Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.arcetri.astro.it/~lt/preprints/p45maser/p45maser.ps.gz Äàòà èçìåíåíèÿ: Tue Sep 11 16:56:44 2007 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 08:00:38 2007 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: photosphere

A&A manuscript no.
(will be inserted by hand later)
Your thesaurus codes are:
08(08.06.2; 09.13.2; 13.09.6)
ASTRONOMY
AND
ASTROPHYSICS
10.12.1996
Letter to the Editor
Detection of Young Stellar Objects with ISO
L. Testi 1;6 , M. Felli 2 , A. Omont 3 , M. P'erault 4 , P. S'eguin 4 , G. Comoretto 2 , and G. Gilmore 5
1 Dipartimento di Astronomia e Scienza dello Spazio, Universit`a degli Studi di Firenze, Largo E. Fermi 5, I­50125 Firenze, Italy
2 Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, I­50125 Firenze, Italy
3 Institut d'Astrophysique de Paris, CNRS, 98bis Boulevard Arago, F­75014 Paris, France
4 Laboratoire de Radioastronomie Millim'etrique, ENS, rue Lhomond 24, 75005 Paris, France
5 Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge, CB3 OHA, United Kingdom
6 visiting astronomer: Institut d'Astrophysique de Paris, CNRS, 98bis Boulevard Arago, F­75014 Paris, France
Received xxxx; accepted xxxx
Abstract. Preliminary 15¯m results obtained with ISO­
CAM are compared with deep K­band and H 2 O maser
observations of a 0:244 deg 2 field in the galactic plane
at l = +45 ffi . A new maser source is found to be asso­
ciated with the infrared source with the largest colour
index, (mK --m 15¯m ) ?
¸ 10, most probably a young star in
its earliest evolutionary phase. Other three objects with
(mK --m 15¯m ) ?
¸ 9, probably young stars, are also found in
the observed region.
Key words: stars: formation -- ISM: masers -- infrared:
stars
1. Introduction
Even though originally found in close association with
diffuse HII regions, most of the interstellar H 2 O masers
(up to 80%) may not be associated with them (Palla et
al. 1993). In fact, H 2 O masers are now believed to trace
the earliest evolutionary phases of massive stars, much
before the onset of an ultra compact (UC) HII region
(Codella et al. 1994; Codella & Felli 1995). This picture
has been confirmed both by high resolution radio observa­
tions and high sensitivity near infrared (NIR) observations
(Felli et al. 1996; Testi et al. 1994; Tofani et al. 1995) as
well as by molecular line and 3mm continuum high reso­
lution observations (Cesaroni et al. 1994a; 1994b; Codella
et al. 1996; Olmi et al. 1996). The NIR observations are
expected to trace not the stellar photosphere, which would
be too extincted to be detectable, but the hottest layers
of the dust cocoon in which the newborn star is embedded
(Felli & Testi 1996).
Send offprint requests to: L. Testi, testi@arcetri.astro.it
H 2 O masers are thus the ideal indicators of the
youngest massive stars which have not yet swept out their
parental cocoon, and are the best targets for the identifi­
cation of regions of recent star formation.
Unfortunately, the best studied star forming regions
are the most luminous ones and may contain several bright
early type stars, making the interpretation more com­
plex. The coexistence of objects in different evolutionary
stages, and the presence of (moderately) evolved objects
that have already strongly interacted with the surrounding
medium, make it difficult to study the youngest and lower
luminosity objects. Comparison of low resolution IRAS
and radio observations may give only global results for
the entire complex. Both high resolution and sensitivity
are needed to separate young stars with different luminosi­
ties and in different evolutionary status (Felli et al. 1996;
Hunter et al. 1995; Palla et al. 1995).
