Content uploaded by Quentin A. Parker
Author content
All content in this area was uploaded by Quentin A. Parker on Aug 14, 2014
Content may be subject to copyright.
Accepted for publishing in MNRAS
The optical emission nebulae in the vicinity of
WR 48 (Θ
ΘΘ
Θ Mus); True Wolf-Rayet ejecta or unconnected
supernova remnant?
M. Stupar,
1,2
Q.A. Parker,
1,2
M.D. Filipović
3
1
Department of Physics, Macquarie University, Sydney 2109, Australia
2
Anglo-Australian Observatory, P.O. Box 296, Epping, SW 1710, Australia
3
University of Western Sydney, Locked Bag 1797, Penrith South DC, SW 1797 Australia
Accepted 2009, October
ABSTRACT
During searches for new optical Galactic supernova remnants (SNRs) in the high resolution,
high sensitivity Anglo-Australian Observatory/United Kingdom Schmidt Telescope (AAO/UKST) Hα survey
of the southern Galactic plane, we uncovered a variety of filamentary and more diffuse, extensive nebular
structures in the vicinity of Wolf-Rayet (WR) star 48 (Θ Muscae), only some of which were previously
recognised. We used the double-beam spectrograph of the Mount Stromlo and Siding Spring Observatory
(MSSSO) 2.3-m to obtain low and mid resolution spectra of selected new filaments and structures in this
region. Despite spectral similarities between the optical spectra of WR star shells and SNRs, a careful
assessment of the new spectral and morphological evidence from our deep Hα imagery suggests that the
putative shell of Θ Mus is not a WR shell at all, as has been commonly accepted, but is rather part of a more
complex area of large-scale overlapping nebulosities in the general field of the WR star. The emission
comprises a possible new optical supernova remnant and a likely series of complex H II regions. Finally, we
present the intriguing detection of apparent collimated, directly opposing, ionized outflows close to Θ Mus
itself which appears unique among such stars. Although possible artifacts or a temporary phenomenon
monitoring of the star is recommended.
Key Words: ISM: supernova remnants, Wolf-Rayet nebulae; ISM: H IIregions; ISM: individual: G304.4-3.1;
ISM: general - stars - Wolf-Rayet; ISM: individual: WR 48; ISM: individual: HD 113904; ISM: individual: Theta (Θ)
Muscae
1 ITRODUCTIO
The optical broad-band or preferably narrow-band imaging
morphologies of emission line structures can provide clues as to
their underlying nature, especially if their shapes are
pathological examples of their generic object types. Usually
however, corroborating optical spectroscopy or other data
provide the basis for firm identification. For example, for optical
filaments or nebular emission to be classified as belonging to
supernova remnants (SNRs), planetary nebulae (PN), H II
regions, Wolf-Rayet shells or other exotica, a variety of optical
emission line ratios and diagnostic plots have traditionally been
used to assist the discrimination between the object classes (e.g.
Sabbadin, Minello & Bianchini (1977), Cantó (1981) and Frew
& Parker (2009)). For SNRs for example, whether they have
confirmed radio counterparts or not, the optical spectra are
expected to show typical diagnostic lines and ratios from the
extant confirmed and optically detected SNRs (see Stupar,
Parker & Filipović (2008) and references therein). SNR spectra
are not generally photo-ionised. Instead the emission arises from
strong, shock heating and collisional excitation as the expanding
blast-wave and ejecta sweeps up and collides with the
surrounding ISM. Usually, this is diagnosed via the presence of a
very strong [S II] doublet at 6717 and 6731Å relative to Hα .
Typically, a ratio of [S II]/Hα
>
0.5 has often been used to
classify nebular emission as likely arising from an SNR,
especially if the associated optical morphology is pathological
and there are other corroborating pieces of evidence such as
radio structure, a central X-ray source or a pulsar. Such optical
emission line ratios are often used to separate SNRs from H II
regions and planetary nebulae (PN). Unfortunately, as lower
surface brightness and more extreme, evolved examples of these
different objects are increasingly uncovered from the next
generation on narrow-band imaging surveys, e.g. Parker et al.
(2006), Miszalski et al. (2008) these optical diagnostics are not
always clear cut (Frew 2008; Frew & Parker 2009). The various
object populations which exhibit optical emission line spectra
can overlap in the so called Cantó diagram (Cantó 1981). Indeed,
high ratios of [S II]/Hα
≤
1.0 have been confirmed in a few
bona-fide, highly evolved PNe which are strongly interacting
with the ISM, which is, in fact, the source of the shock-
excitation (in highly evolved, see Pierce et al. 2004).
Fortunately, there are other optical emission lines such as very
strong [O II] at 3727Å, the Balmer lines, [O III] at 4959 and
5007Å, [N II] at 6548 and 6584Å and especially strong [O I]
at 6300 and 6364Å, that can assist in assigning an SNR
classification as [O I] is not prominent in PN or H II regions.
There is some confusion in these emission line ratio
selection criteria, especially between SNRs and Wolf-Rayet
nebula which can exhibit both morphological similarities in the
form of optical filaments and filamentary shells and similarities
in their optical spectra. Of course the confirmed detection of a
WR star in the centre of such nebulae resolves any ambiguity.
Usually, the WR nebulae (Miller & Chu 1993; Marston et al.
1994; Marston 1997) are in the form of a ring (or significant
fraction of a ring) centred on the star and consisting mainly of
stellar ejecta from the WR star jh65 and swept-up material from
the interstellar medium. In some WR stars, the ratio of [S II] /
Hα emission lines is close to that typically seen for SNRs. This
is understandable, as we know that Wolf-Rayet stars are
evolved, extremely hot and luminous massive stars undergoing
rapid mass loss which can form both shock-excited and photo-
ionised nebula from the material discarded by the central star
(Johnson & Hogg 1965). Fortunately, the progenitor star can
always be identified if the nebula has a WR origin enabling
discrimination against an SNR.
