Assessing the astrometric and photometric performance of Starfinder2 in ELT-METIS simulated images

Abstract

The Mid-infrared ELT Imager and Spectrograph (METIS) will provide diffraction-limited L′-band imaging of crowded stellar fields, where spatial variation in the adaptive optics point-spread function (PSF) complicates astrometry and photometry. We assess StarFinder2 using a 30 min ScopeSim observation of the Galactic Centre generated with a spatially varying library of TIPTOP PSFs and the METIS H2RG detector model. We varied the PSF sampling supplied to StarFinder2 from a single reference PSF to an 11 × 11 grid. Increasing the grid density produced no systematic astrometric gain, whereas a 7 × 7 grid reduced the robust photometric scatter by approximately 60% relative to the single PSF fit. For matched sources, the simulated centroiding scatter reaches approximately 1µas and follows the expected FWHM/SNR trend at intermediate and faint magnitudes. For bright sources near L′ = 11, this fitting term and the remaining astrometric contributions compiled in previous work within our research group give a conditional total of approximately 75µas. Photometric precision is approximately 0.02 mag for 12 ≤ L′ ≤ 17. Saturation limits brighter sources, while structured PSF residuals bias the fluxes and reduce completeness beyond L′ ≃ 20. These results show that a 7 × 7 PSF grid sufficiently samples this simulated field and distinguish the StarFinder2 fitting precision from error terms that must still be calibrated on sky.