Although NIR imaging (especially at K= 2:2 ¯m) has
proved to be a useful tool to investigate these regions, mid
and far infrared high resolution observations are urgently
needed. In fact, it is in this range of wavelengths that the
bulk of the emission from the youngest massive stars is
expected. Even more important, isolated massive young
stellar objects should be studied, in order to unveil the
formation of the first generation of massive stars, without
the difficulties and the complexities that arise in regions
where other stars have already formed. When combined
with near infrared and water maser unbiased surveys,
high resolution and sensitivity ISOCAM 15¯m observa­
tions (Omont 1996; P'erault et al. 1996) which purposely
avoid the brightest IRAS sources (with F 12¯m ? 6 Jy) are
an unique tool to detect young early type stars.
In this paper we report the detection of one source
which very likely belongs to the class outlined above, to­

2 L. Testi et al.: Detection of Young Stellar Objects with ISO
gether with three other potential candidates in a region of
about ¸ 0:244 deg 2 at approximately l = +45 ffi , b = 0 ffi .
2. Observations
All the observations have been carried out in the frame of
a survey at 15¯m of the inner galaxy, the ISOGAL project
(P'erault et al. 1996). In particular the observations used in
this work are those of the field at l = +45 ffi . Both the radio
and NIR observations are part of a preparatory/followup
program for the ISO data. Here we will briefly summarize
the fundamental observational parameters.
2.1. ISO observations
The ISO observations where carried out during April 1996.
ISOCAM was used to observe a ¸ 0:244 deg 2 field around
the position l=+45 ffi b=0:0 ffi . The filter used was the 15¯m
broadband LW3 filter (passband from 12 to 18¯m), and
the pixel field of view 6 00 . After image reduction (P'erault
et al. 1996) about 700 point sources were extracted with
a preliminary photometry. The completeness limiting flux
achieved was approximately 10 mJy. The magnitude at
15¯m (m 15¯m ) has been computed assuming m 15¯m =9 for
a 5 mJy source.
2.2. NIR observations
The K­band observations of a slightly larger field than
that observed by ISOCAM were carried out during Oc­
tober 1995, using ARNICA (Lisi et al. 1996) mounted
at the TIRGO 1 telescope. ARNICA is equipped with a
NICMOS3 detector, the plate scale at the TIRGO is
¸ 0:96 arcsec=pixel. For a comprehensive description of
the instrument performance at the TIRGO see Hunt et
al. (1996a). Photometric calibration was performed by ob­
serving a set of the ARNICA standard stars (Hunt et
al. 1996b), the calibration accuracy is about 10%. The
limiting magnitude achieved in the observations (3oe; 4 00
aperture) was ¸ 15:4 (0:46 mJy).
The K­band data was astrometrically calibrated using
a large set of optical stars extracted from the Digitized Sky
Survey as described by Testi (1993). K stars positions then
were used to astrometrically calibrate the ISO images. The
present astrometric precision achieved is !
¸ 1 00 for the K­
band data and !
¸ 3 00 for the ISO data.
On the 3 rd July 1996, a small region (¸ 2 0 \Theta 2 0 ) around
the position of the H 2 O maser was observed at UKIRT 2
at K and nbL (3:4¯m). The 2:2¯m data are about two
magnitude deeper than the TIRGO data. All the sources
previously detected are confirmed, and a background of
fainter objects suggests that these images are confusion
1 The TIRGO telescope is operated by the C.A.I.S.M.I.--
C.N.R., Firenze, Italy
2 UKIRT is the United Kingdom Infrared Telescope, Mauna
Kea, Hawaii
limited at K¸ 18. In the nbL image ¸ 10 stars corren­
sponding to the brightest K sources are detected, plus a
faint object (m nbL ¸ 11:3) without a bright K counter­
part. The limiting magnitude of the 3:4¯m image is ¸ 12.
2.3. H 2 O maser observations
A new H 2 O maser was discovered during an unbiased sur­
vey of the l = +45 ffi field, which is presently on­going at
the Medicina 3 radio telescope. The observing setup, the
sensitivity and the spectral resolution of the observations
closely resemble those described by Brand et al. 1994. The
Full Width at Half Power (FWHP) of the Medicina an­
tenna at 22 GHz is ¸ 1:9 0 . Up to now ¸ 25% of the field
has been observed. Taking into account the beamwidth
and the pointing accuracy, we estimate an error of ¸ 30 00
on the maser position.
