2009ApJ...694..314A

We have used the visible integral-field replicable unit spectrograph prototype, a new integral field spectrograph, to study the spatially and spectrally resolved Lyman-α emission line structure in the radio galaxy B2 0902+34 at z = 3.4. We observe a halo of Lyman-α emission with a velocity dispersion of ~250 km/s extending to a radius of 50 kpc. A second feature is revealed in a spatially resolved region where the line profile shows blueshifted structure. This may be viewed as either H I absorption at ~-450 km/s or secondary emission at ~-900 km/s from the primary peak. B2 0902+34 is also the only high-redshift radio galaxy with a detection of 21 cm absorption. Our new data, in combination with the 21 cm absorption, suggest two important and unexplained discrepancies. First, nowhere in the line profiles of the Lyman-α halo is the 21 cm absorber population evident. Second, the 21 cm absorption redshift is higher than the Lyman-α emission redshift. In an effort to explain these two traits, we have undertaken the first three-dimensional Monte Carlo simulations of resonant scattering in radio galaxies. We have created a simple model with two photoionized cones embedded in a halo of neutral hydrogen. Lyman-α photons propagate from these cones through the optically thick H I halo until reaching the virial radius. Though simple, the model produces the features in the Lyman-α data and predicts the 21 cm properties. To reach agreement between this model and the data, global infall of the H I is strictly necessary. The amount of gas necessary to match the model and data is surprisingly high, ≥10¹² M_☉, an order of magnitude larger than the stellar mass. The collapsing structure and large gas mass lead us to interpret B2 0902+34 as a protogiant elliptical galaxy. This interpretation is a falsifiable alternative to the presence of extended H I shells ejected through feedback events such as starburst superwinds. An understanding of these gas features and a classification of this system's evolutionary state give unique observational evidence of the formation events in massive galaxies.