Gravitational lensing provides a great tool for studying dark objects, however degeneracies make it very hard to recover their physical parameters (such as lens mass and distance) from observations. Observing astrometric deviations can bring new information and tackle those degeneracies. As astrometric effects are very subtle, it has long been impossible to detect them; however, ongoing and future space missions have great potential to bring new, exciting advances to this field. The Gaia mission has just reached its third data release and continues to provide precise all-sky astrometry. The future Roman Space Telescope mission will survey the Galactic Bulge in the infrared with unprecedented cadence and precision, casting new light on the dense, largely unexplored regions of the sky.
Each of these datasets will pose particular challenges. Gaia brings a unique scanning pattern and 1D astrometry; its low cadence in the Galaxy makes it difficult to constrain complicated microlensing event lightcurves. Moreover, until Gaia Data Release 4, time-series astrometry will not be available. 5-parameter fits from Gaia DR2 and DR3 can be used as an indicator of astrometric lensing signal, but caution must be taken in interpreting such limited datasets. On the other hand, the cadence of Roman observations of the Galactic Bulge will result in large, computationally demanding datasets.
I will introduce open source codes we are developing that could be helpful for simulating and fitting photometric and astrometric microlensing in the context of those missions. In particular, I will present astromet.py (a package for building and fitting astrometric tracks for single stars, binary systems and lensing events, including a close emulation of the Gaia pipeline) and nested_ulens_parallax (application of the nested sampling algorithm to characterising degenerate posterior distributions of microlensing event parameters). I will show simulations and fits of realistic mock Gaia and Roman data obtained using those codes; in case of Gaia, I will also point to ongoing developments with real data.