Characterizing simulated galaxy stellar mass histories [GA]

http://arxiv.org/abs/1406.2967


Galaxy formation simulations can now predict many galaxy properties and their evolution through time. To go beyond studying average stellar mass history properties, we classified ensembles of simulated stellar mass histories, holding fixed their z=0 stellar mass. We applied principal component analysis (PCA) to stellar mass histories from the dark matter plus semi-analytic Millennium simulation and the hydrodynamical OverWhelmingly Large Simulations (OWLS) project, finding that a large fraction of the total scatter around the average stellar mass history for each sample is due to only one PCA fluctuation. This fluctuation differs between some different models sharing the same z=0 stellar mass and between lower (<=3e10 M_o) and higher final stellar mass Millennium samples. We correlated the PCA characterization with several $z=0$ galaxy observables (in principle observable in a survey) and galaxy halo history properties. We also explored separating galaxy stellar mass histories into classes, using the largest PCA contribution, k-means clustering, and simple Gaussian mixture models. For three component models, these different methods often had significant overlap. We provide several classification quantities for the Millennium and OWLS stellar mass histories, which can be compared with other simulation histories as well. These history classification methods provide a succinct and often quick way to characterize changes in histories of a simulated population as physical assumptions are varied, to compare histories of different simulated populations to each other, and to assess the relation of simulated histories to fixed time observations.

Read this paper on arXiv…

J. Cohn and F. Voort
Thu, 12 Jun 14
39/50

Comments: 14 pages, 15 figures. Comments welcomed