Improving ground-based astrometry via turbulence reduction

by Daniel Gomes (University of Pennsylvania)

Thursday 22 Jan 2026, 15:00 → 16:00 US/Eastern

Small Seminar Room (Building 510)

The current barrier to improving ground-based astrometric accuracy beyond the ~10-20mas level is the stochastic displacement of sources due to atmospheric turbulence. These displacements are spatially correlated on scales of a few arcminutes within a single exposure, and therefore can be modeled given a sufficiently dense grid of reference stars from Gaia. I will present recent results I achieved with this method, in which ~90% of the variance from correlated astrometric errors is removed from DECam exposures. Similar performance is expected for LSST, as shown by testing on simulated displacement fields with turbulence parameters inferred from observations near the Rubin site. This effectively lowers LSST’s nominal astrometric floor from ~10 mas to ~3 mas. I will discuss the method, how it differs from two previous approaches to this problem, and the scientific applications that could benefit from this level of improvement in astrometry.