Astron. Astrophys. 362, 715-722 (2000)

3. OB associations and the high-latitude clouds

If the Local bubble is produced by the action of supernovae and stellar winds from massive stars and the high-latitude clouds are formed as condensations in the swept-up interstellar gas (e.g. Breitschwerdt et al. 1996; Elmegreen 1988), then it would be natural to look for any OB associations in the region of the system of high-latitude clouds identified above. In the longitude range and at a distance comparable to the distance estimates ( pc) for the high-latitude clouds, the only OB association is the large Cassiopeia-Taurus association whose nucleus is the more compact Per (Per OB3) association. The Hipparcos measurements give a distance of pc for the Per OB3 association and, with its nucleus and a halo of , is centred at (De Zeeuw et al. 1999). This is very nearly coincident with the positional centre () of the elliptical cavity devoid of high-latitude clouds in this region. The much larger Cas-Tau association surrounds the Per OB3 association and its members have distances ranging between pc and 300 pc (De Zeeuw et al. 1999). We suggest here that the elliptical cavity and the surrounding high-latitude clouds are physically related to the Per OB3/Cas-Tau association. Fig. 3 illustrates this relationship, where the Galactic positions of the high-latitude clouds in the range are plotted together with those of the confirmed members of the Per OB3/Cas-Tau association from De Zeeuw et al. (1999). The elliptical cavity within the inner boundary of the shell-like distribution of the clouds is also shown. If the high-latitude clouds in the region considered above are indeed related to the Per OB3/Cas-Tau association, then the mean distance to these clouds can be taken to be the same ( pc) as that of the OB association, and the linear dimensions of the shell-like structure formed by this group of high-latitude clouds (hereafter Per OB3 shell) are 280 pc 200 pc . The mean shell radius is 120 pc .

3.1. Kinematics of the clouds in Per OB3 shell

The mean radial velocity, with respect to the local standard of rest, of the clouds (from radial velocity data in MHS, HMT and MHHST) is -2.15 km s^-1 and the velocity dispersion is 6.4 km s^-1; if the two intermediate velocity clouds near with velocities km s^-1 are excluded. For the Per OB3 member stars (De Zeeuw et al. 1999) the mean radial velocity is -0.7 km s^-1 and the velocity dispersion is km s^-1. While a majority of the high-latitude clouds in the range are likely members of the Per OB3 shell, a few clouds (like the intermediate velocity clouds mentioned above) may be unrelated to this group. Some of the clouds with , to , having radial velocities km s^-1, similar to that of the near-by () lower latitude Taurus clouds (Taylor et al. 1987), may possibly be related to the Taurus cloud complex. Also, a few high-latitude clouds (especially in the southern Galactic hemisphere) outside of the longitude range and some clouds at lower latitudes may actually be related to the Per OB3 shell. However, in the present discussion we consider only high-latitude clouds within the range as possible Per OB3 shell members. Fig. 4 shows a plot of the radial velocities of the clouds against the Galactic longitude. The intermediate velocity clouds have been excluded from this plot. Also plotted in Fig. 4 are the radial velocities expected for the clouds at their positions () due to differential Galactic rotation following the relation

where km s^-1kpc^-1 is the Oort's constant for kpc (Clemens 1985) and d is the distance to the cloud. For the estimated mean distance (180 pc) and linear dimensions (mean shell radius 120 pc) for the Per OB3 shell of clouds we have plotted the radial velocities due to differential rotation for three values (60, 180 and 300 pc) of distance d. Measured cloud velocities that fall within or close to the range given by the three sets of points in Fig. 4 can perhaps be ascribed entirely to Galactic differential rotation. It can be seen from Fig. 4 that a large number of clouds have radial velocities much in excess of those expected due to differential Galactic rotation. The large observed velocity dispersion ( km s^-1) must be due to large random velocities or systematic motion of the clouds. Some evidence for systematic motions is apparent from Fig. 4. Clouds in the southern Galactic hemisphere have velocities that are on an average lower (more negative) than those of northern clouds. This difference becomes more pronounced if the southern clouds with and velocities ( km s^-1) similar to the Taurus clouds are indeed related to the Taurus cloud complex and are then not members of the Per OB3 shell.

