Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://sn.sai.msu.ru/~sil/preprints/aspb121.pdf Äàòà èçìåíåíèÿ: Sun Jul 12 01:39:46 2015 Äàòà èíäåêñèðîâàíèÿ: Sat Apr 9 23:08:15 2016 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: universe

ISSN 1990-3413, Astrophysical Bulletin, 2014, Vol. 69, No. 2, pp. 121­140. c Pleiades Publishing, Ltd., 2014. Original Russian Text c I.Yu. Katkov, O.K. Sil'chenko, V.L. Afanasiev, 2014, published in Astrofizicheskij Byulleten, 2014, Vol. 69, No. 2, pp. 129­149.

Properties of Stellar Populations in Isolated Lenticular Galaxies
I. Yu. Katkov
1

1, 2 *

, O. K. Sil'chenko1 , and V. L. Afanasiev

3

3

Sternberg State Astronomical Institute, Moscow State University, Moscow, 119992 Russia 2 Physics Department, Lomonosov Moscow State University, Moscow, 119991 Russia Special Astrophysical Observatory, Russian Academy of Sciences, Nizhnii Arkhyz, 369167 Russia
Received December 12, 2013; in final form, March 24, 2014

Abstract--We present the results of observations of a sample of isolated lenticular galaxies, performed at the SCORPIO and SCORPIO-2 spectrographs of the 6-meter BTA telescope of the SAO RAS in the long-slit mode. By direct spectra approximation, using the evolutionary synthesis models, we have measured the radial profiles of the rotation velocity as well as the dispersions of velocities, average age, and average metallicity of stars in 12 objects. The resulting average ages of the stellar population in bulges and discs fill an entire range of possible values from 1.5 to 15 Gyr which indicates the absence in the isolated lenticular galaxies, unlike in the members of groups and clusters, of a certain epoch when the structural components are formed: they could have been formed at a redshift of z > 2 as well as only several billion years ago. Unlike the S0 galaxies in a more dense environment, isolated galaxies typically have the same age of stars in the bulges and discs. The lenses and rings of increased stellar brightness, identified from the photometry of 7 of 11 galaxies, do not significantly differ from the stellar discs by the properties of stellar populations and velocity dispersion of stars. We draw a conclusion that the final arrangement of the morphological type of a lenticular galaxy in complete isolation is critically dependent on the possible modes of accretion of the cold external gas. DOI: 10.1134/S1990341314020011 Key words: galaxies: elliptical and lenticular--galaxies: stellar population--galaxies: evolution

1. INTRODUCTION The problem of the scenarios of formation and evolution of galaxies is the key issue in modern extragalactic astronomy and observational cosmology. The galaxies are formed under the influence of a large number of physical factors, which are often insufficiently known by the theorists in detail. The main issue here is to identify the most important factors which are crucial during the formation and evolution of galaxies of the given morphological type. The type of lenticular (S0) galaxies was proposed by Edwin Hubble as hypothetical in 1936 [1] in order to fill in the intermediate position between the elliptical and spiral galaxies. It was assumed that the objects of this type have large-scale stellar discs as observed in the spiral galaxies but do not have any noticeable star-forming regions and spiral pattern in the stellar discs. Their smooth reddish view and probably an old average age of stars makes them similar to the elliptical galaxies. An intermediate position of lenticular galaxies between the purely spheroidal stellar systems and spiral galaxies, in which the contribution
*

E-mail: katkov.ivan@gmail.com

of the bulge to the total luminosity monotonically decreases with the transition from early to late types (from left to right along the morphological Hubble sequence), gives rise to a natural assumption that the S0 galaxies should possess large bulges. However, the detailed surface photometry of the images of S0 galaxies has shown that the bulges in them can be both very large and tiny [2]. Stemming from these results, the old idea of Van den Bergh [3], stating that the lenticular galaxies should be forming a sequence parallel to the spiral galaxies in the Hubble diagram and the connection between the (close) position in the diagram of the S0 (a, b, c) and spiral galaxies of the corresponding subtypes is given by the "bulge/disc" luminosity ratio [4, 5], becomes increasingly popular. It would seem that such a turn in the understanding of the evolutionary sense of the Hubble sequence would only reinforce the conventional wisdom about the formation of lenticular galaxies via the cessation of star formation in spiral galaxies: the evolutionary stage of transformation of the progenitor galaxy to the resulting S0 galaxy is much easier to imagine when both galaxies have the same bulge/disc ratio. However, it should be noted that if the contribution of the bulge to the total luminosity in an S0 galaxy
121


