Notes on Anita's Talk by Juan de Dios Santander Vela

My notes on Anita's talk. Sorry if they are not what you expect from them, and please correct me if your memory says otherwise!

- Anita Richards: Polarisation
   - Encourage feedback on Polarisation description note.
   - Provide new UCDs/UTypes
   - Most of polarisation data, from radio
   - But all EM, intrinsically polarised
   - Current FITS handling of polarisation
      - FITS codes: I, Q, U, V, RR, LL, RL, LR, XX, YY, XY, YX, POLI, POLA, UNDEF
      - All units, flux density units; in the fundamental plane, Jy;
        for radio images typically, in Jy/beam
   - Pol angle, units of observable angle.
   - Transposition of the codes into unambiguous UCDs (semantics)
   - Changing the phys.polarization UCD because of ambiguity
   - Polarisation handling in the VO
      - Most common axes of polarisation are total intensity or single polarisation
      - But most important currently published data collection is NVSS
      which comprises IQU cubes 
      - Use of additional features in SIAv2 protocol for querying polarisation.
   - XML serialisation
      - Additional characterisation axes for polarisation.
      - In order to get meaningful results from Jy/beam images, the beam
        has to be provided (in resolution).
      - In order to provide meaningful intensities, you need to have at least two
        planes for coverage, including the negative and positive values.
	Certain polarization products (and other quantities like line absorption) can be validly negative so coverage of valid values of the observable encompasses two ranges, for the negative and positive values.

What is not in a FITS header?
NVSS cube: 
The fundamental information is in the FITS header but some more detailed properties of the Q and U planes are in the HISTORY fileand not in "proper" FITS headers 
Typical query: data that is better than a certain sensitivity.
Why XML? 
Bonnarel: This comes from the beginning of the DM WG.
Doug Tody: in the query response mechanism of SIAv2 we have UType characterisation of polarisation.
Anita cannot foresee problems in the proposed UType mechanism to deal with polarisation.
Bonnarel: this comes from mandate of the DM WG to have XML serialisations, that can be made equivalent to UType serialisations.