4. Kinematics of the HR Car nebula
The long-slit echelle data were used to study the physical structure of HR Car's nebula and to determine the radial velocities of the nebula and its background HII region. The brightest feature in the echelle images of the H and [N II ] lines is the extended background component at nearly constant velocity km s-1 The [N II ] 6583 Å/H ratio of the background component is , which is quite common for Galactic H II regions (Shaver et al. 1983).
The echelle images reveal two kinematic components in the HR Car nebula (Fig. 4). One component has a contiguous expansion structure and corresponds to the circular filaments, or the lobes. The other component consists of knots expanding at slower velocities and corresponds to the knots encompassed by the circular filaments. The expanding lobe component has [N II ] 6583 Å/H ratios of , similar to those in the background H II component, while the knots have [N II ] 6583 Å/H ratios approaching 0.9, much higher than those in the lobes or the background H II. Hutsemékers & Van Drom (1991) obtained a [N II ] 6583 Å/H ratio of 0.43, which may reflect a mixture of emission from lobe and knot components in their aperture.
The two kinematic components of the HR Car nebula are individually discussed below.
4.1. The expanding lobes
We have measured the radial velocities in H and the [N II ] 6583 Å lines, and plotted them against positions along the slit in Fig. 5. The origin of the position-axis is where the slit intersects the polar axis (, as measured from the coronographic image). Offsets are positive to the southwest and negative to the northeast. For reference, a dashed line is drawn at the velocity of the background emission, km s-1.
Comparisons between the continuum-subtracted H image (Fig. 3) and the echelle line images (Fig. 4) indicate that the filamentary structure seen in the H image is connected through a faint expanding shell. The circular filament extending to southeast of HR Car corresponds to the rim of the expanding shell, or lobe. The higher surface brightness of the circular filament is caused by the larger emitting depth along the line of sight. The echelle images also reveal that the bipolar lobes of HR Car are larger than what we measured in our continuum-subtracted H image (Fig. 3).
The echelle slit positions were selected in assumption that the polar axis was along . Fig. 5 shows that the largest expansion velocity along the northern slit position occurs at about southwest of the assumed polar axis, and that along the southern position occurs at about northeast of the assumed polar axis. This position difference indicates that the position angle of the polar axis should be smaller than , supporting the measurement made with our H images.
The velocity structure of the two lobes is typical for a bipolar expansion. The northwestern lobe shows the largest receding velocity at km s-1 and the largest approaching velocity at km s-1, while the southeastern lobe shows the largest approaching velocity at km s-1 and the largest receding velocity at km s-1. If we assume point symmetry with respect to the star, these four extreme velocities can be averaged algebraically to determine the systemic velocity of the nebula, km s-1. The interpretation or modeling of the velocity structure observed in the lobes is complicated by three factors: (1) the lobes are not spherically symmetric, (2) the velocity vectors are not necessarily perpendicular to the lobe surface, and (3) the inclination angle of the polar axis is unknown. Given the limited amount of echelle data, we can conclude that both lobes are expanding, but more observations are needed for further modeling of the expansion. Note that the approaching side of the southeast lobe may have different velocities in the H and [N II ] lines. Unfortunately, our data do not have adequate S/N ratios to analyze this result further.
The systemic velocity of the nebula around HR Car is close to the background H II velocity at -10.6 km s-1, but very different from the LSR-velocity of km s-1 reported by Hutsemékers & Van Drom (1991). Nevertheless our radial velocity is within the range of velocities measured by them for different ions in the stellar spectrum. The kinematic distance implied by our systematic velocity of km s-1 is kpc. Since peculiar velocities can affect the kinematic distance, we will adopt the 5 kpc.
The background HII velocity is very similar to the velocity of the Carina nebula (Courtès et al. 1966; Walborn & Hesser 1975). This may be related to the position of HR Car projected onto a faint, outer arc of the Carina nebula. The complex Na I interstellar absorption line profile as discussed by Hutsemékers & Van Drom (1991) with its components between +9 and -99 km s-1 can then be naturally explained by the motions inside the Carina HII complex (Walborn et al. 1984).
4.2. [N II]-bright knots
Beside the main expansion structure, several [N II ]-bright knots were detected in the echelle data. We have identified 5 knots, 4 along the south slit and 1 along the north slit. These knots are plotted as small squares in Fig. 5. Their velocities (see Table 1) scatter between -75 km s-1 and km s-1, always slower than the superimposed expanding lobe component. These knots can be identified in the H image in Fig. 3. All of these knots are inside or close to the bipolar main structure.
Table 1. Positions and velocities of the [N II ]-bright knots.
Knots #1 - #4 are located in the southeastern lobe. Knots #1 and #2 are projected near the center of the southeastern lobe and they expand slower than the lobe. The different expansion velocities and the different [N II ] 6583 Å/H ratios indicate that these knots could not have the same origin as the expanding lobe. It is not known whether the knots are physically inside the lobes or are only projected within. Knots #3 and #4 follow roughly the expansion of the lobe, but a few km s-1 slower than the lobe. The similarities in expansion velocity and projected position may be suggestive of physical interaction between the knots and the expanding lobe.
Knot #5 is projected within the northwestern lobe. Like Knots #1 and #2, Knot #5 has an intermediate velocity that is slower than the expansion of the lobe. There is possibly another knot on the approaching side of the northwestern lobe. Kinematically it is not very different from the lobe, but the [N II ] 6583 Å/H ratio is higher, .
High [N II ] 6583 Å/H ratios in LBV nebulae or knots are quite common. For example, the [N II ] 6583 Å/H ratio for the S condensation of Car nebula maybe as high as a (Davidson et al. 1982). For AG Car, de Freitas Pacheco et al. (1992) found a value of [N II ] 6583 Å/H .
Similar [N II ]-bright knots have been observed in ring nebulae around Wolf-Rayet stars (e.g., RCW 104, Goudis et al. 1988), but their origin is unknown. These knots are also reminiscent of the FLIERS seen in planetary nebulae, which have been suggested to be high-density, collimated material expelled recently from the nucleus (Balick et al. 1994). A high nitrogen abundance has been derived for [N II ]-bright Wolf-Rayet nebulae (M1-67, Esteban et al. 1991; NGC 6888, Esteban & Vílchez 1992), as well as [N II ]-bright knots in planetary nebulae (NGC 6571, Chu et al. 1991). It is very likely that the [N II ]-bright knots in the HR Car nebula are overabundant in nitrogen. The nitrogen enhancement fits well to the interpretation of HR Car being an LBV which lost processed material via outbursts. The overabundance of nitrogen follows naturally from the ongoing CNO cycle in the star, confirming LBVs to be evolved stars (Maeder 1983; Langer et al. 1994).
© European Southern Observatory (ESO) 1997
Online publication: June 30, 1998