Sub-arcsecond [Fe II] spectro-imaging of the DG Tauri jet Periodic bubbles and a dusty disk wind?

Context. The origin of protostellar jets as well as their impact on the regulation of angular momentum and the inner disk physics are still crucial open questions in star formation. Aims. We aim to test the different proposed ejection processes in T Tauri stars through high-angular resolution observations of forbidden-line emission from the inner DG Tauri microjet. Methods. We present spectro-imaging observations of the DG Tauri jet obtained with SINFONI/VLT in the lines of [Fe II]lambda 1.64 mu m, 1.53 mu m with 0 ''.15 angular resolution and R = 3000 spectral resolution. We analyze the morphology and kinematics, derive electronic densities and mass-flux rates and discuss the implications for proposed jet launching models. Results. (1) We observe an onion-like velocity structure in [Fe II] in the blueshifted jet, similar to that observed in optical lines. High-velocity(HV) gas at similar or equal to-200 km s(-1) is collimated inside a half-opening angle of 4 degrees and medium-velocity (MV) gas at similar or equal to-100 km s(-1) in a cone with an half-opening angle 14 degrees. (2) Two new axial jet knots are detected in the blue jet, as well as a more distant bubble with corresponding counter-bubble. The periodic knot ejection timescale is revised downward to 2.5 yrs. (3) The redshifted jet is detected only beyond 0 ''.7 from the star, yielding revised constraints on the disk surface density. (4) From comparison to [O I] data we infer iron depletion of a factor 3 at high velocities and a factor 10 at speeds below -100 km s(-1). (5) The mass-fluxes in each of the medium and high-velocity components of the blueshifted lobe are similar or equal to 1.6 +/- 0.8x10(-8) M(circle dot) yr(-1), representing 0.02-0.2 of the disk accretion rate. Conclusions. The medium-velocity conical [Fe II] flow in the DG Tau jet is too fast and too narrow to trace photo-evaporated matter from the disk atmosphere. Both its kinematics and collimation cannot be reproduced by the X-wind, nor can the conical magnetospheric wind. The level of Fe gas phase depletion in the DG Tau medium-velocity component also rules out a stellar wind and a cocoon ejected sideways from the high-velocity beam. A quasi-steady centrifugal MHD disk wind ejected over 0.25-1.5 AU and/or episodic magnetic tower cavities launched from the disk appear as the most plausible origins for the [Fe II] medium velocity component in the DG Tau jet. The same disk wind model can also account for the properties of the high-velocity [Fe II] flow, although alternative origins in magnetospheric and/or stellar winds cannot be excluded for this component.