During our work (Stupar, Parker & Filipović 2008,
2007a,b) on uncovering new Galactic SNRs from the
AAO/UKST Hα survey of the southern Galactic plane (Parker
et al. 2005), we discovered a number of previously unrecognised
faint nebulae which could be associated with known Wolf-Rayet
stars and have subsequently acquired spectra for some of these
new nebulae. Also, given our previous comments on the
difficulties in unequivocal identifications based only on
available optical emission line ratios, it is possible that some of
our lower confidence, new Galactic SNR candidates reported in
Stupar, Parker & Filipović (2008) from our work, which
currently have weak corroborating evidence as to their nature,
may eventually turn out to be Wolf-Rayet star nebulae or even
PNe, if a fainter Wolf-Rayet star or hot star is later uncovered in
the central vicinity (see later discussion). For WR stars as least
this is considered highly unlikely, since, according to the last
(Version VII) Catalogue of Wolf-Rayet stars (van der Hucht
2001) and the last annexe (van der Hucht 2006) to this VII
Catalogue, none of our significant number of newly uncovered
nebulae have a match in this list of Wolf-Rayet stars or their
nebulae. Also, it should be emphasized that 35% WR stars
(Marston 1997) have an associated nebula. Our deep
AAO/UKST Hα survey has uncovered some additional nebula
that could be associate with known WR stars but detecting
filamentary nebulosities close to a WR does not necessarily
mean they are related, especially if they are highly
asymmetrically distributed. Additional kinematic information
and other evidence may be required. As an example we consider
two objects identified as G283.7-3.8 and G306.7+0.5 in our SNR
candidate Catalogue (Stupar, Parker & Filipović 2008). G283.7-
3.8 was disregarded as a WR star nebulosity due to the observed
morphological form not being typical for this kind of object, e.g.
the presence of long parallel filaments and the fact that the
supposed exciting WR 17 star (van der Hucht 2001) is remote
from the filaments in angular separation (~40 arcmin).
G306.7+0.5, which, although possessing spectral characteristic
and a morphological form that cannot rule out a WR star nebula
origin, has been disregarded as being a WR star nebula (of the
nearby WR star 53) due to the presence of an X-ray source and
pulsar in the vicinity of the observed emission cloud.
Consequently, both nebular structures have been classified as
new optical Galactic SNRs candidates in (Stupar, Parker &
Filipović 2008).
In this paper we present new spectral observations of
several freshly identified nebular components in and around Θ
Mus (WR 48), the second brightest known Wolf-Rayet star. The
findings were made during our search for new optical Galactic
SNRs (Stupar, Parker & Filipović 2008). We compare these
spectral observations with previous observations found in the
literature and discuss the likely origin and nature of the putative
WR 48 shell nebulae which we now reject based on our new Hα
imagery. For the first time we also show the presence of
apparent, almost directly opposing, collimated ionised jets close
in to the WR star itself as uncovered on the quotient image (Hα
divided by SR) from the SuperCOSMOS UKST/AAO Hα
(SHS) data. The blocked down (factor of 16) quotient image for
the relevant SHS survey field HA137 is shown in Fig. 1. It is
clear from this that there is no obvious WR shell or ring around
Θ Mus. There is extensive, scallop shaped diffuse emission
mainly to the South and East that also loops back on itself at the
Western extremity. It is 30 arcmin away from Θ Mus. This
emission also appears connected to more diffuse nebulosity
extending further to the West and indeed across much of the
entire 25 sq.degree of the H-alpha survey field. We have also
uncovered extensive, fine filamentary nebulosity of a completely
different morphological character that does follow some of the
general shape of the diffuse emission but is also found outside of
this emission too and in the general vicinity of Θ Mus (marked
by a red circle on Fig 1). This illustrates the complicated nature
of the extant emission over a large 25 sq.degree region in an
around Θ Mus.
2 METHOD AD OBSERVATIOS
In June 2004 we used the Double Beam Spectrograph (DBS)
1
of
the Mount Stromlo and Siding Spring Observatory attached to
the Australian National University (ANU) 2.3m telescope to
acquire spectra of various nebular components previously
thought to be associated with the Wolf-Rayet star Θ Mus. The
DBS has a dichroic which feeds the blue and red arms of the
spectrograph. In the blue we used a 600 lines mm
-1
grating
which covered wavelengths between 3700 and 5500Å. For the
red arm we used a higher resolution 1200 lines mm
-1
grating
with spectral range between 6100 and 6800Å. This covers the
important diagnostic nebula lines in the red while the resolution
is sufficient for kinematics. The slit width was 2.5” and a
resolution of 2 and 1Å was achieved in the blue and red arms
respectively. Observational details are given in the Table 1. Data
reduction was performed using standard spectral reduction
routines from the IRAF package but supplemented with specially
designed DS9 and IRAF scripts developed for more efficient
extraction of 1D nebula spectra by colleague B. Miszalski
2
. For
spectral flux calibration we used observations of the standard
photometric stars: LTT4302 for June 12 and 14, 2004 and
LTT4816 for June 13, 2004.
1
http://www.mso.anu.edu.au/observing/2.3m/DBS/
2
http://www.aao.gov.au/local/www/brent/pndr/
Figure 1. The full 25 sq.degree area of AAO/UKST H-alpha survey field HA137 as a binned 16 arcsecond per pixel low resolution quotient
map of Hα divided by the matching short red (SR) broad-band image. The WR star Θ Mus (WR 48 aka HD 113904) is marked by a red circle. There is no
obvious WR shell associated with WR 48. The previously identified emission shell supposedly identified with WR 48 from inferior Hα imagery in terms
of sensitivity and resolution is seen to be nothing more than a more prominent arcuate emission component to the South of the star that is connected to
other similar, elongated, diffuse emission structures across much of the survey field.
3 OBSERVIG RESULTS
3.1 The putative Θ
ΘΘ
Θ Muscae (WR 48) nebula: a Wolf-Rayet
shell, SR or H IIregion?
Narrow band, high resolution imaging from the
AAO/UKST Hα survey of the southern Galactic plane (Parker
et al. 2005) clearly showed the existence of various nebula
structures with clearly different morphological characteristics in
the vicinity of Θ Mus. One of the most prominent emission
regions is a scallop-shaped nebula some 40’ in E-W extent about
40 south of Θ Mus that is however, not isolated but connected to
a series of similar diffuse emission structures across much of the
survey field. The new Hα imaging also reveals, for the first
time, the presence of a network of apparently distinct multiple
fine emission filaments typically ~10’ in extent extending across
much of the same vicinity, sometimes overlapping sometimes
not the more diffuse structures (Fig. 2). Prior to the SHS the
evident nebulosity in the region around Θ Mus has generally
been accepted as representing components of a WR shell nebula
(for the first time reported by Heckathorn & Gull 1980). Fainter,
extensive emission and finer detail of the putative WR shell have
now become evident from the improved SHS imagery (see Fig.
