6. Continuous emission in the L483 protostellar jet
In addition to the emission clumps discussed in the last section, the L483 jet also shows continuous emission that extends along the full length of the jet. The emission lines produced in this process are over an order of magnitude less bright than the emission lines in the clumps. Although weak, emission is visible at all positions along the jet from star to bow shock. This low-level emission in the troughs may be residual emission behind the shocks at the peaks; jet edge entrainment (De Young 1986); emission from the warm, partially ionised jet; or unresolved subknots. In this section we discuss each of these alternatives.
H2 emission in the wings of bow shocks continues for some distance behind the leading shock, but can it be sustained over a large enough range to explain the emission in the troughs? Numerical simulations (Suttner et al. 1997) suggest that the 1-0S(1) line intensity should drop by two orders of magnitude within 1400 AU of each jet shock. The models also show significant emission levels following the leading two shocks only. Yet in L483, low level emission is seen over more than 2100 AU between Peaks 2 and 3 (see Fig. 3a). Hatchell et al. (1999) model the temperature structure of the L483 outflow and conclude that the temperature should drop below 1000 K within 800 AU of the bow shock, at which point the ro-vibrational H2 lines are no longer excited. The excitation analysis shows that temperatures remain above 1000 K for more than twice this distance. However, the trough emission could be cooler than the peaks, as expected for bow shock wings - this is consistent with our excitation analysis (Fig. 5) though not definite.
In summary, the length over which the trough emission is sustained in Trough 2 and the fact that it lies behind the third shocked knot counts against a bow shock wing explanation for the trough emission.
The H2 could be in the jet itself, if certain conditions are satisfied. The jet cannot be fully molecular as the cooling time for molecular hydrogen is only a few years, insufficient to produce ro-vibrational H2 emission along the length of the outflow as the travel time in the jet is for a jet. Fully molecular jets also conflict with the observations at optical wavelengths of atomic protostellar jets. If a small fraction of the jet were molecular then the cooling problem would be solved (as the molecular hydrogen radiates energy gained from the atomic fraction). The H2 column densities correspond to a volume density of 20 cm-3 (assuming a jet diameter of 400 AU), much less than the measured (Bacciotti et al. 1995), suggesting 0.2-2% of the jet is molecular hydrogen if the observed emission originates from the jet. Whether a small fraction of molecular hydrogen could survive over outflow timescales in a predominantly atomic jet remains to be determined.
Thirdly, the H2 could be entrained from the ambient medium in a mixing layer along the edges of the jet, as suggested by De Young (1986). The observed H2 column densities of a few are consistent with the entrainment predictions of Taylor & Raga (1995). Predictions for temperatures in mixing layers range from a few K (Lizano & Giovanardi 1995) to K (Cantó & Raga 1991; Taylor & Raga 1995): the observed temperatures are consistent with the lower end of this range. Entrainment models predict falling temperatures with distance from the star, but the uncertainties on our temperature estimates for the troughs are too large to test this prediction.
Whether or not entrainment is the explanation for excited H2 along the length of the jet, it is clear that ro-vibrational H2 traces only a small fraction of the total mass of the outflow. An upper limit on the directly detected H2 mass can be found by assuming the column density of measured at the bow shock extends over the entire 14400 AU outflow length with a (generous) width of 2100 AU ( at the distance of L483 - see map of Fuller et al. 1995), the total mass traced by H2 is , compared to more than traced by the CO outflow. As hot mixing layers should be well traced by the infrared H2 transitions, it is clear that the total mass entrained in mixing layers is much less than the mass which has been swept up by the leading bow shock and cooled to form the CO outflow.
Finally, the H2 could be from intermediate unresolved knots. This effect is seen in another outflow, HH 111, which shows both infrared and optical emission along the jet. The H2 observations (Coppin et al. 1998) highlight knots at certain positions (labelled F, H, L) with what appears to be continuous emission in between, like the trough emission in L483. However, in the higher resolution optical HST images (Reipurth et al. 1997), the H and [SII] emission breaks up into several subknots which are not distinguished in the H2 observations. The troughs in L483 may similarly consist of emission from unresolved subknots. The measured column densities are beam averaged over pixels. Emission regions smaller than this are beam-diluted and have decreased column density when averaged over the pixel area. A factor of 2-4 in beam filling factor could explain the observed differences in column density between peaks and troughs, though not the differences in temperature (Fig. 5). This is suggestive of lower shock velocities, but similar column densities.
In order to distinguish between these alternatives for the continuous emission along the jet, higher resolution observations across and along the jet would be required. Observations along the jet would need sufficient resolution to resolve individual bow shocks. The observations across the jet would need sufficient resolution to resolve the jet, where we would expect to see emission, if the emission were from in the jet, or edge brightened emission if the emission were from a mixing layer along the jet.
© European Southern Observatory (ESO) 1999
Online publication: July 26, 1999