The MultiDrizzle Handbook
3.6 Astrometric Accuracy After Drizzling
Drizzle has been found to impart no additional astrometric error on the images beyond our limitations on the ability to centroid on images which contain power that is not fully Nyquist sampled even when using pixels which are half the original size. Tests have been carried out to characterize the astrometric accuracy of drizzle (Fruchter and Hook 1997). The stellar images described in the previous section (Section 3.5) were again drizzled using the shifts obtained in the HDF WF2 F814W images. Both uniform weight files and cosmic ray masks were used. The positions of the drizzled stellar images were then determined with the IRAF task "imexam", which locates the centroid using the marginal statistics of a box about the star. A box size equal to 6 output pixels, or slightly larger than twice the full-width at half-maximum of the stellar images, was used. A root mean square scatter of the stellar positions of ~0.018 input pixels about the true position was found for the images created both with uniform weight files, and with the cosmic ray masks. However, an identical scatter was produced when the original four-times oversampled images were down-sampled to the two-times oversampled scale of the test images. Thus it appears that no additional measurable astrometric error has been introduced by drizzle. Rather, we are simply observing the limitations of our ability to centroid on images which contain power that is not fully Nyquist sampled even when using pixels half the original size.
For more detailed information on astrometry, see Section 4.3.3 on Astrometric Header Information in Chapter 4.
Space Telescope Science Institute
Voice: (410) 338-1082