A complex stellar line-of-sight velocity distribution in the lenticular galaxy NGC 524

Abstract

We present the detailed study of the stellar and gaseous kinematics of the
luminous early-type galaxy NGC 524 derived from the long-slit spectroscopic
observations obtained with the Russian 6-m telescope and the IFU data from the
SAURON survey. The stellar line-of-sight velocity distribution (LOSVD) of NGC
524 exhibits strong asymmetry. We performed the comprehensive analysis of the
LOSVD using two complementary approaches implemented on top of the nbursts full
spectral fitting technique, (a) a nonparametric LOSVD recovery and (b) a
parametric recovery of two Gaussian kinematical components having different
stellar populations. We discuss the origin of the complex stellar LOSVD of NGC
524.

Full text

Page 1

arXiv:1106.2527v1 [astro-ph.CO] 13 Jun 2011
Baltic Astronomy, vol.XX, XXX–XXX, 2011
A COMPLEX STELLAR LINE-OF-SIGHT VELOCITY DISTRIBU-
TION IN THE LENTICULAR GALAXY NGC 524
I. Katkov1, I. Chilingarian2,1, O. Sil’chenko1, A. Zasov1and V. Afanasiev3
1Sternberg Astronomical Institute, Moscow State University, Universitetskii pr.
13, Moscow, 119992 Russia;
2CDS – Observatoire de Strasbourg, CNRS UMR 7550, Universit´ e de Strasbourg,
11 Rue de l’Universit´ e, 67000 Strasbourg, France
3Special Astrophysical Observatory, Russian Academy of Sciences, Nizhnii Arkhyz,
KarachaevoCherkesskaya Republic, 369167 Russia
Received: 2011 XXXXXX XX; accepted: 2011 XXXXXXX XX
Abstract.
matics of the luminous early-type galaxy NGC 524 derived from the long-slit
spectroscopic observations obtained with the Russian 6-m telescope and the
IFU data from the SAURON survey. The stellar line-of-sight velocity distri-
bution (LOSVD) of NGC 524 exhibits strong asymmetry. We performed the
comprehensive analysis of the LOSVD using two complementary approaches
implemented on top of the nbursts full spectral fitting technique, (a) a non-
parametric LOSVD recovery and (b) a parametric recovery of two Gaussian
kinematical components having different stellar populations. We discuss the
origin of the complex stellar LOSVD of NGC 524.
We present the detailed study of the stellar and gaseous kine-
Key words: Galaxies: kinematics and dynamics – galaxies: individual: NGC
524
1. INTRODUCTION
NGC 524 is a luminous (MB = −21.7 mag) lenticular galaxy settled in the
centre of a rich group (Garcia 1993) containing a X-ray hot gas component with
a slightly lopsided distribution with respect to NGC 524 (Mulchaey et al. 2003).
The galaxy demonstrates nearly circular isophotes, ǫ < 0.05 (e.g. Magrelli et al.
1992), so from the photometric point of view it looks face-on. However, kinematic
measurements revealed quite fast rotation of the galaxy (Sil’chenko 2000, Simien
& Prugniel 2000, Emsellem et al. 2004) that is inconsistent with the photometric
inclination of < 18 deg. Moreover, NGC 524 possesses a roundish disc of ionised
gas extending up to 4 kpc (25 arcsec) from the centre (Macchetto et al. 1996)
rotating even faster than the stars. By comparing the rotation of the ionised gas
and the stars in the centre of NGC 524, Sil’chenko (2000) suggested that its gaseous
disc was inclined with respect to the stellar one. The recent structural analysis of
NGC 524 (Sil’chenko 2009) reveals that the bulge of this giant lenticular galaxy is
rather modest and is seen only inside R ≈ 10 arcsec, while the general structure is

