Instrumental artifacts in photonic lantern spectroastrometry and their mitigation with PLred

Abstract

Spectroastrometry is a powerful spectral-differential technique for probing angular scales below the resolution limit, but it is also well known to be susceptible to instrumental artifacts that can mimic or obscure real signals. In photonic lantern spectroastrometry, recently demonstrated on-sky with Subaru/FIRST-PL, the astrometric signal is encoded in relative flux variations between lantern outputs rather than in centroid shifts along a slit. This changes the artifact landscape: some slit-based spectroastrometric artifacts are avoided, but new artifact mechanisms emerge, including detector nonlinearity and spectral extraction errors, which can produce spurious features on emission or absorption lines. These lessons directly informed the design of PLred, an open source Python package for photonic lantern data reduction and instrument-agnostic spectral-differential image reconstruction. We describe the origin of these artifacts and the key pipeline design choices used to mitigate them.