We present and (2,2) maps sampled at a beam spacing with a FWHM beamwidth of ( pc) over a 120 square-arcminute region toward the high density layer of the W 3 GMC. The map in- cludes the W 3 Main and W 3(OH) star-forming regions. We find that strong emission ( K km s-1 ) is detected toward 9% of the surveyed region. It is only detected toward regions of high-mass star formation, namely the W 3 Main and W 3(OH) clouds, and toward a newly detected core to the southeast of W 3 Main, which we denote W 3 SE.
Toward W 3 West, we find cool, = 45 K, am- monia gas in a region of 1 pc diameter outlined by many other molecular tracers. We find N( = cm-2 and a virial mass of 1100 for this core. This yields an relative abundance of X( = , consistent with predictions based on chemical models.
Toward W 3 East we find only weak and (2,2) emission, but relatively strong (3,3) emission. From the and lines, we determine K and N(H2) = cm-2 . The resulting X( = is an order of magnitude lower than the relative abundance found toward W 3 West or the relative abundance predicted by chemical models. Since W 3 East is the most active star-forming core, we propose that the lack of emission toward the stellar cluster may be due to the enhanced rate of destruction in an environment dominated by the energy input of the emerging young stellar sources.
For the newly detected W 3 SE core, we find a = 25 K, N( cm-2 and a virial mass of . W 3 SE has a relative abundance of , similar to that of W 3 West. The core appears to be marginally resolved in our data, suggesting a core size comparable to the beamsize (0.5 pc). Star formation may be occurring inside of this core as indicated by an elongated, jet-like nebulosity emanating from W 3 SE apparent in our -band images. Recent observations show that this nebula is tracing vibrationally excited emission excited by UV radiation or an outflow.
Toward W 3(OH) we find an extended and cool component with = 27 K, which is not substantially heated by the embedded young stars in W 3(H2O) and the W 3(OH) ultracompact H II region. The strong emission is extended over an area of pc, demonstrating that the W 3(OH) core is much larger than previously thought. The core seems to be composed of a component centered on the W 3(H2O)/W 3(OH) ultracompact H II region and a plume which extends as far as 1 pc from the ultracompact H II region. Toward the central component of W 3(OH), we calculate on average N( = cm-2 and a virial mass of 1500 . We find X( = for the cool extended gas, an order of magnitude lower than than the relative abundance predicted for the hot central core in W 3(OH).
We find that the plume of ammonia reaching from W 3(OH) toward the northeast is a much more extended structure than indicated by previous VLA maps in Using the virial theorem, we find that the plume holds a mass of at least 400 of molecular gas. We find that the plume extends toward two clusters of stars, one of which is located at the edge of the plume. These clusters may be previously unknown sites of star formation associated with the W 3(OH) molecular cloud and merit further investigation. The stellar clusters are found in the westernmost component of a chain of nebulosities which extends eastward toward the W4 H II region.
We detect weak emission over one fourth of the surveyed region. Examination of a representative region suggests that the weak emission is tracing cm-3 gas, which may be in bound clumps. In some cases, the small size of the observed emission regions indicates that beam dilution may also contribute to their weak signal.
We have compared the mass in the surveyed region traced by the CO (1-0) transition to the total mass of the cores detected in the W 3 GMC. We find that the comprises % of the mass traced by CO. This is comparable to the mass fraction of dense cores in Orion B derived by Lada (1992). Although the total mass of dense gas detected in the Orion B survey of Lada (1992) is similar to the total mass of the cores in W 3 GMC, we find that the dense gas in W 3 is concentrated into fewer, more massive cores. Following the observation by Lada (1992) that star formation is concentrated in the most massive cores, we suggest that the higher core masses in the W 3 GMC may be the reason for the higher rate of high mass star formation observed in the W 3 GMC.
Although all sites of recent star formation appear to be associated with the cores on a parsec scale, on a sub-parsec scale we find an anticorrelation between the cores and the sites of recent star formation. This anticorrelation is particularly apparent toward W 3 Main where we find no evidence for NIR clusters or H II regions toward the centers of W 3 West and W 3 SE. One explanation is that the NIR clusters remain hidden in the cores until the cores are dispersed. However, in one case, we find that the anticorrelation is due to a lower relative abundance of the molecule. We propose that this "thinning out" is due to the preferential destruction of molecules by the stars embedded in the cores.
© European Southern Observatory (ESO) 1998
Online publication: July 27, 1998