Intrinsic galaxy alignments in the KiDS-1000 bright sample: Dependence on colour, luminosity, morphology, and galaxy scale

The intrinsic alignment of galaxies is a major astrophysical contaminant to weak gravitational lensing measurements, and the study of its dependence on galaxy properties helps provide meaningful physical priors that aid cosmological analyses. This work studied for the first time the dependence of intrinsic alignments on galaxy structural parameters. We measured the intrinsic alignments of bright galaxies, selected on apparent r-band magnitude r<20, in the Kilo-Degree Survey (KiDS). Machine-learning-based photometric redshift estimates are available for this galaxy sample that helped us obtain a clean measurement of its intrinsic alignment signal. We supplemented this sample with a catalogue of structural parameters from Sérsic profile fits to the surface-brightness profiles of the galaxies. We split the sample on galaxy intrinsic colour, luminosity, and Sérsic index, and we fitted the non-linear linear alignment model to galaxy position–shape projected correlation function measurements on large scales. We observe a power-law luminosity dependence of the large-scale intrinsic alignment amplitude, A_{IA, for both the red and high-Sérsic-index (ns>2.5) samples, and find no significant difference between the two. We measure an ∼1.5σ lower AIA for red galaxies that also have a Sérsic index of ns<4 compared to the expected amplitude predicted using the sample's luminosity. We also probe the intrinsic alignment of red galaxies as a function of galaxy scale by varying the radial weight employed in the shape measurement. On large scales (above 6 Mpc/h), we do not detect a significant difference in the alignment. On smaller scales, we observe that alignments increase with galaxy scale, with outer galaxy regions showing stronger alignments than inner regions. Finally, for intrinsically blue galaxies, we find AIA=−0.67±1.00, which is consistent with previous works, and we find alignments to be consistent with zero for the low-Sérsic-index (ns<2.5) sample.}