A&A 425, L37-L40 (2004)
DOI: 10.1051/0004-6361:200400031
D. L. Block1 - K. C. Freeman2 - T. H. Jarrett3 - I. Puerari4 - G. Worthey5 - F. Combes6 - R. Groess1
1 - School of Computational and Applied Mathematics,
University of Witwatersrand, Private Bag 3, WITS 2050, South Africa
2 -
Mount Stromlo and Siding Spring
Observatories, Research School of Astronomy and Astrophysics, Australian
National University, Australia
3 -
Infrared Processing and Analysis Centre,
100-22, CALTECH, 770 South Wilson Ave, Pasadena, CA 91125, USA
4 -
Instituto Nacional de Astrofísica,
Optica y Electrónica,
Calle Luis Enrique Erro 1, 72840 Tonantzintla, Puebla, México
5 -
Washington State University, 1245 Webster Hall,
Pullman, WA 99163-2814, USA
6 -
Observatoire de Paris, LERMA, 61 Av. de
l'Observatoire, 75014 Paris, France
Received 27 May 2004 / Accepted 8 July 2004
Abstract
Stars with masses in the range from about 1.3 to
pass through an evolutionary stage where they become carbon
stars. In this stage, which lasts a few Myr, these stars are extremely
luminous pulsating giants. They are so luminous in the near-infrared
that just a few of them can double the integrated luminosity of
intermediate-age (0.6 to 2 Gyr) Magellanic Cloud clusters at 2.2 microns. Astronomers routinely use such near-infrared observations to
minimize the effects of dust extinction, but it is precisely in this
band that carbon stars can contribute hugely. The actual contribution
of carbon stars to the outer disk light of evolving spiral galaxies has not
previously been morphologically investigated. Here we report new and
very deep near-IR images of the Triangulum spiral galaxy M 33 = NGC 598,
delineating
spectacular arcs of carbon stars in its outer regions. It is these arcs
which dominate the near-infrared m=2 Fourier spectra of M 33. We
present near-infrared photometry with the Hale 5-m reflector, and propose that the arcs are the signature of accretion of
low metallicity
gas in the outer disk of M 33.
Key words: galaxies: evolution - galaxies: spiral - galaxies: individual: M 33 = NGC 598
Carbon stars are thermally pulsing asymptotic giant branch (TP-AGB)
stars with ages between about 0.6 to 2 Gyr. They are observed in
the intermediate-age globular clusters of the Large and Small
Magellanic Clouds. Although the average number of carbon stars per
intermediate age cluster in the Magellanic Clouds is only about 2.5
(Persson et al. 1983), they radiate such a copious amount of near-IR light
that they contribute about 50% of the bolometric luminosity of the
cluster (Marigo et al. 2003; Maraston 1998). From stellar population
synthesis studies, the near-IR K-band (
m) luminosity is enhanced
in galaxies containing a significant intermediate-age stellar
population (Mouhcine & Lançon 2003) by up to a
factor 2.
In this Letter,
we present new deep near-IR images of the nearby spiral galaxy M 33
which illustrate the dramatic effect that an intermediate-age
population can make to the near-IR light distribution in spiral
systems.
![]() |
Figure 1:
a) Reaching six times deeper than 2MASS is this
|
| Open with DEXTER | |
Astronomers routinely use near-IR observations of spiral galaxies to minimize the effects of dust extinction, but the actual contribution of carbon stars to the near-IR light of evolving galaxies remains poorly quantified. Little has changed observationally since Aaronson noted the importance of the asymptotic giant branch for understanding the stellar content of nearer galaxies (Aaronson 1986). Disks of galaxies appear to form from the inside out (Block et al. 2002). Relative to the inner regions of spiral galaxies, the mean ages of the outer regions are known to be somewhat younger and more metal-poor (Bell & de Jong 2000). We can therefore expect the contribution from the intermediate-age stars to be stronger in these outer regions, and the near-IR surface brightness of the outer disk will be preferentially enhanced by the presence of TP-AGB stars. This brightening of the outer disk in the near-IR may well contribute to the apparent sharp radial truncation observed in the disks of many spiral systems (Kregel et al. 2002).
