Astrophysical Virtual Observatory

Document Sample
Astrophysical Virtual Observatory Powered By Docstoc
					Drivers, Status and Planning

                P.Quinn TIVO June10-14 2002   1
•   The need for a new research
    infrastructure in astronomy
•   AVO
    –   Work Program
    –   Status
    –   Future
•   The International VO Alliance

                          P.Quinn TIVO June10-14 2002   2
       The winds of change
“Internet computing and Grid technologies promise
  to change the way we tackle complex problems.
  They will enable large-scale aggregation and
  sharing of computational, data and other
  resources across institutional boundaries. And
  harnessing these new technologies effectively will
  transform scientific disciplines ranging from high-
  energy physics to the life sciences.”
  Dr. Ian Foster, Co-leader GLOBUS Project

                              P.Quinn TIVO June10-14 2002   3
Particle Physics Problems
        ATLAS                             CMS

   ~8-10 PetaBytes /year/on tape
                The LHC Detectors
     ~1 PetaByte/year/on disk
  ~8000 kSI95 (~300,000 PC2000)


                   P.Quinn TIVO June10-14 2002   4
   What has this to do with
• Astronomy has become a BIG international science
   – Gemini, VLT, ALMA, SKA, NGST, ELTs
• Astronomy projects involve
   – International coordinated research efforts
   – Distributed multi-wavelength teams, resources and data
   – Data volumes with doubling times < 12 months
• Astronomy service organizations need to
   – provide their communities with access to software tools,
     high quality raw and processed data in the face of
     desktop computing power and network bandwidths with
     doubling times > 18 months

                                   P.Quinn TIVO June10-14 2002   5
            Science Data Volume
        Archive Volume (GB)
                                                  T2 < 12 months


  10                                   Data Requested (GB)
   1                          10000



Science Archive Facility        10

                                      1996 1997 1998 1999 2000 2001

                                      P.Quinn TIVO June10-14 2002     6
         Projected Growth
T2 < 12 months

                  P.Quinn TIVO June10-14 2002   7
   Astronomical Strategies

Slow CPU growth     Distributed Computing
Limited storage     Distributed Data
Limited bandwidth   Information Hierarchies
                        - Move only what you need
Data diversity      Interoperability

                         P.Quinn TIVO June10-14 2002   8
    Distributed Computing at Work

                   Total              Last 24 Hours
                                                         •   Virtual and
Users              3675440            5075                   exploration of the
Results received   491854017          1092374

Total CPU time     954229.737 years   1662.448 years

Floating Point     1.502416e+21       4.260259e+18
Operations                            49.31 TFLOPs/sec

                                             P.Quinn TIVO June10-14 2002    9
  An Integrated European Approach
Information Society Technologies
GRID startups (today~40M Euro)
     [3.6B Euro 2003-2007]
Applications                                         EGSO

        GRIA                                                CROSSGRID

             GRIP EUROGRID                                        AVO
Middleware                                  GRIDLAB
& Tools           DAMIEN

          Industry / business                            Science

                                            P.Quinn TIVO June10-14 2002   10
              The AVO Proposal
• EU RTD Proposal submitted 15 February 2001 following OPTICON
  recommendations for a European VO effort (similar effort in the
  US [NVO] proposed by Decadal Report)
• ESO PI (Quinn) + STECF(Benvenuti), CDS (Genova), TERAPIX
  (Mellier), ASTROGRID (Lawrence) and Jodrell Bank (Diamond) +
  NVO affiliates
• Asked for funds for a three year Phase-A study of an
  Astrophysical Virtual Observatory – 7.2 Million € (50% from EU
  and 50% from organizations)
• Focus on
   – Multi-wavelength science case demonstrations (ASTROVIRTEL basis)
   – Interoperability demonstrations (CDS/All)
   – Technology assessments and testbeds (ASTROGRID/ESO)

                                       P.Quinn TIVO June10-14 2002   11
                        AVO STATUS
•AVO approved with EU funds ~2 Million € (total budget ~ 4M €)
• Contract start on 15 November 2001 - 3 Year Phase A study
• 9 NEW POSITIONS for 3 years over 6 institutions - total 18
FTE (~ 50 people)
•Total VO funding AVO+NVO+ASTROGRID = $21 million (US)
•3 Year target :
    •Build VO 1.0 among the 6 partner archive sets by
        •Defining and executing trial science cases
        •Defining, developing and deploying new
        interoperability standards and tools
        •Developing and deploying new Grid-based services

