release version below. as well as data for a selection of SN1a Hosts (MANGA For DR15 Pipe3D version 2.4.3 was run over the TARGET3 target bits 20, 21 and 26, respectively). MaNGA DR15 dataset. The main difference with re- The IC342 program will uniformly mosaic the disk of spect to version 2.1.2 (used in DR14) besides the num- IC 342 using 61 plates, following an initial pilot program ber of analyzed galaxies, was to solve a bug in deriva- of 3 plates that target individual HII regions across the tion of the equivalent widths of the analyzed emission disk. This galaxy serves as a local reference with 30 pc lines. The current Pipe3D VAC provides two different resolution that can inform our understanding of the un- types of data products: (1) A catalog comprising 94 resolved physics in the ∼ 1 kpc resolution main MaNGA different parameters measured for each of 4660 galax- survey. ies (all galaxies in MaNGA cubes for which Pipe3D The M31 ancillary program targets regions in M31 was able to derive the main stellar population, emission where the underlying physical properties are well- line and kinematics properties); and (2) a set of 4660 constrained from resolved stellar population analyses, datacubes manga.Pipe3D.cube.fits presenting a set of provided by the Panchromatic Hubble Andromeda Trea- spatially resolved parameters. The parameters are the sury (PHAT; Dalcanton et al. 2012). The MaNGA obser- same as they were in the DR14 version (S´anchez et al. vations include 18 regions ( 50-100 pc in size) that sam- 2018). More detail is available on the data release web- ple a wide range of environmental conditions, including site https://www.sdss.org/dr15/manga/manga-data/ ancient and recent star formation history, dust column, manga-pipe3d-value-added-catalog/. dust geometry, and metallicity. These observations pro- The major update to FIREFLY with respect to DR14 vide a link between resolved stellar populations and the is the extension of the stellar population modeling inferred properties of unresolved stellar populations, and grid based on the models of Maraston & Str¨omb¨ack can be used to assess the ability of spectral fitting codes (2011). The new catalog uses a finer metallic- to recover key physical parameters. ity grid with the following grid values: [Z/H] = The SNIa Hosts ancillary program will observe Type −2.3, −1.9, −1.6, −1.2, −0.9, −0.6, −0.3, 0.0, 0.3. The Ia supernova (SNIa) host galaxies to investigate causes new version of the VAC also provides geometrical of the intrinsic variation of SNIa. SNIa show a spread information so that maps can be produced directly in absolute magnitude, but can be standardized by tak- from the VAC (a python plotting routine is avail- ing into account relationships like luminosity-decline rate able from the data release website). The entire and SNIa color to reduce the spread to 0.12 mag. Re- VAC is available as either a single fits file contain- search over the past several years indicates that some of ing all measurements, or smaller fits files with se- this remaining spread correlates with global host galaxy lected subsets of the derived parameters. More detail properties such as stellar mass, star-formation history, on the catalog is provided on the data release web- and metallicity (e.g. Lampeitl et al. 2010; Gupta et al. site https://www.sdss.org/dr15/manga/manga-data/ 2011; Hayden et al. 2013; Rigault et al. 2013) caus- manga-firefly-value-added-catalog/ and in God- ing concerns about biases in cosmological measurements. dard et al. (2017a) and Parikh et al. (2018). This project will obtain MaNGA data for roughly 40 SNIa host galaxies in order to look for correlations with 4.5.2. Morphology and Photometry of MaNGA Targets SNIa peak absolute magnitude and host galaxy proper- As part of DR15 we release one photometry VAC and ties like metallicity and star formation rate averaged over two morphology VACs. the whole galaxy and at the location of the SNIa. The PyMorph catalog provides photometric parame- DR15 also incorporates a second Milky Way Analogs ters obtained from Sersic and Sersic+Exponential fits to program (target bit 23), which is similar to the exist- the 2D surface brightness profiles of the MaNGA DR15 ing program described in DR13, but uses morphological galaxy sample. It uses the PyMorph algorithm for deter- information, rather than star formation rates, in combi- mining the fits which has been extensively tested (Meert nation with galactic stellar mass to select analogs. et al. 2013; Fischer et al. 2017; Bernardi et al. 2017), and PyMorph reductions of SDSS DR7 galaxies (Abaza- 14 also see http://www.sdss.org/dr15/manga/ jian et al. 2009) are available (the UPenn SDSS Phot- manga-target-selection/ancillary-targets Dec Catalog: Meert et al. 2015, 2016). We have re-run Data Release 15 17

TABLE 4 Summary of MaNGA Ancillary Programs with Data in DR15

Ancillary Program Observeda BITNAME binary digit Luminous AGN 25c AGN BAT 1 AGN OIII 2 AGN WISE 3 AGN PALOMAR 4 Void Galaxies 3 VOID 5 Edge-On Star-forming Galaxies 20 EDGE ON WINDS 6 Close Pairs and Mergers 57d PAIR ENLARGE 7 PAIR RECENTER 8 PAIR SIM 9 PAIR 2IFU 10 Writing MaNGA (public outreach) 1 LETTERS 11 Massive Nearby Galaxies 23 MASSIVE 12 Milky Way Analogs 4 MWA 13 0 MW ANALOG 23 Dwarf Galaxies in MaNGA 22 DWARF 14 Brightest Cluster Galaxies 24 BCG 17 MaNGA Resolved Stellar Populations 1 ANGST 18 Coma 68 DEEP COMA 19 IC 342 50 IC342 20 M31 18 M31 21 SNIa Hosts 1 SN1A HOST 26 aThese are bundle counts, not always unique galaxies bThese are bundle counts, not always unique galaxies cCount for 1,2,3,4 combined dCount for 7,8,9,10 combined PyMorph for all the galaxies in the MaNGA DR15 sample. models were trained (making use of Galaxy Zoo mor- These re-runs incorporate three improvements: they use phologies, as well as morphologies from Nair & Abra- the SDSS DR14 images, improved bulge-to-disk decom- ham 2010) and tested on SDSS-DR7 images (Dom´ınguez position by slightly modifying our criteria when using S´anchez et al. 2018). The morphological catalog contains PyMorph (see Fischer et al. 2018. for details), and all the a series of Galaxy Zoo like attributes (edge-on, barred, fits in this catalog have been visually inspected for addi- bulge prominence and roundness), as well as a T-Type tional reliability (we recommend using “flag fit”). The and a finer separation between pure elliptical and S0 catalog contains these fits for the g, r, and i bands. One galaxies. important caveat to note is that position angles (PA) are reported relative to the SDSS imaging camera columns, 4.5.3. HI-MaNGA - HI 21cm Follow-up for MaNGA which are not aligned with North, so a correction is The first data release of “HI-MaNGA”, the HI followup needed to convert to true position angles. To convert project for MaNGA is provided as a VAC in DR15. This to the usual convention where North is up, East is left follow-up program is presented in Masters et al. (2018) (note that the MaNGA datacubes have North up, East and is the result of single dish radio 21cm HI obser- right) set PAMaNGA = 90 deg −PAPyMorph − PASDSS, vations of MaNGA galaxies using the Robert C. Byrd where PAPyMorph is the value given in this catalogue, Green Bank Telescope (GBT). The depth of this observ- and PASDSS is the SDSS camera column position angle ing is aimed to be similar to the Arecibo Legacy Fast with respect to North. Arecibo L-band Feed Array (ALFALFA) blind HI survey A curated version of the Galaxy Zoo crowdsourced (Haynes et al. 2018) which covers some of the MaNGA classifications containing an entry for all MaNGA tar- footprint (see Figure 2) with a goal of enabling studies get galaxies is released in DR15. This catalog contains to use HI data from both surveys. In this first release, galaxy classifications previously released in Willett