2015ApJ...809...81R

The W3 giant molecular cloud is a prime target for investigating the formation of high-mass stars and clusters. This second study of W3 within the HOBYS Key Program provides a comparative analysis of subfields within W3 to further constrain the processes leading to the observed structures and stellar population. Probability density functions (PDFs) and cumulative mass distributions (CMDs) were created from dust column density maps, quantified as extinction A_V. The shape of the PDF, typically represented with a lognormal function at low A_V "breaking" to a power-law tail at high A_V, is influenced by various processes including turbulence and self-gravity. The breaks can also be identified, often more readily, in the CMDs. The PDF break from lognormal (A_V(SF) ~ 6-10 mag) appears to shift to higher A_V by stellar feedback, so that high-mass star-forming regions tend to have higher PDF breaks. A second break at A_V 50 mag traces structures formed or influenced by a dynamic process. Because such a process has been suggested to drive high-mass star formation in W3, this second break might then identify regions with potential for hosting high-mass stars/clusters. Stellar feedback appears to be a major mechanism driving the local evolution and state of regions within W3. A high initial star formation efficiency in a dense medium could result in a self-enhancing process, leading to more compression and favorable star formation conditions (e.g., colliding flows), a richer stellar content, and massive stars. This scenario would be compatible with the "convergent constructive feedback" model introduced in our previous Herschel study.