arXiv:astro-ph/9210001v1 2 Oct 1992

1
0
 The Edge of the Galactic Disc
Annie C. Robin1,2 , Michel Crézé2, Vijay Mohan3
1
Observatoire de Besançon, BP1615, F-25010 Besançon Cedex, France
2
CNRS URA1280, Observatoire de Strasbourg, 11 rue de l’Université,
F-68000 Strasbourg, France
3
U.P. State Observatory, Manora Peak, Nainital, 263129 India

February 7, 2020

Abstract
As part of a stellar population sampling program, a series of pho-
tometric probes at various field sizes and depths have been obtained
in a low extinction window in the galactic anticentre direction. Such
data set strong constraints on the radial structure of the disc. At the
forefront of this ”drilling” program, very deep CCD frames probe the
most external parts of the disc. Over the whole effective magnitude
range (18 to 25), all contributions in the statistics which should be
expected from old disc stars beyond 6 kpc vanish, although such stars
dominate by far at distances less than 5 kpc. This is the signature of a
sharp cut-off in the star density: the edge of the galactic disc between
5.5 and 6 kpc. As a consequence, the galactic radius does not exceed
14 kpc (assuming R⊙ =8.5). Colours of elliptical galaxies measured in
the field rule out the risk of being misled by undetected extinction.

Subject headings: Galaxy: structure - Galaxy: stellar content

∗
Based on observations made at Canada-France-Hawaii Telescope (CFHT)

1
1 Introduction
Efforts to sample star distributions in the galactic plane should face both
the problem of overcrowding and the complexe structure of the absorbing
layer. For these reasons the outer part of our own galactic disc is very poorly
known. The radial scale length of the density decrease is controversial and
we have nearly no indication what happens at the end.
Two previous papers (Mohan et al. 1988, and Robin at al. 1992) present
the first results of a stellar population sampling program, including a series
of photometric probes at various field sizes and depths in a low extinction
window in the galactic anticentre direction. Wide field photometry in UBV
in the magnitude range 12-17 is shown to interpret unambiguously in terms
of extinction and stellar density. The interpretation partially uses the model
developped by Robin and Crézé (1986) and Bienaymé et al. (1987). The
model ingredients which play a role in the present anticentre investigation
are the density law of the galactic disc (radially exponential) and the lumi-
nosity function from Wielen et al. (1983). Strong constraints are set on the
radial structure of the disc: the galactic disc scale length is 2.5 kpc. Based
on this scale length and assuming no dramatic change in the luminosity
function one can predict what should normally happen at faintest magni-
tudes: the V, B-V distribution of stars resulting from this investigation is
given in figure 1a and the V counts are plotted in figure 2 (dashed line).

2 Deep observations towards the anticentre

Deep CCD observations in the BV bands have been obtained at the 3.6 meter
CFH Telescope in a low extinction window at low latitude in the direction
of the galactic anticentre. A detailed description of observational and data
analysis aspects is given elsewhere (Robin et al. 1992). Observations cover
four neighbouring fields adding up to 29 square arcminutes. Field charac-
teristics are given in table 1. The detection limit is about magnitude 28 in
V, while the completeness limit corresponding to a photometric accuracy
better than 0.1 is 25 in V and 22.5 to 24 in B depending on the frame. The
resulting V, B-V distribution is plotted in figure 1b down to V = 23. B data
are lacking beyond this limit. Differential star counts down to V = 25 are
plotted on figure 2 together with the associated 1 sigma error bars.