122

KATKOV et al.

is the same as in the spiral galaxy located nearby on the morphological diagram, then this leaves open the possibility of reverse transformation, the transformation of an S0 galaxy into a spiral which would have been quite impossible in the presence of a larger bulge in the S0 galaxy. The literature discusses a considerable number of physical processes that could cease star formation in the disc of the spiral galaxy. Here are some of them: (1) direct collisions of galaxies [6, 7]; (2) tidal effects from the dark halo of the cluster/group [8]; (3) "harassment," or the gravitational tidal influence of galaxies on each other at a sufficiently fast flyby [9]; (4) ram pressure of the hot intergalactic medium [10, 11]; (5) "starvation," or termination of star formation as a result of the disappearance of the external reserves of gas, previously maintaining the gas accretion onto the disc of the galaxy and feeding the processes of star formation [12]. These processes are closely related to the dense environment of galaxies, because only the clusters and rich groups of galaxies with their massive dark haloes can provide the necessary density of the hot intergalactic medium for the ram pressure and close mutual locations of galaxies for the appearance of tidal effects. We know that lenticular galaxies are the dominant population of the nearby clusters of galaxies, where their fraction reaches up to 60% [13]. However, the number of S0 galaxies is quite noticeable among the field galaxies as well: according to the APM survey [14], the fraction of lenticular galaxies in the nearby Universe is about 15%, and they are the second by the frequency of occurrence after the spirals. Furthermore, there are examples of completely isolated lenticular galaxies [15]. There appears a question which has not yet been raised about the origin of such galaxies. Under the effect of which physical mechanisms they were formed, and how do these mechanisms differ from those that work in the dense environment? Despite the apparent deficit of mechanisms responsible for the morphological transformation of isolated galaxies compared to the members of clusters and groups, it is wrong to assume that the isolated galaxies evolve completely independently, as in the "closed box" scenario. Some recent studies point at this. In our work [16] we have demonstrated that a completely isolated early-type spiral galaxy

NGC 7217 over the last 5 Gyr experienced at least two events of companion merger. In a locally isolated S0 galaxy NGC 4124 we also found traces of minor merging, which had apparently occurred 2­3 Gyr ago and provoked a central burst of star formation [17]. Furthermore, it was recently found out that lenticular field galaxies often possess significant amounts of gas. Moreover, it is exactly in the rarefied environment that a galaxy often reveals different kinematics of stars and gas, pointing to the external origin of gas [18]. Thus, the studies of the properties of isolated lenticular galaxies would allow to concentrate on the evolutionary mechanisms, connected either with the internal disc instabilities or just the external accretion of gas and/or companions. It should be noted that the accretion of gas and/or minor merging events may not only cease the star formation in the disc but, on the contrary, provoke it [19, 20]. The objects presented in this paper are the nearby, strongly isolated lenticular galaxies, for which we have conducted the spectroscopic observations with the aim to determine the properties of their stellar populations as well as kinematics of stars and gas. The properties of ionized gas in the investigated objects are discussed in a separate paper [21], while in this paper we concentrate only on the properties of stellar populations. The paper is structured as follows: Section 2 is devoted to the description of the selection of studied objects; the features of the spectral observations, reduction, and data analysis are discussed in Section 3; Section 4 presents the results on each galaxy; Sections 5 and 6 contain the discussion of the results and conclusions of the study. 2. SAMPLE SELECTION We have compiled a sample of the nearby isolated lenticular galaxies using an approach that has recently been developed in the Laboratory of Extragalactic Astrophysics and Cosmology of the SAO RAS by Karachentsev, Makarov, and their coauthors. This approach was proposed and applied to the galaxies of the Local Supercluster and its surroundings with a view of their association into pairs [22], triplets [23], and groups [24], as well as for the identification of isolated galaxies [25]. The data on the radial velocities, apparent magnitudes, and morphological types of galaxies were taken from the updated HyperLeda database1 and the NED database,2 supplemented by the radial velocity measurements from the SDSS, 6dF, HIPASS, and
1 2

