Describing 3D Datasets

1
Describing 3D Datasets

• Igor Chilingarian
• (CRAL Observatoire de Lyon/Sternberg Astronomical
  Institute of the Moscow State University)
• IVOA DM DAL WG VO France Spec WG Euro3D

2
What is 3D spectroscopy?
IFU Spectroscopy

• SCIENCE CASE 1 History of the stellar population
  in nearby galaxies
• Search for stellar population subcomponents and
  their connection with kinematical substructures
  by making analysis of spectra integrated along
  the line of sight (presentation by P. Prugniel in
  the Theory IG session)
• Usage of high-resolution synthetic spectra
• Precise description of the instrumental response
  (spectral resolution) of the spectrograph is
  essential LSF(x,y,l) in case of 3D
• SCIENCE CASE 2 Stellar populations in AGNs
• Making 3D spatial-spectral decomposition of the
  active nucleus will allow to study stellar
  population and star formation in inner regions of
  active galaxies.
• In addition to the spectral resolution we need to
  know spatial PSF built by the telescopespectrogra
  ph.

3
What is 3D spectroscopy?
SCIENCE CASE 1 Studies of gas kinematics in disc
galaxies From 2D velocity field it is possible to
derive a rotation curve in the case of rotating
disc galaxies by taking into account non-circular
motions and perturbations caused by internal
structures in the galaxies (spiral arms, bars)
and then to study the distribution of dark and
luminous matter. By comparing these results to
the simulations one is able to understand the
evolution of the galaxies dynamics in a given
environment (credits O. Garrido) SCIENCE CASE 2
Star-formation complexes and supernovae remnants
in nearby galaxies shells of ionized gas Studies
of supernovae remnants and expanding shells
around star-forming regions (stellar winds) in
the nearby galaxies (e.g. dwarf irregulars)
allows to study gas-removal mechanism and its
rate, chemical composition variations and ISM
distribution, that might shed light on the
evolution of dwarf irregular galaxies. To study
shell-like structures, one creates so-called
position-velocity diagrams planes slicing the
IFP data cube (credits T. Lozinskaya).
Scanning Fabry-Perot Interferometer
Data Processing
Phase Surface
4
Storing 3D Data in FITS

• Key point no information should be lost ?
• one should avoid resampling
• Pure 3D data cube (for IFP data and for some IFU)
• 2D-image (one spectrum per row) binary table ?
• Euro3D Format

• Euro3D Format
• FITS binary data table one row per spectrum
• Binary table describing shape of spatial elements
  (spaxels)
• Some mandatory metadata, including common
  spectral WCS for all spectra, common spatial WCS
  for all spatial elements, meteo parameters during
  the observations, etc.

5
What we propose

• Simple demo at the end of the year
• VO access to two science-ready data archives ESO
  (SINFONI) and SAO RAS (MPFS)
• Retrieving the data
• Visualising them using Euro3D Visualisation Tool
• What we need
• DAL protocol will SSA 1 be sufficient?
• Some modifications of the Euro3D Visualisation
  Tool
• Some upgrade of the Euro3D format (optional)

6
Describing 3D dataset

• Important
• 1. Atmosphere dispersion correction (if not
  made)?
• information about meteo conditions is required
  ?
• characterisation is not sufficient use
  observation DM
• 2. Spectral and spatial resolutions (LSF and PSF)
  depend both on spatial and spectral coordinates
  LSF(x,y,l), PSF(x,y, l)
• 3. Principal question How to characterise image
  slicing?