2006ApJ...639..766P

We present the results of two complementary investigations into the nature of strong (rest equivalent width, W_r>1.0 Å) Mg II absorption systems at high redshift. The first line of questioning examines the complete Sloan Digital Sky Survey Data Release 3 set of quasar spectra to determine the evolution of the incidence of strong Mg II absorption. This search resulted in 7421 confirmed Mg II systems of W_r>1.0 Å, yielding a >95% complete statistical sample of 4835 absorbers (systems detected in S/N>7 spectral regions) spanning a redshift range 0.35<z<2.3. The redshift evolution of the comoving line-of-sight number density, ℓ_Mg(X), is characterized by a roughly constant value at z>0.8, indicating that the product of the number density and gas cross section of halos hosting strong Mg II is unevolving at these redshifts. In contrast, one observes a decline in ℓ_Mg(X) at z<0.8, which we interpret as a decrease in the gas cross section to strong Mg II absorption and therefore a decline in the physical processes relevant to strong Mg II absorption. Perhaps uncoincidentally, this evolution roughly tracks the global evolution of the star formation rate density. Dividing the systems in W_rsubsamples, the ℓ_Mg(X) curves show similar shape with lower normalization at higher W_rvalues and a more pronounced decrease in ℓ_Mg(X) at z<0.8 for larger W_rsystems. We also present the results of a search for strong Mg II absorption in a set of 91 high-resolution quasar spectra collected on the ESI and HIRES spectrographs. These data allow us to investigate the kinematics of such systems at 0.8<z<2.7. In this search a total of 22 systems of W_r>1.0 Å were discovered. These systems are characterized by the presence of numerous components spread over an average velocity width of Δv~200 km/s. Also, absorption due to more highly ionized species (e.g., Al III, C IV, Si IV) tends to display kinematic profiles similar to the corresponding Mg II and Fe II absorption. We consider all of these results in light of two competing theories previously introduced to explain strong Mg II absorption: post-starburst, supernova-driven galactic winds and accreting gas in the halos of massive galaxies. The latter model is especially disfavored by the absence of evolution in ℓ_Mg(X) at z>1. We argue that the strong Mg II phenomenon primarily arises from feedback processes in relatively low mass galactic halos related to star formation.