Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://mavr.sao.ru/hq/balega/PUBL_BAL/PUB_2006/2005_k3-50a_dina4.ps Äàòà èçìåíåíèÿ: Tue May 5 16:20:23 2009 Äàòà èíäåêñèðîâàíèÿ: Tue Aug 18 07:21:04 2009 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: http astrokuban.info astrokuban

MPIfR
High resolution near­infrared interferometric imaging
of the ultracompact H II region K3­50 A
S. Kraus (1) , Y. Y. Balega (2) , K.­H. Hofmann (1) , T. Preibisch (1) , G. Weigelt (1)
(1) Max­Planck­Institut f˜ur Radioastronomie, Bonn, Germany
(2) Special Astrophysical Observatory, Karachai­Cherkesia, Russia
Abstract
We present K ­band bispectrum speckle interferometry of the ultra­
compact H II region K3­50A. Our image resolves the central 1 ## â 1 ##
region into at least 7 point­like objects. We find K ­band counter­
parts for all but one of the N­band sources discovered by Okamoto et
al. (2003), and there are additional K ­band sources which were un­
resolved in the N­band images. Our reconstructed image also reveals
the fine­structure of the cone­shaped nebulosity extending to the south.
The brightest K ­band source is located exactly at the tip of the cone­
shaped nebulosity. The nebula shows several arcs and the orientation of
its main axis agrees very well with the direction of the CO outflow from
K3­50A. This nebulosity therefore very likely represents the clumpy inner
surface of a partially evacuated cavity excavated by the strong outflows.
Observations and data analysis
The speckle interferograms of K3­50 A were recorded on 24 Septem­
ber 2002 with the SAO 6 m telescope in Russia using a K ­band filter.
The exposure time per frame was 300 ms. Our data set consists of 980
speckle interferograms of K3­50 A and of the unresolved reference star.
The seeing was # 1.1 ## (FWHM). An image with a resolution of 106 mas
(Fig. 1) was reconstructed using the bispectrum speckle interferometry
method (Weigelt 1977; Lohman et al. 1983).
Results
Our speckle image shows a bright, point­like source at the northern tip
of a cone­like nebula extending towards the south, in which a number
of fainter, point­like sources (in the following denoted with the numbers
1 to 10) are embedded.
Comparing our image with the 0.4 ## resolution seeing­limited K­band
image presented by Okamoto et al. (2003; their Fig. 3 and 4) allows a
reliable cross­identification of the sources. Our brightest source 1 cor­
responds to the brightness peak in the seeing­limited K­band image of
Okamoto et al. (2003), and our sources 8 and 10 can be reliably iden­
tified with point­like sources in their image. The rest of our sources,
i.e. numbers 2 to 7 and number 9, are not resolved in the K­band
images of Okamoto et al. (2003).
Since Okamoto et al. (2003) suggested that the brightest K­band
source corresponds to the 10 µm source OKYM 3, we assume here
that our source 1 is positionally coincident with OKYM 3. This cross­
identification also allows us to search for counterparts of the N­band
sources in our K­band speckle image (see Table 1).
1" 2.09 µm
Figure 1: Pseudocolor representation of our K # ­band image of
K3­50 A, reconstructed with the bispectrum speckle interferometry
method. This image has a resolution of 106 mas, the field of view
is # 5.4 ## â 5.4 ## . North is up and east is to the left.
Only for the 10 µm source STHO 1 we can see no counterpart in
our K­band image. Our speckle image directly confirms the presence
of multiple sources in the core of K3­50 A, and confirms the indication
from the N­band data that OKYM 2 and OKYM 3 are separate sources.
We can use the observed magnitude di#erences to check the suggested
spectral types of the main sources in the K3­50 A core (Okamoto et
al. 2003): if we assume that the relative K # ­band magnitudes of the
sources are roughly representative of their relative luminosities, we find
that the 4.5 times smaller flux of the second brightest source is consis­
tent with a spectral type # B0 if the brightest source is of spectral type
# O8. Therefore, the spectral type estimates of the ionizing sources in
K3­50 A derived by Okamoto et al. (2003) are supported by our data.
It is also interesting to compare the K # ­band and N­band fluxes of
the sources. Inspection of the N­band spectra (Okamoto et al. 2003)
of OKYM 3 (our source 1) and OKYM 4 (our source 10) shows that
these two sources have quite similar fluxes at # 10 µm. In the K ­band,
however, OKYM 4 is more than 4 magnitudes fainter than OKYM 3.