3. Results and discussion
In this section we will briefly discuss the four sources in
our field that are possible candidates for young deeply
embedded (proto­)stars. All of the sources have a colour
index (mK --m 15¯m ) greater than 9, calculated as the differ­
ence between the observed magnitudes. Such large excess
is probably due to the reprocessing of the stellar radiation
by optically thick dust cocoons. In fact, for both unred­
dened main sequence and evolved stars (giant branch and
asymptothic giant branch stars), the value of the colour
index should be in the range 0 \Gamma 3. Reddening in the galac­
tic plane cannot increase this value up to 9 even for very
distant objects.
3.1. The H 2 O maser source
In Fig. 1 the spectrum of the H 2 O maser observed at
Medicina is reported. The velocity of the two main compo­
nents (\Gamma95; \Gamma20 km s \Gamma1 ) are on the lowest side of the val­
ues expected for this galactic longitude. A ``mean'' velocity
of ¸ \Gamma57 km s \Gamma1 could be obtained for objects at large dis­
tance ( ?
¸ 15 kpc). In the Columbia 12 CO survey (Cohen et
al. 1986) and in the radio recombination line observed to­
ward the nearby HII region G45:45+0:06 (Lockman 1989)
the bulk velocity is ¸ 55 km s \Gamma1 . No H 2 O emission was
detected at this velocity (in Fig. 1 only a small part of
the spectrum is shown). In Table 1 the parameters of the
observed H 2 O maser and of the closest ISO source to­
gether with its associated K­band and 3:4 ¯m source are
reported.
Due to the extremely large number of K­band sources
in the field (¸ 34000) and the relatively large astrometric
error, a reliable association between 15¯m sources and K­
band sources is a major problem in these regions, which
is presently being studied in a thorough statistical way. In
3 The Medicina telescope is operated by the I.R.A.--C.N.R.,
Bologna, Italy

L. Testi et al.: Detection of Young Stellar Objects with ISO 3
Table 1. Parameters of the H2O maser, and of the associated
infrared source
ff ffi Fpeak F
(2000.0) (Jy) (Jy km s \Gamma1 )
22 GHz 19 : 12 : 46 10 : 45 : 30 17 \Sigma 2 110 \Sigma 10
ff ffi mag Fš
(2000.0) (mJy)
15¯m 19 : 12 : 46:5 +10 : 45 : 55 4:34 370 \Sigma 37
3:4¯m -- a -- a 11:3 8:5 \Sigma 0:9
2:2¯m 19 : 12 : 46:5 +10 : 45 : 56 14:7 0:87 \Sigma 0:08
a ) No astrometry performed at nbL. The source is coincident
with the K band source.
Fig. 1. Spectrum of the H2O maser observed at Medicina. The
intensity scale is in Jansky.
42 s
44 s
46 s
48 s
50 s
19 12 52
h m s
30"
+10 45'00"
o
30"
+10 46'00"
o
30"
+10 47'00"
o
Fig. 2. Greyscale 15¯m image. The small ellipse represents the
IRAS­PSC error box, the circle marks the position of the H2O
maser, the diameter of the circle is approximately equal to the
Medicina pointing error box. The epoch of the coordinates is
2000.
this case we have considered all the K sources detected in
our image within 10 00 from the ISO source, finding only
three, very faint candidates. The source given in Table 1
is the closest to the ISO source (¸ 1:3 00 ), the other two
are at a distance of ¸ 6 00 and ¸ 9 00 respectively. All have a
magnitude greater than 14:4 in K, implying that, regard­
less of which of them is associated with the ISO source,
the (mK --m 15¯m ) colour index is in any case greater than
10. Note that even taking into account the flux calibra­
tion uncertainties of the 15¯m and the K­band data, the
resulting colour index must exceed 9.5 (which remains one
of the highest measured).
Four other 15¯m sources are detected close to but out­
side the error box of the H 2 O maser position. The closest is
at ¸ 40 00 , and the others are at ?