Page 2

2
I. Katkov et. al.
G-band (log-scale)
-100 -500 50 100
-∆α, arcsec
-100
-50
0
50
100
∆δ, arcsec
PA=217 deg
major axis
SAURON
-0.3
0.4
-0.2
-0.1
-0.0
0.1
0.2
0.3
H3
-60 -40 -20
radial distances, arcsec
0204060
-0.1
0.0
0.1
0.2
0.3
H4
Fig.1. Left: The positions of the SCORPIO slits and the SAURON mosaic field of
view are overlapped onto the g-band image of NGC 524 obtained at CFHT. Right: The
Gauss-Hermite coefficients characterizing the asymmetry of the stellar LOSVD.
dominated by two exponential stellar discs with the scalelengths of 0.9 and 3 kpc
respectively.
2. OBSERVATIONS AND DATA REDUCTION
We used two datasets derived from long-slit and integral-field spectroscopy.
The long-slit spectroscopic observations were obtained with the SCORPIO
universal spectrograph (Afanasiev & Moiseev 2005) at the prime focus of the
Russian 6-m BTA telescope of the Special Astrophysical Observatory. We observed
NGC 524 in six slit positions (P.A. = 137,217,240,295,307,317 deg, see Fig. 1
left) going through the galaxy centre.
spectrum obtained along the kinematical major axis (P.A. = 217 deg). We used
the “green” (480–550 nm) and “red” (610–710 nm) spectral setups covering strong
stellar absorption features as well as the emission lines of the ionised gas, Hβ,
[Oiii] (λ = 4959,5007˚ A), Hα, [Nii] (λ = 6548,6583˚ A), [Sii] (λ = 6716,6731˚ A)
providing the spectral resolution of 0.22 nm and 0.31 nm correspondingly.
Data reduction for the long-slit spectral data of NGC 524 was identical to that
of the lenticular galaxy NGC 7743 presented in Katkov et al. (2011a). Briefly,
the data reduction steps included: bias subtraction, flat fielding, cosmic ray hit
removal, building the wavelength solution using arc-line spectra, constructing the
spectral line spread function (LSF) variation model using twilight spectrum and
night sky spectrum subtraction taking into account the LSF variation along and
across the wavelength direction (Katkov & Chilingarian 2010), and adaptive bin-
ning along the slit in order to achieve the minimal value of the signal-to-noise ratio
S/N = 20 per spatial bin.
We also used the data obtained with the integral-field spectrograph SAURON
(Bacon et al. 2001) at the 4.2-m William Herschel Telescope. NGC 524 was
observed in three positions of the SAURON lenslet array with 0.94 arcsec sampling
(see Fig 1 left). The covered spectral range was 480–540 nm with the spectral
resolution 0.48 nm. For our analysis we used the science-ready data cube kindly
provided by E. Emsellem that we binned adaptively using Voronoi tessellation
(Cappellari & Copin 2003) to the minimal S/N ratio of 100 per bin.
But here we present only the long-slit

Page 3

A complex stellar LOSVD in NGC 524
3
-40-20
radial distances, arcsec
02040
-600
-400
-200
0
200
400
600
Velocity, km/s
1800 2100 240027003000
velocity, km/s
r = 25"
r = 0"
r = -25"
Fig.2.
profile obtained using one-component model. Red and blue crosses - ”bulge” and ”disc”
components in the two-component model. Right: The LOSVD at radii at -25,0,25 arcsec.
Left: The position-velocity diagram. White line presents a radial velocity
3. DATA ANALYSIS
3.1. SSP-EQUIVALENT PARAMETERS AND EMISSION LINE KINEMATICS
We derived the parameters of internal kinematics and stellar populations of
NGC 524 by fitting high-resolution PEGASE.HR (Le Borgne et al. 2004) simple
stellar population (SSP) models against our spectra using the nbursts full spectral
fitting technique (Chilingarian et al. 2007a,b). We determined SSP-equivalent ages
T and metallicities [Z/H] of the stellar population as well as the stellar kinematics
using the Gauss-Hermite parametrization up-to the 4th moment, i.e. v, σ, h3and
h4 (van der Marel & Franx 1993). The derived parametric LOSVD exhibits a
strong asymmetry leading to the non-physical values of h3 and h4 (see Fig 1.
right) corresponding to significantly negative LOSVD “wings”.
The emission line spectrum at every spatial bin was obtained by the subtraction
of the stellar contribution (i.e. the best-fitting model) from the observed spectrum.
Then we fitted it with pure Gaussians convolvedwith the LSF in order to determine
the kinematics of the ionised gas and emission line ratios.
3.2. NON-PARAMETRIC LOSVD
We propose a non-parametric recovery technique based on the full spectral
fitting requiring no a priory LOSVD knowledge. The logarithmically rebinned
model spectrum, F(λ) is the convolution of the assumed normalized LOSVD,
L(v), with the rest-frame SSP model, Fr(w):
F(w) =
?umax
umin
Fr(w − u)L(u)du,
where w = ln(λ), u = ln(1+v/c). We used the output SSP model of the nbursts
fitting as a template spectrum Fr. This equation can be considered as a linear
inverse problem whose solution is very sensitive to the noise in the data. Hence,
we chose to regularize the problem be requiring the LOSVD to be smooth. In
order to do so, we use the cubic penalization P(L) = LT· DT· D · L, where D -
is the third-order difference operator. The function to be minimized is given by
χ2+ λP(L). For discussion on the choice of λ see Press et al. (2007).
Using this technique, we confirmed a strong asymmetry of the NGC 524 LOSVD.