If the intrinsic properties of both objects were identical, this di#erence
in K # - N color would indicate that OKYM 4 su#ers from about 140
mag more optical extinction than OKYM 3. Alternatively, OKYM 4
may also be surrounded by a much larger amount of warm (T # 500
K) circumstellar material than OKYM 3, causing a strong mid­infrared
excess.
This might suggest that source 10 is surrounded by a much larger
amount of warm circumstellar material than source 1.
source # RA # Dec # K Ident. with Okamoto et al. (2003)
number [mas] [mas] [mag] K­band N­band
1 0 0 0 yes OKYM 3
2 135 ­108 1.74 saturated/unresolved
3 13 ­256 1.65 saturated/unresolved OKYM 2
4 405 ­283 3.13 unresolved
5 567 ­445 2.83 unresolved OKYM 1
6 742 ­513 3.36 unresolved
7 297 ­567 3.84 unresolved
8 ­216 ­2146 2.61 yes
9 202 ­1674 3.65 ?
10 1593 ­1431 4.08 yes OKYM 4
Table 1: Relative positions and photometry for the point like
sources in our K # ­band bispectrum speckle image. The J2000 coor­
dinates of source 1 determined from the 2MASS K­band image are
R.A. 20:01:45.7, Dec +33:32:44
The di#use emission
The main axis of the di#use cone­shaped nebulosity surrounding the
point­like sources is P.A.# 175 # . This agrees quite well with the position
angle of P.A.=160 # that was found for the high­velocity bipolar radio
outflow by DePree et al. (1994). The full opening angle is 90 # . Other
very interesting features are the straight edges of the di#use nebulosity
and the X­like structure around the brightest source. The direction of
the short northern extensions corresponds well with the directions of
the edges in the southern direction. This structure therefore seems to
represent a well defined cone with source 1 at its center.
These results suggest the following interpretation of the nebulosity:
what we see is the clumpy inner surface of a partially evacuated cav­
ity in the circumstellar envelope around the central source, which has
probably been excavated by the strong outflow.
The fact that the southern part of the cavity is much brighter than
the norther part, suggests that the southern outflow structure is tilted
towards the line­of­sight and that we look ``into'' the southern cavity.
The di#use emission has a very clumpy structure with several promi­
nent knots and some rather sharp and straight features. We note that
the general morphology of the di#use structures in our K­band speckle
image is very similar to that seen in the N­band image of Okamoto et
al. (2003).
Figure 2: Greyscale representation of the central part of our K­
band speckle image with annotation of the objects discussed in the
text. The red crosses mark the 10 µm sources detected by Okamoto
et al. (2003), the red circles give the positional uncertainty. The
blue crosses mark the point­like sources in our K­band speckle im­
age for which we performed aperture photometry. The direction of
the CO outflow is indicated by the arrow.
One of the most peculiar features is the V­shaped structure at the
southern end. The western part of the V­shaped structure corresponds
to the northern part of ``arc 1'' defined by Okamoto et al. (2003). In our
speckle image this part of the arc appears very sharp and remarkably
straight.
It is also interesting to note that all the point­like K­band sources in
our image are located within (or very close to the edges of) the southern
cavity. As it seems quite unlikely that all members of a stellar cluster as­
semble in a cone­like volume and not in a roughly spherical distribution
around the central source, this may imply that we can detect only those
objects that are located in the cavity, because the extinction along the
line­of­sight through the cavity is much lower than through other parts
of the cloud.
Summary and conclusions
Our high­resolution imaging resolves the central K ­band emission of
K3­50 A into 10 point­like sources. The brightest K'­band source dom­
inates the near­infrared emission and since it is centered on the bipolar
cavity structure, it also seems to be the dominant driving source of
the molecular outflow. However, the magnitudes of the other point­like
sources indicate that some of them are also rather massive (probably
early B­type) stars. This suggests that K3­50 A is not dominated by a
single massive star, but by a small cluster of massive to intermediate­
mass stars. Our results demonstrate the importance of high spatial
resolution observations for revealing the true nature of massive YSOs.
References
[1] DePree, C.G., Goss, W.M., Palmer, P., Rubin, R.H. 1994, ApJ428670
[2] Hofmann, K.H., Balega, Y.Y., Preibisch, T., Weigelt, G. 2004, A&A417981
[3] Lohmann, A.W., Weigelt, G., Wirnitzer B. 1983, Appl. Opt.224028
[4] Okamoto, Y.K., Kataza, H., Miyata, T., et al. 2003, ApJ584368
[5] Weigelt, G. 1977, Opt Commun.2155
Poster presented at the IAU Symposium 227, May 16­20 2005, Acireale, Italy
``Massive Star Birth: A Crossroads of Astrophysics''
contact e­mail: skraus@mpifr­bonn.mpg.de