Dynamics of Lyman Break Galaxies and Their Host Halos

Abstract

We present deep two-dimensional spectra of 22 candidate and confirmed Lyman break galaxies (LBGs) at redshifts 2<z<4 in the Hubble Deep Field (HDF) obtained at the Keck II telescope. The targets were preferentially selected with spatial extent and/or multiple knot morphologies, and we used slitmasks and individual slits tilted to optimize measurement of any spatially resolved kinematics. The median target magnitude was I_814=25.3, and total exposure times ranged from 10 to 50 ks. We measure redshifts, some new, ranging from z=0.2072 to z=4.056, including two interlopers at z<1, and resulting in a sample of 14 LBGs with a median redshift z=2.424. The morphologies and kinematics of the close pairs and multiple knot sources in our sample are generally inconsistent with galaxy formation scenarios postulating that LBGs occur only at the bottom of the potential wells of massive host halos; rather, they support ``collisional starburst'' models with significant major merger rates and a broad halo occupation distribution. For 13 LBGs with possible kinematic signatures, we estimate simple dynamical masses ranging from 4e9 h^-1 M_sun to 1.1e11 h^-1 M_sun for individual galaxies and from <10e10 h^-1 to ~10^14 h^-1 M_sun with a median value 1e13 h^-1 M_sun for host dark matter halos. Comparison with a recent numerical galaxy formation model implies that the pairwise velocities might not reflect true dynamical masses. We compare our dynamical mass estimates directly to stellar masses and find no evidence for a strong correlation. The diversity of morphologies and dynamics implies that LBGs represent a broad range of galaxy or proto-galaxy types in a variety of evolutionary or merger stages rather than a uniform class with a narrow range of mass.