NGC3603-a

Lots of Small Stars Born in Starburst Region

Decisive Study of NGC 3603 with the VLT and ISAAC

An international group of astronomers [1] has used the ESO Very Large
Telescope (VLT) at Paranal (Chile) to perform unique observations of
an interstellar nebula in which stars are currently being born.

Thanks to the excellent imaging properties of the first of the four
8.2-m VLT Unit Telescopes, ANTU, they were able to demonstrate, for
the first time, the presence of large numbers of small and relatively
light, new-born stars in NGC 3603, a well-known "starburst" region in
the Milky Way Galaxy.

Until now, it has only been possible to observe brighter and much
heavier stars in such nebulae. The new observations show that stars of
all masses are being born together in the same starburst event, a
fundamental result for our understanding of the very complex process
of star formation.

Background of the project

The present research programme was granted observing time with VLT
ANTU in April 1999. Its general aim is to investigate collective,
massive star formation, in particular the coalescence of high- and
low-mass stars in the violent environments of starburst regions. These
are areas in which the processes that lead to the birth of new stars
are particularly active just now.

Several fundamental questions arise in this context. A very basic one
is whether low-mass stars form at all in such environments. And if so,
do they form together with the most massive stars in a starburst event
or do they form at different times, before or after or perhaps on
different timescales? Are low-mass stars born with any "preferred"
mass that may possibly give further clues to the ongoing processes?

All of this is most important in order to understand the detailed
mechanisms of star formation. Most current theoretical scenarios
explain how single stars form in an isolated, contracting gas cloud,
but most stars in the Universe did not form in that simple way. Once
some massive stars have formed in some place and start to shine, they
will quickly affect their environment, but how much? At this moment,
nobody knows for sure what determines the actual masses of individual
stars that are formed in a very massive and turbulent gas cloud,
although some ideas can now be tested with these new observations.

The NGC 3603 region

The new VLT observations are the key part of a larger research
programme that also includes observations of the stellar cluster in
the famous Tarantula Nebula in the Large Magellanic Cloud (LMC) with
the NICMOS instrument on the Hubble Space Telescope (HST), as well as
adaptive optics observations with ground-based telescopes of more
quiescent, star-forming regions in the Galaxy.

However, the team considered the starburst region NGC 3603 as the best
target for this kind of investigation. It is situated in the far
southern constellation Carina (The Keel) and can only be observed from
the South. Moreover, such a study has to focus on the densest part of
the cluster that can only be resolved with a very sensitive infrared
(IR) instrument under the best seeing conditions. The VLT ANTU
telescope and the multi-mode ISAAC facility are ideally suited for
this purpose.

NGC 3603 is located in the Carina spiral arm in the Milky Way galaxy
at a distance of about 20,000 light-years (6 - 7 kpc).  It is the only
massive, galactic "HII-region" (so denoted by astronomers because part
of its hydrogen is ionized) in which a central cluster of strongly
UV-radiating stars of types "O" and "B" that ionize the nebula can be
studied at visual and near-infrared wavelengths. This is because the
line-of-sight is reasonably free of dust in this direction; the
dimming in near-infrared radiation due to intervening matter between
the nebula and us is only about a factor of 2 (contrary to 80 in
visible light).

The total mass of the hot O- and B-stars in NGC 3603 is over 2000
solar masses. Together, the more than fifty heavy and bright O-stars
in NGC 3603 have about 100 times the ionizing power of the well-known
Trapezium cluster in the Orion Nebula. In fact, the star cluster in
NGC 3603 is in many respects very similar to the core of the large,
ionizing cluster in the approx. eight times more distant Tarantula
Nebula in the LMC.
  
 Caption: ESO PR Photo 38a/99 is a composite "false-colour" infrared
 image of the starburst region NGC 3603 that is composed from three
 exposures obtained with the multi-mode ISAAC instrument at the
 Nasmyth focus of the first 8.2-m VLT Unit Telescope (ANTU) in April
 1999. Three near-infrared filters were used, Js (wavelength 1.24
 µm; here reproduced in blue), H (1.65 µm; green) and Ks (2.17
 µm; red). The intensities are scaled in logarithmic units and the
 field measures 3.4 x 3.4 arcmin2, or about 20 x 20 light-years2 at
 the distance of the nebula. North is up; East to the left.

 The central cluster is the densest concentration of massive stars
 known in the Milky Way (this area is enlarged in ESO PR Photo 38b/99;
 the field shown is about 2.5 x 2.5 light-years2). It hosts more than
 50 hot O-type stars. The brightest star in the field is the red
 supergiant IRS4; it is located about 80 arcsec NE of the
 center. About 18 arcsec N of the center are the ring nebula and the
 bipolar outflows around the blue supergiant Sher25. The photo also
 shows three proplyd-like objects [2] that have been recently
 discovered; they are similar to those seen in Orion Nebula, but 20-30
 times more extended. About 1 arcmin SSE of the central cluster are
 seen the brightest members of the deeply embedded proto cluster IRS9.

 The nebulosities to the South and West of the center appear to be red
because of strong emission in the Bracket-gamma spectral line from
hydrogen atoms at 2.166 µm.


