SKA Configuration studies

1
SKA Configuration studies

• Steven Tingay
• Swinburne University of Technology
• (Chair SKA Simulations Working Group)
• SKA wide-field imaging workshop, June 2005
• ASTRON

2
Outline

• Configuration simulations guidelines for SKA
  site proposal process
• A real-life realisation of these guidelines for
  an Australasian SKA
• Some notes on implications of such an array for
  wide-field imaging.

3
Configuration simulations guidelines (for site
proposals), generated by the SKA Configuration
Simulations Task Force (CSTF)

• Large-N Small-D (LNSD) 125 stations
• Small-N Large-D (SNLD) 62 stations
• One configuration per concept.
• Core
• LNSD diameter 1.4 km (25 stations 20 of
  collecting area)
• SNLD diameter 1.7 km (12 stations 20 of
  collecting area)
• Mandatory arrangement of stations in core (all
  proposers must use these Core configurations)
• Central area
• Inner diameter 1 km, outer diameter 5 km
• LNSD 40 stations 32 of collecting area
• SNLD 20 stations 32 of collecting area
• Mandatory arrangement of stations in central area
  5-arm symmetric logarithmic spiral
  configuration (all proposers must use these
  Central area configurations)
• Beyond the Central area
• Stations populate to at least 3000 km from the
  Core
• Logarithmic spiral configuration

4
LNSD Asymmetric array
SNLD Asymmetric array
5
LNSD symmetric array
LNSD asymmetric array
6

• Sky visibility
• Stations have 30o elevation limit
• Any given RA/DEC is visible if observable by
  all stations in the array for a period of 4 hrs
  in 24hr
• Some relief from this definition for stations
  beyond 3000 km
• Visibility of sources of key scientific interest
• (a) Nearby dark clouds Taurus,
  Ophiucus, Chamaeleon
• (b) 70 coverage of the inner Galaxy,
  including the Galactic Centre
• (c) 20,000 square degrees of sky out of
  the Galactic Plane including the coldest
  regions of sky

7

• (u,v) coverage analysis
• Uniformity of coverage
• Circularity of coverage
• Ellipticity of beam

Configuration simulations guidelines Available
from the International SKA Project
Office Configuration software
http//astronomy.swin.edu.au/ska/SSWG/CSTF-scripts
8
Implications for wide-field imaging An
Australasian (Australia New Zealand) example
9
Fractional bandwidth 25 Integration time 1
minute Source declination -30 deg (Array core _at_
-27 deg) Source at transit (Array core _at_ 117
deg. east) Assume that usable FoV is
limited by temporal and spectral resolution.
Assume that non-coplanar effects are
corrected using suitable algorithms e.g.
W-projection
10
Each point contributing to this histogram
represents 7.5 MHz of bandwidth and 1 minute of
integration time.
11

• What sort of FoV/resolutions are required?
• Single beam field of view at 1 GHz 6 arcmin
  (200 m station diameter)
• 100 beams gives moderate total FoV of deg.
• Many more beams must be produced by inner array
  stations for 200 sq. deg. fast survey capability
• Maximum resolution at 1 GHz 20 mas (3000 km
  baseline)
• Frequency and time averaging considerations?
• No averaging correlator output determined on
  all baselines by the longest baselines (typical
  case currently)
• Just averaging in time correlator integration
  time is dependant on baseline length (i.e. Space
  VLBI observations with the VLBA correlator)
• Averaging in both time and frequency frequency
  averaging trickier e.g. probably not
  possible/desirable for spectral line
  observations.

12

• Assume time and frequency averaging to give 1
• degradation at FWHM points of station beam.
• Single beam
• All Stokes products
• 32 bit floats
• 25 bandwidth
• 1 minute observation

13

• Correlated data volumes, estimated based on
• Number of beams required (total FoV)
• Fractional bandwidth
• Total integration time
• Maximum resolution (driven by confusion limit,
  science)
• Average time/freq 1min, 1beam 21 GB
• 1 hr, 100beams 130 TB
• No averaging 1min, 1beam 296 GB
• 1 hr, 100 beams 1800 TB
• Average just time 1min, 1beam 41 GB
• 1 hr, 100 beams 250 TB

14
Summary

• A set of guidelines for configuration simulations
  exists as part of the site proposal process
  (contact the ISPO for further information)
• A real-life example of these guidelines, as
  applied to the proposed Australian SKA site has
  been presented (used as part of the Australian
  site selection process)
• Many factors to consider when looking at the
  implications of wide-field imaging for SKA
  configurations big consideration is the total
  data volume generated and implications for
  post-correlation analysis
• Number of beams (total FoV) as a function of
  baseline length
• Maximum baseline length confusion limit at any
  given frequency
• Bandwidth and observation time
• Science requirements continuum/spectral line,
  spatial scales required.
• Telescopes like the xNTD and KAT will confront
  problems approaching this order of magnitude
  soon