et al. data are provided for 331 MaNGA galaxies observed (2013, which was selected from the SDSS DR7 galaxy in the 2016 GBT observing seasons under project code catalog) as well as new unpublished classifications for AGBT16A 95. Total HI masses, and line widths (measured MaNGA targets missing from that list. All morphologi- with five different common techniques) are provided for cal identifications are provided based on the citizen sci- all detections, while HI mass upper limits (assuming a entist input using the improved technique for aggrega- line width of 200km/s) are provided for non-detections. tion and debiasing described in (Hart et al. 2016). This HI-MaNGA has observed an additional ∼ 1800 MaNGA accounts better for redshift bias in the detailed classifica- galaxies in the 2017 observing season (under project code tions of spiral arms, bars than the version used in Willett AGBT17A 12); these data will be released in a future VAC. et al. (2013). For a simple conversion between Galaxy The sky distribution of all MaNGA galaxies observed by Zoo classifications and modern (bulge-sized based) T- this program is shown in Figure 2. types see details in Willett et al. (2013), which also in- cludes general advice on how to best use Galaxy Zoo 4.5.4. GEMA-VAC; Galaxy Environment for MaNGA Value classifications for science. Added Catalog A second morphology catalog is provided that has been The environment in which a galaxy resides plays an obtained with the help of “Deep Learning” models. The important role in its formation and evolution. Galaxies evolve as a result of intrinsic processes (i.e. their nature - 18 SDSS Collaboration this includes processes such as internal secular evolution, SDSS DR15 includes no new APOGEE data. The cur- feedback of various kinds etc), but they are also exposed rently available set of APOGEE Survey data consists of to the influences of their local and large-scale environ- the first two years of SDSS-IV APOGEE-2 (Jul 2014- ments (i.e. how they are “nurtured”). We present a Jul 2016) as well as the entirety of SDSS-III APOGEE-1 the Galaxy Environment for MaNGA Value-Added Cat- (Aug 2011-Jul 2014) and is an exact duplicate of that alog (GEMA-VAC) which provides a quantification of data which was released in DR14. DR15-associated the the local and large-scale environments of all MaNGA APOGEE documentation builds upon that from DR14 galaxies in DR15. There are many different definitions of with extended explanations and the addition of infor- environment and there are also several ongoing projects mation and relevant text (e.g., a description of the within the MaNGA team exploring the influence the en- New Mexico State University (NMSU) 1.0m Telescope). vironment on galaxy properties. With this VAC we aim Note that the DR15 APOGEE data model has remained to join and coordinate efforts so that the entire astro- largely the same with only slight revisions to the text for nomical community can benefit from the products. The clarity. Described below are the APOGEE technical pa- GEMA-VAC catalog will be described in more detail in pers that contain details which should assist users in the M. Argudo-Fern´andezet al (in prep). We describe the exploitation of APOGEE data as well as provide further contents of the VAC briefly below. understanding as to data quality (Zasowski et al. 2017; We estimate the tidal strength parameter for MaNGA Holtzman et al. 2018; J¨onssonet al. 2018; Pinsonneault galaxies in pairs/mergers (B. Hsieh et al., in prep.), et al. 2018; Wilson et al. 2018). Additionally, details are the tidal strengths exerted by galaxies in the catalog provided on the recently-generated Value-Added Catalog of galaxy groups in Yang et al. (2007), and tidal forces (VAC) from Donor et al. (2018), which contains a cata- exerted by nearby galaxies in two different fixed aper- log of identified APOGEE open cluster members. There ture volume limited samples (namely 1 and 5 Mpc pro- are currently 4 VACs that rely upon APOGEE DR14 jected distances within a line-of-sight velocity difference in order to extend and enhance the standard APOGEE of ∆ v ≤ 500 km s−1 Argudo-Fern´andez et al. 2015). data release products (DR14 APOGEE TGAS Catalog, Estimations of the local densities with the distance to APOGEE Red Clump Catalog, APOGEE DR-14 Based the 5th nearest neighbor are also provided. The local Distance Estimations, and OCCAM).15 density within N nearest neighbors include the correc- tions explained in Goddard et al. (2017b) following the 5.1.1. Technical Papers methodology described in Etherington & Thomas (2015). Two new APOGEE-related technical papers are high- To have a more general picture of the environment, we lighted below: the instrument paper from Wilson also provide a characterization of the cosmic web envi- et al. (2018) that relays an extensive description of ronment which can be used to identify galaxies in clus- the APOGEE spectrographs, and the APOKASC pa- ters, filaments, sheets, or voids, as explained in Zheng per from Pinsonneault et al. (2018) that details the et al. (2017). The full details of the reconstruction of APOGEE spectroscopic follow-up of Kepler stars. these density and tidal fields are described in Wang et al. (2009) and Wang et al. (2012). APOGEE Instrument Paper The forthcoming publication from Wilson et al. (2018) 4.5.5. MaNGA Spectroscopic Redshifts describes the design and performance of the near in- We present a Value-Added Catalog that contains the frared, fiber-fed, multi-object, high resolution APOGEE best-fit spectroscopic redshift and corresponding model spectrographs. Since 2011, the first APOGEE instru- flux for each MaNGA spectrum that has sufficient signal- ment has been in operation on the 2.5-m Sloan Tele- to-noise. We provide the mean of the spectroscopic red- scope at the Apache Point Observatory in New Mex- shifts sampled within the inner high signal-to-noise re- ico, USA (a Northern hemisphere site). Several key gion of the MaNGA galaxies, which can be compared innovations were made during the development of the to the single-valued NSA catalog redshift. Since the APOGEE instrument which include a multi-fiber con- MaNGA instrument uses the BOSS spectrograph, our nection system known as a ’gang-connector’ which allows results are derived by iterative application of the BOSS for the simultaneous disconnection and reconnection of pipeline’s spec1d software (Bolton et al. 2012) to pre- 300 fibers; hermetically sealed feedthroughs to permit cisely measure the redshift using the higher signal-to- fibers to pass through the cryostat wall continuously; the noise measurements as a prior to the algorithm, which first cryogenically-deployed mosaic volume phase holo- allows us determine good redshifts on spectra that would graphic grating; and, a massive refractive camera that not otherwise have sufficient signal-to-noise to result in is comprised of large-diameter mono-crystalline silicon a good redshift. Since the MaNGA survey uses an IFU, and fused silica elements. Specifically for the North- the radial velocity profile of the galaxy can significantly ern spectrograph, Wilson et al. (2018) reports on the impact the redshift of each spectra. Thus the spectro- following: the performance of the 2.5-m Sloan Founda- scopic redshifts can both be a benefit to galaxy kine- tion Telescope in the near infrared wavelength regime; matic measurements and improve the accuracy of spectra the cartridge and fiber systems; the optical and optome- modeling and analysis. The authors of this Value-Added chanical systems; the detector arrays and electronic con- Catalog have used the spectroscopic redshifts to search trols; the cryostat; the instrument control system; cali- for background emission lines to discover strong gravita- bration procedures; instrument optical performance and tional lenses in MaNGA (Talbot et al. 2018). 