ec1ad7aeaf0d5baa.tex
1: \begin{abstract}
2:   The interpretation of molecular-line data in terms of hydro
3:   dynamical simulations of planet-disk interactions fosters new hopes
4:   for the indirect detection of protoplanets. In a model-independent
5:   approach, embedded protoplanets should be found at the roots of
6:   abrupt Doppler flips in velocity centroid maps. However, the largest
7:   velocity perturbation known for an unwarped disk, in the disk of HD\,100546,
8:   leads to a conspicuous Doppler flip that coincides with a thick dust
9:   ring, in contradiction with an interpretation in terms of a
10:   $\gtrsim 1 \,M_{\rm jup}$ body.  Here we present new ALMA
11:   observations of the $^{12}$CO(2-1) kinematics in HD\,100546, with a
12:   factor of two finer angular resolutions. We find that the disk
13:   rotation curve is consistent with a central mass
14:   $ 2.1 < M_\star /M_\odot < 2.3$, and that the blue-shifted side of
15:   the Doppler flip is due to vertical motions, reminiscent of the disk
16:   wind proposed previously from blue-shifted SO lines. We tentatively
17:   propose a qualitative interpretation in terms of a surface
18:   disturbance to the Keplerian flow, i.e. a disk eruption, driven by
19:   an embedded outflow launched by a $\sim 10\,M_{\rm earth}$
20:   body. Another interpretation involves a disk-mass-loading hot-spot
21:   at the convergence of an envelope accretion streamer.
22:   
23: \end{abstract}
24: