Are strong Mg II absorbers the signature of outflows?

BackgroundGas, and in particular, cold gas is seen locally in many different locations (disks, halos, merger tails, and even in outflows). Which one is traced by QSO absorption lines, and in particular by the Mg II metal line? Given that QSO absorption lines probe galaxy halos out to 100 kpc and have no bias on the host galaxy luminosity, they are crucial to our understanding of galaxy evolution in general and feedback in particular. Strong Mg II absorbers with equivalent width Wr2796>1 Å are in fact made of several (up to 20) sub-components or clouds. As a result, W_r is a measure of the line-of-sight velocity dispersion (Δ v) of the clouds. Most models for QSO absorbers, except the outflow scenario, predict that the host-galaxy mass and the velocity dispersion Δ v should correlate (by the Virial theorem). GoalIf one could measure the mass for QSO absorbers, one could test whether mass and Δ v are correlated. ResultUsing a statistical technique that does not require any spectroscopic follow-up, we find that the halo-mass M_h of the host-galaxy is anti-correlated with the equivalent width W_{Δ v}. This shows that the Mg II clouds are not virialized in the gaseous halos of the host-galaxies. If Mg II absorbers were virialized in galaxy halos, a positive mass-velocity correlation would have been observed. In addition, the kinematics of super-winds are most consistent with our M_h- W_r anti-correlation.