15 More information regarding all available APOGEE VACs in- 5. OTHER SURVEY DATA AND PRODUCTS cluding brief descriptions and the corresponding authors may be found in the SDSS online documentation (http://www.sdss.org/ 5.1. APOGEE-2 dr15/data_access/value-added-catalogs/). Data Release 15 19 stability; and, lessons learned. The final sections of Wil- son et al. (2018) provide similar details on the second APOGEE spectrograph located at the 2.5-m du Pont Telescope at Las Campanas Observatory in Chile. This second (Southern hemisphere-based) instrument, a close copy of the first, has been operating since April 2017. Wilson et al. also contains multiple appendices for the interested user. The Second APOKASC Catalog Over both the APOGEE-1 and APOGEE-2 Surveys, a joint effort known as the APOGEE Kepler Asteroseis- mic Science Consortium (APOKASC) APOGEE has en- gaged in a spectroscopic follow-up of stars in the Kepler field. Pinsonneault et al (submitted) presents the sec- ond APOKASC Catalog of stellar properties for a sam- ple of 6681 evolved stars with APOGEE spectroscopic parameters and Kepler asteroseismic data analyzed us- ing five independent techniques. The APOKASC data includes evolutionary state, surface gravity, mean den- sity, mass, radius, age and the spectroscopic and aster- oseismic measurements used to derive them. As shown in Figure 7, the APOKASC catalog asteroseismic log g values and evolutionary state classifications allow for a clear distinction between Red Giant Branch (RGB) and Red Clump (RC) members. Pinsonneault et al. (2018) employ a new empirical approach for combining astero- seismic measurements from different methods, calibrat- ing the inferred stellar parameters, and estimating un- certainties. With high statistical significance, they find that asteroseismic parameters inferred from the different pipelines have systematic offsets that are not removed by accounting for differences in their solar reference values. Pinsonneault el al. includes theoretically motivated cor- rections to the large frequency spacing (∆ν) scaling rela- Fig. 7.— Spectroscopic effective temperature (from APOGEE tion as well as calibrates the zero point of the frequency DR14) versus asteroseismic surface gravity (log g) in the of maximum power (νmax) relation to be consistent with APOKASC sample by asteroseismic evolutionary state. Red clump masses and radii for members of star clusters. For most (RC; core He-burning) stars are signified in blue while Red Giant Branch (RGB; H-shell or double shell burning) stars are shown targets, the parameters returned by different pipelines in red. The two populations have a clear offset in this plot, with are in much better agreement than would be expected RC stars having higher surface gravity at the same temperature from the pipeline-predicted random errors, but 22% of compared to RGB stars. them had at least one method not return a result and a much larger measurement dispersion. This supports the the handling of telluric absorption and persistence in the usage of multiple analysis techniques for asteroseismic DR13/DR14 versions of APOGEE-2 data as opposed to stellar population studies. DR12. Holtzman et al. (2018) details the derivation and calibration of stellar parameters, chemical abundances, In the SDSS DR14 data release paper (Abolfathi et al. and respective uncertainties, along with the ranges over 2018), brief references were made to the Holtzman et al. which calibration was performed. The work reports some (2018) and J¨onssonet al. (2018) publications. For the known issues with the public data related to the calibra- benefit of users, concise descriptions of each are now pro- tion of the effective temperatures (DR13), surface gravity vided. Please note that in addition to the DR15 Docu- (DR13 and DR14), and C and N abundances for dwarfs mentation, users should refer to these publications for de- (DR13 and DR14). Holtzman et al. (2018) also discusses tailed information regarding the Data Reduction Pipeline how results from The Cannon (Ness et al. 2015) are in- (DRP) and the APOGEE Stellar Parameter and Chem- cluded in DR14 and compares those with the values from ical Abundance Pipeline (ASPCAP) as well as to under- ASPCAP. stand data quality and performance. Comparison of SDSS/APOGEE DR13 and SDSS/APOGEE DR13 and DR14 Pipeline Pro- DR14 Values to Optical Results cessing and Data Description J¨onsson et al. (2018) evaluates the ASPCAP perfor- Holtzman et al. (2018) describes the data and analy- mance for both the DR13 and DR14 APOGEE datasets sis methodology used for the SDSS/APOGEE Data Re- with 160,000 and 270,000 stars, respectively. A compari- leases 13 and 14 as well as highlights differences from the son of the ASPCAP-derived stellar parameters and abun- DR12 analysis presented in Holtzman et al. (2015). For dances is done to analogous values inferred from optical example, the work demonstrates some improvement in spectra and analysis with a subset of several hundred stars. For most elements, J¨onssonet al. (2018) find that 20 SDSS Collaboration the DR14 ASPCAP results have systematic differences Palanque-Delabrouille et al. 2016) will also describe the with the comparison samples of less than 0.05 dex (me- LRG and quasar samples for the final eBOSS sample. dian) and random differences of less than 0.15 dex (stan- Several new algorithms for the spectroscopic data reduc- dard deviation). These departures are attributed to a tions were implemented in DR14 (Hutchinson et al. 2016; combination of the uncertainties in both the comparison Jensen et al. 2016); we will further improve sky subtrac- samples as well as the ASPCAP-analysis. Specifically, tion with higher order models to the fiber-to-fiber sky in comparison to the optical data, J¨onsson et al. (2018) model, flux calibration with new models for standard find that magnesium is the most accurate alpha-element stars, and spectral extraction to account for cross-talk derived by ASPCAP while nickel is the most accurate such as that found in (Hemler et al. 2018, submitted) for Fe-peak element (excluding iron). the final sample. A new method to improve the classi- Additionally in Abolfathi et al. (2018), detailed in- fication of galaxy spectra (Hutchinson et al. 2016) was formation was provided regarding the recently-published implemented in DR14 and new methods for classifying APOGEE-2 Targeting Paper from Zasowski et al. (2017) emission line galaxies (ELG) and quasars are being con- Users are encouraged to consult Zasowski et al. for spe- sidered for the final sample. cific details and insight regarding APOGEE-2 targeting. 