Relative pointing for consecutive images is correct to within 10% of the commanded shift, or 1", whichever is larger.
The driver for this requirement is ? suvernova field dithering ?
Same as for Point-R1
At each position for data taken in Point-R1 (but only for airmass < 1.5), command an offset of xxx arcsec in RA and/or DEC and take another exposure. Compute the astrometric solutions for each exposure and difference, comparing the offset with the commanded shift.
Maximum offset relative to telescope pointing information must agree to within the stated accuracy of the requirement.
(What is the driver for this requirement?)