 |  |
Astron. Astrophys. 347, 194-202 (1999)
3. Results
3.1. Line shapes
The discussion in this section, except in Sect. 3.5, is based on
the spectra obtained towards the map centre position. The
12CO(2 1) line profile has
the highest S/N-ratio in our data (see Fig. 1), and a significantly
higher S/N-ratio than the line profiles presented by Loup et al.
(1990), Trams et al. (1990), and van der Veen et al. (1993). The line
profile is well described by a weakly double-peaked central part (the
Intermediate Velocity Feature, IVF, covering the central
100 )
with extended line wings (the High Velocity Wings, HVWs), and two
narrow (in a relative sense) features at the extremes of the line
wings (the High Velocity Features, HVFs, occasionally referred to as
bHVF and rHVF, where b and r indicate blue- and redshifted emission
with respect to the systemic velocity, respectively). The line profile
is close to symmetric around its centre, but the line wings become
somewhat asymmetric at velocities beyond
80 of
the centre velocity. The bHVF is also narrower
(FWHM 15 )
than the rHVF
(FWHM 20 ).
The 12CO(1 0) line profile
is very similar to the
12CO(2 1) line profile
except that the central part may be more flat-topped. The S/N-ratio of
our 12CO(1 0) spectrum is
comparable to that obtained by van der Veen et al. (1993). The
13CO line profiles, in particular the high S/N-ratio
13CO(2 1) spectrum, clearly
reveal an additional feature; a narrow component at the centre of the
line profile (the Low Velocity Feature, LVF). Furthermore, the
13CO spectra have, in a relative sense, weaker extended
line wings than the 12CO spectra, and the two horns seen in
the central part of the
12CO(2 1) spectrum are not
seen in the 13CO(2 1) line.
When comparing with the spectra of van der Veen et al. (1993) we
confirm the existence of their extreme blue- and redshifted features
(features d and e in their nomenclature) and the inner
peaks (features b and c ), but not the intermediate
high-velocity features f and g , which we believe can be
attributed to the limited S/N-ratio in their spectra. The central
feature a is not prominent enough to be safely identified in
our 12CO spectra, but it has the same characteristics as
the LVF, which is clearly seen in the 13CO spectra.
3.2. Centre velocities
The horns of the IVF part of the
12CO(2 1) line profile
appear at -4 2 and
88 2 ,
i.e., they are symmetrically placed around a velocity of
42 3 .
The bHVF and rHVF have centre velocities of
-92 1 and
173 1 ,
respectively, as determined by fits of Gaussians profiles, i.e., they
are symmetrically placed around a velocity of
41 2 .
The LVF of the 13CO(2 1)
line lies at
41 1
(as estimated by a fit of a Gaussian). Thus, we estimate a systemic
velocity of
41 2
from the CO data (corresponding to a heliocentric velocity of
50 2 ).
Based on this we find that the blue- and redshifted
velocity-integrated intensities of the
12CO(2 1) line profile
differ by 1% and those of the
13CO(2 1) line profile by
2%. That is, the emission is very
symmetric with respect to the systemic velocity, which is also
indicated by the close agreement between the intensity-weighted centre
velocities and the estimated systemic velocity, see Table 2.
3.3. Line widths
Most likely none of the line components in the CO spectra originate
in a symmetrically expanding circumstellar envelope, thus all inferred
expansion velocities are projected velocities. The widths of the LVF
in the 13CO lines correspond to an expansion velocity of
7 . The
full width at half intensity of the
13CO(2 1) IVF is
100 ,
and the two horns in the IVF of the
12CO(2 1) profile are
separated by 92 (i.e., an expansion
velocity of
50 can
be inferred). The HVWs probably extend to about
145 on
each side of the systemic velocity, and the HVFs both expand with a
(projected) velocity of
132 2 .
3.4. Line intensities and optical depths
In order to analyze the emission in more detail we have calculated
the integrated intensity over seven velocity intervals: the LVF, the
IVF regions on both sides of the LVF, the HVWs, and the HVFs, see
Table 3. A measure of the uncertainties in the integrated
intensities are obtained as 30 , where
is the peak-to-peak noise in spectra
with a velocity resolution reduced to
30 . Note that the integrated
intensities in Table 3 do not necessarily come from the
individual features, since there is considerable overlap in velocity
space between these features, and we have made no attempt to separate
them.
![[TABLE]](img37.gif)
Table 3. Line intensities and line intensity ratios in seven velocity ranges
The CO emission is essentially unresolved at a resolution of
20". Thus, the high and nearly
constant
2 1/1 0
12CO intensity ratios, 3-4,
over the entire velocity range is most likely attributed to beam
dilution, assuming that the 1 0 and
2 1 line-emitting gas coincide (a not
unreasonable assumption considering the similarity of the line
profiles), rather than being an effect of low optical depths, in which
case differences in the excitation conditions in the different regions
should show up. The 13CO lines also have a fairly constant
intensity ratio, only marginally higher than the 12CO
ratio, 3-5, indicating relatively high
optical depths also in these lines. There is a clear trend in the
12CO/13CO line intensity ratio in the sense that
the emission in the LVF and IVF has a significantly lower ratio than
the HVW emission; this applies especially to the LVF. The HVFs may
also have a lower ratio (in particular, if one subtracts the line wing
emission in this velocity range). This suggests lower optical depths
in the high velocity emission.
3.5. The 12CO(2 1) brightness distribution
We will here concentrate on the
12CO(2 1) map data since
they have a much higher quality than the
12CO(1 0) map data. We have
produced brightness distribution maps in
20 -intervals centred at -90, -70, ...,
170 , Fig. 2. In each velocity interval
the emission is unresolved with a HPBW=23". The maxima of the
brightness distributions in the central part of the emission, -20 to
100 , lie within a radius of 1" of each
other, Fig. 3 [this is not an effect of the way we centred the maps
with respect to each other; we will in the following discussions
assume that the central star coincides in position with the
(0,0)-position)]. However, for the higher-velocity emission maxima
there is a clear trend: they are displaced with respect to the centre
along a PA 90o, with the
blueshifted emission to the W and the redshifted emission to the E in
a very symmetric way. The separation between the HVFs is
9". Furthermore, the velocities are
clearly (essentially linearly) increasing (in an absolute sense with
respect to the centre velocity) with increasing offset, Fig. 4.
![[FIGURE]](img44.gif) |
Fig. 2. 12CO(J= ) brightness distributions in 20 -intervals centred at -90, -70, ..., 170 . The contours represent 10%, 20%, ..., 90% of the maximum intensity in each panel
|
![[FIGURE]](img52.gif) |
Fig. 3. Positions of the 12CO(J= ) brightness maxima, in 20 -intervals centred at -90, -70, ..., 170 , with respect to the position of the total brightness maximum
|
![[FIGURE]](img60.gif) |
Fig. 4. Offsets of the 12CO(J= ) brightness maxima, in 20 -intervals centred at -90, -70, ..., 170 , with respect to the position of the total brightness maximum [positive (negative) offset for a point lying on the east (west) side of a line with PA=0o]
|
© European Southern Observatory (ESO) 1999
Online publication: June 18, 1999
helpdesk.link@springer.de  |