5.3. Optical Emission Line Properties and Black Hole 5.1.2. New Value-Added Catalog - OCCAM Mass Estimates for SPIDERS DR14 Quasars The Open Cluster Chemical Analysis and Mapping This VAC, released in DR15, contains optical spectral (OCCAM) Survey generates a VAC of open cluster mem- properties for all X-ray selected SPIDERS quasars re- bers as targeted in both APOGEE-1 and APOGEE-2 leased in DR14. The SPIDERS DR14 catalog is based on fields. To establish membership probabilities, the cata- a clean sample of 9399 sources from the Second ROSAT log combines APOGEE DR14 DRP-derived radial veloci- All-Sky Survey catalog (2RXS; Boller et al. 2016) and ties (RV) and ASPCAP-derived metallicities with proper 1413 sources from the first XMM-Newton Slew survey motion (PM) data from Gaia DR2. This first VAC from catalog (XMMSL1; Saxton et al. 2008) with optical spec- the OCCAM Survey includes 19 open clusters, each with tra available. X-ray sources were matched to ALL- 4 or more APOGEE members. The OCCAM VAC con- WISE infrared counterparts using the Bayesian algo- sists of two components: a set of bulk cluster properties rithm “NWAY” (Salvato et al. 2018), which were then which include motions (RV, PM) as well as robust av- spectroscopically identified using SDSS (Dwelly et al. erage element abundance ratios; and, a set of member- 2017). Visual inspection results for each object in this ship probabilities for all stars considered in the analysis sample are available from a combination of literature of the 19 open clusters. For further information on the sources and the SPIDERS group, which provide both OCCAM Survey please consult Donor et al. (2018). reliable redshifts and source classifications. A spectral fitting code has been produced which fits the spectral 5.2. eBOSS, TDSS and SPIDERS regions around the Hβ and MgII emission lines and pro- There are no new reduced eBOSS data included in this vides both line and continuum properties, bolometric lu- data release; the VACs which are released are based on minosity estimates, as well as single-epoch black hole previously released eBOSS spectra. The final eBOSS mass estimates. This VAC includes X-ray flux measure- spectroscopic sample will be released in DR16. For more ments, visual inspection results, optical spectral proper- details on what’s coming in DR16 see §6.1. ties, black hole mass estimates, and additional derived DR14 marked the first cosmological sample from quantities for all SPIDERS DR14 quasars. For more de- eBOSS, consisting of spectra predominantly of luminous tails see Coffey et al. (2018). red galaxies (LRG) and quasars. These data enabled the 6. FUTURE PLANS first baryon acoustic oscillation (BAO) measurement in the 1 < z < 2 redshift range from quasars (Ata et al. SDSS-IV has a full 2 years of operations remaining, and 2018) and a 2.6% precision constraint on the distance is planning a further two public data releases. The next scale using the clustering of LRG’s (Bautista et al. 2018). data release, DR16, is now scheduled for December 2019 These measurements reflect the two primary goals for and will comprise data taken by both the APOGEE-N early eBOSS science, yet are only a subset of the results and APOGEE-S instruments through July 2018 as well from the two-year eBOSS sample. The large-scale struc- as being the final complete data release for eBOSS op- ture catalogs for both of these studies were released in erations. The final, complete release, DR18 (which will July 2018 and were not described in the 14th Data Re- follow an internal only DR17) is planned for December lease publication. These value-added catalogs can now 2020. 16 be accessed from the DR14 site and from this new re- 6.1. eBOSS lease in the parallel location. These catalogs contain all necessary information such as the window function, sys- The eBOSS schedule was recently accelerated in order tematic quantities, completeness estimates and correc- to achieve its cosmological goals earlier than previously tions for close-pairs and redshift failures to reproduce planned. This acceleration began on January 1, 2018 those clustering measurements, similar to the catalogs and continues through February 16, 2019, at which time from the final BOSS sample (Reid et al. 2016). eBOSS will complete its program significantly ahead of The publications that document the DR14 target se- the start of the Dark Energy Spectroscopic Instrument lection algorithms (Myers et al. 2015; Prakash et al. 2016; (DESI) survey (DESI Collaboration et al. 2016). Un- der the original schedule, eBOSS and MaNGA divided 16 https://data.sdss.org/sas/ebosswork/eboss/lss/ the dark time roughly equally. Under the new sched- catalogs/DR14/ ule, eBOSS will control all the dark time in January and Data Release 15 21

February 2018, will return to the original schedule for (e.g., see MacLeod et al. 2018), and the more recently March through July 2018, and will control all the dark initiated TDSS Repeat Quasar Spectroscopy (RQS; also time from August 2018 through February 16, 2019. see MacLeod et al. 2018) program – thereby also have The total time scheduled for eBOSS in this period been accelerated toward completion, in practice this ad- amounts to 1540 hours. Historical weather patterns in- vance is such that SDSS-IV data collection for the TDSS dicate that 50% of this time will allow for open dome RQS program in particular is now nearing completion17. with seeing and transparency conducive to spectroscopy. The TDSS RQS program obtains multi-epoch spectra Based on the 1158 plates completed from July 1, 2014 for thousands of known quasars (and with larger sam- until June 30, 2018, we project the final data sample to ple size, and greater homogeneity and less a priori bias include the spectra from observations covering 302 ELG to specific quasar subclasses than the TDSS FES pro- plates and more than 1200 LRG/quasar plates. These grams), all of which have at least one earlier epoch of observations will define the complete ELG sample fol- SDSS spectroscopy already available in the SDSS archive. lowing the selection algorithms in Raichoor et al. (2017). The RQS program especially addresses quasar spectral The final LRG and quasar samples will cover a volume variability on multi-year timescales, and in addition to roughly 2.3 times larger than the two-year cosmology its own potential for new discoveries of phenomena such samples released and analyzed in DR14. as changing-look quasars or broad absorption line (BAL) The final eBOSS sample will enable precision measure- variability and others, will also provide a valuable (and ments of BAO in the clustering of galaxies, quasars, and timely) resource for planning of yet larger scale multi- the Lyman-α forest. The final sample will also enable epoch quasar repeat spectral observations anticipated for new measurements of redshift space distortions in the the Black Hole Mapper program in the future SDSS-V anisotropic clustering of galaxies and quasars over the (see §6.6 below). From data taken for the RQS SDSS-IV redshift range 0.6 < z < 2.2. The next data release program to date, we expect RQS to add another recent has been scheduled around the expected time that these epoch of spectroscopy for ∼16000 SDSS quasars, sam- analyses will complete. This data release will be the last pling across a broad range of properties including red- to include new eBOSS data. Also included will be the shift, luminosity, and quasar subclass type. value-added catalogs that allow others to reproduce the final cosmology measurements. 6.4. MaNGA MaNGA will continue to take observations for the next 6.2. SPIDERS two years of SDSS-IV operations. The time trade with At the completion of the eBOSS survey, SPIDERS will eBOSS has slowed the rate of observations during 2018; have only obtained spectra from the ongoing followup however, it will provide an overall increase in the to- program of ROSAT and XMM-Newton sources. Contin- tal observing time allocation for MaNGA by 8%. The uing at the current pace, at the end of the survey SPI- projected final survey footprint, assuming we continue DERS will have collected about 12000 new spectra of X- nominal survey operations through July 2020, is shown ray selected AGN and 40000 spectra of member galaxies for two different expectations for weather at the tele- of about 5000 clusters over the final eBOSS area. scope and overlaid on other relevant surveys in Figure 2. The delayed launch of the eROSITA satellite (Predehl We expect to exceed our original goal of 10,000 galaxies et al. 2014), combined with the accelerated program for slightly under nominal weather conditions. obtaining eBOSS spectra mean that it will not be possi- ble to obtain redshifts for eROSITA targets during rou- 6.5. APOGEE-2 tine eBOSS operations. The eROSITA Performance Ver- ification data set is currently planned to be available by The APOGEE-2 Survey continues to acquire obser- early-mid 2019 and should consist