Astron. Astrophys. 363, 917-925 (2000)

10. Comparison with radio and optical tracers

The celestial coordinates of the knots have been used for identification of the sources in observations at other wavelengths. Special emphasis has been placed on the catalogue of optically identified H II regions (Pellet et al. 1978) and the radio continuum as tracer for H II regions. We have used the 610 MHz catalogue of Bystedt et al. (1984) which is flux limited and provides a list of well identified sources. In order to separate the thermal and nonthermal radio emission, maps at 2 cm or 6 cm are required but are not yet available. Furthermore, catalogues of supernova remnants (SNRs) (Dickel & D'Odorico 1984) and dark clouds (Hodge 1980), and surveys of CO (Dame et al. 1993) and H I (Brinks & Shane 1984) have been looked up for comparison. We limited our comparison to these catalogues, because they cover the complete area of M 31, where the FIR knots are situated. The results of this search are listed in Column (15) of Table 1. These tracers do not simply correlate with classification of the FIR knots. Nevertheless, some statistical trends can be recognized: A large number of our warm knots (SED types III, and also II) are identified as H II regions. This is expected, as nearly all the 60 µm point sources coincide with H II regions (Xu & Helou 1996). The 20 cm map of Beck et al. (1998) has been looked up and yields coincidence with nearly every knot independant of the SED type. However, the cold knots of type I are not as prominent at 20 cm as the warm knots of type III.

Although the optically dark dust lanes in M 31 generally match the FIR ring quite well, only a weak correlation with the dark clouds (Hodge 1980) is found. The high number density of these dark clouds, however, considerably increases the likelihood of finding one at any randomly picked position within our spatial resolution of 90". In this respect, it appears more interesting that a number of even large dark clouds like D76, D144, D425 or D517 are not found in the 175 µm map. They probably do not have sufficient mass to provide a prominent 175 µm emission. This picture is consistent with the lack of any correlation between optical extinction maps as created by Nedialkov (1998) and our FIR sources. A moderate amount of just 1 mag of visual foreground extinction already results in a dark lane.

SNRs are found close to the position of some of the knots, but they are not sufficiently coincidental to provide evidence for a correspondence. However, SNRs are only found for warm (and medium) knots with temperatures of the cold dust component above 17 K (SED types II and III). Most of the warm knots correspond with H II regions, hence they contain high mass stars, i.e. the precursors of SNRs. Therefore SNRs are also most likely to be found in warm knots.

In Fig. 7 the coincidence of the FIR knots with the other tracers is illustrated as a histogram for the three SED types. Whereas the knots of type II more or less follow the general distribution, the warm knots are more frequently represented by the SNRs and H II regions as discussed above, and the cold knots of type I correlate with detections in CO and H I.

Fig. 7. Statistical coincidence of the FIR knots with detections by other tracers separated for each SED type. The cold knots (type I) are more related to detections in H I and CO, whereas the warm knots (type III) are dominant at SNRs and H II regions. The medium knots (type II) follow more a mean distribution.

For the southwest part of M 31, new CO observations with higher resolution and sensitivity have been presented by Neininger et al. (1998) and Loinard et al. (1999). A comparison of the knots of this area with the new CO data reveals, that all of them have a faint CO counterpart. The same will probably be the case for the other radio tracers, where so far no better data are available. The catalogues used are intensity limited. Hence, the coincidences of the FIR knots with the other tracers must not be interpreted in the way, that e.g. the warm knots do not contain any CO, but that they emit too faintly to be detected in the survey of Dame et al. (1993). On the other hand, also the cold knots do probably contain H II, but the emission is too faint to be detected. However, this does not alter the general conclusion, that the cold knots are more related to CO and H I, the warm knots more to H II.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk@link.springer.de