http://leda.univ-lyon1.fr/ http://ned.ipac.caltech.edu/ Vol. 69 No. 2 2014

ASTROPHYSICAL BULLETIN


PROPERTIES OF STELLAR POPULATIONS IN ISOLATED LENTICULAR GALAXIES

123

ALFALFA surveys. Clustering was carried out for the galaxies with the radial velocity relative to the Local Group of VLG < 4000 km s-1 and galactic latitude of |b| > 10 . A peculiarity of the clustering algorithm proposed by the authors is to account for the individual characteristics of galaxies, specifically, the indicator of the galaxy mass--the K -band luminosity. Uniting the galaxies in systems pairwise, the authors suppose that each virtual pair has to satisfy the negative total energy condition and the condition of finding its components within the "zero velocity sphere," i.e., the galaxies of the pair do not recede from each other as a result of the Hubble expansion of space. The clustering algorithm involves an iterative revision of all galaxies of the original sample for their further integration of the bound pairs of galaxies that have common members in the groups or clusters. The details of the algorithm are given in the above papers. One of the intermediate products of the algorithm is a list of paired isolation indices between any two galaxies of the sample. The isolation index (II ) of two galaxies is the value characterizing the dynamic mutual influence of both components on each other. In the case of an unbound pair, log II is positive and equal to the logarithm of the number of times by which it is necessary to increase the mass of one of the components so that the pair would meet the given association criteria. And conversely, in the case of a bound pair, log II is negative and equal to the logarithm of the number of times by which the mass has to be reduced to become unbound. The same value of the isolation index may be realized for a wide pair of galaxies of comparable luminosity and for a tight pair consisting of a giant galaxy and a nearby but faint companion. The authors of this approach have kindly provided us the information about mutual isolation indexes for all galaxies of the Local Supercluster and its environs. Using this information, we have selected for our study the lenticular galaxies having II > 2.5 both for the galaxies of higher and lower luminosity as compared to the considered one. We have conducted the spectroscopic observations of 12 galaxies from the resulting list in order to study the properties of stellar populations. Table 1 shows the five most influential neighbors of higher and lower luminosity for each studied galaxy as well as the information about the morphology, radial velocities, and absolute magnitudes. In the cases of the NGC 16 and NGC 3098 galaxies, the restrictions on the isolation from the potential companions of lower luminosities are not met. Both galaxies have a lowluminosity neighbor with II = 1.9, but due to a very large luminosity difference MK 5 of values (the masses differ by about 100 times), we believe that
ASTROPHYSICAL BULLETIN Vol. 69 No. 2 2014

the companions cannot have a significant gravitational effect on large galaxies, so these galaxies do not egress from the generally accepted selection criteria for isolated galaxies. 3. OBSERVATIONS AND DATA ANALYSIS

3.1. Observations
The observations of the sample of studied galaxies were carried out in the prime focus of the 6-m BTA telescope of the SAO RAS over the period of 2011­2012. The spectral data for all the galaxies, except NGC 6615 and NGC 6654, were obtained using the SCORPIO-2 focal reducer [26] in the longslit mode with the slit size of 6 â1 . In the observations we have used the VPHG 1200@540 holographic grating (the grism), which provides a spectral ° resolution of FWHM 4 A in the operating range ° of 3800­7300 A. This spectral range includes both the strong absorption lines, like Mg, Fe, G-band, and a number of emission lines (H, H , [O III], [N II], etc.) that allows to explore both the kinematics, age, and chemical composition of the stellar component and the kinematics of ionized gas, and at the same time to diagnose the excitation mechanisms of ionized gas. The detector used was an E2V CCD 42-90 with the chip sized 2kâ4k which at the 1â2 binning mode readout provides a spatial scale along the slit of 0 . 357 per pixel and disper° sion of 0.86 A per pixel. Unlike the other galaxies, NGC 6615 and NGC 6654 were observed in another instrumental configuration, namely, with the SCORPIO instrument [27] and the VPHG 2300G holographic grating that provided a spectral resolu° ° tion of 2.2 A in the range of 4800­5600 A, and the use of the EEV CCD 42-40 detector with the chip sized 2kâ2k gave the same scale along the slit with the ° dispersion of 0.37 A per pixel. During the observations the slit was oriented along the major axes of the galaxies. The log of observations (Table 2) gives the dates of observations, total exposures, average seeing during the exposure of each galaxy, and the position angle of the slit.

