Galaxy formation and evolution

Abstract:

We report high quality near-infrared spectroscopy of 12 ultra-luminous infrared galaxy mergers (ULIRGs). Our new VLT and Keck data provide ~0.5" resolution, stellar and gas kinematics of these galaxies most of which are compact systems in the last merger stages. We confirm that ULIRG mergers are 'ellipticals-in-formation'. Random motions dominate their stellar dynamics, but significant rotation is common. Gas and stellar dynamics are decoupled in most systems. ULIRGs fall on or near the fundamental plane of hot stellar systems, and especially on its less evolution sensitive, r(eff)-sigma projection. The ULIRG velocity dispersion distribution, their location in the fundamental plane and their distribution of v(rot)*sin(i)/sigma closely resemble those of intermediate mass (~L*), elliptical galaxies with moderate rotation. As a group ULIRGs do not resemble giant ellipticals with large cores and little rotation. Our results are in good agreement with other recent studies indicating that disky ellipticals with compact cores or cusps can form through dissipative mergers of gas rich, disk galaxies while giant ellipticals with large cores have a different formation history.

Abstract:

Bottom-up hierarchical formation of dark matter haloes is not as monotonic as implicitly assumed in the Press-Schechter formalism: matter can be ejected into tidal tails, shells or low density `atmospheres'. The implications that the possible truncation of star formation in these tidal `debris' may have for observational galaxy statistics are examined here using the ArFus N-body plus semi-analytical galaxy modelling software. Upper and lower bounds on stellar losses implied by a given set of N-body simulation output data can be investigated by choice of the merging/identity criterion of haloes between successive N-body simulation output times. A median merging/identity criterion is defined and used to deduce an upper estimate of possible star formation and stellar population losses. A largest successor merging/identity criterion is defined to deduce an estimate which minimises stellar losses. In the N-body simulations studied, the debris losses are short range in length and temporary; maximum loss is around 16%. The induced losses for star formation and luminosity functions are strongest (losses of 10%-30%) for low luminosity galaxies and at intermediate redshifts (1 < z < 3). This upper bound on likely losses is smaller than present observational uncertainties. Hence, Press-Schechter based galaxy formation models are approximately valid despite ignoring loss of debris, provided that dwarf galaxy statistics are not under study.