The DiskMass Survey. II. Error Budget

Abstract

We present a performance analysis of the DiskMass Survey. The survey uses collisionless tracers in the form of disk stars to measure the surface-density of spiral disks, to provide an absolute calibration of the stellar mass-to-light ratio, and to yield robust estimates of the dark-matter halo density profile in the inner regions of galaxies. We find a disk inclination range of 25-35 degrees is optimal for our measurements, consistent with our survey design to select nearly face-on galaxies. Uncertainties in disk scale-heights are significant, but can be estimated from radial scale-lengths to 25% now, and more precisely in the future. We detail the spectroscopic analysis used to derive line-of-sight velocity dispersions, precise at low surface-brightness, and accurate in the presence of composite stellar populations. Our methods take full advantage of large-grasp integral-field spectroscopy and an extensive library of observed stars. We show that the baryon-to-total mass fraction (F_b) is not a well-defined observational quantity because it is coupled to the halo mass model. This remains true even when the disk mass is known and spatially-extended rotation curves are available. In contrast, the fraction of the rotation speed supplied by the disk at 2.2 scale lengths (disk maximality) is a robust observational indicator of the baryonic disk contribution to the potential. We construct the error-budget for the key quantities: dynamical disk mass surface-density, disk stellar mass-to-light ratio, and disk maximality (V_disk / V_circular). Random and systematic errors in these quantities for individual galaxies will be ~25%, while survey precision for sample quartiles are reduced to 10%, largely devoid of systematic errors outside of distance uncertainties.