Astronomical News

Report on the ESO–ESA Workshop Science Operations 2015: Science Data Management

held at ESO Headquarters, Garching, Germany, 24–27 November 2015

Martino Romaniello1 Specific topics included quality assurance and ESO to corroborate the growing role Chistophe Arviset 2 of science data and related cali­brations, of archive science in the astronomical Bruno Leibundgut1 data reduction and analysis, and science landscape. Examples cited were: more Danny Lennon 2 archives (content and user services). than 50 % of the refereed publications Michael Sterzik1 that use Hubble Space Telescope (HST) Some 45 talks and 20 posters were pre­ data are based on archival data; the user sented. All of the talks and a selection of base of processed data products from 1 ESO the posters are accessible via the work­ the ESO Science Archive Facility has 2 ESAC, Villafranca de Castillo, Madrid, shop web page1 (individual talks have grown to a grand total of more than 1200 Spain links to the Zenodo platform2 to provide individuals (and counting). permanent Digital Object Identifiers [DOIs] that are citable and discoverable). During four intense days, more than The programme was structured around Archives and data centres 100 astronomers, software engineers, broad groups of topics, or tracks. Here science operation and data manage- we will follow these tracks, rather than Felix Stoehr introduced ALMA science ment experts gathered for the second the purely chronological order of the data management, which may be con­ installment of the ESO–ESA workshop presentations. (In the electronic version sidered as an evolution of the Very Large series “Science Operations: Working of this article, clicking on the name of the Telescope (VLT) approach, but with together in support of science”. Two presenter leads to the corresponding important differences (e.g., the Science years ago, the inaugural meeting of the presentation.) Archive is at the heart of the ALMA data­ series at the European Space Astron- flow: quality assessment is made and omy Centre in Madrid provided an over- based on the actual science goals of the view of all of the different aspects of Introductory overviews PIs, rather than on relying on the instru­ Science Operations. This year’s gather- ment characterisation and modelling). ing was focused on science data The proceedings were opened by wel­ Stoehr also offered thought-provoking ­management, and an overview of the come addresses from Andreas Kaufer, ideas on the evolution of astronomy as a presentations and a summary of the ESO Director of Operations, and Martin science, from being ­photon-starved to discussions is provided. Kessler, Head of the Operations Depart­ having to deal with a deluge of data. ment at the European Space Astronomy Centre (ESAC). In presenting broad over­ Roland Walter brought in experience Introduction: Science data management views of the goals and perspectives of from high-energy astrophysics in demon­ for European astronomy the two organisations, they highlighted strating the power of legacy datasets the central role that science data man­ to generate new discoveries, especially ESO and the European Space Agency agement plays in enabling the scientific when analysed in ways that the original (ESA) generate a significant fraction of exploitation of the missions/observato­ researchers could not have anticipated. science data from ground and space for ries. Christophe Arviset from ESA and A range of examples was given, from the the European astronomical community Martino Romaniello from ESO expanded new discovery of an accretion event by (and beyond). These data feed both direct on this introduction. In particular, Arviset the black hole in the galaxy NGC 4845, principal investigator (PI) demands and presented the new approach that ESA is to a detailed study of the wind velocity use by the community-at-large through taking to integrate the development of all field of a nearby star. An archival stacked powerful science archives. The objective archives under the same organisational image corresponding to an exposure time of the workshop was to present and dis­ unit in order to foster close collaboration of no less than 35 years was particularly cuss the various approaches to science between astronomers and software engi­ impressive! data management, with the goals of: neers. This approach also allows individ­ – comparing and improving processes ual mission archives to be embedded In a series of talks spread over different and approaches; within a more general context,­ enabling sessions, Denis Mourard, Françoise – fostering innovations; and facilitating multi-mission scientific ­Genova, Giovanni Lamanna and Johannes – enabling a more efficient use of archive searches (e.g., the ESASky inter­ Reetz elaborated on the wider European resources; face, described below). Romaniello context of science data management. – establishing and intensifying collabora­ described ESO’s efforts towards an inte­ With initiatives like ASTRONET, the Hori­ tions, specifically exploring ways to grated approach to science data man­ zon 2020 Astronomy European Strategy enhance the value of the data through agement, ensuring that instruments are Forum on Research Infrastructures common strategies and practices; working properly, that the science con­ (ESFRI) and Research Infrastructure Clus­ – establishing ways to collect community tent can be extracted from the data and, ter (ASTERICS) project and the European feedback and gauge the success and finally, the science data delivered to our data initiative (EUDAT), it is clear that limitations of implemented solutions; users, PIs and archive researchers alike. attention, at a continental level, is high – exploring synergies and mutual support with regard to astronomical science data of science operations for ground and Impressive statistics on archive access management and the quest for synergies space missions. and data usage were shown for both ESA across disciplines. This found a nice echo

