Mapping water in protostellar outflows with Herschel

Abstract

Context. Water is a key probe of shocks and outflows from young stars because it is extremely sensitive to both the physical conditions associated with the interaction of supersonic outflows with the ambient medium and the chemical processes at play. Aims. Our goal is to investigate the spatial and velocity distribution of H2O along outflows, its relationship with other tracers, and its abundance variations. In particular, this study focuses on the outflow driven by the low-mass protostar L1448-C, which previous observations have shown to be one of the brightest H2O emitters among the class 0 outflows. Methods. To this end, maps of the o-H2O 110–101 and 212–101 transitions taken with the Herschel-HIFI and PACS instruments, respectively, are presented. For comparison, complementary maps of the CO(3–2) and SiO(8–7) transitions, obtained at the JCMT, and the H2 S(0) and S(1) transitions, taken from the literature, were used as well. Physical conditions and H2O column densities were inferred using large velocity gradient radiative transfer calculations. Results. The water distribution appears to be clumpy, with individual peaks corresponding to shock spots along the outflow. The bulk of the 557 GHz line is confined to radial velocities in the range ±10–50 km s-1, but extended emission at extreme velocities (up to vr ~ 80 km s-1) is detected and is associated with the L1448-C extreme high-velocity (EHV) jet. The H2O 110–101/CO(3–2) ratio shows strong variations as a function of velocity that likely reflect different and changing physical conditions in the gas that is responsible for the emissions from the two species. In the EHV jet, a low H2O/SiO abundance ratio is inferred, which could indicate molecular formation from dust-free gas directly ejected from the proto-stellar wind. The ratio between the two observed H2O lines and the comparison with H2 indicate averaged Tkin and n(H2) values of ~300–500 K and 5 × 106 cm-3, respectively, while a water abundance with respect to H2 of about 0.5–1 × 10-6 along the outflow is estimated, in agreement with results found by previous studies. The fairly constant conditions found all along the outflow imply that evolutionary effects on the timescales of outflow propagation do not play a major role in the H2O chemistry. Conclusions. The results of our analysis show that the bulk of the observed H2O lines comes from post-shocked regions where the gas, after being heated to high temperatures, has already been cooled down to a few hundred K. The relatively low derived abundances, however, call for some mechanism that diminishes the H2O gas in the post-shock region. Among the possible scenarios, we favor H2O photodissociation, which requires the superposition of a low-velocity nondissociative shock with a fast dissociative shock able to produce a far-ultraviolet field of sufficient strength.