In this context, we present new near-infrared images of the Local Group
spiral M 33 from a special set of 2MASS observations. The integration
time for these images was increased by a factor of six, extending
approximately 1 mag deeper than the standard survey.
Three mosaics are constructed for M 33, corresponding to the J (1.2
m),
H (1.6
m), and "K-short'',
(2.2
m) bands.
At
,
the image resolves larger-area features
as faint as
22.5 mag arcsec-2.
Foreground Milky
Way stars were removed statistically, using the
histogram of two
control star-fields to the East and West of M 33 (but at the same
galactic latitude) as a template. Of the
7000 sources in
the original image,
2300 were removed as foreground stars.
M 33 is famous for an optically bright pair of spiral
arms, but surprisingly, it is not
these two arms which dominate the near-infrared m=2 Fourier spectra.
The near-infrared images reveal remarkable arcs of red stars in the
outer disk of M 33, spanning
in azimuth angle. The northern
arc is dominant although a very faint southern counterpart arc, forming
a partial ring, can also be seen (Fig. 1). A hint of the
northern swath can be seen in the study by Regan & Vogel (1994).
Fourier analysis of the
light distribution (Fig. 1d) shows that the dominant m=2 peak
corresponds to the giant arcs. The pitch angle of the dominant m=2structure is
58 degrees.
![]() |
Figure 2: An overlay of the integrated cluster light in the Magellanic Clouds (from Fig. 4 of Persson et al. 1983), with the integrated colors for plume stars in the NE sector (green dots) and SW sector (red dots) of M 33. The open symbols principally show the Searle, Wilkinson & Bagnuolo (1980) (SWB) Magellanic cluster types I, II and III; these do not contain carbon stars. The filled symbols are SWB types V, VI and VII. The red Magellanic carbon-star-bearing clusters lie above and to the right of the black dashed lines, in the Magellanic "IR-enhanced'' cluster regime identified by Persson et al. (1983); these are predominantly SWB types V and VI. Red bars show our 1-sigma uncertainty in the color for each point in this figure; the black error bars are from the Persson et al. (1983) analysis. The thin blue dotted and solid lines show the Bessell & Brett (1998) tracks for giants and dwarfs respectively. The overlay provides strong evidence for a carbon star population in the M 33 plumes. |
| Open with DEXTER | |
![]() |
Figure 3: A J-H, H-Ks color-color diagram contrasting the nuclear colors with those of northern inner spiral arm and the NE and SW plume populations. The colors were derived from the integrated flux of regions centered on the N inner arm and NE and SW plumes. The photometry consisted of circular, 10 arcsec radius aperture photometry carried out on the 2MASS J, H and Ks images, cleaned of foreground stars and corrected for disk inclination. The evolved giant and main-sequence dwarf tracks are shown with green dashed and solid lines, following Bessell & Brett (1998). |
| Open with DEXTER | |
![]() |
Figure 4:
Observations with the Hale 5m reflector, through a
sector of the northern red arc. The sample has been cleaned of
all low detection sources, with a S/N < 10.
The extinction reddening vector is indicated with the red
arrow. The evolved giant and main-sequence dwarf tracks are
from Bessell & Brett (1998).
In this color-color plot, no foreground MW stars
were statistically removed, as in the other color-color plot.
Here foreground stars appear with blue colors
(
|
| Open with DEXTER | |
The very red color of the arcs is not due to dust. Wilson (1991)
derived
mag, which includes both foreground
(Milky Way) and internal M 33 extinction. The K-band extinction
A(K) is estimated
0.09 mag.
The color of the northern arc extends to
.
Very old M giants of solar abundance can reach
(see e.g. Fig. 2
in Bessell & Brett 1998), and even redder if they are
super-metal-rich (Frogel & Whitford 1987). However, as reviewed by Pagel &
Edmunds (1981), there is a strong radial abundance
gradient (Searle 1971) in M 33 (
dex/kpc in O/H); the
outer regions are relatively metal-poor, and solar abundance is reached
only within the inner 1'' of M 33. In regions of lower abundance, the
giant branch is bluer. If stars with
are found in the outer
low-metallicity regions, they cannot be M-giants. Figure 2 presents
the integrated colors of the plumes in the NE and SW sector of M 33, overlayed
with the integrated colors of the clusters in the Magellanic
Clouds (Persson et al. 1983). Figures 2 and 3
provide strong evidence for a
carbon star population in the extended M 33 plumes. We conclude that
the near-IR arcs are probably dominated by a population of very red
carbon stars.
