The Illumination Function

The aperture plane amplitude image is shown in figure 6. This is the night-time image; the daytime image is very little different.

Figure 6: The aperture plane illumination function, from the June 12 observation

Figure 7: The azimuthally-averaged radial illumination function. The solid line is the function predicted for a feed whose taper is 12 dB at the edge of the antenna (r=50m)

Three factors dominate the amplitude image:

• The feed pattern.

  This should be a smooth, azimuthally symmetric function, falling to low values at the edge of the antenna. Figure 7 shows the radial illumination function computed by averaging the amplitude image around concentric annuli. The predicted function for a feed with a 12 dB taper at the 50m radius is also shown, and is in good agreement.

• The blockage shadows, and their associated Gibbs echoes.

• Phase gradients.

  Our aperture plane distribution is the true distribution convolved with the transform of the sampling window - in effect, the vector average over a patch, roughly 1 $m^{2}$ in size. Phase variations within the patch will therefore show up as amplitude variations, even if though the amplitude (set by the reflectivity of the panel) has not changed. This is very evident in the haloes surrounding the four calibration panels.

  In addition to the panels however, there are the signatures of a number of other phase gradients in the amplitude map - a number of rings which coincide with the panel rings - ring 14 in particular. These are examined in more detail in the surface error maps.