1) which show the previously identified, supposed WR partial
shell to be actually clearly connected to an extensive network of
diffuse emission structures across much of the SHS survey field
- see Fig. 1. There does not appear to be a distinct WR emission
shell. Furthermore, careful examination of this region in the
radio, particularly in the PMN survey (Condon, Griffith &
Wright 1993) at 4850 MHz clearly uncovered a scallop-shaped
region of radio emission in a similar form to that seen for the
finer emission filaments seen in Hα light that overlaps and
appears to follow the similar shaped more diffuse emission but
may be unconnected or interacting with it. Fig. 3 shows this
nebula at 4850 MHz. Note that the Hα filaments seen in Fig. 3
are not distinguished from the radio emission, most probably due
to the low resolution of the PMN survey (~5’). However, this
newly noted radio structure can also just be recognised as a very
low surface brightness feature barely above the noise in the
SUMSS 843 MHz (Cram, Green & Bock 1998) survey data.
For our spectroscopic follow-up we focused on several
of the fine-scale newly uncovered Hα filaments (Figs. 2 and 4)
as such structures are typical of optical supernova remnant
detections. The observations were made as part of our
spectroscopic follow-up programme of new optical SNR
candidates (as example, see filaments similarity in Blair et al.
1999; Ghavamian et al. 2001. Spectra were taken at two different
filament locations (see Table 2). The position and orientation of
the slits is shown in Fig. 4.
Figure 2. The area ~50’ south the WR star Θ Mus again from the AAO/UKST H
α
Survey as a binned 16 arcsecond per pixel low resolution
quotient image. The fine filaments and surrounding diffuse (scallop shaped) nebula are clearly seen. The X-Ray sources 1RXS J130924.1-655355 and
1RXS J130551.0-655117 are also marked (with X) and lie very close to the fine H
α
filaments (see later discussion). What can also be noticed on this
quotient image are apparent, ionised, opposing jets close to the Θ Mus star itself (shown in greater detail in a later figure). This unique discovery is
elaborated on in Section 3.3. In the right corner of the figure is an enlarged part of the image north-east from Θ Mus illustrating the fine detail in the
arcuate filaments and cloud emission.
Examination of the resultant 1-D flux calibrated
spectra of the two observed filaments in the vicinity of Θ Mus
(see Figs. 5 and 6) showed that both spectra and resultant
diagnostic emission line ratios actually fit well inside the normal
criteria for optical SNRs (Fesen, Blair & Kirshner 1985) which
generally distinguishes SNRs from H II regions and most PN.
Many of the same optical emission lines are present in all 3
object types and although there are strong similarities the
physical processes that give rise to the ionisation are different.
According to Fesen, Blair & Kirshner (1985), the main factor
which characterize optical spectra of SNRs, confirming a shock,
is the ratio of [S II] 6717 and 6731Å/ Hα
> 0.5, though others
consider a ratio > 0.7 to be better (Mathewson & Clarke 1973).
Such an identification can additionally be supported by the
presence of [O I] , [O II] and [O III] lines as well as [N II]
lines at 6548 and 6584Å and the Balmer lines.
However, in addition to such line ratios being typical
for optically detected SNR and not usually in PN and never in H
II region emission spectra, we also have the nebulae of WR
stars. Here, we have both cases: the spectral features can be very
similar to the optical spectra of SNRs so that they can be very
hard to distinguish (see later discussion). This is to be expected
as both ejecta processes involve the creation of shocks. In SNRs
the root of the shock is in the supernova explosion, while in WR
star nebulae it is in the speed of stellar wind ejecta from the WR
star. Besides, the observed morphological structures of WR
shells can be connected with a specific evolutionary stage of the
Wolf-Rayet nebula (see Marston (1995) for details). Large,
slowly expanding shells which are mostly connected with the
initial phase of the WR star, when the star is in the O phase, are
also often seen in the IR and as H I holes. Faster, expanding
shells are also present from prior to the WR phase (when the star
is a red supergiant or Luminous Blue Variable) where expansion
is probably connected with initial mass ejection or due to
acceleration when a later WR star is formed at the centre. In a
WR wind-blown shell, filamentary structures can be created
when a fast WR wind overtakes the shell of surrounding ejecta.
Also, on the basis of Hα
and [OII] imaging morphology of
currently known WR shells, they have been classified into a few
basic types (Type I to Type IV) depending on the spatial
displacement of the front of Hα
and [O III] emission (Gruendl
et al. 2000) which can be most probably applied to the shock
front of SNRs for the cases when [O III] is detected.
A basic analysis of the 1-D optical emission lines
shown in Fig. 2 and 3 is presented in Table 2. Due to non-
photometric nights, F (λ) instead of the extinction corrected flux
I (λ) is given. Both slit positions provide [S II]/Hα
ratios of 0.86
and 0.71 (well inside that typical of SNRs). The usual Balmer
and nitrogen lines are also present. The [O II] and [O III] lines
are present but [O I] at 6300 and 6364Å are not seen. This is
quite common when reducing low resolution spectra as the [O I]
6300 and 6364Å lines are quite prominent in the night sky
spectrum which can make proper sky subtraction difficult unless
there is a significant kinematic component from the source. Even
if these lines are detected they often do not present the canonical
3:1 ratio demanded theoretically due to such sky-subtraction
errors. The lack of [O I] at 6300 and 6364Å in both spectra can
also be explained by contamination of the underlying SNR
spectrum in this region with emission from a H II region (e.g.
Fesen, Blair & Kirshner (1985) for the case of G166.2+2.5) or
simply an interaction between the ionized gas of the scallop-
shaped feature and the filaments. One more example can be
found in Fesen, Blair & Kirshner (1985) where, for two slit
positions for the confirmed SNR G206.9+2.3 only one spectrum
showed low level [O I] emission at 6300Å.
