The vertical velocity dispersion profile of the Galactic thick disk

Abstract

We present the results of radial velocity measurements of 770 thick disk red giants toward the South Galactic Pole, vertically distributed from 0.5 kpc to 5 kpc with respect to the Galactic plane. We find a small gradient in the vertical velocity dispersion (sigma_W) of 3.8+/-0.8 km/s kpc. Even more noteworthy, our values of $\sigma_W$ are small compared to literature values: in the middle of the vertical height range we find sigma_W(z=2kpc)=30 km/s. We found no possible explanation for this small value of sigma_W in terms of sample contamination by thin disk stars, nor by wrong assumptions regarding the metallicity distribution and the derived distances.

Full text

Page 1

arXiv:0712.3528v1 [astro-ph] 20 Dec 2007
T   “XII L-A R   IAU (2007)”
RevMexAA(SC)
THE VERTICAL VELOCITY DISPERSION PROFILE OF THE GALACTIC THICK DISK
C. Moni Bidin1,2, T. M. Girard2, G. Carraro3, R. A. M´ endez1, W. F. van Altena2, V. I. Korchagin2, and
D. I. Casetti-Dinescu2
RESUMEN
Presentamos resultados de mediciones de velocidad radial para 770 estrellas gigantes rojas del disco grueso hacia el
Polo Gal´ actico Sur, distribu´ ıdas verticalmente desde 0.5 kpc hasta 5 kpc de distancia con respecto al plano Gal´ actico.
Encontramos un peque˜ no gradiente en la dispersi´ on de velocidad vertical (σW) de 3.8±0.8 km s−1kpc−1. Aun mas
notable,nuestrosvaloresdeσWsonpeque˜ noscomparadosconvaloresdeliteratura: enel mediodelrangodedistanciadel
plano encontramos σW,z=2kpc=30 km s−1. No es posible explicar este peque˜ no valor de σWen t´ erminos de contaminaci´ on
de la muestra por estrellas del disco delgado,y tampocoa trav´ es de una erradadistribuci´ onde metalicidady las distancias
deducidas.
ABSTRACT
We present the results of radial velocity measurements of 770 thick disk red giants toward the South Galactic Pole,
vertically distributed from 0.5 kpc to 5 kpc with respect to the Galactic plane. We find a small gradient in the vertical
velocity dispersion (σW) of 3.8±0.8 km s−1kpc−1. Even more noteworthy, our values of σW are small compared to
literaturevalues: inthemiddleoftheverticalheightrangewefindσW,z=2kpc=30kms−1. Wefoundnopossibleexplanation
for this small value of σWin terms of sample contamination by thin disk stars, nor by wrong assumptions regarding the
metallicity distribution and the derived distances.
Key Words: Galaxy: disk — Galaxy: fundamental parameters — Galaxy: kinematics and dynamics
1. INTRODUCTION
We are undertaking a spectroscopic study of nearly
1,200 thick disk red giant stars toward the South Galac-
tic Pole, to analyze the chemical and kinematical vertical
structure of the thick disk (Carraro et al. 2005).
The sample is vertically distributed with respect to
the Galactic plane, probing the Galactic thick disk with
unprecedented detail from 0.5 up to 5 kpc from the
Galactic plane.Details regarding target selection in
the K vs. (J-K) plane using 2MASS photometry and
the distance estimation procedure were presented by
Girard et al. (2006), who studied the proper motions of
the sample from the SPM3 catalog (Girard et al. 2004).
We collected high-resolution Echelle spectra for 770
stars (two thirds of the sample) during two observing
seasons in 2005 and 2006 at various instruments. Ra-
dial velocities (RVs) were measured through a cross-
correlation technique (Tonry & Davis 1979), using as
templates three red giant RV standard stars observed in
all runs. Using synthetic templates and twilight solar
spectra acquired each night we corrected all RVs from
systematic errors due to a number of factors, as for ex-
1Departamento de Astronom´ ıa, Universidad de Chile, Casilla 36-D,
Santiago, Chile (mbidin@das.uchile.cl).
2Yale University, Dept. of Astronomy, P.O. Box 208101, New
Haven, CT 06520-8101.
3European Southern Observatory, 3107 Alonso de Cordova, Santi-
ago, Chile.
ample stars not perfectly centered in the slit/fiber and
RV variations of the templates. Errors in the measure-
ments are in the range 0.4–1.0 km s−1, they were eval-
uated quadratically summing up the contribution of all
relevant sources (for more details, see Moni Bidin 2008,
PhD Thesis, in preparation). The comparison with liter-
ature RVs for 162 stars with published values reveals an
excellentagreementwithnosystematic trend,andtheRV
dispersion measured in this subsample is the same as that
deduced from the literature RVs.
2. RESULTS
Our results are shown in Figure 1, where the verti-
cal velocity dispersion (σW) is plotted as a function of
distance z from the Galactic plane. In this plot we con-
servatively excluded all stars with |W| ≥120 km s−1to
avoid ant significant residual contaminationof halo stars.
Our values are systematically lower when compared to
those in the literature. Various values of σWfor the thick
disk have been proposed in the last two decades, vary-
ing in the range 40–70 km s−1(see for example Table 1
in Casertano et al. 1990), although recent determinations
tendtoprefervaluesintheloweredgeofthis interval(see
for example Soubiran et al. 2003, and the value adopted
by Bensby et al. 2003). In contrast, we find a σW al-
ways smaller than 40 km s−1at all Galactic heights, with
σW,z=2kpc=30 km s−1. In the presence of a vertical gra-
dient (see below), the local extrapolated value would be
1

