Requirements on SKA Low beam model accuracy - further work needed

SKA1-LFAA-170 "Aperture array DDE" reads:

The LFAA shall have direction dependent models for the beams of each station (or sub-station) with an accuracy of 1% (TBC) of peak gain in the beam center in amplitude and 0.01 radians (TBC) in phase at the half-power points to be used for calibration and imaging.

This is derived from L1 requirement SKA1-SYS_REQ-2724 "Aperture array DDE" which reads:

The SKA1_Low shall have a direction dependent model for the aperture array primary beam with an accuracy of 35dB at the half-power points to be used in calibration and imaging.

Further work is needed here in a few respects:

• The TBCs in this SKA1-LFAA-170 need to be resolved.
• It has been suggested that SKA1-LFAA-170 should also specify what the model looks like. This is an important (implementation) detail as it allows the possibility to have a parameterised model whose parameters need to be determined with array level (self-)calibration. As shown using simulations (see section 8.1 of SKA1 Low Station Calibration Report SKA-TEL-SKO-0001088), the first order differences between the nominal station beam (calculated by assuming an ideal system) and the actual station beam can be described by a pointing offset and beam elongation. Those effects can be captured with just a few parameters. Further simulations are also needed to assess whether such a first order model will reach the desired accuracy.
• The ‘accuracy of 35dB’ in parent requirement SKA1-SYS_REQ-2724 needs to be disambiguated.
• It needs to be considered and recorded how both of these requirements will be verified in practice, bearing in mind, for example, that the measurement uncertainty of the recent UAV measurements as part of AAVS2.0 is higher than the accuracy mentioned in SKA1-
  LFAA-170. Another approach would be to try get an idea of possible errors in the EEPs themselves (see section 5 of SKA1 Low Station Calibration Report).