3.2. Initial Data Reduction
Initial data reduction was performed using the original programs, written in the IDL environment, and contained the following steps: accounting for the registration system bias by subtraction of the averaged zero exposure frame from all the images; accounting for the uneven illumination and inhomogeneities in the CCD sensitivity by the spectra of the flat-field calibration lamp; removal of the traces


124

Table 1. Table of the environment of the studied galaxies Morph. type Vsy
s K

Name M Name M
K K

Neighbors "from the top" V 278 31.0 PGC 046033 -3.949 -48 II Name M V II 4.2

Neighbors "at the bottom"

HyperLeda NGC 4814 S0 NGC 5322 NGC 5430 UGC 08237 IC 0356 S0-a S0 NGC 1184 IC 0342 NGC 6951 NGC 7817 -0.190 E-S0 SAB0 edge-on NGC 7619 NGC 7331 NGC 2339 S0-a S0/a 1774 -22.725 NGC 2365 NGC 3031 NGC 4472 IC 0342
- +

NED 1.034 S0 2.242 1.491 -224 150.4 NGC 5109 993 126.9 NGC 4964 -0.851 -1.563 2913 -22.713 NGC 5218

IC 0875

1.175 -121 107.9 SDSSJ 1324... -3.577 -126 10.3 268 57.3 660 78.6

0.315 -86 179.2 SDSSJ 1259... -4.268 -83 84.0 1.793 1131 158.6 UGC 12247 0.925 -264 311.8 UGC 12921 -0.361 1997 376.8 UGC 12160 1.074 517 473.3 UGC 12069 745 214.2 PGC 000446 -4.960 298 46.2 -4.322 -461 64.6 -2.427 -477 73.5 -1.135 404 74.2 -2.052 -405 106.8 -5.443 -140 1.9 -5.043 -212 22.8 -5.736 -513 191.4 1.071 -689 1296.4 PGC 001153 0.410 2179 1339.5 UGC 00285 1.168 -396 135.1 UGC 03775 0.681 -429 400.0 UGC 03691 0.584 1672 402.6 PGC 097214 2.552 -6.733 -636 312.1 -3.498 872 459.0 -4.638 -244 12.1 -0.708 -327 81.7 -2.601 -298 102.4 901 635.7 PGC 2807004 -5.452 0.618 1534 1054.5 ­ 40 172.1 -5.371 -154 470.6 KATKOV et al.

IC 1502

2237 -23.704 NGC 3031 -0.394 2135 280.1 UGC 12504

NGC 0016

3300 -24.511 NGC 0253 -0.100 3024 1029.7 UGC 12873 NGC 3031 -1.201 3198 1212.8 PGC 087206

ASTROPHYSICAL BULLETIN

Vol. 69

NGC 2350

No. 2

2014


Table 1. (Contd.) Morph. type Vsy
s K

Name M Name M
K K

Neighbors "from the top" V 117 29 27.1 IC 2520 271 40.4 UGC 05588 -1.545 -2.090 4.1 PGC 2806869 -4.981 II Name M V II

Neighbors "at the bottom"

HyperLeda NGC 3190 S0-a 1.216 0.955 1.033 1.618 1.808 0.782 1.568 1.307 1.970 0.520 116 168 544 53.3 NGC 3026 edge-on NGC 2964 NGC 3379 NGC 3190 S0 NGC 3245 NGC 3227 NGC 3379 NGC 6574 S0-a SB0 ? NGC 6548 NGC 6501 NGC 6587 S0-a
+

NED 1.456 S0 NGC 3227 1305 -22.170 NGC 3245

ASTROPHYSICAL BULLETIN

58 1.9 151 24.1 95 33.0 10 37.5

NGC 3098

44 49.0 PGC 029347 -2.527

-1.433 -109 38.5 252 13.2

PROPERTIES OF STELLAR POPULATIONS IN ISOLATED LENTICULAR GALAXIES

Vol. 69 No. 2 2014

2.5 PGC 166105 -5.029

NGC 3248

S0

1356 -21.818 NGC 3301

6.3 PGC 2806870 -5.424 -115 16.0

80 15.9 PGC 030270 -3.603 -78 24.1 321 21.4 UGC 05588 -1.738 145 35.6 47 36.9 396 21.7 UGC 11214 -1.687 34 2.5 594 38.5 PGC 031387 -2.524

NGC 6615

2868 -23.779 UGC 11057 -0.299 -161 92.4 PGC 061685 -2.895 -198 2.9 0.244 498 110.9 PGC 061658 -1.804 -319 13.5 0.432 -224 123.7 UGC 11168 -0.981 381 22.4 0.390 -429 138.0 PGC 061621 -0.776 -274 29.4 NGC 6643 -0.203 459 21.1 PGC 062387 -3.539 757 189.7 NGC 6654A NGC 6951 0.948 484 225.3 UGC 10892 -3.756 -4.568 NGC 3031 -0.520 2102 273.3 PGC 062173 -3.925 NGC 6911 0.517 -576 470.4 UGC 11295 101 18.4 380 25.6 26 33.3 534 51.3 -4.629 -427 57.8