46 The Messenger 163 – March 2016 Figure 1. Delegates of the Science Operations 2015 Sauvage. With the aim of characterising future (4MOST on VISTA, WEAVE on the (SciOps2015) ESO/ESA Workshop pose for a photo the properties of the accelerating Uni­ William Herschel Teles­cope [WHT]) facili­ on the bridge connecting buildings at ESO Head­ quarters in Garching, Germany. verse in its components (dark matter and ties. Key to its success is to have people dark energy) and the dominant large- in the team who actively do science with scale force (gravity), the mission will col­ the data as this is the only way to under­ in the talks by Fréderic Hemmer from lect shape parameters for 1.5 billion gal­ stand the relevant details. the European Organization for Nuclear axies and 30 million photometric redshifts Research (CERN), where managing the to an accuracy of Δz of 0.001. Prepara­ The activities of the Wide Field Astro­ data is a significant challenge at all levels, tions are well underway, so that imple­ nomy Unit (WFAU) in Edinburgh were pre­ and a crucial one for its exploitation, and mentation can begin for the start of the sented by Nicholas Cross. Most of by David Schade and Séverin Gaudet. nominal mission in 2021. The topic of the data received is processed at CASU, The latter presented the nationwide ­distributed data management systems, and WFAU specialises in building and Canadian Advanced Network for Astro­ or data federations, was also the focus operating survey science archives. nomical Research (CANFAR) e-infrastruc­ of the presentation of Edwin Valentijn. ­Imaging ­surveys carried out on WFCAM ture and the transformational effect it has The case in point is the AstroWise infor­ on UKIRT, VIRCAM at VISTA and had on the role of data centres like the mation system, currently in operation with ­OmegaCAM on the VST account for most Canadian Astronomy Data Centre (CADC). OmegaCAM on the VLT Survey Tele­ of its current work, which also includes scope (VST) and the Multi Unit Spectro­ operating the archive for the Gaia–ESO The requirements, and associated scopic Explorer (MUSE) on the VLT. The spectroscopic Public Survey. Again, the ­challenges, of processing, calibrating to multi-disciplinary versatility of AstroWise importance of staff understanding the an astrometric precision of order 10 was highlighted, though its appli­cations science, the data and the archive system microarcseconds, and archiving the 1012 to the Low Frequency Array for Radio itself was highlighted as critical for its ulti­ measurements for one billion stellar astronomy (LOFAR) telescope, life sci­ mate success. sources and spectra delivered by the ence projects and business applications. ESA Gaia satellite were the focus of talks The Centro de Astrobiología (CAB) Data by Antonella Vallenari, Fred Jansen, Continuing the topic of data centres, Centre is the most important Spanish José Luis Hernandez-Munoz (presented Mike Irwin traced the multi-decennial astronomical data centre and was the by William O’Mullane), Marco Riello ­history of the Cambridge Astronomical focus of the presentation by Enrique and Juan Gonzalez-Núñez. In order to Survey Unit (CASU), from digitising Solano. Among others, it contains the meet these challenges, a pan-European United Kingdom Schmidt Telescope scientific archives of the Gran Telescopio cooperation, the Data Processing and (UKST) and Palomar Observatory Sky Canarias (GTC) and the Calar Alto Obser­ Analysis Consortium (DPAC), has been Survey (POSS) photographic plates to vatory (CAHA). It draws its raison d’être put in place, through which over 1000 supporting surveys on cutting-edge cur­ and success from a tight connection with staff years have been channelled since rent facilities (e.g., VIRCAM on the VLT its community of data providers, pro­ 2006, mostly supported through national Infrared Survey Telescope for Astronomy fessional astronomers and amateurs and funding. [VISTA], OmegaCAM on the VST, WFCAM the general public, providing them with on the United Kingdom InfraRed Tele­ added-value services on top of the data. A similar scheme of distributed data cen­ scope [UKIRT], MegaCAM on the Canada José Manuel Alacid introduced the GTC tres for processing, archiving and dis­ France Hawaii Telescope [CFHT], Public Archive. It provides access to tributing data is being put in place for the SuprimeCAM on the Subaru telescope, both raw and science-ready data, which next big ESA astronomy mission, Euclid, DECAM on the Cerro Tololo Inter-­ are provided both by the community the topic of the presentation by Marc American Observatory [CTIO], etc.) and and generated in-house. The archive has