¸ 50 00 , from the maser po­
sition. Although the possibility that the H 2 O maser may
be associated with one of these sources cannot be com­
pletely ruled out, this seems unlikely, and in the following
we will assume that the maser is associated with the clos­
est 15¯m source. Certainly higher resolution observations
of the H 2 O maser are needed to settle this point.
Somewhat surprisingly, there is no IRAS point source
closely related to this source. The error ellipse of the
closest IRAS­PSC source (IRAS19104+1040) is shown in
Fig. 2, and it is almost coincident with a 15¯m source
which is not associated to masers and does not show a
large infrared excess (mK --m 15¯m ¸ 2). Given the large
infrared excess of our source one would expect a steeply
rising spectrum toward the longer wavelengths. Inspection
of the HIRES IRAS maps shows indeed that there are two
sources in a 1:5 0 \Theta 1:5 0 area, and that there is a peak at
25 ¯m at the position of the ISO source associated with
the water maser.
One important question to be addressed is if the maser
is associated with a late type star or with a young stellar
object. In fact, sources with large H 2 O spectra and as­
sociated with IRAS have been suggested to be late type
stars (see e.g. IRAS19134 + 2131 in Engels et al. 1984,
and IRAS16342 \Gamma 3814 in Likkel & Morris 1988). How­
ever, we believe that the ISO source is probably a YSO,
even though only detection of OH maser or CO emission
could settle this point. The extreme colour index observed
for the infrared source can be explained only in terms of
warm dust emission, most probably heavily extincted at
K. From the fluxes reported in Table 1 it can be clearly
seen that the spectral energy distribution of this source is
steeply rising toward the long wavelengths, with a spec­
tral index greater than 2, as expected for young stellar
objects still embedded in optically thick cocoons (Shu et
al. 1987).
No radio continuum emission was detected from this
source at 20 cm in the VLA survey of Zoonematkermani et
al. (1990). This implies that no HII region has formed or
that, if present, is too faint or compact (and self­absorbed)
to be detectable at 20 cm, thus confirming the young na­
ture of the object.

4 L. Testi et al.: Detection of Young Stellar Objects with ISO
3.2. Other Young Stellar Objects candidates
In the 0:244 deg 2 field observed at 15¯m we found a
surprisingly large number (210) of sources with a (mK --
m 15¯m ) colour index greater than 6, corrensponding to
¸ 900 sources per square degree. Among these, three have
a colour index greater than 9. None of these three sources
are near the position of the HII region G45:45+0:06, which
was not observed with ISOCAM to avoid saturation. A
remarkable difference with this bright radio HII region,
is that none of the 15¯m sources have a radio continuum
counterpart in the 20 cm VLA observations of Zoonematk­
ermani et al. (1990), which excludes the presence of an
evolved HII region, leaving only the possibility of an asso­
ciated UCHII region, optically thick at 20 cm.
In order to check that the large colour index observed is
not due to a spurious association with a faint background
K­band star, but due to an intrinsic infrared excess of the
sources, we have searched around the position of the can­
didate 15¯m sources for the presence of bright nearby K
sources, but with no success. The main difference in the
infrared morphology between these three sources and that
associated with the maser is that all of them show faint ex­
tended emission (on arcminute scale) at 15¯m. The nature
of this extended emission is currently under investigation.
All the three sources coincide with an IRAS point source,
but only in one the IRAS colours are those of a typical
UCHII regions (Wood & Churchwell 1989). None of these
sources have yet been observed in the H 2 O maser line.
4. Conclusions and future work
We have presented preliminary results obtained by the
ISOGAL team during the 15¯m survey of the inner galaxy.
An infrared source with (mK --m 15¯m ) colour index greater
than 10 is found and is associated with a newly discov­
ered H 2 O maser. This finding confirms that ISO can de­
tect young massive (proto­)stars in their earliest evolu­
tionary stages. In these stages, massive stars have not yet
developed an HII region around them, and, being deeply
embedded in their parental dusty cocoon, can be studied
only at infrared and sub­millimiter wavelength.