Page 4

4
I. Katkov et. al.
Velocity, km/s
-60 -40 -200 20 40 60
-300
-200
-100
0
100
200
300
N-ES-W
stars
Hα
[NII]
Dispersion, km/s
-60 -40 -200204060
0
50
100
150
200
250
300
Age, Gyr
-60 -40 -200204060
5
10
15
20
[Z/H], dex
-60 -40 -20
radial distances, arcsec
0204060
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
-30-150 15 30
radial distances, arcsec
-300
-200
-100
0
100
200
300
N-E S-W
-30 -150 15 30
radial distances, arcsec
0
100
200
300
400
500
-30-150 1530
radial distances, arcsec
5
10
15
20
-30-1501530
radial distances, arcsec
-0.6
-0.4
-0.2
0.0
0.2
Fig.3. SCORPIO long-slit spectroscopic data. First row – one-component spectra
fitting with S/N ratio equal to 20 in the green spectral domain.
component spectra fitting in green domain (S/N = 40). Black points correspond to
”disc” component, blue points to ”bulge” one, red line shows Hα kinematics.
Second row – two
Fig. 2 displays the result of the LOSVD recovery for the long-slit data along the
kinematical major axis as a position-velocity diagram (left). The LOSVD profiles
at radii −25, 0 and 25 arcsec are shown to the right.
3.3. TWO-COMPONENT PARAMETRIC LOSVD RECOVERY
Another approach we use is a full spectral fitting using a two-component model
where different stellar population components have two different pure Gaussian
LOSVDs. An optimal template is represented by the linear combination of two
SSPs each convolved with its own LOSVD, hence the χ2value is computed as
follows:
(Fi− Pp·?j=2
χ2=
?
Nλ
j=1kj· S(Tj,Zj) ⊗ L(vj,σj))2
δF2
i
,
where L(v,σ) - pure Gaussian LOSVD; Fi and δFi are observed flux and its
uncertainty; S(Tj,Zj) is the flux from the j-th synthetic spectrum of SSP with
given age Tjand metallicity Zj; Ppis multiplicative Legendre polynomials of order
p for correcting the continuum which determined at each step of minimization
loop by solving the linear least-square problem; kj is the j-th component weight
(normally found by the linear minimization). The important point in this study
is that we fixed the relative SSP contributions kj to the values derived from the
light profile decomposition.
4. RESULTS AND DISCUSSION
4.1. KINEMATICS
At all radii the LOSVD of NGC 524 clearly demonstrates the presence of at
least two components clearly visible in Figs. 1–2 confirmed independently by the
non-parametric and parametric LOSVD reconstruction on SAURON and SCOR-
PIO datasets. The inner region of the galaxy that corresponds to the small ex-
ponential pseudobulge identified by Sil’chenko (2009) is very hot in a dynamical

Page 5

A complex stellar LOSVD in NGC 524
5
Fig.4. SAURON integral-field spectroscopic data. From left to right: velocity, veloc-
ity dispersion, SSP-equvalent ages and metallicities. First row correspond to one com-
ponent fitting spectra. Second and third rows present a ”bulge” and ”disc” parameter
maps for two component spectra decomposition.
sense with the velocity dispersion exceeding 300 km s−1.
This component disappears at about R ≈ 20 arcsec, where we see a sharp
drop in the velocity dispersion and radial velocities increasing outwards suggest-
ing the counter-rotation with respect to the main disc (see blue profiles in the
bottom panels of Fig. 3). At large radii (R > 20 arcsec), the LOSVD recovered
non-parametrically also becomes strongly bi-modal with the secondary component
clearly visible in Fig. 1.
4.2. STELLAR POPULATIONS AND EMISSION LINE RATIOS
The radial variations of the stellar population parameters derived from the SSP
fitting of the long-slit data are shown in Fig. 3 (right-top). The mean stellar age
is very old (T ? 15 Gyr) being at the limit of the SSP model grid in the central
20 arcsec. The metallicity in the centre is close to the Solar value decreasing down
to [Z/H] ≈ −0.4 dex in the outer disc. The results of spectrum fitting using a two-
component model are presented in Fig. 3 (right-bottom). “Bulge” (blue points)
and “disc” (black points) components do not have significant differences of the age
radial profile being both very old with the “bulge” reaching the limiting age in
the models. The “disc” metallicity decreases from [Z/H] ≈ 0.1 dex in the centre
to -0.2 dex at large radii. Our analysis of the virtually noiseless (S/N > 100)
SAURON dataset (see Fig. 4, right columns) agree very well with the long-slit
data. The mean stellar parameters of the “bulge” component are T > 15 Gyr
and [Z/H] ≈ −0.1 dex, while for the “disc” component they are T ≈ 14 Gyr and
[Z/H] ≈ 0.1 dex.

