Astronomical spectroscopy and the Virtual Observatory: a user-oriented perspective

1
Astronomical spectroscopy and the Virtual
Observatory a user-oriented perspective

• Matteo Guainazzi1, Andrea Laruelo, Jesus Salgado,
  Isa Barbarisi, Pedro Osuna
• ESA-VO, Science Operations Department
• European Space Astronomy Center of ESA
• Villafranca del Castillo, Spain
• 1also EURO-VO astronomer

2
Outline

• Why do we care?
• Spectroscopic surveys in the past and in the
  future
• Which added value can VO bring to the analysis of
  spectroscopic data?
• What does the VO currently offer?
• Data models
• Data access protocols
• Spectral services
• Spectral VO applications
• Adds-on
• Access to atomic and molecular lines databases
• Access to theoretical models

3
Spectroscopic surveys the past
(Yip et al. 2007)
Number of redshifts in surveys as a function of
time
4
Evolution of survey source numbers
(Cropper 2007)
5
What do we gain with the VO?

• We should be able to largely gain in efficiency
  (from it would take my whole lifetime to do it
  to it can be done by my Ph.D. student) for
  tasks like
• Building spectral energy distributions
  (spectraphotometry) by combining large datasets
  from different sources
• Comparing spectra and images (footprint,
  contamination )
• Analysis and clever visualization of large
  datasets
• Correlation of spectrally-derived parameters from
  different sources
• Automatic classification (and outliers discovery)
• Multi-parameter connectedness (M.Cropper)
• Selection of sub-samples on the basis of a
  multi-parameter space
• Access to ancillary information useful for the
  interpretation of the data existing object
  classification, publications

6
What should VO tools be also able to do?

• Basic display tools superposition, rebinning,
  smoothing, dynamical spectra
• Radial velocity measurements and superposition of
  spectral region in the velocity parameter space
• Rectification
• Fits (Gaussian, Maxwellian, Voigt profiles)
• Measurement of line profile asymmetry and
  broadening
• Time analysis (periodograms, Fourier analysis,
  Doppler imaging)
• Spectral deconvolution in energy and Fourier
  space
• Comparison between observed spectra and spectral
  templates
• Comparison between observed spectra and
  theoretical models
• Not all functionalities must be simultaneously
  present in a given tool!

7
What does the VO offer?

• Data models and data access protocols
• Simple Spectral Access (SSA Tody et al. 2007)
• Based on the concept of a 1-D spectrum
  spectral coordinates, flux, error, quality flag,
  etc,
• Virtual data generation cut-out, reprojection,
  dynamic extraction etc.
• Support to multiple data format VOTable, FITS
  binary, CSV, native XML, HTML
• Spectrum Data Model (Mc Dowell et al. 2007)
• Spectral services
• Applications
• Specview (STScI)
• SPLAT (AstroGrid)
• VOSpec (ESA-VO)

8
What does the VO offer?

• Data models and data access protocols
• Simple Spectral Access (SSA Tody et al. 2007)
• Based on the concept of a 1-D spectrum
  spectral coordinates, flux, error, quality flag,
  etc,
• Virtual data generation cut-out, reprojection,
  dynamic extraction etc.
• Support to multiple data format VOTable, FITS
  binary, CSV, native XML, HTML
• Spectrum Data Model (Mc Dowell et al. 2007)
• Spectral services
• Applications
• Specview (STScI)
• SPLAT (AstroGrid)
• VOSpec (ESA-VO)

9
EURO-VO spectroscopic tools
VOSpec
SPLAT-VO
http//star-www.dur.ac.uk/pdraper/splat/splat-vo/
splat-vo.html
http//esavo.esa.int/vospecapp
10
Basic working scheme
Collect spectra
Build SED
Prepare spectra
Analyse spectra
Publish results

• Local files
• Remote servers

• Data screening
• Rebinning
• Filtering
• De-noising
• De-reddening
• Units conversion
• Spectra arithmentic
• Convolutions

• Paper-quality plots
• Export analysis

• Units?