In the same field we observed more than 200 point
sources with (mK --m 15¯m ) ?
¸ 6, three of which with a colour
greater than 9. They could be young stars still embedded
in dusty envelopes. These results confirm the finding of
P'erault et al. (1996), who found ¸ 1300 objects with sim­
ilar colours per square degree in a field at l = \Gamma45 ffi . In our
case, the association of the source with the largest colour
index with a water maser confirms that we are indeed
detecting a YSO. The low 15 ¯m fluxes of these objects
(when compared with bright IRAS YSOs) suggests that,
at least statistically, they might be intermediate luminos­
ity YSOs not detectable by IRAS.
These findings prove the effectiveness of the ISOGAL
observations in detecting new very young stars away from
the luminous, well known star forming regions, when com­
pared with deep K­band observations.
Clearly the objects presented here, and in particular
that associated with the H 2 O maser, deserve to be studied
in detail at other wavelengths with high resolution and
sensitivity, in order to ascertain their nature. ISOCAM
observations at other wavelengths would also be advisable,
due to the high resolution and sensitivity achievable in
spectral regions not easily accessible from the ground.
Acknowledgements. LT acknowledges financial support re­
ceived from the CNRS and thanks the IAP for their hospitality.
References
Brand J., et al. 1994, A&AS, 103, 541
Cesaroni R., Churchwell E., Hofner P., Walmsley C.M., Kurtz
S., 1994a, A&A, 288, 903
Cesaroni R., Olmi L., Walmsley C.M., Churchwell E., Hofner
P., 1994b, ApJ Lett., 435, L137
Codella C., Felli M., Natale V., Palagi F., Palla F., 1994, A&A,
291, 261
Codella C., Felli M., 1995, A&A, 302, 521
Codella C., Testi L., Cesaroni R., 1996, A&A submitted
Cohen R.S., Dame T.M., & Thaddeus P., 1986, ApJS, 60, 695
Engels D., Habing H.J., Olnon F.M., Schmid­Burgk J., Walm­
sley C.M., 1984, A&A, 140, L9
Felli M., Testi L., 1996, in ``The impact of large­scale Near­IR
Sky Surveys'', Ed. P. Garzon Lopez, Kluwer, in press
Felli M., Testi L., Valdettaro R., Wang J.­J., 1996, A&A, in
press
Hunt L.K., Lisi F., Testi L., Baffa C., Borelli S., et al., 1996a,
A&AS, 115, 181
Hunt L.K., Migliorini S., Testi L., Stanga R.M., Aspin C., et
al., 1996b, AJ, submitted
Hunter T.R., Testi L., Taylor G.B., Tofani G., Felli M., Phillips
T.G., 1995, A&A, 302, 249
Likkel L. & Morris M. 1988, ApJ, 329, 914
Lisi F., Baffa C., Biliotti V., et al., 1996, PASP, 108, 364
Lockman F.J., 1989, ApJS, 71, 469
Olmi L., Cesaroni R., Neri R., Walmsley C.M., 1996, A&A,
315, 565
Omont A., 1996, in ``The impact of large­scale Near­IR Sky
Surveys'', Ed. P. Garzon Lopez, Kluwer, in press
Palla F., Cesaroni R., Brand J., Caselli P., Comoretto G., Felli
M., 1993, A&A, 280, 509
Palla F., Testi L., Hunter T.R., et al. 1995, A&A, 293, 521
P'erault M., Omont A., Simon G., et al. 1996, A&A, 315, L165
Shu F.H., Adams F.C., Lizano S., 1987, ARAA, 25, 23
Testi L., 1993, Technical Report 10/93, Arcetri Astrophysical
Observatory
Testi L., Felli M., Persi P., Roth M., 1994, A&A, 288, 634
Tofani G., Felli M., Taylor G.B. Hunter T.R. 1995, A&AS, 112,
299
Wood D.O. & Churchwell E., 1989, ApJ, 340, 265
This article was processed by the author using Springer­Verlag
L A T E X A&A style file L­AA version 3.