Data Characterisation

In characterising a data product we have to describe

In the following only the spatial description is considered. Similar work on the other axes is more than welcome.

Notes:

  1. Polarisation not included.
  2. Probably too biased towards the optical regime; please, X, Radio, Interferometry experts: intervene!
  3. My apologies if I have missed something, please send me corrections.
  4. Suggested names are likely to change after similar work on other axes is carried out (aim: uniform description of different axes).
  5. I still have to read Arnold's STC Metadata document! It should really be included in the comparison here.
    Arnold, maybe you could help me ... ?

Spatial Characterisation: Comparison of various models

In the following the Spatial Characterisation is devided into 3 parts:
  1. Spatial Coverage: area covered, in both scalar (simple FoV for query purposes) and vectorial context (for full description, intersections, etc)
  2. Spatial Sampling: resolution, bins, nyquist, undersampling, etc.
  3. Spatial Sensitivity: to do with sensitivity maps, pixel response function, vignetting, etc.

Spatial Coverage

  Concept Resource Metadata V0.82 CVO (Pat's) Model Doug's Requirements Suggested
C
o
v
e
r
a
g
e 		Detailed vectorial representation of spatial sky coverage (eg, polygon) Coverage.Spatial spatial_bound

Polygons, in equatorial, galactic and/or ecliptic coordinates

 circular or rectangular region or aperture on the sky
Full WCS should be defined elsewhere		 DM.coverage.spatial.shape
Estimate of the size of the field of view

Useful to identify products with eg large enough FoV

 Coverage.RegionOfRegard

(half the FoV in degrees)

 spatial_bound.size()

spatial_bound includes methods for computing various things, eg:

  • center() the coordinates of the pointing
  • size() the diameter of the minimum bounding circle (ie. the largest ~linear dimension),
  • area() the area covered by the polygon,
  DM.coverage.spatial.fov

FoV Diameter, but also (FoVx, FoVy)?
Estimate of the fraction of sky represented in the observation Coverage.SkyFraction [0..1] spatial_fill

The fraction of area actually sampled by the detector or spectrograph entrance aperture.

   DM.coverage.spatial.fill_factor
Astrometric Accuracy Uncertainty.Spatial (float) the error in spatial_bounds

which is assumed to be a systematic error appied to all vertices.

   DM.coverage.spatial.uncertainty (degrees)

Spatial Sampling

  Concept Resource Metadata V0.82 CVO (Pat's) Model Doug's Requirements Suggested
S
a
m
p
l
i
n
g 		Product Resolution (including seeing, optics, detector, readout) Resolution.Spatial (degrees)

vague definition

 spatial_resolution

The instrumental spatial resolution (FWHM).

 spatial_bandpass.loValue

Doug also considers a hiValue, a fill_factor in order to describe the range of spatial frequency in the data.

 DM.coverage.spatial.sampling.resolution (degrees)

Overall resolution, covolution of Seeing, PSF, and detector pixel size projected onto the sky
Heuristic formula:
sqrt( seeing**2 + PSF_fwhm**2 + ps**2 )
where, for diffraction limited instruments, PSF_fwhm = 1.22*Lambda/D

Seeing Component N/A     DM.coverage.spatial.sampling.atmospheric (seeing)

(avgSeeing, errSeeing)
Point Spread Function Characterisation N/A     DM.coverage.spatial.sampling.optics (psf)

(loValue, refValue, HiValue)
Bin Size (pixel scale on the sky after readout, ie, including possible binning) N/A spatial_sample

The spatial sampling is the pixel size or bin size in decimal degrees. For slit spectra this represents the spatial sampling perpendicular to the direction of the dispersion. For fiber spectroscopy this is simply the diameter of the input fiber as projected on the sky.

   DM.coverage.spatial.sampling.bin_size (array)

eg. (pixel_scale_x, pixel_scale_y)
Number of Bins (after readout) N/A spatial_bins

The number of spatial bins.

   DM.coverage.spatial.sampling.num_of_bins (array)

eg. (naxis1, naxis2)
Binning Factor (at readout, or via processing) N/A     DM.coverage.spatial.sampling.binning_factor (array)

eg. (1x1, 2x2 etc)
Undersampled vs oversampled;
is Nyquist respected ?		 N/A spatial_Nyquist

The Nyquist ratio (spatial_resolution/spatial_sample). Values less than 2.5 are undersampled.

   DM.coverage.spatial.sampling.nyquist_index

resolution/bin_size

Spatial Sensitivity

Full characterisation via sensitivity maps (TBD). Here just a first order approximation.

  Concept Resource Metadata V0.82 CVO (Pat's) Model Doug's Requirements Suggested
S
e
n
s
i
t
i
v
i
t
y 		Estimate of the saturation level (in flux units) N/A   flux_bandpass.hiValue

Doug's idea is to describe the range of fluxes in the data. Not clear what the refValue should be.

 DM.coverage.spatial.sensitivity.saturation

maybe both in counts and flux ?
Estimate of the limiting flux level (in flux units) for point sources at an assigned signal to noise ratio level Coverage.Depth (jansky)

(incomplete definition though: at which S/N ?)

 flux_SN10

the flux of a (hypothetical) point source with S/N of 10

 flux_bandpass.loValue DM.coverage.spatial.sensitivity.limiting_flux

(limiting_flux, S/N ratio at which lim. flux is computed or measured)
      flux_bandpass.fill_factor  
Photometric Accuracy Uncertainty.Photometric (float)     DM.coverage.spatial.sensitivity.uncertainty (jansky)

Ambient

  Concept Resource Metadata V0.82 CVO (Pat's) Model Doug's Requirements Suggested
A
m
b
i
e
n
t 		Air Mass N/A     DM.coverage.ambient.airmass

(only for ground based)