Figs. 5 and 6 and also Table 2 show that both spectra
have emission of [Ne III] at 3867Å while the spectrum from
June 12, has additional lines of He+H at 3888Å and [Ne III] +H
at 3970Å. These lines, thought not common, are occasionally
seen in some observed SNR filaments (e.g. Fesen, Blair &
Kirshner (1985) for the case of G180.0-1.7) and also in WR
nebula (Esteban et al. 1990). Besides the spectra of the newly
discovered fine filaments acquired for this work and the
intriguing morphological structure of the filaments which are
typical of an SNR, more circumstantial evidence for our
proposition that we are actually dealing with a new SNR and not
WR ejecta can be seen in Fig. 2 where the X-ray sources 1RXS
J130924.1-655355 and 1RXS J130551.0-655117 are situated in
close proximity to the fine Hα
filaments (see later discussion).
Figure 2. The radio image of the Θ Mus nebula region from
the PMN survey data at 4.85 GHz. The inset white shape shows the
position and extent of the H
α
filament where the spectra were taken.
Due to the low resolution of the PMN survey (~5’), the optical filament
is not recognized as a separate feature in the radio. However, the overall
similarity in the form of the diffuse nebula between this radio image and
the optical ( H
α
) image shown on Fig. 2 is clear.
Our spectra of the fine optical filaments are compared
with what may (or may not) be published spectra of the same
filaments of Θ Mus in de Castro & Niemela (1998).
Unfortunately, the exact RA/DEC values of their slit position for
their spectra of the putative emission and knots from Θ Mus
were not provided. However, it is clear that at least one of them
3
was taken very close (1.5’ south-east) to where spectra have
been taken for this work since the size of the individual
filaments is typically ~10’. The basis for the spectral comparison
were [S II] 6717 and 6731Å and their ratio against Hα. de
Castro & Niemela (1998) used the diagnostic diagram of [S II]
6717/6731Å as a function of log (Hα/[S II]) from Sabbadin,
Minello & Bianchini (1977) to show that spectra from their
points (knots) are actually inside the area populated by H II
regions (see Fig. 7).
We believe this is because their slits were sampling the
more diffuse emission components present. For this work spectra
targeted specifically on the newly discovered fine filaments in
the same vicinity showed the opposite: i.e. very strong [S II]
lines, with their ratio against Hα
well inside the region normally
occupied by SNR. Actually, Fig. 7 is an adapted diagram used
by de Castro & Niemela (1998) (from Sabbadin, Minello &
Bianchini (1977)) confirming strong likely SNR originated
shocks in the observed filaments. The ratio of [S II]
6717/6731Å against log(Hα
/[S II]) in this figure was only for
3
de Castro & Niemela (1998) took spectra from 7 locations in the
general area and it is slit position 3 in their work which appears closest to
our slit position.
comparison purposes as we use the generally accepted Fesen,
Blair & Kirshner (1985) criteria for SNR classification in our
work.
We also performed a check for how our spectral
observations fit into the new Frew & Parker (2009) diagnostic
diagrams log I Hα /I [N II] versus log I Hα /I [S II] where all
major nebulosity groups (SNRs, PNs, Wolf Rayet shells etc.) are
shown and generally well separated. However, the new plot
shows that there is some overlap between classes for more
extreme, evolved examples of object types. Again, however,
both spectra from Figs. 5 and 6 are located well inside in the
domain of evolved supernova remnants.
Chu & Treffers (1981) classified the scallop-shaped,
more diffuse optical nebula prominent in previous Hα
imagery
and seen in better context in Fig. 2 as a radiatively excited
nebula of the R
s
type (Chu 1981) which means that it is a shell
structured H II region. On the basis of images of this area from
the low-quality emission line survey of Parker, Gull & Kirshner
(1979), Chu & Treffers (1981) concluded that this nebula is
embedded in a more general H II region, though it is not
mentioned which H II region. The only one found in the
literature that covers this area is the large H II region G305.1-
1.9 (size 360
×
135 arcmin) as listed in the comprehensive
review of optical Galactic H II regions of Maršálková (1974).
Note that identification of this H II region came from the low
resolution Hα survey data of Dottori & Carranza (1971). Now,
the whole of this region can be seen in far greater clarity from
the AAO/UKST Hα
data of survey field HA137 (shown in Fig.1)
a which demonstrates the highly complex nature of the extensive
emission structures which far exceed the reported dimensions of
G305.1-1.9. Examining this survey, that of Parker, Gull &
Kirshner (1979) (which was the basis for the conclusions of Chu
& Treffers (1981)) and the arcminute-resolution SHASSA
survey Gaustad et al. (2001) we concluded that this could have
easily been identified as a single massive H II region from the
point of view of these low resolution observations. However,
with the availability of our high resolution SHS images (Parker
et al. 2005), G305.1-1.9, where the scallop-shaped WR shell
nebula is supposed to be embedded (according to Chu & Treffers
(1981)), we show that this area can not be accepted as a simple
H II region. The SHS data (shown in Fig. 2 and partially at large
scale in Fig. 1) present this area as an extremely large complex
area of emission with a concentration of different objects as
might be expected for areas close to the Galactic mid-plane (at b
= -1.9). Some separated emission clouds could be individual H II
regions. Furthermore, we failed to find in the literature any
spectral observations of the 360
×
135 arcmin G305.1-1.9
emission cloud to accept it as a single, large H II region.
Spectroscopically, the spectra of H II regions (low
excitation photo-ionised nebulae) are usually easily
distinguished from SNRs, Wolf-Rayet shells and other nebulous
objects including all except very low excitation PNe (due to
different physical processes; see examples in Frew & Parker
2009). We conclude that accepting an area as a H II region only
on the basis of low resolution optical imaging and without
spectral confirmation is not satisfactory.
Figure 4. High resolution images of the area of Θ Mus from the AAO/UKST H
α
survey field HA137 showing the detection of a network of
fine optical filaments some 10 in extent. The white lines show the position and orientation of the 2.3-m MSSSO telescope DBS slits. Spectra were obtained
on June 12, 2004 at RA(2000)=13
h
07
m
38
s
and δ = -65° 54’ 37” (left image) and June 13, 2004 at RA(2000.0)=13
h
08
m
26
s
and δ = -65° 56’ 06” (right
image).
Table 1. Spectral observation log for the various optical components of nebulae in the vicinity of WR 48. All exposure were of 1200 seconds.
Object Telescope Date Grating Spectral Slit Slit
(nebula) (lines mm
-1
) range (Å) RA Dec.
