6. X-ray "hot spots" and extended emission
In addition to point sources, the Monoceros and Rosette PSPC images (Figs. 2) show evidence for extended emission in two ways. First, we see "hot spots" of X-ray emission, regions of limited extent with tightly clustered sources. These appear near the center of the Monoceros field, and to the NNW around the inner window support structure in the Rosette field. Second, emission significantly more extended than point sources is seen towards the NW edge of the Rosette field. We seek to discriminate emission from unresolved sources and intrinsically diffuse emission.
6.1. The Monoceros X-ray hot spot and molecular core
This X-ray hot spot is located around the dense molecular core at the PSPC field center, which has been extensively studied in the near-IR. A cluster of nearly 200 sources within a area is present (Carpenter et al. 1997). These IR sources are pointlike with positions determined within . In contrast, the X-ray image shows both emission from a few point sources (with FWHM 30") and extended emission, on a scale of a few square arcminutes. Eight Monoceros hot spot sources coincide with sources found by Carpenter et al. (1997; see Sect. 4.2).
The near-IR photometry of the Monoceros sources permits estimation of individual source extinctions, (see CMFA for details), given in Table 5. With between 2.5 and 6.75, it is clear that these sources are significantly embedded in the cloud. Fig. 9 shows the X-ray/J-band relation for these stars after extinction correction. The correlation is comparable to that found previously in the Oph core. We conclude that the Monoceros core X-ray hot spot sources must be of a very similar nature to those of Oph core.
A further insight into the nature of the ROSAT sources in the Monoceros core can be gained through the inspection of the positions of these sources with respect to the core material, as deduced from molecular tracers. Maps of , (Montalbán et al. 1990) and (Gonatas et al. 1992) have been obtained, as well as a map of the cold dust in the millimeter continuum (Henning et al. 1992). Except for the map, which suffers from a large beam, the maps have angular resolutions comparable to that of our X-ray observations. Because it is less sensitive to often subtle interstellar chemistry effects, the best material tracer for star formation is the dust. Fig. 11 shows the superposition of the X-ray map of Fig. 4 (white contours), near-IR sources (, star symbols, Carpenter et al. 1997; , asterisks, Aspin & Walther 1990), and the grey-scale 1300 m map of Henning et al. (1992). This superposition shows that the X-ray sources and their associated "bright J " IR sources lie away from the central dust condensation, while the K -band sources, mostly invisible in X-rays, are concentrated within this condensation.
The X-ray hot spot is not identified with bright IR sources, and we suggest it may be associated with fainter sources. (It does coincide with the optical reflection nebula, but no physical mechanism could conceivably link X-rays with this nebula.) To gain insight into this possibility, we consider how the Oph core, which has dozens of low luminosity IR sources, would appear in X-rays if it lay at the distance of the Monoceros cloud. The result (Fig. 11) shows that the observed and simulated images are morphologically very similar, showing a cluster of partially resolved sources several arc minutes in extent.
We conclude that the ROSAT point source population in the Monoceros cloud has two components: well-resolved X-ray sources associated with bright optical and IR sources, which are likely to be moderate-mass pre-main sequence stars; and the hot spot of partially-resolved X-ray sources associated with a densely concentrated embedded population of lower mass pre-main sequence stars such as seen in the Oph core. More widely distributed low mass stars are likely present but are too faint to be detected in our ROSAT exposures. These two populations are consistent both with the conclusions drawn from the luminosity function (Fig. 8, showing the similarity to Cha I and Herbig Ae/Be stars on a large scale), and from the relationship between X-ray and J -band emission (Fig. 9) for embedded sources in the core region.
6.2. The Rosette X-ray hot spot and CO observations
The Rosette X-ray hot spot is about 10' in extent around = , , NNW of the field center, with its peak coinciding with Rosette X-11. Its morphological appearance is strikingly similar to the Monoceros hot spot. We have simulated a Oph core-like cluster of ROSAT sources at the distance of the Rosette: Fig. 12 shows the similarity between the Rosette hot spot and the embedded Oph cluster as found for the Monoceros hot spot.
This view is independently supported by the new analysis of CO observations (Sect. 2.2) which confirms (bound) localized molecular cores surrounded by a very extended (unbound) diffuse emission (Williams et al. 1995). Fig. 13 shows the superposition of the contours, with a resolution of , the ROSAT contours of Fig. 3b, and the location of the IR clusters of Phelps & Lada (1997). The X-ray hot spot is found to coincide, within the positional uncertainties, with the location of the molecular core, also visible as the largest clump in the map of Blitz & Stark (1986), and encompasses, in a broad sense, the IR clusters 4, 5, and 6. Another dense clump, in the vicinity of the HII region, contains the IR cluster 2, and is associated with a weak, extended X-ray emission. Other molecular clumps are not associated with X-ray emission, yet harbor IR clusters (1, 7). Most likely, these contain only low-mass, X-ray faint stars. Clusters associated with X-ray emission (Table 6) probably contain stars of higher mass, such as Herbig Ae/Be stars, the presence of which was already suggested above (Sect. 5.4 and Fig. 8). This could be verified by near-IR photometry of the clusters. It is also noteworthy that our ROSAT observation was pointed at the peak of the CO contours existing at the time (Montalbán et al. 1990). The fact that we do not find a cluster of X-ray sources at this location, but at the location of the CO peak obtained later with an improved analysis, is an additional argument supporting a close association between star-forming molecular cores and X-ray emission, as first found in the Oph core.
6.3. Diffuse emission associated with the Rosette HII region
At the NW edge of our Rosette cloud lies a region of extended emission coincident with the sharp boundary between the inner "hole" of the nebula and the outer HII region seen in visible light (Fig. 3b). Far off-axis, the ROSAT telescope angular resolution is strongly degraded (), and the observed feature in size could be produced by a handful of point sources. The X-ray luminosity of this feature, integrated over all its area, is erg s-1, assuming keV and . This is comparable to the integrated luminosity of the point sources, where the average luminosity per source is erg s-1. The extended feature could thus be produced by tens of low-mass young stars which formed prior to the massive stars responsible for the HII region, and/or as a result of the shocks created by their strong stellar winds.
As an alternative, we examine the case where the X-ray emission is, at least in part, extended. 2 The extended X-ray feature is morphologically consistent with the shock boundary between the strong stellar winds of the early O stars exciting the nebula and the dense surrounding HII region. Because the wind velocities are very high (a few thousand km s-1), diffuse keV X-ray emission from wind energy dissipation is expected at this boundary. When taking into account the non-linear conduction downstream of the shock (Dorland & Montmerle 1987, DM hereafter), we expect ( being the total mechanical wind luminosity ). The angle centered in the inner hole subtended by the extended feature is , or a solid angle . From DM, we estimate the X-ray luminosity radiated by the stellar winds in the ROSAT field: . For the Rosette nebula, the mass loss is dominated by one O4V and one O5V star with and yr-1 and and 2800 km s-1, respectively (Dorland, Montmerle, & Doom 1986; Lamers & Leitherer 1993). This yields erg s-1, compatible to the observed extended emission. Given the theoretical uncertainties (see DM), such a contribution of wind shocks from massive stars is plausible.
© European Southern Observatory (ESO) 1998
Online publication: February 4, 1998