of 120 sq deg, with vations from both the Northern and Southern hemi- 100-140 targets per sq deg. To address at least part of spheres. SDSS-IV Data Release 16 will contain the first the original goals of SPIDERS involving eROSITA fol- APOGEE-2 data from the Southern instrument. For lowup we plan to dedicate a special set of 12 plates for DR16, a variety of improvements are planned to both the these targets, however this plan cannot be confirmed un- DRP and the APOGEE Stellar Parameters and Chem- til February 2019. ical Abundance Pipeline (ASCAP). A new atomic line list will be generated (which will include transitions of Ce II and Nd II) and a new molecular list will be as- 6.3. TDSS sembled(which will be more extensive in size and will The accelerated pace for eBOSS discussed above cor- incorporate FeH features). An expansion of the stel- respondingly accelerates TDSS, which also relies on the lar atmosphere model grid is underway which will entail BOSS spectrographs, using a small portion (about 5%) the inclusion of higher surface gravities, lower carbon of the optical fibers piggybacking on eBOSS plates. abundances, and higher nitrogen abundances. Note that TDSS observations will thus effectively also conclude Model Atmospheres in Radiative and Convective Scheme with eBOSS data collection in about mid-February of (MARCS18) models (Gustafsson et al. 2008) will be em- 2019, and with SDSS-IV/TDSS data to be included in ployed for both the M and GK grids, ensuring a smooth the future DR16. Although all three main components of transition across an effective temperature range of ap- TDSS – the optical spectroscopic follow-up of PS1 pho- proximately Teff = 2500 − 6000K. Additionally, some tometric imaging variables (e.g., see Morganson et al. 2015, Ruan et al. 2016), repeat Few-Epoch Spectroscopy 17 This is primarily just because RQS piggybacks on a subset of (FES) of selected subclasses of stars and quasars antic- eBOSS plates which received recent heavy emphasis ipated or suspected to reveal spectroscopic variability 18 http://marcs.astro.uu.se 22 SDSS Collaboration tweaking of the data processing and derivation proce- Python system with a database which records the dure will occur. Planned modifications include the im- hundreds of millions of file paths, file sizes, and file provement of the LSF determination (with the poten- verification checksums. This system is currently tial employment of on-the-fly LSF derivation), a better being re-implemented to allow a more seamless and methodology for the extraction of the individual element high speed data access. abundances, and an improved technique for filling holes in the stellar atmosphere model grid. • Migrating the SDSS Software Repository to GitHub: The SDSS subversion software reposi- 6.6. SDSS-V tory, currently served along-side the SAS, will be Preparations for the fifth generation of SDSS are un- replaced by repositories copied into a GitHub orga- derway, with SDSS-V anticipated to begin operations in nization (https://github.com/sdss), with GitHub 2020 (Kollmeier et al. 2017) . SDSS-V will collect data Teams created to manage repository access con- at both APO and LCO using the existing APOGEE and trol, with public release of software including open BOSS spectrographs on the 2.5-meter telescopes, as well source licensing, starting with DR15 (e.g. Marvin as new optical spectrographs dedicated to integral field and the underlying code “Marvin’s Brain”). spectroscopy on new smaller telescopes. The current • SDSS Software Framework Development: SDSS plugplate system will be replaced with robotic fiber The Data and Operations Teams are currently positioners in the focal planes of the 2.5-meter telescopes. designing a new software framework to provide SDSS-V comprises three primary projects: the Milky Python-based tools, including improved data ac- Way Mapper, the Black Hole Mapper, and the Local cess, database access, data model documentation Volume Mapper. The Milky Way Mapper will use the and machine-readability. APOGEE and BOSS spectrographs to observe 4-5 mil- lion stars in the Milky Way and Local Group, probing • SDSS Software Containers: Portable images questions of galaxy formation and evolution, stellar as- of SDSS systems have been developed and imple- trophysics, and stellar system architecture. The Black mented on docker hub, and are currently used at Hole Mapper will use the BOSS spectrographs to mea- NERSC for the Science Archive Mirror and JHU for sure masses for ∼1200 supermassive black holes via rever- Science Archive Webapp development (e.g. Marvin beration mapping (e.g. Grier et al. 2017), determine spec- at JHU). The data team is now looking at develop- tral variability for ∼25,000 quasars, and provide identi- ing a new wave of such virtual machines to replicate fications and redshifts for ∼400,000 X-ray sources de- the experience of working on an SDSS computer at tected by eROSITA (Predehl et al. 2014). The Local the University of Utah. Volume Mapper will collect integral field spectroscopy 6.7.1. Modernizing SkyServer using new, R ∼ 4000 optical spectrographs coupled to small telescopes at APO and LCO. These spectra will The SkyServer has been the primary online web portal span ∼3000 deg2 of sky in the Milky Way midplane, to the Catalog Archive Server (CAS) since the beginning the Magellanic Clouds, and other Local Group galax- of SDSS, and although it underwent a significant facelift ies at high spatial resolution, with the goal of tracing in 2007, it is now woefully outdated in terms of its layout ISM physics and stellar-ISM energy exchange on differ- and the user experience, and generally in terms of its us- ent physical scales in a range of galactic environment. ability and accessibility. The SkyServer has been due for a rewrite with modern web technology for several years 6.7. Long Term Sustainability of the SDSS Archives now, and we are finally undertaking this daunting task Starting in 2017, the Science and Catalog Archive as we wind down SDSS-IV and look forward to SDSS-V. Teams have been proceeding on a roadmap toward a sus- One of the biggest constraints that makes this a difficult tainable data archive,19 designed to protect the legacy of enterprise is the large user base that the SkyServer has SDSS Data. built over the past 15+ years. We do not want to com- Some of the steps on this roadmap, which are currently pletely rearrange the site in such a way that users do receiving attention, include: not recognize it any more, and more importantly, we do not want to break all the functionality that works very • Archival Quality Storage: The Science Archive well currently in spite of the outdated interface. In short, Server (SAS) file system was not designed to last we want to adopt a philosophy of going from “working beyond warranty of the disks, and disk corruption to working” versions as we modernize the site. We list issues require meticulous and time intensive repair. below the specific changes we are currently working on. The SAS Team is currently implementing a ceph- based archival quality object storage system (Weil • Upgrade Technology: First and foremost, we et al. 2006) similar to that used by organizations are upgrading the web technology underneath the specializing in big data (e.g. Google and Amazon) SkyServer. This includes everything from the ver- providing complete internal redundancy, support sion of HTML and CSS that it was originally writ- for geographical distribution, internal failure de- ten in, to the way that the SkyServer website code tection and self recovery, and inexpensive backup is logically organized. We are going toward the in cloud-based big data object storage systems. MVC (model-view-controller) paradigm that mod- ern websites use to produce modular, reusable and • Science Archive Database: The census of what robust web applications. is contained on the SAS is managed through a • Portability: The SkyServer has been a Microsoft 19 Funded by a dedicated grant from the Sloan Foundation Windows application developed with the .NET Data Release 15 23