Page 2

2MONI BIDIN ET AL.
01234
0
10
20
30
40
z [kpc]
-100-500 50100
0
10
20
RV [km/s]
Fig. 1. Left panel: Vertical velocity dispersion profile of the sample: The dispersion in W-velocity is plotted as a function of distance
from the Galactic plane z. The solid line indicates the least square fit, the dashed line is the fit obtained exluding the first two data
points (suspected to be contaminated by thin disk stars). Right panel: RV distribution of the 200 stars in the range z=1.5-2.5 kpc.
even lower. We note that no effort was made to identify
and removebinary systems from the sample, a correction
that would further reduce the velocity dispersion.
We find a small vertical gradient of 3.8±0.8 km s−1
kpc−1, but it is mainly due to the first two bins, corre-
sponding to the nearest points, where a residual contam-
ination by thin disk red giants can not be excluded. In
fact (see Figure 1), after excluding these two first dis-
tance bins from the fit, the resulting gradient is much
shallower, and within error bars the data are consistent
with a flat profile.
We find no plausible source of thin disk contamina-
tion that can account for our small values. Thin disk
red giants contaminate (if any) only the nearest bins.
Dwarfs wereefficientlyexcludedbya cut in magnitude,a
(conservative) cut in proper motions (Girard et al. 2006)
and a further inspection of all the stellar spectra. More-
over, Girard et al. (2006) showed that in the range z=1–4
kpc the density profile of the sample is well described
by a single exponential with a scale height of 783 pc,
thus demonstrating that it is dominated by thick disk
stars. We found that different assumptions on the metal-
licity distribution, which can lead to a wrong distance
estimate, hardly change the results: Assuming extreme
fixed values for the whole sample, we find σW,z=2kpc=31
km s−1for [Fe/H]=−0.5 and σW,z=2kpc=27 km s−1for
[Fe/H]=−1.1.
It is worth noting that our sample selection include
stars in a rather wide metalicity interval, but higher
metallicities are preferred (see the isocrones in Figure 1
of Girard et al. 2006). Hence, we conclude that either
the thick disk is, as a whole, a stellar population kine-
matically cooler than believed so far, or that there exists
some σW-metallicity relation, in the sense of the metal-
rich component being kinematically cooler. This conclu-
sion was already proposedby Schuster et al. (2006), who
argued for the existence of two thick disk components,
although their derived σW for the kinematically cooler
population is still high with respect to ours. Moreover, it
is becomingevidentthat bona-fidethick disk stars extend
at least up to solar metallicities (Bensby et al. 2007).
CMB attendance to the meeting was funded by Conicyt and
by the LOC. RAM acknowledge support by the Chilean Centro
de Astrof´ ısica FONDAP (No. 15010003).
REFERENCES
Bensby, T., Feltzing, S., & Lundstr¨ om, I. 2003, A&A, 410, 527
Bensby, T., Zenn, A. R., Oey, M. S., & Feltzing, S. 2007, ApJ,
663, L13
Carraro, G., van Altena, W. F., Moni Bidin, C., Girard, T. M.,
Mendez, R. A., Dinescu, D., Korchagin, V., 2005, BAAS
37, p. 1378
Casertano, S., Ratnatunga, K. U., & Bahcall, J. N. 1990, ApJ,
357, 435
Girard, T. M., Dinescu, D. I., van Altena, W. F., Platais, I.,
Monet, D. G., & L´ opez, C. E. 2004, AJ, 127, 3060
Girard, T. M., Korchagin, V. I., Casetti-Dinescu, D. I., van Al-
tena, W. F., L´ opez, C. E., & Monet, D. G. 2006, AJ, 132,
1768
Schuster, W. J., Moitinho, A., M´ arquez, A., Parrao, L., & Co-
varrubias, E. 2006, å, 445, 939
Soubiran, C., Bienaym´ e, O., & Siebert, A. 2003, å, 398, 141
Tonry, J., & Davis, M. 1979, AJ, 84, 1511