Astron. Astrophys. 350, 985-996 (1999)

5. Distance - absorption features in the (V-I) - V diagram

The reducing effect of the extinction associated with the cometary tails on the stellar surface density is most obvious in a box defined by the pixel ranges [5002000] for X and [15002000] for Y, see Fig. 1. The color - magnitude diagram for this 585"195" box is shown in Fig. 5. The faint magnitude limit is approximately as for the complete sample but compared to Fig. 2 the blue limitation is shifted to the red by approximately (V-I) = 05 at all V magnitudes from 15 to 21, the (V-I)₀ shift is caused by reddening. Such a common minimum reddening may either be caused by a physical entity or more probably be the accumulated effect of the patchy interstellar medium sampled over large distances. In the CG 31 tail direction the blue envelope is possibly a combination of the two. The dashed curve is a fit by eye of the (V-I)₀ - M_V relation to the blue edge. The shift in V is 149 and that in (V-I)₀ is 10. A marginally inferior fit is obtained with 14.5 and 0.9, indicating the uncertainty 05 and 01 in the (V - M_V) and (V - I)₀ shifts respectively. The translation we perform is in both coordinates and the success depends on two things. First we rely on the structure in the (V-I)₀ - M_V relation, it wiggles a little, in Fig. 5 the maximum amplitude of the bluest wiggle is located at (V-I) = 1.6 after the shift. The second requirement is to nature, if a reddening feature has to be revealed a set of stars having the feature's reddening must be located at the same common distance emphasizing the importance of including the intrinsically faint stars of large spatial density. The existence of such a set means that the distances we derive are not only upper limits, as most often is the case for cloud distances based on color excesses; our resulting distance is the distance to the dust feature, almost like the distance to an open cluster when derived from main sequence fitting. First it is of course an upper distance limit since the absorption is sampled in front of the stars on the shifted main sequence; second it is also a lower distance limit because if the feature's dust was located nearer than indicated by the translation stars with the same (V-I) shift but a smaller V-M_V shift must exist, and as seen on the reddest demarcation on Fig. 8 they do not. The intrinsically reddest stars are spatially so frequent that they are found anywhere within any 20-40 pc distance interval in the local disk, 20-40 pc is the error estimate from the fit to a feature at 200 pc, see Sect. 5.3.

Fig. 5. Color - magnitude diagram for the sight lines with X(pix) in the range and Y(pix) in the interval . This box mainly covers the tail of CG 31, see Fig. 1. The dotted curve is the (V-I)0 - MV relation discussed in the introduction of Sect. 5 shifted to fit the blue envelope. Giants pertaining to the shifted main sequence defined by the blue envelope, and reddened by the same amount, are located in the long dashed box. Giants at the same distance but reddened by twice the reddening of the blue envelope are in the short dashed box. The solid line is a reddening vector corresponding to = 1.0 starting at (V - I, V) = (2.0, 16.0)

Our interpretation of the (V-I)₀ shift is that it is entirely caused by reddening: (V-I). Together with the (V-I)₀ shift the V - M_V translation determines the distance of the feature. The ratio A_I/A_V enters in the distance determination, Rieke & Lebofsky (1985) quotes 0.50 whereas He et al. (1995) finds 0.58, both these values assumes R_V = 3.2. We however adopt the A_I/A_V ratio equal to 0.64 from Thoraval et al. (1997) implying the (V-I)₀ shift of 10 to correspond to A_V = 2.8. The value by Thoraval et al. is preferred since it is derived from a molecular environment where the spatial filling factor may play a role as may be the case for the region we study. Our I band has = 798.2 nm and = 829.2 nm, FWHM = 142.6 nm. We do not expect any troubles from differences to the (V-I)_C quoted in the Hipparcos and Tycho Catalogues, it is a well known fact that broad bands as V and I have sufficient redundancy when transformed to standard systems even if they differ from the standard definition of the band.

For (V - I) 09 V and I photometry alone is not able to distinguish dwarf stars from giants in the general field. Egret et al. (1997) did a study of G3 - M3 giants in the (V-I) - M_V plane. Giants (LC III and brighter) have M_V +1 and (V-I)₀ 1.0, this means that giants in Fig. 5, assuming that they pertain to the shifted main sequence represented by the dotted curve, will have V  16 and (V-I)  2.0, these dividing lines are indicated in Fig. 5 (long dashes). If = 2 we would find the giants with V 188 and (V-I)  3.0 still assuming that the distance indicated by the blue envelope is valid (short dashed lines in Fig. 5). Stars to the bright and red side of the reddening vector connecting (2.0, 16.0) and (3.0, 18.8) in Fig. 5 may thus be giants reddened by more than A_V  3 mag. For the tail region of CG 31 we may conclude that few giants indeed seem be present, see also Fig. 8 and the discussion in Sect. 6. Note we do not claim that the distance of the stars on the blue envelope relates to the tail distance, part of the absorption of course does.