framework all along. This has made it very dif- CAS data and access tools will at that point be well- ficult to port even the front end to any other plat- positioned to be readily integrated into existing data cen- form. Now, with the availability of .NET Core, we ters for a minimal incremental cost. have the opportunity to migrate the website to a portable platform that can not only run in Linux, 7. ACKNOWLEDGEMENTS but can also be Dockerized. Funding for the Sloan Digital Sky Survey IV has been provided by the Alfred P. Sloan Foundation, the U.S. • Usability: Usability standards have changed con- Department of Energy Office of Science, and the Partici- siderably since the time when the SkyServer origi- pating Institutions. SDSS-IV acknowledges support and nally came online. In spite of significant upgrades resources from the Center for High-Performance Com- to different parts of the SkyServer over the years, puting at the University of Utah. The SDSS web site is there has not been a comprehensive reexamination www.sdss.org. of the usability aspects. We aim to rectify this as SDSS-IV is managed by the Astrophysical Research we redesign the user interface. Usability changes Consortium for the Participating Institutions of the will include effective presentation of information, SDSS Collaboration including the Brazilian Partici- compatibility with all browsers, responsiveness of pation Group, the Carnegie Institution for Science, page loads, and consistency of display modes (e.g, Carnegie Mellon University, the Chilean Participa- opening a new tab for results from a query and/or tion Group, the French Participation Group, Harvard- bringing the results pane to the front). Smithsonian Center for Astrophysics, Instituto de As- trof´ısica de Canarias, The Johns Hopkins University, • Accessibility: Accessibility pertains to the versa- Kavli Institute for the Physics and Mathematics of tility of the website and how responsive and easy the Universe (IPMU) / University of Tokyo, Ko- it is to use and work with for users that are re- rean Participation Group, Lawrence Berkeley National stricted in various ways. This ranges from users Laboratory, Leibniz Institut f¨urAstrophysik Potsdam on mobile devices to users with restricted access to (AIP), Max-Planck-Institut f¨urAstronomie (MPIA Hei- the internet as well as users with impaired vision or delberg), Max-Planck-Institut f¨ur Astrophysik (MPA other handicaps. Incorporating modern web design Garching), Max-Planck-Institut f¨ur Extraterrestrische standards and technologies will mostly take care Physik (MPE), National Astronomical Observatories of of these aspects, but we will pay special attention China, New Mexico State University, New York Uni- to make sure that the SkyServer can be used by versity, University of Notre Dame, Observat´ario Na- as many people as possible anywhere in the world cional / MCTI, The Ohio State University, Pennsylva- there is internet access. nia State University, Shanghai Astronomical Observa- tory, United Kingdom Participation Group, Universidad • Integrate SkyServer and Voyages: The Sky- Nacional Aut´onomade M´exico,University of Arizona, Server has an extensive educational sections that University of Colorado Boulder, University of Oxford, contains several levels of classroom exercise based University of Portsmouth, University of Utah, Univer- on SDSS data. These are known collectively as sity of Virginia, University of Washington, University of the SkyServer Projects. Voyages is a SkyServer Wisconsin, Vanderbilt University, and Yale University. “spinoff” website that has become quite popular This publication uses data generated via the Zooni- and presents several virtual “voyages” through the verse.org platform, development of which is funded by SDSS data for non-scientist audiences. The Voy- generous support, including a Global Impact Award from ages website is a much more modern web applica- Google, and by a grant from the Alfred P. Sloan Foun- tion that is based on a content management system dation. (CMS) - WordPress. This allows new pages and This research was made possible through the use of the functionality to be added to Voyages much more AAVSO Photometric All-Sky Survey (APASS), funded easily than SkyServer. As part of the SkyServer by the Robert Martin Ayers Sciences Fund. modernization, we are migrating all the SkyServer The Pan-STARRS1 Surveys (PS1) and the PS1 public student projects to Voyages and using the same science archive have been made possible through con- CMS (WordPress) for SkyServer too. We are also tributions by the Institute for Astronomy, the Univer- integrating Voyages further with SkyServer so that sity of Hawaii, the Pan-STARRS Project Office, the it uses the SkyServer API to run queries on the Max-Planck Society and its participating institutes, the SDSS data. Max Planck Institute for Astronomy, Heidelberg and the Max Planck Institute for Extraterrestrial Physics, • Streamline CAS ⇔ SAS Interface: There are Garching, The Johns Hopkins University, Durham Uni- hooks currently between SkyServer/Voyages and versity, the University of Edinburgh, the Queen’s Uni- the Science Archive Server, but they are awkward versity Belfast, the Harvard-Smithsonian Center for As- at best. The SAS API has recently been upgraded, trophysics, the Las Cumbres Observatory Global Tele- and the points of access to SAS data that currently scope Network Incorporated, the National Central Uni- exist in SkyServer and Voyages will be updated to versity of Taiwan, the Space Telescope Science Institute, use the proper SAS API calls. the National Aeronautics and Space Administration un- der Grant No. NNX08AR22G issued through the Plan- These steps will help ensure the availability of SDSS etary Science Division of the NASA Science Mission Di- data to astronomers for years to come, and long beyond rectorate, the National Science Foundation Grant No. the current funded plans for SDSS-IV and SDSS-V. The AST-1238877, the University of Maryland, Eotvos Lo- 24 SDSS Collaboration rand University (ELTE), the Los Alamos National Lab- ing this paper) and was organized by Jennifer Sobeck and oratory, and the Gordon and Betty Moore Foundation. Jos´e S´anchez-Gallego and Anne-Marie Weijmans and This work presents results from the European Space attended by Amy Jones, Ben Murphy, Bonnie Souter, Agency (ESA) space mission Gaia. Gaia data are be- Brian Cherinka, David Stark, David Law, Dan Lazarz, ing processed by the Gaia Data Processing and Analy- Gail Zasowski, Joel Brownstein, Jordan Raddick, Julie sis Consortium (DPAC). Funding for the DPAC is pro- Imig, Karen Masters, Kyle Westfall, Maria Argudo- vided by national institutions, in particular the insti- Fern´andez,Michael Talbot, Rachael Beaton, Renbin Yan tutions participating in the Gaia MultiLateral Agree- and Sten Hasselquist (as well as Becky Smethurst, Rita ment (MLA). The Gaia mission website is https:// Tojeiro, Ben Weaver and Ani Thaker via video link). www.cosmos.esa.int/gaia. The Gaia archive website This research made use of astropy, a community- is https://archives.esac.esa.int/gaia. developed core python (http://www.python.org) We would like to thank the e-Science Institute at package for Astronomy (Robitaille et al. 2013); ipython the University of Washington, Seattle for their hospi- (P´erez& Granger 2007); matplotlib (Hunter 2007); tality during DocuVana 2018. This event held in May numpy (Walt et al. 2011); scipy (Jones et al. 2001); 2018 kick started the documentation for DR15 (includ- and topcat (Taylor 2005).