WR 48 2.3-m 12/06/2004 600
a
3700-5500 13 07 38 -65 54 37
2.3-m 12/06/2004 1200 6100-6800 13 07 38 -65 54 37
2.3-m 13/06/2004 600 3700-5500 13 08 26 -65 56 06
2.3-m 13/06/2004 1200 6100-6800 13 08 26 -65 56 06
a
For the 600 and 1200 lines mm
-1
gratings the rms error in the dispersion solution (in Å) was between 0.05Å and 0.02Å and the relative
percentage error in the flux estimate was between 10% and 17% for the 600 lines mm
-1
grating and ~20% for the 1200 lines mm
-1
grating.
We also re-evaluate the previously identified shell
around the Θ Mus Wolf-Rayet star and the Chu & Treffers
(1981) classification in the context of our deep, high resolution
Hα imagery. Inspection of the derived quotient images, confirms
the existence of the previously noted bright, extended scallop-
shaped arc to the south-west of Θ Mus (see Fig. 2). Therefore, if
we define a shell around a Wolf-Rayet star as comprised of ''arcs
of nebulosity centered on and ionized by the Wolf-Rayet star"
(Chu (1981) and references therein) and accept from Chu,
Treffers & Kwitter (1983) that ''a WR ring nebula is an
identifiable symmetric nebula around a WR star", then the
observed and clearly connected south-west extension to the
putative Θ Mus shell does not fit with the original morphological
classification and is not what we see around other Wolf-Rayet
stars (which can exhibit one or more rings). This conclusion is
based on examination of the morphological structures of Wolf-
Rayet shells in Miller & Chu (1993), Marston (1995) or Gruendl
et al. (2000) as examples, where it is clearly seen that the shells
are more or less in a circular or oval coherent form. In the low
resolution, but high sensitivity SHASSA survey (Gaustad et al.
2001), even though not as clearly seen as in the SHS survey, the
bright, south-west extensions from the putative WR shell can
also be recognised. We conclude that the low sensitivity of
previous surveys gave rise to the scallop-shaped nebula being
incorrectly associated with the Wolf-Rayet star Θ Mus (WR 48).
Unfortunately, de Castro & Niemela (1998) took spectra of the
emission cloud south-east of the scallop-shaped nebula
(emission seen in the lower left corner of Fig. 2) but not on the
purported partial ring itself (see their position 5). The resultant
spectra is typical for a H II region. Although, de Castro &
Niemela (1998) claimed that this cloud belongs to a different
component of the Θ Mus shell, it is very hard to validate their
connection of this emission with the scallop-shaped nebula. On
the other hand, Heckathorn, Bruhweiler & Gull (1982) claimed
that, according to their imaging, in the scallop-shaped nebula [S
II] (and Hβ) are not present. This is in contradiction with Fesen,
Blair & Kirshner (1985) where some value of [S II] should be
present (at least [S II]/Hα <0.5) to define and separate the
emission of H II regions from, say, SNRs and PNs.
The nebula structures south of Θ Mus star are part of a
very complex area of diffuse and filamentary emission and
cannot solely be explained as resulting from ionised WR star
ejecta. Indeed, cht81 concluded from kinematic and photometric
distance estimates for the star itself (1.3 and 0.9 kpc
respectively), from the physical diameter of the main diffuse
partial shell 40 arcmin to the south (11 pc at the distance of 1
kpc), via determination of the expansion velocity of 7 km s
-1
,
and from the resultant lifetime of the nebula of about 10
6
years,
that this could be in fact a stellar wind nebula or supernova
remnant. This conclusion follows as the lifetime of the nebula is
greater then the estimated lifetime of star's Wolf-Rayet phase.
Table 2. Characteristic spectral lines for the first slit position in the vicinity of Θ Mus.
June 12, 2004
Line λ F(λ) Line λ F(λ)
(Å) Hβ =100 (Å) Hα =100
[O II] 3727 1073 [O I] 6300
[Ne III] 3867 15 [O I] 6364
He+H 3888 19 [N II] 6548 31
[Ne III] +H 3970 10 Hα 6563 100
Hδ 4101 24 [N II] 6583 106
Hγ 4340 42 [S II] 6717 47
Hβ 4861 100 [S II] 6731 39
[O III] 4959 98
[O III] 5007 271
F( Hβ )=2.62
×
10
-14
erg cm
-2
s
-1
Å
-1
F( Hα )=1.40
×
10
-14
erg cm
-2
s
-1
Å
-1
[N II] / Hα =1.37
[S II] / Hα = 0.86
[S II] (6717/6731 = 1.20
June 13, 2004
Line λ F(λ) Line λ F(λ)
(Å) Hβ =100 (Å) Hα =100
[O II] 3727 2921 [O I] 6300
[Ne III] 3867 168 [O I] 6364
Hδ 4101 [N II] 6548 18
Hγ 4340 50 Hα 6563 100
Hβ 4861 100 [N II] 6583 68
[O III] 4959 232 [S II] 6717 40
[O III] 5007 740 [S II] 6731 31
F( Hβ )=4.69
×
10
-16
erg cm
-2
s
-1
Å
-1
F( Hα )=1.39
×
10
-13
erg cm
-2
s
-1
Å
-1
[N II] / Hα =0.86
[S II] / Hα =0.71
[S II] (6717/6731)=1.27
Figure 5. Flux calibrated 1-D blue (top image) and red (bottom
image) spectra for the slit position from the June 12, 2004 observation. The
high [S II] / Hα ratio of 0.9 would classify this object as being a likely
SNR. In the blue spectrum, the strongest lines are of [O III] at 5007 and
4959Å and especially [O II] at 3727Å. Several Balmer lines are also
recognized together with [Ne III] at 3867Å, He+H at 3888Å and [Ne III]
+H at 3970Å sometimes seen in the spectra of SNRs.
Figure 6. Flux calibrated 1-D blue (top image) and red (bottom
image) spectra for the slit position from June 13, 2004. At this slit position,
[S II] / Hα =0.7 was obtained, again consistent with that from an SNR. In
the blue part of the spectrum, [O III] at 5007 and 4959Å are seen together
with extremely strong [O II] at 3727Å which is not expected for a H II
region.
