Where do galaxies end? Comparing measurement techniques of hydrodynamic-simulation galaxies' integrated properties

Using the suite of high-resolution zoom re-simulations of individual haloes by Martig et al., and the large-scale simulation MassiveBlack-II, we examine the differences in measured galaxy properties from techniques with various aperture definitions of where galaxies end.We perform techniques popular in the literature and present a new technique of our own, where the aperture radius is based on the baryonic mass profiles of simulated (sub)haloes. For the average Milky Way-mass system, we find the two most popular techniques in the literature return differences of the order of 30 per cent for stellar mass, a factor of 3 for gas mass, 40 per cent for star formation rate, and factors of several for gas accretion and ejection rates. Individual cases can show variations greater than this, with the severity dependent on the concentration of a given system. The average difference in integrated properties for a more general galaxy population is not as striking, but is still significant for stellar and gas mass, especially for optical-limit apertures. The large differences that can occur are problematic for comparing results from various publications. We stress the importance of both defining and justifying a technique choice and discourage using popular apertures that use an exact fraction of the virial radius, due to the unignorable variation in galaxy-to-(sub)halo size. Finally, we note that technique choice does not greatly affect simulated galaxies from lying within the scatter of observed scaling relations, but it can alter the derived best-fitting slope for the Kennicutt-Schmidt relation.