2
3 Interpretation
The faint star count predictions deviate strongly from the observations at
the faint end and there is a clear excess of blue stars in the down left part of
figure 1a. The bulk of disc contributors in this magnitude range is made of
disc dwarfs beyond 5.5 kpc (in figure 1a stars closer than 5.5 kpc are plotted
as full circles while farther stars are crosses).
All disagreements in the V, B-V diagram vanish if the stellar galactic
disc ends abruptly at 5.5 kpc as also shown in the V star counts in figure
2 (solid line). Other explanations could come from the extinction along the
line of sight, the scale length used in the model or the luminosity function.
We show below that none of these hypotheses resists to investigation.
Interpretation of the cutoff seen in the V, B-V diagram in terms of ab-
sorbing cloud would need a total extinction AV = 2.4 over the line of sight,
including a cloud of AV = 1.2 at 5.5 kpc that is about 300 pc from the
galactic plane. The extinction measured from the UBV diagrams and HI
column density gives a total extinction to 4 kpc of AV = 1.2. The colours of
4 elliptical galaxies measured in the fields give an estimate of the total ex-
tinction inside the Galaxy of 1.4, incompatible with the existence of a deeply
absorbing cloud a 5.5 kpc. This is also in agreement with the fact that the
fields are inside Special Area 23 selected by Kapteyn as a low extinction
window.
Density distributions with scale lengths larger than the adopted 2.5 kpc
would impose a still closer cutoff (at 3.5 kpc from us if h = 3.5 kpc) while
shorter scale lengths are hardly compatible with the observations of bright
stars in the same region (Mohan et al. 1988).

4 Discussion
Stars that should be expected to appear in the sample to V=23 at distances
larger than 5.5 kpc are common dwarfs of absolute visual magnitudes be-
tween 3 and 7, a part of the luminosity function which is well known from the
study of the solar neighbourhood. Moreover these stars appear at brighter
V magnitudes when they are closer. Changing their luminosity function
would lead to strong disagreements with Schmidt plates data of Mohan et
al. at magnitudes 12 to 16 and with the CCD data at magnitudes 17 to 21.
Rather than invoking a sudden change of the luminosity function of stars
of magnitude 3 to 7 at 5.5 kpc from us the data are better interpreted as

3
a sharp cutoff of the density distribution for all stars at this distance. This
observation is indicative of the end of the star formation process, possibly
related to the external Lindblad resonnance.
Most external disc galaxies show a radial truncation. These cutoffs seem
to arise within 1 kpc or less (van der Kruit 1988). From a sample of 7 edge-
on galaxies this drop has been found between 3.4 and 5.3 times the old disc
scale length (van der Kruit and Searle 1981, 1982). When reanalysing the
Wevers’ sample (1986) of 20 face-on spirals, van der Kruit (1988) showed
that the mean ratio between Rmax and scale length h is 4.5 ± 1.0 for 16
galaxies and larger than 6.0 for the remaining 4 spirals. However Barteldrees
and Dettmar (1990) sample give a mean ratio of 2.8 ± 1. It should be noted
that the scale lengths of external spirals may be uncertain by a factor two
(Knappen and van der Kruit 1991).
In our Galaxy star counts in the anticentre (Robin et al. 1992) give a
radial scale length of 2.5 ± 0.3 kpc. Together with the presently determined
cutoff it implies a ratio Rmax /h of 5.6 ± 0.6 if R⊙ = 8.5 kpc or 5.2 ± 0.6 if
R⊙ = 7.5 kpc, in agreement with the observed ratio in external galaxies.
Indications for a cutoff in the disc distribution has already been found in
other data. From the IRAS Point Source Catalogue Habing (1988) found a
cutoff at distances between 1 and 2.5 kpc. However Habing stressed that this
very short cut-off distance depends strongly on the assumed disc scale length
(4.5 kpc), which on the other hand is inconsistent with our wide field star
count results. Habing did not test any model with a shorter scale length for
the old disc components. In the view of the present result the combination
of short scale and larger distance cutoff should be tested against IRAS data.
The edge of the disc was also determined in the gaseous component of
the Galaxy at larger distances. Wouterloot et al. (1990) got a sharp decline
of the CO cloud density between 18 and 20 kpc, a signature that star for-
mation stops at about this distance. However these distances being derived
from kinematics are subject to errors due to perturbations of the velocity
field. HII regions observed by Fich et al. (1989) also end at galactocentric
distances of about 15 kpc.
Our determination of the radial extent of the old disc does not conflict
with the possibly larger extent of young stars or star forming regions if an
evolutionnary scenario like the one of Larson (1976) (where the star forma-
tion propagates from the centre of the Galaxy to the outer part) is realistic.
In this case one expects to find only recent star formation in the outer part
of the Galaxy and no old disc stars.