Figure 7. The diagnostic diagram [S II] 6717/6731 versus
log H
α
/[S II] taken from de Castro & Niemela (1998) (originally from
Sabbadin, Minello & Bianchini 1977), which was used by them to
classify Θ Mus as possessing an external WR nebula exhibiting H II
region characteristics. However, the spectra of two newly discovered
fine filaments obtained for this work fall clearly inside the area occupied
by SNRs (rotated triangles). It is possible that the original spectra,
lacking the high resolution quality imagery with which to select regions
for study, were taken across the diffuse emission regions and not on the
fine filaments now clearly visible for the first time in the high sensitivity,
high resolution SHS data. Unfortunately, precise slit positions from the
previously published spectra are not provided but this could easily
explain the differences in spectral signatures observed.
It is also possible that the many fine filaments in the
region and the adjacent much more diffuse scallop-shaped
nebula (see Fig. 1), although different morphologically, are
interacting. The filamentary structures, which are very
reminiscent of those associated with optical SNR detections, but
also seen in Dufour, & Buckalew (1999) as part of Wolf-Rayet
nebula, are first found immediately to the east and west of the
star but also occur further south and west in a broad sweep of
filaments. These certainly seem to follow the general curvature
of the much more diffuse broad nebulosity arc before then
extending further west. Perhaps this is happening as a
consequence of different processes of nebula in this area and the
possible SNR what is implied by our spectral observations.
3.2 Proposed new optical components of a new
Galactic SR G304.4-3.1
The fine SHS structures in Fig. 2 and Fig. 4 seen for
the first time in such detail south of Θ Mus have given us the
opportunity to more clearly resolve the various morphologically
distinct filaments and diffuse emission structures that appear to
be present and which most probably arise from different physical
processes. We believe that the existing multi-wavelength optical,
radio and X-ray data now available, together with our new
spectral evidence supports the notion that these fine filaments
and associated spectra indicate the presence of a previously
unrecognised SNR separate from the diffuse emission regions.
There is no doubt that the filaments (see also Hester (1987)),
seen on Fig. 4 as "ropes" (another possibility are thin "sheets"
seen edge-on) are typical for the shock front of supernova
remnants.
Further, the work of Marston (1996) uncovered several
large IRAS 'bubbles' in the area of known Wolf-Rayet stars,
including a bubble in the Θ Mus region which is clearly seen in
the IRAS 60µm image presented in Fig. 8 (left plate)
4
with
marked approximate position of Hα filaments from Fig. 4.
Keeping in mind that the IRAS 60µm and PMN 4.85 GHz are
both low resolution surveys (2 and 5 arcmin typical resolution
respectively) we overlaid PMN contours over the IRAS 60µm
bubble (right on Fig. 3.2) and found an excellent positional
match on the south-east side of the image between the IRAS and
PMN data
5
. Although the PMN signal is fragmented, a match
between emissions can also be noticed on the south-west side.
This identification of an IR bubble and almost the same
component in the radio can also support our proposal that this is
a case of a new supernova remnant. Regarding the structure of
the radio emission at 4.85 GHz, shown on Fig. 8, it is in
accordance with our analysis of many Galactic SNRs in the
PMN survey (partially elaborated in Stupar et al. (2005)) and
confirms that this morphological structure of fragmented radio
signals (often very little over the level of noise) is very common
between known but senile SNRs registered in the PMN. IR
studies of SNRs are mostly connected with highly evolved (and
large) remnants with weak radio detection expected due to low
shock velocities (Saken, Fesen & Shull (1992) and references
therein). One more item of support comes from the presence of
two X-ray sources (Fig. 2) where 1RXS J130551.0-655117 is
nearest to the center of this new proposed SNR (see Fig. 8)
which we estimate to be around R.A.(2000)=13
h
05
m
31
s
and
δ = -65° 55’ 47” or at G304.4-3.1. The arcuate filaments from
Fig. 4 are partially opened towards the X-ray source 1RXS
J130924.1-655355. However, this source cannot easily be
accepted as the centre of this newly proposed SNR due to their
small extension of ~10 arcmin for these filaments compared with
the overall size of G304.4-3.1 (~1.7°; see Fig. 8). If there is no
connection between these X-ray sources and G304.4-3.1, some
link with the pulsar J1306-6617 (Fig. 8) is possible, though the
pulsar is not close to the apparent centre of the SNR due to the
expected (proper/transverse) motion of the star. Furthermore,
very old SNRs seen in the optical (mostly Hα), are usually
fragmented inside their radio borders (if any radio emission is
detected) in different forms, from filaments to emission clouds,
and are probably regularly shaped by local processes within the
ISM. There is clear similarity between these filaments, (east and
south-east from Θ Mus, see Fig 2) and randomly distributed
optical filaments in other SNRs, e.g. G279.0+1.1 (see Stupar &
Parker 2009), inside ~2° radio border, and G332.5-5.6 (Stupar et
al. 2007c) inside ~30’ radio border.
A systematic investigation is required to get the final
conclusion about the true nature of this complex of nebular
structures. However, our new data do support re-classification of
the extensive fine, newly detected optical filaments as resulting
from an SNR. More optical spectra for additional components of
the diffuse nebula and the fine Hα filaments are required. The
nature of the X-ray source 1RXS J130551.0-655117 (and 1RXS
J130924.1-655355) should also be examined together with the
relevance of the nearby pulsar J1306-6617 and the possible
confirmation of non-thermal nature of the weak radio detections.
All this give additional weight to the proposal for an SNR in the
region which we believe we have detected in the form of fine
optical Hα filaments, confirmatory optical spectra as well
detection in the IR and radio regimes.
4
This bubble structure can also be noticed at the other IRAS
wavelengths.
5
Unfortunately, a check in the SUMSS 843 MHz data showed that only
part of the examined area is covered in this survey.
Figure 8. IRAS 60µm image showing a large dust hole around the Θ Mus area (left image) with the approximate position of the Hα filaments
from Fig. 4 indicated. The right image is also the IRAS 60µm image but overlaid with PMN 4.85 GHz radio contours (up to 0.08 Jy beam
-1
). Although the
signal is fragmented, on the south-east side there is a clear match between the IRAS and PMN signals as well as partial match of radio emission on the
south-west side of the IRAS bubble. One can notice the position of Θ Mus close to the edge of the shell. If we really have a Wolf-Rayet shell, Θ Mus
should be, if not in the centre of the shell, then close to the centre which is not the case.
