We present an analysis of the Mg ii \lambda \lambda 2796 , 2803 and Fe ii \lambda \lambda 2586 , 2600 absorption line profiles in individual spectra of 105 galaxies at 0.3 < z < 1.4 .
The galaxies , drawn from redshift surveys of the GOODS fields and the Extended Groth Strip , fully sample the range in star formation rates ( SFRs ) occupied by the star-forming sequence with stellar masses \log M _ { * } / M _ { \odot } \gtrsim 9.5 at 0.3 < z < 0.7 .
Using the Doppler shifts of the Mg ii and Fe ii absorption lines as tracers of cool gas kinematics , we detect large-scale winds in 66 \pm 5 \% of the galaxies .
High-resolution Hubble Space Telescope /Advanced Camera for Surveys imaging and our spectral analysis indicate that the outflow detection rate depends primarily on galaxy orientation : winds are detected in \sim 89 \% of galaxies having inclinations ( i ) < 30 ^ { \circ } ( face-on ) , while the wind detection rate is only \sim 45 \% in objects having i > 50 ^ { \circ } ( edge-on ) .
Combined with the comparatively weak dependence of the wind detection rate on intrinsic galaxy properties ( including SFR surface density ) , this suggests that biconical outflows are ubiquitous in normal , star-forming galaxies at z \sim 0.5 , with over half of the sample having full wind cone opening angles of \sim 100 ^ { \circ } .
We find that the wind velocity is correlated with host galaxy M _ { * } at 3.4 \sigma significance , while the equivalent width ( EW ) of the flow is correlated with host galaxy SFR at 3.5 \sigma significance , suggesting that hosts with higher SFR may launch more material into outflows and/or generate a larger velocity spread for the absorbing clouds .
The large ( > 1 Å ) Mg ii outflow EWs typical of this sample are rare in the context of Mg ii absorption studies along QSO sightlines probing the extended halos of foreground galaxies , implying that this wind material is not often detected at impact parameters > 10 kpc .
Assuming that the gas is launched into dark matter halos with simple , isothermal density profiles , the wind velocities measured for the bulk of the cool material ( \sim 200 - 400 { km s ^ { -1 } } ) are sufficient to enable escape from the halo potentials only for the lowest- M _ { * } systems in the sample .
However , the highest-velocity gas in the outflows typically carries sufficient energy to reach distances of \gtrsim 50 kpc , and may therefore be a viable source of cool material for the massive circumgalactic medium observed around bright galaxies at z \sim 0 .