5.1. Imprints of the diffuse dust in the galactic disk

To see the variation of the color magnitude diagram in the surveyed field we present the data from a box with X(pix) ranging [13002100] and Y(pix) in the interval [-700300] from the sub-area with the largest number of stars per unit area and consequently the smallest reddening, Fig. 6. In this sub-area most stars are bluer than the blue confinement of the CG 31 tail box. Our interpretation is that these stars are less reddened and more distant. The dashed - dotted curve of Fig. 6 has the same shift 149 in V but in (V-I)₀ only 06 and again we note how well the shifted main sequence seems to follow the details of the blue envelope of Fig. 6. Our suggested interpretation is that stars along the dotted curve is at a distance of 2.6 kpc (2630 pc) and those along the dashed-dotted curve at 4.4 kpc (4365 pc) and the absorption in V are 2.8 and 1.7. The corresponding distances below the galactic plane are 75 and and 125 pc respectively meaning that the blue envelope stars of Fig. 6 are in the outskirts of the dust layer. For a recent discussion of the scale heights of the absorbing layer see Jonch-Sorensen (1994). With 4-6 diffuse, Knude (1981a, 1981b), clouds per kpc along the line of sight each with A_V = 4.20.027 = 0113 implies A_V(total) = 4.4(0.452-0.678) = 1989-2983, exceeding the measured minimum value A_V 17 by 03-15. For a sight line of length only 2.6 kpc the expected values are 1175 and 1763 for 4 and 6 diffuse clouds respectively both much smaller than the 28 measured. In both cases the parameters valid for the diffuse medium predict absorption values that deviate from the proposed minimum reddening. Obvious reasons for these discrepancies could be that these fields, also that outside the area with the cometary globules, may contain other interstellar structures than the complex of globules. The missing extinction in the area outside the globules may in fact have its origin in the low density interior of the Gum Nebula. See also the discussion in the next section on the distance to the Gum Nebula.

Fig. 6. Color - magnitude diagram for the sight lines with X(pix) in the range and Y(pix) in the interval . This box is mainly located in the less obscured part to the south of the complex of globules, see Fig. 1. The upper dashed curve is again the (V-I) - MV relation discussed in the introduction of Sect. 5 shifted to fit the blue envelope in the tail box, see Fig. 5. The lower dashed - dotted curve has a (V-I)0 shift of 06 but the same shift in V as the dashed one

5.2. Imprint of cometary tails and other ISM features

Our working hypothesis is thus that the bluest envelope of any sub-area may be caused mainly by the diffuse medium, with the addition that if it is shifted more to the red than the blue confinement of the total area some common absorption apart from the diffuse takes place. Such an interpretation seems applicable when we compare Fig. 5 and Fig. 6 where the latter, supposed to be virtually unreddened, has the envelope shifted in (V-I) approximately by 0.5 less than the former.

(V - I) - V features are sought after by pushing around a box in the surveyed area; the box size may be varied depending on the stellar surface density.

In the X(pix) and Y(pix) sub-area in the north western part of Fig. 1 covering a major fraction of the CG 31 tails the color magnitude data also seems to have an "intermediate or internal confinement" to the red side, the (V-I)₀ shift for this is 10 but the V shift only 119 compared to the blue envelopes 149. The upper dashed curve in Fig. 7 shows the location of this proposed envelope. With these shifts in distance and color the stars on the envelope still have A_V = 28 but are located at 660 pc and not at 2.6 kpc as the blue confinement. The possible significance of this distance is discussed in the next section. There are some stars to the bright and red side of the 660 pc envelope, according to their magnitude they are too faint to be giants pertaining to the (D, A_V) = (660, 2.8) feature, rather they are dwarfs nearer than 0.6 kpc and with an absorption larger than A_V = 28.

Fig. 7. Color - magnitude diagram for the sight lines with X(pix) in the range and Y(pix) in the interval , same sub-area as in Fig. 5. The upper long dashed curve is the (V-I)0 - MV relation discussed in Sect. 5.2 shifted to fit an "intermediate or internal" red envelope to the main sequence formed by the stars in the tail box, see Fig. 5. It is at 650 pc and is suggested pertaining to the forefront of the Gum Nebula. The lower dotted curve is the same as in Fig. 5

In order to have slightly more stars for discussion and to search for larger absorptions we have opened up the tail box to X(pix): and Y(pix): . Our original blue envelope for the tail still fits the enlarged data set reasonably well, see Fig. 8. For this expanded tail box a new red envelope appears. Its shift parameters result in A_V = 36 at a distance of a mere 206 pc. The 206 pc feature is shown as the dash - dot curve in Fig. 8. The stars brighter and redder than the reddest envelope may be giants, but very nearby red dwarfs can not be completely excluded. The reddest demarcation of the dwarfs pertains to the nearest and most reddened stars and because it is the maximum absorption measured by the main sequence fitting we suggest that the A_V = 36 feature originates in the tails of the cometary globules, mainly in that of CG 31A.