For now this proposed new remnant, G304.4-3.1, is
only seen in the radio in the PMN survey at 4850 MHz. Clearly
establishing non-thermal emission from this remnant is the best
way of confirmation. However, in such case as this, when the
remnants are old and evolved, they are usually highy fragmented
in the radio and mixed in with background noise and not easily
recognised. Therefore, we need a range of high sensitivity radio
frequency observations to confirm non-thermal emission. It is
clear from Fig. 4 that these fine filaments are completely
consistent with those seen in other evolved supernova remnants
(see previous discussion) but not to young SNRs, where [S II]
ratio of 6717/6731Å should be ~ 0.5 and not for low density
~ 1.2 what follows from our spectral observations and what is
common for evolved SNRs.
3.3 Possible optical jets from Θ
ΘΘ
Θ Muscae
During examination of the complex nebulae around
the Θ Mus star from the high resolution SHS images, unusual,
low surface brightness emission spikes were seen emanating
from the star as directly opposing, quite well collimated "jets"
with narrow cone angles in approximate SE and NW directions.
An enlarged 5×5 arcminute quotient image of the full resolution
data centred on Θ Mus of Hα over SR is shown in Fig. 9 which
illustrates this clearly. The NW jets (four relatively strong and
one lower in brightness) have a radial maximal extension some
140 arcseconds from the centroid of the stellar image though
there is a suspicious drop off in jet intensity at the edge of the
diffraction halo for the NW emission though the SE jet extends
past the main diffraction halo of the star before the intensity
drops off. The extension of the three SE jets is shorter, only
some 78 arcseconds with a slightly broader opening cone angle.
There are also a couple of weaker but equally narrow jets at PA
of ~20° in an essentially easterly and westerly direction from
the star's centre. To clarify the uniqueness of these jets, we
checked all other southern Wolf-Rayet stars visible in the SHS,
and did not notice any similar effects. One of us (QAP) has 25
years of experience with UKST imagery and has extensively
studied the SHS and its associated quotient imaging and has not
encountered an equivalent feature before. There are several
points to appreciate here. Firstly, the Hα emission revealed in
collimated form close-in to Θ Mus is a unique feature not seen in
any other of the hundreds of stars of similar magnitude across 25
sq.degrees of the survey field in question. Secondly, of all the
stars in the field it is the WR star that is exhibiting the feature
which would be a cruel co-incidence if it is indeed some weird
photographic artifact or result of some dust speck on the
photographic detector at the star's position. Thirdly, the position
angle of the narrow emission cones has an intriguing connection
with the modeling of what is in fact a triple star system involving
Θ Mus undertaken by Hill, Moffat & St-Louis (2002). They
describe a cone shaped emission region that partly wraps around
the OB star. It is well known that disks, filaments and jets are
common with Wolf-Rayet stars (Underhill 1994) supporting
emission-line spectra but this is the first time that apparent WR
star jets have been seen optically. We tried to fit these jets in the
model of Hill, Moffat & St-Louis (2002) where they observed
dramatic variations of C III λ5696 emission-line profiles and
made geometrical models that assume that emission arises from
two regions: an optically thin spherical shell around the WR star
and a cone-shaped region with a position angle of θ=51°,
strikingly similar to the position angle values for the optical
emission cone seen. They concluded that the cone-shaped region
is the result of collision between the WR star ejecta/winds and
an OB companion star making a binary system (Θ Mus is
actually a triple system; see Sugawara, Tsuboi & Maeda 2008).
Beeckmans et all. (1982), observing the C IV line in the UV,
also found variations of mass loss from the system most
probably due to the positions of the stellar components in the Θ
Mus triple system. We need improved observations to establish
the possible connection between the Θ Mus triple system and
any mass loss seen in the form of jets in Fig. 9.
Figure 9. An enlarged quotient image (Hα divided by SR) of
the Θ Mus star as provided from the SHS data. Several narrow jet spokes
are seen to emanate from the star in almost diametrically opposing SE
and NW directions. The NW jet has an apparent extension (at the angle
of ~49°) of ~140 arcseconds from the edge of the photographic core of
the stellar image. The extension of the jet(s) on the SE side is somewhat
less at ~78 arcseconds. Faint further narrow emission rays emanating
from the central star are also seen in an approximately E-W direction.
However, we stress that it is likely that the apparent
angular jet extent may be a diffraction effect related to actual
intense, collimated Hα emission emanating very close in to the
bright star. The observed jets and their apparent angular extent
are not a true physical representation of the jets that we believe
may be present. This supposition is re-inforced by the visual
examination of two B-grade exposures of the same SHS survey
field taken weeks before on the UKST where no evidence of
these jets was seen. Although the image quality of these B-
grades was inferior the observed angular extent of the jets
implies longevity and these jets should have been visible on the
B-grades if real in a physical dimension sense. If a diffraction
effect caused by intense emission then this emission could have
been due to a specific outburst and does not appear to be present
all the time if real. Narrow-band CCD monitoring, perhaps with
an occulting disk to mask the light from the very bright star
itself, is required to establish the true nature of this intriguing
structure which appears unique amongst such stars. We are
unable to obtain such repeat Hα imaging with the same
instrumental set up on the UKST as this mode of operation is no
longer supported. We present these intriguing result in the hope
the community is able to shed light on this possible discovery.
4 COCLUSIOS
We find that the putative Θ Mus shell previously
identified in the literature is not supported by our deeper Hα
imaging and related data. Careful study reveals a complex
picture of extensive, overlapping and possibly interacting optical
emission components from the highly diffuse to the extremely
filamentary over most of the 25 sq.degree of the relevant SHS
survey field HA137, much of it associated with the previously
identified H II region G305.1-1.9 which itself is shown to be
more extensive and incoherent. This makes it difficult to
untangle the various components that are present and to
determine which, if any, of the nebulae are associated with the
WR star itself. Filaments and cloud emissions could originate
from an SNR or the WR star or, in the case of the extensive
diffuse components part of a major H II region complex. Our
spectral observations confirm the presence of shocked material
from the newly discovered fine emission filaments which is
common only in supernova remnants though also present in
some WR nebulae and evolved PN. Based on our new data we
conclude that Θ Mus (WR 48) should be reclassified as a Wolf-
Rayet star without a clearly identified associated WR nebula. We
also propose the existence of a new, optically detected SNR in
the final dissipation phase and have compiled corroborating
fresh spectroscopic, morphological, radio, X-ray and infrared
evidence to support this newly uncovered Galactic SNR G304.4-
3.1 which is at least 45 arcminutes across.
