The Crescent Nebula (NGC 6888) around the WR
136 star is Wolf-Rayet bubble. In most Wolf-Rayet bubbles the outermost
edge of the [O III] emission leads the H alpha emission. Gruendl et al suggest that these
offsets, when present, are due to the shock from the Wolf-Rayet bubble expanding into the
circumstellar envelope. This image enhaced to show the H alpha (red) and [O III] (green)
shows a clear offset between the H and [O III] emission. The [O III] emission seems to be
centered on the WR star, while the H shell is off-center.
What follows is a quote from Gruendl et al.
Physical Significance of H and [O III]
Morphologies
H and [O III] line images of a nebula
provide useful diagnostics of its physical structure. H is a recombination line and thus
shows the overall distribution of ionized material, but because the line strength of H
decreases with temperature its sensitivity drops for temperatures greater than 104 K. On
the other hand, the [O III] line originates in a forbidden transition whose upper level is
populated by collisional excitation; therefore its intensity increases with temperature,
and [O III] line images trace high-excitation regions with temperatures 104 K.
Consequently, we expect the H and [O III] morphologies to be different, especially over
regions where physical conditions change rapidly.
Behind a shock front, radiative cooling
takes place; as the temperature drops, the density increases. Thus, a displacement between
[O III] emission and H emission occurs (Cox 1972). The magnitude of the displacement
depends on the postshock temperature, which is determined by the shock velocity, and the
cooling rate, which is dependent on the density and metallicity. In cases in which a shock
propagates into a dense medium, the cooling rate behind the shock front is high, the
cooling zone is narrow, and the offset between H and [O III] emission peaks is minimal.
For a shock propagating in a tenuous medium, the cooling rate is lower, the cooling zone
is wider, and thus the offset between the H and [O III] emission peaks is larger and may
be observable. If the ambient density is low enough, it is possible that [O III] emission
is observable behind the shock, while H emission is too faint to be detected. Finally, if
a shock propagates through a high-density medium and then into a low-density medium, it is
possible to see bright H emission associated with the dense medium and a leading [O III]
front associated with the recently shocked low-density medium. This could produce the
largest displacement between the leading edges of H and [O III] emission.