The Messenger 163 – March 2016 47 Astronomical News Romaniello M. et al., Report on the ESO–ESA Workshop "SciOps2015"

been designed in compliance with the expert and standard users are provided ment path was identified and imple­ standards defined by the International in the form of virtual machines and web mented to take into account what was Virtual Observatory Alliance (IVOA) to interfaces, respectively. being learnt from the actual data, as guarantee a high level of data accessi­ opposed to the expectations from pre- bility and handling and, indeed, receives Ivan Zolotukhin presented an interesting operation simulations. more accesses through Virtual Obser­ example of citizen science, through vatory protocols than through web which a new web application was built to Steven Crawford presented the data ­interfaces. efficiently expose the 3XMM-DR5 cata­ management activities at the Southern logue from the XMM-Newton mission. By African Large Telescope (SALT). The Mark Allen presented the activities of providing convenient access to previously 11-metre-diameter telescope is operated the Centre de Données astronomiques existing data products, in its few months entirely in queue mode and is designed de Strasbourg (CDS), which since 1972 of public operations the interface has to explore the time domain. The chal­ has been collecting useful­ data on celes­ already tackled different science cases lenge of operating it on a small budget tial objects in electronic form, and then and led to a variety of results, including is met by leveraging existing software improving the data by critical evaluation the discovery of new cataclysmic varia­ and user expertise, ultimately yielding a and combination, before distributing the bles, of the first non-recycled pulsar and rate of refereed publications comparable results to the international community of the second cooling neutron star. to other 8–10-metre-class telescopes. and conducting research using these data. The CDS has grown to be a refer­ Providing a glimpse of the ground-based ence data centre for astronomy that Science data management future, Gijs Verdoes Kleijn’s talk focused delivers critically evaluated, professionally on the science data management of the curated information and whose services, Science data management at ESO was European Extremely Large Telescope’s e.g., SIMBAD, Vizier, Aladin and X-Match, discussed in some detail in the talks (E-ELT) first-light instrument, the Multi- provide added value to scientific content by Steffen Mieske and Burkhard Wolff Adaptive Optics Imaging Camera for in order to support the astronomy (data quality assurance and instrument Deep Observations (MICADO), an optical research community. trending), Reinhard Hanuschik (quality- and near-infrared imager and spectro­ controlled generation of science data graph. Key to being able to handle data The archives of several ESA astronomy products for archive publication), Wolfram from such a system, composed not and planetary missions were presented Freudling (user science data reduction), only of the instrument itself, but also a in the talks by Santa Martinez and Iñaki Jörg Retzlaff (Phase 3, i.e. the publication highly complex, time-variable teles­cope Ortiz de Landaluce (BepiColombo), of science data products through the and Earth’s atmosphere, is to model it ­Deborah Baines (the new European HST ESO Science Archive Facility) and thoroughly. In this way the instrument, archive), Eva Verdugo (the Herschel Sci­ ­Nausicaa Delmotte (validation of Phase 3 telescope and atmosphere status can ence Archive), Xavier Dupac (the Planck data). Isabelle Pércheron elaborated on be determined at once, and then the Legacy Archive), Tanja Lim (ExoMars) the experience of bringing the PIONIER ­science information extracted from the and Michele Armano (Laser Interfer­ VLT Interferometer instrument from a astronomical source. ometer Space Antenna [LISA] Pathfinder). (very successful) visitor instrument to one Bruno Merín showcased the brand-new that is offered to the whole community, Arguably the ultimate outcome in the life ESASky astronomy multi-mission inter­ fully integrated into ESO dataflow opera­ cycle of science data is that of generating face, which was recently released in its tions. Marina Rejkuba described how refereed publications. The NASA Astro­ beta version, and which brings access to ESO s­upports its community in preparing physics Data System (ADS) has long all ESA astronomy missions under one and designing observations in order to been used as a source of information roof, allowing for easy access to multi- maximise their science return as well as about the scientific literature in astronomy wavelength targets across the entire sky. the overall efficiency of the observatory. and physics. Alberto Accomazzi reported ­Pascal Ballester reported on scientific on recent efforts at ADS to widen the Vicente Navarro introduced the activities software development at ESO, discussing range of available scientific resources to carried out at ESAC to provide long-term the lessons learned and evolution of the include datasets linked to the publica­ preservation of data analysis software process for the next generation of tools tions, observing proposals and software for four missions that are now in their leg­ and observing facilities. used in refereed astronomy papers. Per­ acy phase, namely ISOPHOT Interactive haps not surprisingly, well-linked data Analysis (PIA) for the Infrared Space Peter Weilbacher presented the MUSE papers, which are easy to discover by a Observatory (ISO) from 1995, the Science instrument, a wide-field integral field broad audience, ultimately have a higher Analysis Software (SAS) for the X-ray spectrograph that started science oper­ impact by receiving more citations. Multi-Mission Mission (XMM) Newton ations about one year ago at the VLT. ­satellite from 1999, the Herschel Interac­ The availability of a mature science pipe­ Tracking the publication record of ESO tive Processing Environment (HIPE) from line from the beginning has allowed facilities was the focus of the presenta­ 2009 and EXIA for the European X-ray ­science results to be generated from the tion by Uta Grothkopf on the ESO Teles­ Observatory SATellite (EXOSAT), dating very complex data that the instrument cope Bibliography (telbib). telbib is a back to 1983. Tailored solutions for produces. At the same time, a develop­ database of refereed papers published by