For the first time, we have also found evidence of
collimated emission cones from Θ Mus apparently in agreement
with the modeling of Hill, Moffat & St-Louis (2002). These
could be a cruel photographic artifact or a temporary
phenomenon. Regular monitoring of the star and additional CCD
narrow-band data is required to confirm the veracity of this
intriguing possibility.
ACKOWLEDGMETS
We are grateful referee A.P. Marston for valuable
comments and suggestions that have significantly improved the
paper, and to David Frew for comments given during the
preparation of this paper. We are also thankful to the Mount
Stromlo and Siding Spring Observatory Time Allocation
Committees for enabling the spectroscopic follow-up to be
obtained. We thank the WFAU of the Royal Observatory
Edinburgh for the provision of the SHS data on-line and
inspection of original SHS films at Plate Library of the Royal
Observatory Edinburgh. We also thank Sue Tritton and Mike
Read of the WFAU for the visual checking of the B-grade
exposures of the Hα survey field containing Θ Mus.
REFERECES
Beeckmans F., Macchetto F., Grady C.A., van der Hucht K.A.,
1982, in Wolf-Rayet stars: Observations, physics, evolution,
IAU Proc. 99, edit. C.W.H. de Loore and A.J. Willis, Dordrecht,
D. Reidel Publishing Co., p. 311
Blair W.P., Ravi S., Raymond J.C., Knox S.L., 1999, AJ, 118,
942
Cantó J., 1981, in Kahn F.D. ed., Astrophys. Space Sci. Library,
Investigating the Universe. Reidel, Dordrecht, p. 95
Chu Y.-H., 1981, ApJ, 249, 195
Chu Y.-H., Treffers R.R., 1981, ApJ, 250, 615
Chu Y.-H., Treffers R.R., Kwitter K.B., 1983, ApJS, 53, 937
Condon J.J., Griffith M.R., Wright A.L., 1993,AJ, 106,, 1095
Cram L.E., Green A.J., Bock D.C.-J., 1998, 15, 64
de Castro G.C., Niemela V.S., 1998, MNRAS, 297, 1060
Dottori H., Carranza G., 1971, Ap&SS, 13, 180
Dufour R.J., Buckalew B., 1999, Proc. of the 193rd symposium
of the International Astronomical Union, Eds. Karel A. van der
Hucht, Gloria Koenigsberger, and Philippe R. J. Eenens. San
Francisco, Calif.: Astronomical Society of the Pacific, p.350
Esteban C., Vilchez J. M., Manchado A., Edmunds M.G., 1990,
A&A, 227, 515
Fesen R.A., Blair W.P., Kirshner R.P.,1985, ApJ, 292, 29
Frew D.J., 2008, PhD Thesis, Macquarie University, Sydney
Frew D.J., Parker, Q.A., PASA, in press
Gaustad J.E., McCullough P.R., Rosing W., Buren D. Van,
2001, PASP, 113, 1326
Ghavamian P., Raymond J., Smith R.C., Hartigan P., 2001, ApJ,
547, 995
Gruendl R. A., Chu Y.-H., Dunne B. C., Points S. D., 2000, AJ,
120, 2670
Heckathorn J.N., Gull T.R., 1980, BAAS, 12, 458
Heckathorn J.N., Bruhweiler F.C., Gull T.R., 1982, ApJ, 252, 30
Hester J.J., 1987, ApJ, 314, 187
Hill G.M., Moffat A.F.J., St-Louis N., 2002, MNRAS, 335,1069
Johnson H.M., Hogg D.E., 1965, ApJ, 142, 1033
Marston A.P., Yocum D.R., Garcia-Segura G., Chu Y.-H.
Marston A.P., 1995, AJ, 109, 1839
Marston A.P., 1996, AJ, 112, 2828
Marston A.P., 1997, ApJ, 475, 188
Maršálková P., 1974, ApSS, 27, 3
Mathewson D.S., Clarke J.N., 1973, ApJ, 180, 725
Miller G.J., Chu Y.-H., 1993, ApJS, 85, 137
Miszalski, B, Parker, Q.A, et al., 2008, MNRAS,
Moffat A.F.J., Seggewiss W., 1977, A&A, 54, 607
Pannuti T.G., Schlegel E.M., Lacey C.K., 2007, AJ, 133, 1361
Parker R.A.R., Gull T.R., Kirshner R.P., 1979, NASA-SP434
Parker Q.A., Phillipps S., Pierce M.J., Hartley M., Hambly N.C.,
Read M.A. et al., 2005, MNRAS, 362, 689
Parker Q.A., Acker A., Frew D.J. et al., 2006, MNRAS, 373, 79
Pierce M.J., Frew D.J., Parker Q.A., Köppen J., 2004, PASA, 21,
334
Sabbadin F., 1977, A&A, 54, 915
Sabbadin F., Minello S., Bianchini, A., 1977, A&A, 60, 147
Saken J.M., Fesen R.A., Shull J. M., 1992, ApJS, 81, 715
Shull J.M., Fesen R.A., Saken J.M., 1989, ApJ, 346, 860
Smith L.F., Batchelor R.A., 1970, AuJPh, 23, 203
Stupar M., Filipović M.D., Parker Q.A., White G.L., Pannuti
T.G., Jones, P. A., 1995, Ap&SS, 307, 423
Stupar M., 2007, PhD Thesis, Macquarie University, Sydney
Stupar M., Parker Q.A., Filipović M.D., 2007a, MNRAS, 374,
1441
Stupar M., Parker Q.A., Filipović M.D., 2007b, AAO
Newsletter, 112, 12
Stupar M., Parker Q.A., Filipović M.D., Frew D. J., Bojičić I.,
Aschenbach B., 2007c, MNRAS, 381, 377
Stupar M., Parker Q.A., Filipović M. D., 2008, MNRAS, 390,
1037
Stupar M., Parker Q.A., 2009, MNRAS, 391, 179
Underhill A.B., 1994, ApSS, 221, 383
Sugawara Y., Tsuboi Y., Maeda Y., 2008, A&A, 490, 259
van der Hucht K.A., 2001, NewAstr. 45, 135V
van der Hucht K.A., 2006, A&A, 458, 453