REFERENCES Abazajian, K. N., Adelman-McCarthy, J. K., Ag¨ueros,M. A., DESI Collaboration, Aghamousa, A., Aguilar, J., et al. 2016, et al. 2009, ApJS, 182, 543 ArXiv e-prints, arXiv:1611.00036 Abolfathi, B., Aguado, D. S., Aguilar, G., et al. 2018, ApJS, 235, Dom´ınguezS´anchez, H., Huertas-Company, M., Bernardi, M., 42 Tuccillo, D., & Fischer, J. L. 2018, MNRAS, 476, 3661 Ahn, C. P., Alexandroff, R., Allende Prieto, C., et al. 2012, ApJS, Donor, J., Frinchaboy, P. M., Cunha, K., et al. 2018, AJ, 156, 142 203, 21 Drory, N., MacDonald, N., Bershady, M. A., et al. 2015, AJ, 149, Aihara, H., Allende Prieto, C., An, D., et al. 2011, ApJS, 193, 29 77 Alam, S., Albareti, F. D., Allende Prieto, C., et al. 2015, ApJS, Dwelly, T., Salvato, M., Merloni, A., et al. 2017, MNRAS, 469, 219, 12 1065 Albareti, F. D., Allende Prieto, C., Almeida, A., et al. 2017, Eisenstein, D. J., Weinberg, D. H., Agol, E., et al. 2011, AJ, 142, ApJS, 233, 25 72 Argudo-Fern´andez,M., Verley, S., Bergond, G., et al. 2015, A&A, Etherington, J., & Thomas, D. 2015, MNRAS, 451, 660 578, A110 Evans, D. W., Riello, M., De Angeli, F., et al. 2018, A&A, 616, A4 Ata, M., Baumgarten, F., Bautista, J., et al. 2018, MNRAS, 473, Falc´on-Barroso,J., S´anchez-Bl´azquez,P., Vazdekis, A., et al. 4773 2011, A&A, 532, A95 Bailer-Jones, C. A. L., Rybizki, J., Fouesneau, M., Mantelet, G., Finkbeiner, D. P., Schlafly, E. F., Schlegel, D. J., et al. 2016, & Andrae, R. 2018, AJ, 156, 58 ApJ, 822, 66 Bautista, J. E., Vargas-Maga˜na, M., Dawson, K. S., et al. 2018, Fischer, J.-L., Bernardi, M., & Meert, A. 2017, MNRAS, 467, 490 ApJ, 863, 110 Fischer, J.-L., Dom´ınguezS´anchez, H., & Bernardi, M. 2018, Beifiori, A., Maraston, C., Thomas, D., & Johansson, J. 2011, MNRAS submitted, arXiv:1811.02580 A&A, 531, A109 Frieman, J. A., Bassett, B., Becker, A., et al. 2008, AJ, 135, 338 Bernardi, M., Fischer, J.-L., Sheth, R. K., et al. 2017, MNRAS, Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al. 2016, 468, 2569 A&A, 595, A2 Blanton, M. R., Kazin, E., Muna, D., Weaver, B. A., & —. 2018, A&A, 616, A1 Price-Whelan, A. 2011, AJ, 142, 31 Goddard, D., Thomas, D., Maraston, C., et al. 2017a, MNRAS, Blanton, M. R., Bershady, M. A., Abolfathi, B., et al. 2017, AJ, 466, 4731 154, 28 —. 2017b, MNRAS, 465, 688 Bohlin, R. C., M´esz´aros,S., Fleming, S. W., et al. 2017, AJ, 153, Grier, C. J., Trump, J. R., Shen, Y., et al. 2017, ApJ, 851, 21 234 Gunn, J. E., Siegmund, W. A., Mannery, E. J., et al. 2006, AJ, Boller, T., Freyberg, M. J., Tr¨umper, J., et al. 2016, A&A, 588, 131, 2332 A103 Gupta, R. R., D’Andrea, C. B., Sako, M., et al. 2011, ApJ, 740, 92 Bolton, A. S., Schlegel, D. J., Aubourg, E.,´ et al. 2012, AJ, 144, Gustafsson, B., Edvardsson, B., Eriksson, K., et al. 2008, A&A, 144 486, 951 Bowen, I. S., & Vaughan, Jr, A. H. 1973, Applied Optics, 12, 1430 Hart, R. E., Bamford, S. P., Willett, K. W., et al. 2016, MNRAS, Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000 461, 3663 Bruzual A., G. 1983, ApJ, 273, 105 Hayden, B. T., Gupta, R. R., Garnavich, P. M., et al. 2013, ApJ, Bundy, K., Bershady, M. A., Law, D. R., et al. 2015, ApJ, 798, 7 764, 191 Cappellari, M. 2017, MNRAS, 466, 798 Haynes, M. P., Giovanelli, R., Kent, B. R., et al. 2018, ApJ, 861, Cappellari, M., & Copin, Y. 2003, MNRAS, 342, 345 49 Cappellari, M., & Emsellem, E. 2004, PASP, 116, 138 Hemler, Z., Grier, C., Brandt, W., et al. 2018, ApJ Chambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016, ArXiv Høg, E., Fabricius, C., Makarov, V. V., et al. 2000, A&A, 355, L27 e-prints, arXiv:1612.05560 Holtzman, J. A., Shetrone, M., Johnson, J. A., et al. 2015, AJ, Cherinka, B., S´anchez-Gallego, J., Andrews, B., & Brownstein, J. 150, 148 2018, sdss/marvin: Marvin Beta 2.2.5 [Computer Software], Holtzman, J. A., Hasselquist, S., Shetrone, M., et al. 2018, AJ, https://doi.org/10.5281/zenodo.1230529,,, 156, 125 doi:10.5281/zenodo.1230529 Hunter, J. D. 2007, Computing In Science & Engineering, 90 Coffey, D., Salvato, M., Merloni, A., et al. 2018, A& A submitted Hutchinson, T. A., Bolton, A. S., Dawson, K. S., et al. 2016, AJ, Conroy, C. 2013, ARA&A, 51, 393 152, 205 Conroy, C., & Gunn, J. E. 2010, FSPS: Flexible Stellar Jensen, T. W., Vivek, M., Dawson, K. S., et al. 2016, ApJ, 833, Population Synthesis, Astrophysics Source Code Library, , , 199 ascl:1010.043 Jones, E., Oliphant, T., Peterson, P., & Others. 2001, SciPy: Dalcanton, J. J., Williams, B. F., Lang, D., et al. 2012, ApJS, Open source scientific tools for Python, , , [Online; accessed 200, 18 2016-01-15] Dawson, K. S., Kneib, J.-P., Percival, W. J., et al. 2016, AJ, 151, J¨onsson,H., Allende Prieto, C., Holtzman, J. A., et al. 2018, AJ, 44 156, 126 Data Release 15 25

