High-resolution near-infrared observations of Herbig-Haro flows - II. Echelle spectroscopy

We use line profile data to probe the physical conditions associated with molecular hydrogen features in four Herbig-Haro (HH) flows: HH 7-11, 33/40, 26 and 212. We compare these kinematic data with new H2 images and proper motion measurements presented in a companion paper (Paper I) by Chrysostomou et al. We find in these combined data evidence for bow shocks and turbulent mixing layers. HH 7 and 33 represent spectacular examples of resolved bow shocks; double-peaked H2 emission profiles are observed in the flanks of both targets. HH 26C is also thought to be a bow shock, although one that has undergone considerable fragmentation during its lifetime [based on its current proper motion (from Paper I), this bow has a dynamical age of roughly 600 (±130) yr]. HH 40 and 26A instead seem to represent turbulent boundary layers between the HH flows and their ambient surroundings; both features have very low proper motions. However, although the H2 profiles in HH 40 are narrow and symmetric, as one might expect from a turbulent spectrum of unresolved shocks, in HH 26A we see complex structure in position-velocity space. This suggests that shocks generated in the HH 26A turbulent boundary are resolved in these data. The associated curved shocks thus generate more asymmetric profiles and, in some cases, double-peaked profiles. In HH 212 we see narrow, symmetric profiles in the knots along the flow axis, as well as clear evidence of acceleration along the jet. The spatial symmetry evident in images of this bipolar jet is also reflected in the velocities of the knots. The bow shocks NB1/2 and SB1 in HH 212 also possess bow-shock-like profiles in our position-velocity plots. Transverse velocity gradients in the knots provide some evidence for jet rotation. The rotation is consistent with the necessary extraction of angular momentum from the underlying rotating disc to enable the continued accretion of material on to the protostar. Lastly, we return to the issue of whether H2 shock features accelerate molecular gas to form the massive bipolar outflows usually traced in CO. Comparison of the mass fluxes measured in each HH object (from our H2 data) with mass outflow rates typical of 'CO' outflows suggests that this is indeed the case.