• EW
• Flux
• Fit
• Bisectoring
• Mirroring
• Interpolation
• Line identification
• Theory

11
SED generation and graphics

• Access to spectra from
• Local files (FITS, ASCII, NDF )
• Files available on the Internet
• Simple Spectral Access Protocol compatible
  servers
• (CDF-S, EUVE, FUSE, Giraffe, HST, HUT,
  Integral/OMC, ISO, IUE, SDSS, XMM-Newton/RGS,
  WHS, WUPE )
• Units handling
• Most of the standard units on the
  frequency/energy/wavelength axis, and on the flux
  density axis can be hadled automatically
• VOSpec also handles automatically any units via
  dimensional analysis
• Plotting
• Single- and multi-spectra plotting, with various
  standard graphical facilities
• SPLAT allows the co-existence of different
  plotting windows, a drag the file name function
  to display a new spectrum, and movie generation
• VOSpec allows interactive mouse-driven zooming

12
Analysis tools SPLAT

• Conversion to velocity, redshift or Lorentz
  factor space
• Spectral arithmetic for spectra (limited to
  strictly overlapping frequency ranges)
• Pairs of spectra
• Spectra and a constant
• Cut and removal of spectral ranges
• Fit of continua and emission/absorption lines
• Generation of a spectrum from an interpolated
  line (needed to evaluate the continuum level
  underneath a fitted line)
• Filtering
• Flipping/translating
• SPEFO (legacy software from Ondreov Observatory)

13
VOSpec spectral analysis functionalities I.

• Arithmetic and convolution between spectra
• Line fitting
• Best Fit of theoretical models

Fit of an absorption line profile with different
models (Gaussian, Loretntzian Voigt)
Fit of theoretical models to an observed spectrum
(see later)
14
VOSpec spectral analysis functionalities II.

• Mirroring, bisector
• Interactive integrated flux and equivalent width
• Filtering
• Discrete wavelet analysis
• Wavelength-to-velocity conversion
• Dereddening
• Calzetti law
• Cardelli-ODonnel law
• LMC law
• Calculation of luminosity differences between
  observational spectra and models

15
The power of interoperability I.
Aladin multipanel imaging
AstroScope catalogue search
GB
2MASS
SDSS
ROSAT
Chandra
VOSpec generation of SEDs
TOPCAT tabular data plotting/manipulation
16
Community efforts I. EZ and GOSSIP
GOSSIP Spectrophotometric analysis through SED
fitting (not VO yet)
EZ Automatic calculation of spectroscopic
redshifts
(Franzetti 2007 Garilli 2007)
http//cosmos.iasf-milano.inaf.it/pandora
17
Simple Line Access through VOSpec

• VOSpec provides access to atomic and molecular
  line Databases throuh a specific VO access
  protocol (SLAP)
• NIST
• CHIANTI (X-rays)
• CIELO (XMM-Newton)
• IASD (ISO)
• LERMA (molecular)
• This allows on-the-fly identification of
  transitions associated to a given observed
  feature
• SPLAT includes a hardcoded line identification
  facility on the basis of standard sets of lines

18
Easy access to theoretical models (TSAP)
SVO web server implementation (http//svo.laeff.in
ta.es)
ESA-VO VOSpec implementation
19
Thoretical models fit step by step

• Selecting TSA server

20
A typical session
21
Will we ever have X-ray data in the VO?

• Traditional wisdom no!
• Spectra in X-ray archives are in instrumental
  counts
• Conversion into physical units is strongly model
  dependent due to non-linearity of the
  redistribution matrix
• We X-ray astronomers know how to do it ask us!

XMM-Newton EPIC redistribution function
?
22
Yes! XMM-Newton/RGS
(Gonzalez-Riestra Rodriguez-Pascal 2007)

• However, high-resolution instruments are now
  flying XMM-Newton RGS (R300-1800)
• The redistribution function is very close to
  linear
• Model independent fluxed spectra can be produced
• Soon availabe as VO server (the first all-sky
  spectral X-ray server ever)
• Pilot project on Seyfert 2 galaxies available
  since 1 year (CIELO-AGN)
• No general solution yet for moderate-resolution
  spectra (XMM-Newton/EPIC, Chandra/ACIS,
  Suzaku/XIS, Swift )

RGS redistribution function
ABDor
23
Conclusions wher do we stand?

• On-the-fly SED generation
• Spectra access from SSAP-compatible server
• Automatic handling of units
• Spectral analysis tools
• Growing suite of spectral analysis tools
  available in the mainstream tools
• Multi-component fit possible
• Community-based web-services start appearing for
  more specialized tasks
• Interoperability
• All tools use PLASTIC
• First steps to export via PLASTIC also the
  analysis results
• Usage on massive sets of data
• Mainly designed to handle one source at the time
  ? highly interactive

24
Science highlight from spectroscopic surveys
VVDS Discovery of high-redshift dusty galaxy
population with strong star formation
XMM-Newton
SDSS High-redshift QSOs
SDSS, z0.1
(Cooper et al. 2006)
SDSS New stellar streams and satellites of the
Milky Way
DEEP2, 0.75ltzlt1.05
SDSSDEEP2 Red galaxy environment