                                  P.Quinn TIVO June10-14 2002    12
           AVO Work Program
                                         Work Area O
                                  AVO Programme Management
                                    Deputy : Benoit Pirenne

                          WP0.1                            WP0.2
                      AVO Integration                   Phase-B Plan
                   WP Manager M.Dolensky            WP Manager : M.Dolensky
                   AVO Systems Engineer

    Work Area 1                        Work Area 2                    Work Area 3
      Science                       Interoperability                  Technology
WA Manager P.Benvenuti            WA Manager F.Genova             WA Manager A.Lawrence
       STECF                              CDS                         ASTROGRID

            WP1.1                             WP2.1                            WP3.1
         AVO SRM                        Archive Inclusion                Grid Technologies
 WP Manager : AVO Scientist TBD      WP Manager : F.Ochsenbein          WP Manager : G.Rixon

            WP1.2                             WP2.2                            WP3.2
   Science Requirements                Test and Evaluation          Storeage and Compute Tech .
 WP Manager : AVO Scientist TBD        WP Manager : M.Allen            WP Manager : A. Wicenec

            WP1.3                           WP2.3                               WP3.3
  Implement Science Cases              New Functionality                Database Technology
 WP Manager : AVO Scientist TBD       WP Manager : X.Derriere            WP Manager : C.Page

                                        Tool Evaluation
                                      WP Manager : F.Genova

                                                         P.Quinn TIVO June10-14 2002              13
                         AVO Science WG
Lofti BEN JAFFEL         Institute d’Astrophysique, Paris
Andrea CIMATTI           Arcetri Observatory, Florence                          Selection of preliminary
Richard DE GRIJS
                         ESO Garching
                         Institute of Astronomy, Cambridge                      science cases in June’02
Daniel EGRET             Observatoire de Strasbourg
Giuseppina FABBIANO      Center for Astrophysics, Cambridge, MA
Gerry GILMORE            Institute of Astronomy, Cambridge
Ana I. GOMEZ DE CASTRO   Universidad Complutense de Madrid
Krzysztof GORSKI         ESO Garching
                                                       Patrick LEAHY                 Jodrell Bank , Manchester
Preben GROSBØL           ESO Garching
                                                       Bruno LEIBUNDGUT              ESO, Garching
Bob HANISCH              ST ScI, Baltimore             Tommaso MACCACARO             Brera Observatory, Milan
Ulrich HOPP              Universitäts-Sternwarte MunichJanet MATTEI                  American Assoc. of Variable Star Observers
Florian KERBER           ST-ECF
                                                       Yannick MELLIER               Institute d’Astrophysique, Paris
Mary KONTIZAS            University of Athens          Tom OOSTERLOO                 NFRA (ASTRON), Dwingeloo
                                                            Patricio ORTIZ           Osservatorio di Capodimonte

  See Piero’s talk on
                                                            Fabio PASIAN             Osservatorio Astronomico, Trieste
                                                            Emanuela POMPEI          ESO Paranal

  ASTROVIRTEL                                               Philippe PRUGNIEL
                                                            Timo PRUSTI
                                                                                     Observatoire de Lyon
                                                                                     ESA / ESTEC
  development                                               Wolfgang VOGES           MPE Garching
                                                            Nicholas WALTON          Isaac Newton Group, La Palma
                                                            Hans ZINNECKER           Astronomical Institute, Potsdam

                                                                           P.Quinn TIVO June10-14 2002                    14

• Major step forward : VOTable 1.0
• See F.Ochsenbein talk
                     P.Quinn TIVO June10-14 2002   15
• Scalable storage and computing (see talk
  by Andreas Wicenec on NGAST)

• Web services prototypes (ASTROGRID
  talk by Andy Lawrence)
• Tool prototypes : QUERATOR (talk by F.