Fig. 8. Color magnitude diagram for the sight lines with X(pix) in the range and Y(pix) in the interval . The rightmost dashed - dotted curve is the (V-I)0 - MV relation discussed in Sect. 4 shifted to fit a partial red envelope, the shift corresponds to a distance 206 pc and AV = 36. The lower dotted curve is the same as in Fig. 5. The central curve has AV = 05 and are at 400 pc and may delineate stars at a distance similar to the Vela OB2 association

5.3. Interpretation

In the sub-area covering part of CG 31's tail we may have identified two reddening structures with shifts V-M_V, (V-I) equal to (11.9, 1.0) and (10.2, 1.3) respectively. In order to translate these shifts into absorption A_V and distance of a dust feature the ratio must be known. As mentioned we have chosen to use 0.64 for this ratio following the discussion by Thoraval et al. (1997), the range they propose is from 0.50 to 0.70. This range introduces an uncertainty 5. A ratio 0.64 also result from a linear regression analysis on the few stars forming the bright and red confinement to Fig. 9. If we conservatively estimate the uncertainty of fitting the (V-I)₀ - M_V relation to 0.5-1.0 mag in V-M_V and 0.2 in (V-I) implying that from the fit alone become 10 - 20.

Fig. 9. As Fig. 2. But the lower dotted line indicates where the red clump giants will be located when V-MV = 149 and (V-I) = 10. The upper red confinement is the reddening vector for red clump stars at 1 kpc displaying AV from 00 to AV 62. The lower dashed line is the reddening locus for red clump giants removed 9.3 kpc. With RGC = 8.5 kpc and Rgalaxy = 14.5 kpc the clump giants along the dashed line are at the edge of the galactic disk

With the ratio 0.64 the shift (11.9, 1.0) represents an A_V = 28 feature at 660 pc and the (10.2, 1.3) shift a nearby structure with A_V = 36 at only 200 pc. After Sahu & Blaauw (1993) the 660 pc feature might be part of the near side of the Gum Nebula. The deduced value for A_V may not be representative of the absorption in the Gum shell alone but should be understood as the accumulated absorption along the line of sight and being deduced from a red envelope it represents a maximum absorption; the last contribution to this accumulation along the sight lines comes from the Gum shell itself but we can not estimate how much that is. The 200 pc feature with A_V = 36 is so near and the feature is most pronounced in the CG 31 tail region that it must be related to the cometary globules. It is hard to imagine an absorption 3^m present in front of the globules but without leaving a trace in the area outside the globules (see Fig. 6). We suggest accordingly that CG 30 and CG 31 is at 20040 pc with a somewhat conservative error estimate. With (V-M_V) = 0.5 and [(V-I)] = 0.1 the distance is accurate to 20 pc, essentially the 10 Hipparcos precision for stars within 50 pc demonstrated in Fig. 3c.

Since we identify the absorption features from the extreme confinements of the color - magnitude diagrams one could think that the envelopes were a result of the error propagation, several stars will be scattered into the wings of the error distribution even with our small errors but large number of stars. If the envelopes are due to the error distribution the stars forming them should have large color errors, but this is not found to be the case. We may on the other hand identify small sub-areas where we exclusively observe stars on the blue (14.9, 1.0) envelope. The 300 by 300 (pix²) box centered on (X, Y) = (2150, 2150) in the north western outskirts of the tail region is such an example and it is improbable that all stars in a small fraction of a CCD frame should have the same large error. With the DFOSC pixel scale this corresponds to 2´2´ area. The rectangle X: [-7001000], Y: [+7001100] almost discrepant with the box displayed in Fig. 8 display a relatively more populated (10.2, 1.3) feature than Fig. 8.

So far we have not discussed any shift indicating the presence of an interstellar feature located at 400 pc, the previously assumed CG 30 distance identical to the distance to what have been labelled the Vela sheet, Henning & Launhart (1998). A recent determination of the Vela OB2 distance is 410 pc, de Zeeuw et al. (1999). If we use the width of the main sequence valid for Vela OB2 members as a reddening indicator, Fig. 12 of de Zeeuw et al. (1999), we estimate A_V 0.4 - 0.5 see Fig. 8 central curve.

© European Southern Observatory (ESO) 1999

Online publication: October 14, 1999
helpdesk.link@springer.de