Fig. 4. Plot showing the radial velocity as a function of Galactic longitude for the Per OB3/Cas-Tau shell of high-latitude clouds. Expected radial velocities due only to the differential Galactic rotation are shown by dots for three values (60, 180 and 300 pc) of distances to the clouds. Filled squares denote clouds with , while open squares represent clouds with

The observed velocity distribution and the shell-like structure formed by the Per OB3 shell of high-latitude clouds can be understood if these clouds formed from an expanding shell of swept-up gas around the Per OB3 association. For an expansion speed of the order of times the one-dimensional radial velocity dispersion (6.4 km s^-1), i.e. km s^-1, and a mean distance ( 120 pc) of the clouds from the centre of the Per OB3 association, the kinematic age of the shell can be estimated to be 11 Myr . Expansion from a common centre is also supported by Fig. 5 where the observed radial velocities of the clouds are plotted against , being the angular distance of the cloud from the centre of expansion (). () is the maximum angular separation. If the clouds are distributed in a shell (thin) and expansion velocity is , then the observed radial velocity would be given by

where is the systemic velocity of the whole group (see Rajagopal 1997). For a thin-shell distribution the points in the plot would lie along two straight lines as shown in Fig. 5. If the clouds are distributed in a thick shell volume, then the points would lie within the envelope defined by the two lines. It can be seen from Fig. 3 and Fig. 5 that most of the clouds lie at projected angular distances from the centre and have radial velocities that are more or less uniformly distributed between the two lines, while there is a relative paucity of clouds with larger angular distances having velocities that fall between the two lines. This may indicate the presence of a thick shell of clouds at from the centre surrounded by a thin shell of much larger angular radius (). The fit shown in Fig. 5 (with of the points contained within the envelope defined by the two lines, and a single isolated cloud at , in Fig. 3 not considered for the fit in Fig. 5) for the Per OB3 clouds gives: km s^-1 and km s^-1. For an outer shell radius of pc corresponding to used for the fit shown in Fig. 5, this implies an expansion age Myr . We will adopt an expansion velocity of 15 km s^-1 and an expansion age of 10 Myr in the following discussion.

Fig. 5. Expansion of the Per OB3/Cas-Tau shell of high-latitude clouds: The radial velocity is plotted against

It would be interesting to look for other signatures of expanding gas in the Per OB3 shell. From the kinematics of the high-latitude clouds in this region we have estimated an expansion velocity km s^-1. At lower latitudes (say ) evidence for expansion may be present in the Galactic HI surveys (eg. Weaver and Williams 1973, 1974; or the Leiden/Dwingeloo survey: Hartmann & Burton 1997). An expanding shell centred near would appear as a small disk of HI emission in measurements made at velocities close to the expansion velocity km s^-1, the approaching polar cap at negative velocities and the receding polar cap at positive velocities. Unfortunately, in the Galactic longitude range of the Per OB3 shell the HI gas of the general ISM along the line of sight is approaching us and dominates emission at negative velocities. At positive velocities near km s^-1 some emission is seen in the Leiden/Dwingeloo survey, but this could also be due partly to HI gas along the line of sight having large velocity dispersion. No clear pronounced disk of emission is evident. It is also possible that the receding part of the shell is much weaker because there might have been too little gas on that side of the Per OB3 association. At larger angular distances (and higher Galactic latitudes) from the centre HI emission would be enhanced in a shell at velocities near zero. Possible existence of such emission may be seen in the Leiden/Dwingeloo HI maps at velocities km s^-1, although the picture is rather complicated with several HI arcs and filaments superimposed in projection in this region.

3.2. Relationship of Per OB3 shell with the Gould's Belt

The Per OB3/Cas-Tau association is also at the centre of the Gould's Belt (Gould 1874), a flat system of young stars and interstellar matter within pc of the Sun that is tilted by to the Galactic plane. At lower latitudes the interstellar gas related to the Gould's Belt is distributed in an elliptcal ring with semiaxes pc 210 pc , while the stellar component is more extended with semiaxes pc 700 pc , and the system is expanding with an expansion age estimated to be 35 Myr (e.g. Olano, 1982; Poppel 1997 and references therein). The Per OB3 shell of high-latitude clouds under consideration here has smaller dimensions (semiaxes pc 100 pc) and seems to be nearly at right angles to the plane of the Gould's Belt. Thus, geometrically, the relatively compact Per OB3 association is surrounded by the more dispersed Cas-Tau association, which in turn is surrounded by the shell of high-latitude clouds. These structures, at lower latitudes, reside in the inner cavity of the Gould' Belt that is largely free of diffuse gas and interstellar clouds (e.g. Olano 1982; Ramesh 1994) in an elliptical region with semiaxes pc 210 pc .