                          P.Quinn TIVO June10-14 2002   16
              VO Challenges
              AVO: the promise
          Bob Fosbury JENAM 2001
(1) To take the data collected by teams for specific projects
    and manage/massage it in such a way as to be of general
Comments -
•   If the data in an archive are neither well-calibrated nor
    well-described, they are of little value for the AVO
•   If the data are in good order, the AVO is unnecessary —
    it is just a layer of middle-men
•    If I were a funding agency, why should I support the
    ‘middle-man’ when I could fund the original science

                                  P.Quinn TIVO June10-14 2002   17
           A VO Roadmap
              Quinn ADASS’01

• To address Bob’s comments the
  international VO effort needs to define a
  unified and astronomically accepted
  roadmap for the next three years
• Without a publicly visible direction,
  accountability and receptiveness to
  requirements we will fail -
• This roadmap must be in defined/produced
  by the next international VO meeting

                         P.Quinn TIVO June10-14 2002   18
The International Virtual
 Observatory Alliance

              P.Quinn TIVO June10-14 2002   19
• IS an alliance of existing and future
  national and international projects to
  – Define the common ground needed to make an
    operational and scientifically effective IVO
  – Reach agreement on standards and
  – Allow the international scientific communities
    resident in the national projects, and
    elsewhere, a global channel for comment,
    criticism and guidance

                             P.Quinn TIVO June10-14 2002   20
     An Agreed Roadmap - 1
January 2002     Initiate international dialog on
                 interoperability. OPTICON
                 Interoperability Working Group
                 meeting, Strasbourg.
                 + Discussion/revision draft VOTable std.

April 15, 2002   Reach agreement on VOTable 1.0.

June 10-14, 2002 Formation of IVOA

                               P.Quinn TIVO June10-14 2002   21
                        Early Progress
• VOTable: A Proposed XML Format for Astronomical Tables
•   Roy Williams, California Institute of Technology, USA                   NVO
•   François Ochsenbein, Observatoire Astronomique de Strasbourg, France    AVO
•   Clive Davenhall, University of Edinburgh, UK                            AVO
•   Daniel Durand, Canadian Astronomy Data Centre, Canada                   CADC
•   Pierre Fernique, Observatoire Astronomique de Strasbourg, France        AVO
•   David Giaretta, Rutherford Appleton Laboratory, UK                      STARLINK
•   Robert Hanisch, Space Telescope Science Institute, USA                  NVO
•   Tom McGlynn, NASA Goddard Space Flight Center, USA                      NVO
•   Alex Szalay, Johns Hopkins University, USA                              NVO
•   Andreas Wicenec, European Southern Observatory, Germany                 AVO

•   Version 1.0 (15 Apr 2002)

•   Document repository:
•   Comments:

                                                        P.Quinn TIVO June10-14 2002    22
    An Agreed Roadmap - 2
January 2003      Coordinated initial science
                  demonstrations by IVOA
January, 2003     IVOA agreement on initial suite
                  of interoperability Standards and
May 2003          Working Published Web Services
August 2003Coordinated intermediate science
            demonstrations including
            international access at IAU
            General Assembly

                             P.Quinn TIVO June10-14 2002   23
    An Agreed Roadmap - 3
October 2003   Astronomical Query Language
January 2004   Coordinated intermediate
               demonstrations + Grid
May 2004       Resource Discovery 1.0
July 2004      IVO 2005+ roadmap
October 2004   Compound Web Services
               and Ontology Service 1.0
January 2005   Coordinated complex science

                         P.Quinn TIVO June10-14 2002   24
            An Invitation
The ASTROGRID, AVO and NVO projects
 take the opportunity of the Munich VO
 meeting to formally announce the IVOA
 and would like to extend an invitation to all
 VO projects to join this alliance for the
 pursuit   of    an   international   virtual
 observatory and the expansion of
 astronomical research capabilities in the
 21st century.

                          P.Quinn TIVO June10-14 2002   25
             ESO Archive Policy
ESO is fully aware of the importance of international collaborations in achieving the
   scientific goals and break-throughs necessary for astronomy to advance in the 21st
   century. The scientific potential of facilities like ALMA and OWL will only be
   realized by coordinated international efforts and the free exchange of astronomical
   ideas and data. To this end, ESO is preparing to provide open international access
   to all its archival data. Before this can happen, the ESO data holdings must be of a
   uniform high quality, in a form appropriate for recalibration and archival research
   and be supported by adequate operational manpower. ESO is taking the lead in
   Europe in defining data quality and interoperability standards through its
   coordination and funding efforts for an Astrophysical Virtual Observatory. The
   aim of the AVO is to provide the European and global astronomical community
   with an archival and data handling resource that can meet the needs of 21st century
   astronomy. The AVO program aims to join data from ESO telescopes with space
   and ground archives from Europe in an AVO Phase A facility by the end of 2003.
   From this time, ESO data will start to become freely available to the international
   astronomical community in a phased manner, following standards for data centers
   in the global virtual observatory and the availability of new VLT instruments.

                                                 P.Quinn TIVO June10-14 2002         26

Shared By: