We use large-scale cosmological simulations to estimate the mass-to-light ratio of galaxy systems as a function of scale , and compare the results with observations of galaxies , groups , clusters , and superclusters of galaxies .
We find remarkably good agreement between observations and simulations .
Specifically , we find that the simulated mass-to-light ratio increases with scale on small scales and flattens to a constant value on large scales , as suggested by observations .
We find that while mass typically follows light on large scales , high overdensity regions — such as rich clusters and superclusters of galaxies — exhibit higher M / L _ { B } values than average , while low density regions exhibit lower M / L _ { B } values ; high density regions are thus antibiased in M / L _ { B } , with mass more strongly concentrated than blue light .
This is true despite the fact that the galaxy mass density is unbiased or positively biased relative to the total mass density in these regions .
The M / L _ { B } antibias is likely due to the relatively old age of the high density regions , where light has declined significantly since their early formation time , especially in the blue band which traces recent star formation .
Comparing the simulated results with observations , we place a powerful constraint on the mass density of the universe ; using , for the first time , the entire observed mass-to-light function , from galaxies to superclusters , we find \Omega = 0.16 \pm 0.05 .