Understanding the physics of low-temperature atmospheres is a key goal of research in the brown dwarf and exoplanet communities. Model-independent distances and temperatures are needed to inform and validate existing theory. Achieving such measurements robustly and with high precision is challenging and has been attempted with a variety of ground- and space-based instrumentation, notably including the recently deactivated Spitzer Space Telescope. Homogeneous, long-duration data sets from Spitzer have been particularly helpful for studies of the coolest Y dwarfs but are often limited by uncorrected systematic astrometry errors. Such errors may be introduced in various ways but commonly arise due to the combination of imperfect knowledge of focal plane distortion and a lack of field stars with precisely known positions in the relevant bandpasses. Measurements using Spitzer/IRAC are further complicated by the under-sampled nature of its point spread function.
We are developing a novel software pipeline for the estimation of positions and proper motions designed to mitigate systematic errors in key ways. Our approach adapts error correction techniques developed for the wide-field transiting exoplanet surveys (such as the Trend Filtering Algorithm and Sys-Rem) and applies them to infrared position measurements. In effect, our method simultaneously examines the residuals of astrometric fits to many objects in the field and removes deviations common to many sources. This minimally parametric technique can effectively compensate for a wide range of systematic errors with only the assumption that highly correlated residuals are unlikely to be astrophysical. Our pipeline further allows for joint fitting of data from multiple instruments. Our solution process iterates between joint fitting and independent residual analysis to separate instrumental effects from true astrophysical signal. The resulting, empirically determined astrometric residuals can be used to find objects exhibiting excess astrometric noise that may indicate the presence of additional components in the system. We apply this pipeline to several known Y dwarfs with long-term Spitzer observations to derive updated physical parameters and illustrate the usefulness of our technique.