Images of the NGC 3603 region were obtained in three near-IR filter
bands (Js, H and Ks) with the ISAAC instrument at the ANTU
telescope. The observations were made in "service" mode on April 4 - 6
and 9, 1999, during selected periods when the (optical) seeing was
equal to or better than 0.4 arcsec. This was a most essential
requirement in order to achieve sufficient angular resolution (image
sharpness) that would allow to do accurate photometric measurements of
individual stars in this crowded cluster. This particular observing
mode, during which ESO observers at ANTU kept careful track of the
actual atmospheric conditions, contributed greatly to securing the
very high quality images needed for this programme.

In view of the many comparatively bright stars in the field, the
observing strategy was to use the shortest possible exposure time
(1.77 sec) to keep the number of over-exposed (saturated) stellar
images to a minimum. As the minimum time required to stabilize the
telescope's active optics control system and guarantee the optimum
optical quality was about 1 min, thirty-four short exposures were made
at each sky position and then co-added to an effective one-minute
exposure. After each such series, the telescope pointing was offset in
a random pattern up to 20 arcsec from the center; this enlarged the
imaged sky area somewhat and facilitated the subtraction of the
infrared emission from the sky background.

The individual 1-min exposures were then very carefully co-aligned to
obtain the highest possible spatial resolution and co-added. The
resulting images cover a sky field of 3.4 x 3.4 arcmin2 with a pixel
size of 0.074 arcsec. The effective exposure times of the final
broad-band images in the central 2.5 x 2.5 arcmin2 area are 37, 45,
and 48 min in the Js, H and Ks filters, respectively.

The final step involved the computer-aided detection of the individual
stars in the frames, the measurement of their brightness as seen in
the different wavebands and hence their infrared colours. About 20,000
intensity peaks were detected in each waveband at the same pixel
location. However, after the rejection of very faint and spurious
images and recording only objects that were detected independently in
all three wavebands within the same pixel, the resulting list of
measured stellar images was reduced to 6967 objects, still a
substantial number, though.

The brightness and colours of a star are an indication of its mass and
age. By comparing the measured brightness and colours with computer
simulations, the astronomers were therefore able to deduce the numbers
of stars with different ages and masses in NGC 3603.

Detecting the low-mass stars in NGC 3603

The new VLT observations are the most sensitive ones made to date of
this densely packed starburst region. They allowed the team to
investigate in unprecedented detail the low-mass stellar population in
this area.

Although the low-mass stars in NGC 3603 are not exceedingly faint -
they are in fact about 3 magnitudes brighter than ISAAC's detection
limit - it is extremely difficult to detect them and to measure their
brightness accurately because of the enormous range of brightnesses
(more than a factor of 10,000) among the densely crowded stars in the
inner region of the cluster. Unless high angular resolution, high
optical stability and high overall sensitivity is achieved, the
fainter images of the low-mass stars will "drown" in the light of the
adjacent, much brighter stars. Only a powerful telescope/instrument
combination like ANTU/ISAAC can successfully perform such a critical
observation.

The sensitivity limit obtained - set by the requirement that a star
must be detected in all three infrared wavebands - corresponds to
about one-tenth of a solar mass for young stars (in the astronomical
sense) aged only 700,000 years, and still in the initial contraction
phase. Thus, for the first time, it was possible to reach the
necessary angular resolution and sensitivity to study a starburst
region on a star-by-star basis down to this low mass limit. For
comparison, the most sensitive observations of the more distant
Tarantula Nebula only reach down to a limit of about 1 solar mass.

A most important conclusion of this study is that there are lots of
sub-solar mass stars in NGC 3603, i.e., contrary to several
theoretical predictions, these low-mass stars do form in violent
starbursts!

The overall age of stars in the contraction phase that are located in
the innermost region of NGC 3603 was found to be 300,000 - 1,000,000
years. The counts clearly show that this cluster is well populated in
sub-solar mass stars.

The next steps

The team describes these new results in a scientific article
("Low-mass stars in the massive HII region NGC 3603 - Deep NIR imaging
with ANTU/ISAAC") that will appear in the European research journal
Astronomy & Astrophysics in December 1999. Further information about
related work on NGC 3603 is available at a dedicated webpage.

The present VLT data will now be used for continued studies during
which the limits of detection and measurement will be further pushed
by means of advanced image processing and analysis.

It will also be interesting to look further into possible variations
of the number of stars with a given mass over the observed field, not
least, to compare the new results with other ongoing studies of
different regions (although less massive), e.g. with the Hubble Space
Telescope and its infrared instrument NICMOS or with ground-based
Adaptive Optics instruments.

Notes

[1] The team consists of Bernhard Brandl (Principal Investigator;
Cornell University, Ithaca, New York, USA), Wolfgang Brandner
(University of Hawaii, Honolulu, USA), Frank Eisenhauer
(Max-Planck-Institut für Extraterrestrische Physik, Garching,
Germany), Anthony F.J. Moffat (Université de Montreal, Canada),
Francesco Palla (Osservatorio Astrofisico di Arcetri, Florence, Italy)
and Hans Zinnecker (Astrophysikalisches Institut Potsdam, Germany).

[2] Proplyd is an astronomical term that stands for "proto-planetary
disk", i.e. disks around young stars in which planets may later
form. However, although they look like the proplyds found in the Orion
Nebula, the "proplyd-like" objects in NGC 3603 are not likely to
develop into planets.



How to obtain ESO Press Information

ESO Press Information is made available on the World-Wide Web (URL:
http://www.eso.org/outreach/press-rel/). ESO Press Photos may be
reproduced, if credit is given to the European Southern Observatory.