Only integrated fluxes with a S/N greater than 10 are shown in Fig. 3; hence, the accuracy of the photometry is comparable (to within a factor of two) to that of Persson et al. (1983). Note that individual M 33 stars are not resolved by 2MASS (unlike the Palomar observations as discussed below), and so the colors best represent the "local'' area (300 sq. arcsec) of the targeted regions. Individual AGB and carbon stars will "redden'' the J-H and H-Ks colors, but the overall effect is mitigated by the stellar population in the local region. Consequently, the colors represent a lower limit to the color of the evolved population.
Why are the Fourier spectra not dominated by the inner, very robust spiral arms? In a spiral galaxy with a sustained star formation history, in the first few Myr red supergiants will dominate the infrared light. In the 0.5-2 Gyr range carbon stars dominate, especially in a somewhat metal-poor regime. After this time, the carbon stars die and ordinary RGB/AGB stars dominate. In the first 2 Gyr, the C-stars clearly dominate, as accreted gas in the outer disk is fresh and not yet well mixed by M 33's near-infrared oval/bar. In the inner disk, there presumably are older populations present that dominate in light, and the Ks-band enhancement of the TP-AGB carbon stars is weakened.
One of us (THJ) recently imaged a section of the northern
plume of M 33
with the 5 m-Hale reflector, using the
array near-infrared
camera
WIRC (see Fig. 4). The field of view is
with
0.25 arcsec pixels,
reaching a limiting surface brightness at
of 23.7 mag per arcsec-2.
A plethora of stars is seen toward the C-Star regime in the
upper right corner of the figure, with colors exceeding
mag.
In their modelling of the HI envelope,
Corbelli & Schneider (1997) find that the phase changes
at a radius of
20 arcmin, which is precisely the
outer domain of the arcs reported in this study.
It seems highly plausible therefore, that fresh, low-metallicity gas is
being fed to the host
galaxy M 33, via external accretion. The outer disk from which mass is
accreted
is inclined to the inner disk of M 33, and has a different angular
momentum.
The accretion of gas
from gaseous filaments is expected to be asymmetric, since
it
comes from one side
only - gas is being accreted and only later becomes bound and
finds its way
to
circular
orbits. It takes a dynamical time to relax. In the present epoch, we
expect the
signature of accreted gas (the external arcs) to be asymmetric.
We believe that it is the signature of gas flowing
inward and accreting at
20 arcmin, from which
the very red,
and relatively metal-poor stars have been formed. This
is also fully consistent with the recent observations elucidated
by Tiede et al. (2004).
It is tempting to liken the ring in M 33 to the one recently reported
in the outer disk of our Milky Way
(e.g. Ibata et al. 2003).
For the Milky Way ring,
the rotation period is
600 Myr. For M 33, the rotation
time-scale at the radius of the ring would be
240 Myr.
Any carbon-bearing
clusters of age 0.6 Gyr would have only undergone
2 orbits.
Our discovery of a dominant carbon-star-bearing stellar population in the outer regions of a spiral galaxy (M 33) shows how imperative it is to use stellar population synthesis models which include carbon stars with their correct bolometric contributions (Maraston 1998), when analysing the integrated light of spiral galaxies. The outer arcs in M 33 are not revealed on standard near-infrared surveys, such as 2MASS, because their integrated surface brightness is fainter than the 2MASS threshold. It is a sobering thought that it takes a groundbased class 4-5 m telescope to resolve individual carbon stars in the outer disk of M 33, our second closest spiral.
Acknowledgements
We thank the anonymous referee for his/her insightful comments. D.L.B. thanks the RSAA, ANU for visitor status and support in Canberra. D.L.B., K.C.F. and I.P. are indebted to the Anglo American Chairman's Fund, Mrs. M. Keeton and the Board of Trustees. I.P. acknowledges support from the Mexican foundation CONACyT under project 35078-E. 2MASS is a joint project of the Univ. of Massachusetts and the Infrared Processing and Analysis Center, funded by NASA and the NSF.