4
Acknowledgements. This research was partially supported by the Indo-
French Centre for the Promotion of Advanced Research / Centre Franco-
Indien Pour la Promotion de la Recherche Avancée.

References

Barteldrees, A., & Dettmar, R. J. 1990, in International Conference on

Dynamics and Interactions of galaxies, ed R. Wielen (Springer), p. 348
Bienaymé, O., Robin, A. C. & Crézé, M. 1987, Astron. Astrophys. 180, 94
Fich, M., Blitz, L. & Stark, A. A. 1989, Astrophys. J. 342, 272
Habing, H. J. 1988, Astron. Astrophys. 200,40
Knappen, J. H. & van der Kruit, P. C. 1991, Astron. Astrophys. 248, 57
Larson, R. B. 1976, Mon. Not. R. astr. Soc. 176, 31
Mohan, V., Bijaoui, A., Crézé, M. & Robin, A. C. 1988, Astron. Astrophys.
Suppl. 73, 85
Robin, A. C. & Crézé, M. 1986, Astron. Astrophys. 157, 71
Robin, A. C., Crézé, M. & Mohan, V. 1992, Astron. Astrophys. in press
van der Kruit, P. C. 1988, Astron. Astrophys. 192, 117
van der Kruit, P. C. & Searle, L. 1981, Astron. Astrophys. 95, 105
van der Kruit, P. C. & Searle, L. 1982, Astron. Astrophys. 110, 61
Wevers, B. M. H. R., van der Kruit, P. C., & Allen, R.J. 1986, Astron.
Astrophys. Suppl. 66, 505
Wielen, R. Jahreiss, H. & Krüger, R. 1983, In The Nearby Stars and the
Stellar Luminosity Function, I.A.U Coll. 76, p. 163 , Davis Philip and Up-
gren (eds), L. Davis Press (Schenectady, NY).
Wouterloot, J. G. A., Brand, J., Burton, W. B. & Kwee, K. K. 1990, Astron.
Astrophys. 230, 21

5
Table 1: Field coordinates and area in square arc minutes.

Field RA (1950) Dec (1950) l b Area

1 5 53 06.5 +30 39 43 179.70 2.88 6.7
2 5 52 45.4 +30 40 52 179.64 2.83 6.5
3 5 52 54.7 +30 36 13 179.73 2.82 7.7
4 5 52 46.4 +30 38 24 179.68 2.81 6.6

Figure captions

Figure 1: (V, B-V) diagram of anticentre stars. (a) Model predicted dis-
tribution. Dots are stars closer than 5.5 kpc while crosses are stars beyond
5.5 kpc. (b) Observed distribution. The solid line is a guide to identify the
zone where most stars are beyond 5.5 kpc.

Figure 2: V star counts to magnitude 25 in the anticentre direction. Di-

amonds: Observed counts with 1 sigma error bars (Poisson noise only).
Dashed line: Predicted counts assuming no cutoff in the radial distribution
of stars. Solid line: Predicted counts assuming a cutoff at 5.5 kpc from the
sun.

6
Authors postal addresses

Michel Crézé, Observatoire de Strasbourg, 11 rue de l’Université, F-68000

Strasbourg, France
Vijay Mohan, U.P. State Observatory, Manora Peak, Nainital, 263129 In-
dia
Annie C. Robin, Observatoire de Besançon, BP1615, F-25010 Besançon
Cedex, France