The kinematic age of 10 Myr estimated here for the Per OB3 shell of high-latitude clouds is to be compared with the kinematic age of 35 Myr for the expanding system of young stars and interstellar matter associated with the Gould's Belt and the age of 50 Myr for Per OB3 association (Meynet et al. 1993). The shell of high-latitude clouds around Per OB3 is therefore a much younger feature than the ring of expanding interstellar matter associated with the Gould's Belt and the Per OB3/Cas-Tau association. Stellar winds and supernovae from the massive stars of the Cas-Tau association centred around Per (Per OB3), 35 Myr ago, could have produced the expanding ring of the Gould's Belt (Blaauw 1956; Olano 1982). This event would also have produced shells of gas at high galactic latitudes as two caps that oscillate perpendicular to the galactic plane under the influence of gravity due to matter in the galactic disc (Olano 1982). Half period for the vertical oscillation depends on the mass density in the disc, and is estimated to be Myr (e.g. Rampino & Stothers 1984). By now, the material of the two caps would have therefore returned to the Galactic midplane. Also, this material would follow trajectories like in a fountain and is unlikely to show a shell-like pattern (in the Galactic latitude distribution) centred around its point of origin. The Per OB3 shell of high-latitude clouds is perhaps produced by a more recent (10 Myr ago) supernova event in the Per OB3 association which swept out the backfalling material of the high-latitude gas-caps belonging to the Gould's Belt expansion event of 35 Myr ago. It is to be noted that the Per OB3/Cas-Tau association still has several stars earlier than spectral type B3 (De Zeeuw et al. 1999) that can explode as supernovae.

3.3. High-latitude clouds associated with other OB associations

The group of high-latitude clouds clustering around may similarly be associated with the Sco-Cen OB association (Sco OB2) and originate in the gas shell swept up by stellar winds and supernovae taking place in this young association having sub-groups with ages in the range 5-15 Myr (Blauuw 1991). Unlike Per OB3 association, that has a single centre and a clear shell-like distribution of high-latitude clouds, Sco OB2 has several sub-groups of OB stars (Sco OB2-2: , distance pc; Sco OB2-3: , distance pc; Sco OB2-4: , distance pc; De Zeeuw et al. 1999) with different ages and the shell of the high-latitude clouds is rather ill-defined. However, the system of high-latitude clouds can be seen to envelope the extended OB association.

The system of high-latitude clouds around the Sco OB2 association may also be expected to be expanding. Only 26 of the 44 high-latitude clouds in this poorly surveyed region have radial velocity measurements (MHS, HMT, MHHST). The mean radial velocity is +0.2 km s^-1 and the velocity dispersion is 4.2 km s^-1, which is much larger than that can be expected to arise due to differential Galactic rotation. Fig. 6 shows a plot of radial velocities against for the Sco OB2 group of high-latitude clouds, being the angular distance of the cloud from , representing roughly the centre of this rather extended OB association with several subgroups. In Fig. 6, = . Unlike the case of the Per OB3 clouds (Fig. 5), no clear evidence for expansion from a common centre is apparent from Fig. 6 for the Sco OB2 clouds. More complete CO line surveys of this region to search for high-latitude molecular clouds and measure their radial velocities would be required to investigate the kinematics of clouds associated with Sco OB2.

Fig. 6. The radial velocity is plotted against for clouds associated with Sco OB2

A number of high-latitude clouds are also found in the southern Galactic hemisphere in the longitude range . It is to be noted that some members of the much dispersed Cas-Tau association do extend eastward into this region and may be responsible for the formation of the high-latitude clouds there. However, some of these clouds (especially those at relatively lower latitudes) may be related to the Orion and Taurus cloud complexes. The Galactic dark clouds are generally confined to the Galactic plane within . In some directions, however, clouds, other than the local high-latitude clouds being discussed here, can be seen at relatively larger (up to ) latitudes, often making shell-like structures. These clouds could be physically similar to the local high-latitude clouds, but much farther away. An example is the expanding system of gas and clouds in the Gum-Vela complex (Galactic longitudes in the range ) centred around () the Vela OB2 association at a distance of pc (e.g. Sahu 1992; Sridharan 1992; Rajagopal 1997; De Zeeuw et al. 1999). This cloud complex has linear dimensions 210 pc and is expanding with a speed km s^-1, similar to the Per OB3 shell of high-latitude clouds. Were the Vela OB2 association as close to us as the Per OB3 association (180 pc), its clouds would be at galactic latitudes as large as compared to the observed .

Fig. 7 shows the galactic positions of prominent OB associations within pc of the sun (taken from De Zeeuw et al. 1999) together with the clouds of Fig. 1. As disussed earlier the clustering of high-latitude clouds around Per OB3 and Sco OB2 is clear. At relatively lower latitudes also the positional coincidence between OB associations and groups of clouds showing signs of shell-like distribution and extensions to higher latitudes is apparent. This can be seen for the Vela OB2 (as discussed above), Ori OB1 (; distance pc), Per OB2 (; distance pc), Lac OB1 (; distance pc) and Cep OB6 (; distance pc) associations. Being more distant than the nearby Sco OB2 and Per OB3 associations, the angular sizes of the shell-like cloud distributions around these OB associations are smaller.

Fig. 7. Prominent OB associations within pc are shown together with the Galactic distribution of molecular clouds of Fig. 1. The OB associations are shown by the large open star shaped symbols whose sizes are in inverse proportion to their distances

© European Southern Observatory (ESO) 2000

Online publication: October 24, 2000
helpdesk@link.springer.de