Page 6

6
I. Katkov et. al.
The analysis of emission line ratios using the classical diagnostics [Oiii]/Hβ
vs [Nii]/Hα (Baldwin et al. 1981) rules out the ongoing star formation in the
disc of NGC 524 leaving the space for the shockwave excitation or the nuclear
activity. However, the latter mechanism cannot explain the large spatial extent of
the emission line region.
4.3. DISCUSSION
A very good agreement between the dynamically cold “disc” component of the
stellar LOSVD and the ionised gas kinematics suggest that the gas is rotating in
the plane of the main stellar disc. A small discrepancy of the rotation (20 km s−1)
can be explained by the asymmetric drift and well corresponds to the observed
stellar velocity dispersion of the disc component of ∼ 100 km s−1. Old stellar
population and the emission-line diagnostics rule out both, recent and ongoing
star formation in the gaseous disc. The second kinematical component (blue pro-
files in Fig. 3) probably corresponds to two different structures at different radii.
The dynamically hot inner part without much rotation is a manifestation of the
compact central pseudo-bulge, while at R > 20 arcsec we see the presence of a
counter-rotating disc component that is supported by the drops in the velocity
dispersion (300 to <100 km s−1) and metallicity (0.0 to −0.3 dex) profiles.
The origin of NGC 524 has to be investigated in detail using state-of-art nu-
merical simulations. Right now we can speculate about its evolution based on the
observational results we have. NGC 524 might have originated from the face-on
collision of two initially counter-rotating co-planar giant disc galaxies. The metal-
licity difference between the two components suggests the merger mass ratio of
about 1:4. The gas in the main stellar disc of NGC 524 might have survived from
the original galaxy or collected later from mergers with low-mass satellites or from
the accretion from the cosmic filaments. However, its surface density is still below
the threshold and therefore it prevents the start of the star formation.
REFERENCES
Afanasiev, V. L., Moiseev, A. V. 2005, Astronomy Letters, 31, 193
Bacon, R., Copin, Y., Monnet, G. et al. 2001, MNRAS, 326, 23
Baldwin, J. A., Phillips M. M., Terlevich R., 1981, PASP, 93, 5
Cappellari, M., Copin, Y. 2003, MNRAS, 342, 345
Chilingarian, I. V., Prugniel, P., Sil’chenko, O. K., Afanasiev, V. L. 2007a, MN-
RAS, 376, 1033
Chilingarian, I. V., Prugniel, P., Sil’chenko, O. K., Koleva, M. 2007b, In: Stellar
Populations as Building Blocks of Galaxies, Proc. of the IAU Symp. no 241,
eds. A. Vazdekis and R. F. Peletier, p. 175, arXiv:0709.3047
Emsellem E. et al. 2004, MNRAS 352, 721
Garcia A.M. 1993, A& A Suppl. Ser. 100, 47
Katkov, I.Yu., Moiseev, A.V., Sil’chenko, O.K. 2011a ApJ, submitted
Katkov, I.Yu., Chilingarian I.V. 2011, arXiv:1012.4125
Le Borgne, D., Rocca-Volmerange, B., Prugniel, et al. 2004, A&A, 425, 881
Macchetto F., Pastoriza M., Caon N., Sparks W. B., Giavalisco M., Bender R.,
Capaccioli M. 1996, A& A Suppl. Ser. 120, 463
van der Marel, R.P. & Franx, M. 1993, ApJ, 407, 525
Magrelli G., Bettoni D., Galletta G. 1992, MNRAS 256, 500

Page 7

A complex stellar LOSVD in NGC 524
7
Mulchaey J., Davis D., Mushotzky R., Burstein D. 2003, ApJ Suppl 145, 39
Press, W.H., Teukolsky S.A., Vetterling, W.T., Plannery, B.P. 2007, Numerical
Recipes: The Art of Scientific Computing, 3d edn, Cambridge Univ. Press
Sil’chenko O.K. 2000, AJ 120, 741
Sil’chenko O.K. 2009, In: The Galaxy Disk in Cosmological Context, Proc. of the
IAU Symp. no 254, eds. J. Andersen, J. Bland-Hawthorn, and B. Nordstrom,
p. 173
Simien F., Prugniel Ph. 2000, A& A Suppl. Ser. 145, 263