AT M4 aberrations

M4 is not in an image plane, but is close enough to show a very substantial change in the footprint for off-axis stars. This component comes before the STS, and before the laser metrology corner cube, so systematic error terms cannot be eliminated using the normal calibration procedures. The surface of this mirror must be modelled to $nm$ accuracy in order to reach the astrometric accuracy required.

Property	Value
f / ratio	?
Plate scale	$\sim 0.48 \mbox{ mm arcsec}^{-1}$
Surface roughness	?
Beam diameter	?