Sylvie Cabrit 1, S. Bontemps 2, L. Nordh 2, G. Olofsson 2, P. André 3, O. Boulade 3, C. Cesarsky 3, P.O. Lagage 3, M. Sauvage 3, F. Boulanger 4, F. Sibille 5, & R. Siebenmorgen 6,
1 DEMIRM, Observatoire de Paris, 61 Avenue de l'Observatoire, F-75014 Paris
2 Stockholm Observatory, S-133 36 Saltsjobaden, Sweden
3 DSM/DAPNIA/Service d'Astrophysique, C.E. Saclay, F-91191 Gif-sur-Yvette
4 IAS, Universite Paris XI, 91405 Orsay, France
5 Observatoire de Lyon, 69230 St. Genis Laval, France
6 ISO Science Operations Centre, Astrophysics Division of ESA, Villafranca del Castillo, Spain
We present new results from the ISOCAM central program study of bipolar molecular outflows. The data include complete maps of flows in broad-band filters at 6.7 and 15 microns, that locate the total mid-IR radiative loss from the outflow, and maps in narrow band filters centered on several H2 pure rotational lines and on the [Ne II] line, which allow us to characterize the shock physics. On the brightest objects, CVF scans were also obtained to sample the full series of H2 lines from 12.3 down to 5.0 microns and to estimate the hydrogen ortho/para ratio in the shock.
We find that mid-IR emission in bipolar outflows from low luminosity sources is dominated by pure H2 rotational emission from warm gas at a temperature around 500-800 K. The temperature and intensities are consistent with mild MHD shocks of velocities in the range 10-30 km/s and pre-shock densities around 1e3-1e5 cm-3. The ortho/para ratio is found to vary among the flows from 3 (high-temperature value) down to 1.5.
In most of the outflows observed, the emission is highly localized in ``hot spots'' along the flow axis, that coincide spatially with previously known regions of ro-vibrational H2 emission at 2 microns. We present in particular spectro-imaging results on the HH 54 object, on which previous LWS spectra revealed a wealth of molecular lines consistent with an MHD shock. Our CVF data reveal a clear curved shock front with internal excitation gradients, highly suggestive of a bowshock. In the more evolved flow L1551, however, some emission is also observed toward the edges of the flow cavity.
The [Ne II] line was never detected, implying either that the wind/jet density is low (<1e3 cm-3), or that the working surface is very small, or that the wind is propagating into an already moving environment, i.e. the true jet head is located beyond the end of the CO flow. Finally, the kinetic energy flux rate inferred for the detected shocks is compared to that necessary to set the CO outflow in motion. Whenever they are comparable, it favors a scenario where the whole outflow is accelerated in these H2 `hot spots' and then pushed sideways (jet wind); When our measured shock kinetic energy flux is too low, it rather suggests that the flow is accelerated by undetectable low-velocity (< 10 km/s) shocks, possibly associated with the cavity edges (wide-angle wind). In principle, such measurements thus provide independent constraints on the degree of intrinsic wind collimation in young stellar objects driving outflows.