48 The Messenger 163 – March 2016 the ESO user community that links data The great response to the workshop, with ties with the one that saw classical in the ESO Science Archive Facility with more than 100 people attending, testifies observing being more and more replaced the published literature, and vice versa. to the timeliness of a broad discussion by service or queue mode for ground- After careful curation, the rich metadata on science data management in astron­ based observatories. As was the case for provide parameters for reports and omy. It provided a forum for people to the observing transition, the success ­statistics that allow the performance of exchange ideas and compare approaches of the data transition will critically depend ESO’s facilities­ to be investigated and to through which both the similarities and on two pillars: building the trust of the help understand trends and develop­ the differences were identified, not only user community in the data they will ments in the publishing behaviour of the concerning the problems in this particular receive; and providing a streamlined, user community. telbib is an invaluable area that we are facing as a community, ­science-oriented user experience. Data tool to guide future developments. but also in the ways in which they can be and archive service providers will have to addressed. The workshop will hopefully adapt to and participate in defining these give momentum to a dialogue, or, better changes. Concluding discussion said, dialogues, that will carry on and deepen in the future and foster collabora­ After the introductory general workshop Andreas Kaufer and Danny Lennon tions for the benefit of the astronomical two years ago (Primas & Hanowski, wrapped up the workshop by offering user community at large. 2013), the success of this ESO–ESA points for discussion on the themes that workshop, focused on a specific topic, emerged during the workshop itself: confirms the validity and potential of – Quality assurance is critical in building Prospects the formula. Discussions have already trusted content, which, in turn, is critical started on possible topics for the next to having the community embrace Astronomy as a science is going through installment in the series: see you in two archives for science use. a very interesting phase of change, driven years at ESAC! – There is likely room for both centralised by the fast-growing amount and com­ and distributed data reduction, depend­ plexity of data. Analysis of multi-wave­ ing on the project. The role of obser­ length, multi-messenger data will no Acknowledgements vatories to support this is probably doubt grow in importance, as will the We would like to warmly thank the Local Organising going to increase as data volume and need to minimise the transfer of large vol­ Committee (Véronique Ziegler, Adam Dobrzycki, complexity increase. umes of data by processing them in situ. Michael Naumann and Rein Warmels) and Simon – The content of science archives should In an increasing number of cases the Lowery and Judy Asher for IT support. ESA’s sup­ be trusted and ready for science. focus is shifting from primarily working port in providing food and refreshments to sustain the long hours of (fruitful!) discussions is gratefully – Archive services should be oriented to with one’s own data obtained directly acknowledged. Special thanks go to Alberto users in order to increase the ease at the telescope to complementing, or ­Accomazzi for taking time off from his Thanksgiving of utilisation and the discoverability of even replacing this data entirely, with holiday to be with us for his presentation. data, with tools to enhance the scien­ data accessed from powerful archives tific return of the (ground and space filled and curated by dedicated profes­ References based) facilities. sionals. In some cases the archives are filled with data originally obtained for Primas, F. & Hanowski, N. 2013, The Messenger, The debate that followed was too varied observing proposals on specific individ­ 154, 67 to be represented here in any detail, but ual projects, while in other cases, as for it surely highlighted the richness and the Sloan Digital­ Sky Survey (SDSS) Links enormous scientific potential that science or the Large Synoptic Survey Telescope 1 data management increasingly has in (LSST), data are conceived from the Workshop web page: https//www.eso.org/sci/ meetings/2015/SciOps2015.html the present and future landscape in onset for archive use by the community 2 Zenodo DOI platform: http://zenodo.org/ astronomy. at large. This transition has many similari­

A 360-degree panorama of the Very Large Tele­ scope Control Room at night.

The Messenger 163 – March 2016 49