This will be measured on precursor data (almost certainly from HSC, via DM-2380) using the following procedure:

• Run visit-level processing (i.e. ProcessCcdTask, probably via the new HSC driver ported on DM-3368).
• Run relative astrometric calibration (i.e. meas_mosaic, DM-2674).
• Select bright stars and match across visits (new scripts, mostly delegating to existing code). Just selecting the stars used for PSF determination would probably work.
• Generate pairs of objects with separation near the target (5' for this issue, 20' or 200' for DM-3064 and DM-3071). Bin widths will have to be determined, as that's not included in the requirement specification. Will require new code.
• Plot separation deltas (difference from per-pair mean separation) vs. magnitude; measure RMS and outliers in magnitude bins.

This requirement is specified only for r and i band.

The same procedure will be used for the AM1 (DLP-310), AM2 (DLP-311, DM-3064), and AM3 (DLP-312, DM-3071) measurements for this cycle.