Koleva, M., Prugniel, P., Bouchard, A., & Wu, Y. 2009, A&A, Reynolds, A. P., de Bruijne, J. H. J., Perryman, M. A. C., 501, 1269 Peacock, A., & Bridge, C. M. 2003, A&A, 400, 1209 —. 2011, ULySS: A Full Spectrum Fitting Package, Astrophysics Rigault, M., Copin, Y., Aldering, G., et al. 2013, A&A, 560, A66 Source Code Library, , , ascl:1104.007 Robitaille, T. P., Tollerud, E. J., Greenfield, P., et al. 2013, A&A, Kollmeier, J. A., Zasowski, G., Rix, H.-W., et al. 2017, ArXiv 588, A33 e-prints, arXiv:1711.03234 Ruan, J. J., Anderson, S. F., Green, P. J., et al. 2016, ApJ, 825, Kurucz, R. L. 1979, ApJS, 40, 1 137 Kurucz, R. L., & Avrett, E. H. 1981, SAO Special Report, 391 Salvato, M., Buchner, J., Budav´ari,T., et al. 2018, MNRAS, 473, Lampeitl, H., Smith, M., Nichol, R. C., et al. 2010, ApJ, 722, 566 4937 Law, D. R., Yan, R., Bershady, M. A., et al. 2015, AJ, 150, 19 S´anchez, S. F., P´erez,E., S´anchez-Bl´azquez,P., et al. 2016, Law, D. R., Cherinka, B., Yan, R., et al. 2016, AJ, 152, 83 RMxAA, 52, 171 Leitherer, C., Schaerer, D., Goldader, J. D., et al. 1999, ApJS, S´anchez, S. F., Avila-Reese, V., Hernandez-Toledo, H., et al. 123, 3 2018, RMxAA, 54, 217 Li, N., & Thakar, A. R. 2008, Computing in Science and S´anchez-Bl´azquez,P., Peletier, R. F., Jim´enez-Vicente, J., et al. Engineering, 10, 18 2006, MNRAS, 371, 703 Luo, A.-L., Zhao, Y.-H., Zhao, G., et al. 2015, Research in Saxton, R. D., Read, A. M., Esquej, P., et al. 2008, A&A, 480, 611 Astronomy and Astrophysics, 15, 1095 Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, MacLeod, C. L., Green, P. J., Anderson, S. F., et al. 2018, AJ, 525 155, 6 Smee, S. A., Gunn, J. E., Uomoto, A., et al. 2013, AJ, 146, 32 Majewski, S. R., Schiavon, R. P., Frinchaboy, P. M., et al. 2017, Stoughton, C., Lupton, R. H., Bernardi, M., et al. 2002, AJ, 123, AJ, 154, 94 485 Maraston, C. 2005, MNRAS, 362, 799 Talbot, M. S., Brownstein, J. R., Bolton, A. S., et al. 2018, Maraston, C., & Str¨omb¨ack, G. 2011, MNRAS, 418, 2785 MNRAS, 477, 195 Masters, K., Stark, D., Pace, Z., et al. 2018, MNRAS submitted Taylor, M. B. 2005, in Astronomical Data Analysis Software and Meert, A., Vikram, V., & Bernardi, M. 2013, MNRAS, 433, 1344 Systems XIV —. 2015, MNRAS, 446, 3943 Thakar, A. R. 2008, Computing in Science and Engineering, 10, 9 —. 2016, MNRAS, 455, 2440 Thakar, A. R., Szalay, A., Fekete, G., & Gray, J. 2008, Morganson, E., Green, P. J., Anderson, S. F., et al. 2015, ApJ, Computing in Science and Engineering, 10, 30 806, 244 Trager, S. C., Worthey, G., Faber, S. M., Burstein, D., & Myers, A. D., Palanque-Delabrouille, N., Prakash, A., et al. 2015, Gonz´alez,J. J. 1998, ApJS, 116, 1 ApJS, 221, 27 Vazdekis, A., S´anchez-Bl´azquez,P., Falc´on-Barroso,J., et al. Nair, P. B., & Abraham, R. G. 2010, ApJS, 186, 427 2010, MNRAS, 404, 1639 Ness, M., Hogg, D. W., Rix, H.-W., Ho, A. Y. Q., & Zasowski, G. Wake, D. A., Bundy, K., Diamond-Stanic, A. M., et al. 2017, AJ, 2015, ApJ, 808, 16 154, 86 Oke, J. B., & Shipman, H. L. 1971, in IAU Symposium, Vol. 42, Walt, S. v. d., Colbert, S. C., & Varoquaux, G. 2011, Computing White Dwarfs, ed. W. J. Luyten, 67 in Science & Engineering, 13 Palanque-Delabrouille, N., Magneville, C., Y`eche, C., et al. 2016, Wang, H., Mo, H. J., Jing, Y. P., et al. 2009, MNRAS, 394, 398 A&A, 587, A41 Wang, H., Mo, H. J., Yang, X., & van den Bosch, F. C. 2012, Parikh, T., Thomas, D., Maraston, C., et al. 2018, MNRAS, 477, MNRAS, 420, 1809 3954 Weil, S. A., Brandt, S. A., Miller, E. L., Long, D. D. E., & Penny, S. J., Masters, K. L., Weijmans, A.-M., et al. 2016, Maltzahn, C. 2006, in Proceedings of the 7th Symposium on MNRAS, 462, 3955 Operating Systems Design and Implementation, OSDI ’06 P´erez,F., & Granger, B. E. 2007, Computing in Science and (Berkeley, CA, USA: USENIX Association), 307–320 Engineering, 9 Wilkinson, D. M., Maraston, C., Goddard, D., Thomas, D., & Pinsonneault, M. H., Elsworth, Y. P., Tayar, J., et al. 2018, Parikh, T. 2017, MNRAS, 472, 4297 ApJSS submitted (arXiv:1804.09983) Willett, K. W., Lintott, C. J., Bamford, S. P., et al. 2013, Prakash, A., Licquia, T. C., Newman, J. A., et al. 2016, ApJS, MNRAS, 435, 2835 224, 34 Wilson, J., Hearty, F., Skrutski, M., et al. 2018, PASP submitted Predehl, P., Andritschke, R., Becker, W., et al. 2014, in Worthey, G., & Ottaviani, D. L. 1997, ApJS, 111, 377 Proc. SPIE, Vol. 9144, Space Telescopes and Instrumentation Yan, R., Chen, Y., Lazarz, D., et al. 2018, ApJ submitted 2014: Ultraviolet to Gamma Ray, 91441T Yan, R., Bundy, K., Law, D. R., et al. 2016a, AJ, 152, 197 Raddick, M. J., Thakar, A. R., Szalay, A. S., & Santos, R. D. C. Yan, R., Tremonti, C., Bershady, M. A., et al. 2016b, AJ, 151, 8 2014a, Computing in Science Engineering, 16, 22 Yang, X., Mo, H. J., van den Bosch, F. C., et al. 2007, ApJ, 671, —. 2014b, Computing in Science Engineering, 16, 32 153 Raichoor, A., Comparat, J., Delubac, T., et al. 2017, MNRAS, Yanny, B., Rockosi, C., Newberg, H. J., et al. 2009, AJ, 137, 4377 471, 3955 York, D. G., Adelman, J., Anderson, Jr., J. E., et al. 2000, AJ, Reid, B., Ho, S., Padmanabhan, N., et al. 2016, MNRAS, 455, 120, 1579 1553 Zasowski, G., Cohen, R. E., Chojnowski, S. D., et al. 2017, AJ, 154, 198 Zheng, Z., Wang, H., Ge, J., et al. 2017, MNRAS, 465, 4572