How Single-Star Guiding affects HST's Pointing Stability

TL;DR

This study evaluates the pointing stability of the Hubble Space Telescope when using a single guide star and a gyro, finding that drift is minimal and manageable for most scientific observations.

Contribution

It provides the first quantitative assessment of HST's pointing accuracy in reduced gyro mode with a single guide star, demonstrating its viability for scientific imaging.

Findings

Drift in 1GS mode is less than 0.2 pixels over a full orbit in most cases.

PSF quality in 1GS mode is comparable to 2GS for exposures under 500 seconds.

Post-processing with a perturbed PSF effectively corrects drift-induced residuals.

Abstract

HST is designed to use two guide stars (GSs) in the fine-guidance sensors (FGSs) to maintain its pointing and tracking during exposures. The primary GS holds the boresight fixed and the secondary GS keeps the orientation fixed. However, HST is also able to track using only a single GS by fixing the boresite on one star and maintaining the orientation using the available gyro(s). We evaluate the pointing quality achieved in this latter case, when one GS and one gyro (RGM, a.k.a., reduced gyro mode) are used. We find that in 1GS-RGM, there is indeed more drift during the course of an orbit than when two guide stars are used, but the drift is much smaller than was seen in previous times of sub-optimal gyro performance. We quantify the 1GS-RGM drift seen in recent archival GO data and in images from a test calibration program and find that (a) for exposures < 500 sec, the PSF quality in 1GS is indistinguishable from that of 2GS and (b) over the course of full orbits (~2500 sec), the drift in four of five cases was less than 0.2 pixel and ~0.4 pix in the other case. Such a drift is marginally detectable in observations, but it should have a marginal impact on most science programs, since the variation in the PSF caused by drift is smaller than the PSF variations with focus and with location on the detector. For observers who wish to correct drift effects, we show that the use of a perturbed PSF during post-acquisition data analysis removes essentially all astrometric residuals, even for drift levels up to ~0.5 pix, as well as most of the photometric residuals.