NGC 6654

(R )SB0/a(s) 2204 -23.830 NGC 6340 -0.651

125


126

Table 1. (Contd.) Morph. type Vsy
s K

Name M Name M
K K

Neighbors "from the top" V 41 -0.813 31.1 UGC 11457 -4.740 -10 II Name M V II

Neighbors "at the bottom"

HyperLeda NGC 6764 -0.001 S0 1.934 -1091 232.1 NGC 6757 1.045 0.376 -853 505.2 PGC 063313 0.083 3.488 2389 577.0 NGC 6796 90 344.4 UGC 11502 NGC 6703 NGC 6829 NGC 6946 NGC 0253 S0 NGC 3031 IC 1459 NGC 7507 NGC 2712 S0 NGC 2639 NGC 3031 NGC 2681 NGC 4472 S0 S0? 1782 -21.710 NGC 5353 NGC 5611 NGC 5582 NGC 5194 S0? 1794 -22.633 NGC 2768 1.827 0.334 2.387 SAB0^0? 1077 -20.923 NGC 7727 S0 2741 -23.520 NGC 6824

NED

6.1 75 36.9

NGC 6798

0.047 -335 112.3 0.604 -1241 193.8 -0.424 263 303.2 61 -7 2.5 3.4 23 11.9 -37 169.8 0.579 -809 626.0

801 137.0 PGC 069415 -1.826 975 1006.1 PGC 1028063 -3.405

NGC 7351

3.637 -881 899.2 PGC 069224 -1.777 4.208 -599 1042.6 PGC 982181 -5.031 3.518 -558 1163.6 PGC 069293 -50 328 66.2 UGC 04659

KATKOV et al.

-2.959 69.3 PGC 023834 -3.120 71.1 UGC 04543

17 26.7 -10 27.5 -1.791 -189 58.5

UGC 04551

2.259 -1458 0.677 0.009 3.567

1691 127.7 UGC 04648 1048 159.4 UGC 04922 909 88.7 NGC 5727

-3.851 -138 93.2 -1.653 -228 168.9 -2.425 206 42.1 3.466 -457 131.0 PGC 2080256 -3.641 -196 69.1 0.804 -243 152.5 PGC 052694 -2.617 0.986 2.284 247 159.8 UGC 09597 1221 194.7 NGC 5798 -3.886 -0.596 132 72.0 -30 123.5 -88 127.6

ASTROPHYSICAL BULLETIN

Vol. 69

UGC 09519

No. 2

2014


PROPERTIES OF STELLAR POPULATIONS IN ISOLATED LENTICULAR GALAXIES

127

of cosmic particles by using the L.A.Cosmic [28] algorithm, implementing the Laplacian filter for particle detection, and by addition of spectra; construction of a two-dimensional dispersion equation by the spectrum of the helium-neon-argon calibration lamp and the further linearization of spectra with the characteristic ° accuracy of 0.03­0.06 A depending on the grating used; subtraction of the background spectrum of the night sky; conversion of the instrumental fluxes in the absolute ones using the spectra of spectrophotometric standard stars. During the observations in semidarkness the night sky contribution is large enough and varies with time. In such cases we performed the night sky subtraction before the spectrum addition to fit the parameters of the procedure for each spectrum the best way possible. The final result of the primary data reduction were not only the spectra of the objects but also the frames of the error level that were calculated based on the Poisson photon statistics and readout noise and were transformed along with the images of the object spectra at each reduction stage. In addition to the spectra of objects and spectrophotometric stellar standards, the spectra of dawn or twilight sky were also analyzed in the process of data reduction, which, in the essence, represent the solar spectrum convoluted with the instrumental profile of the spectrograph. Therefore, the analysis of the spectra of the dawn or twilight sky allows to determine the behavior of the instrumental profile of the spectrograph along the spectrograph slit and along the direction of dispersion. The former is important to subtract the contribution of the night sky, and the latter allows to accurately determine the parameters of the kinematics of galaxies. The spectra of the dawn sky were reduced in the same manner as the spectra of the objects. The reconstruction of the instrumental profile and its use in analyzing the galaxies are described in detail in the following Section 3.3.

Table 2. Log of observations No. 1 2 3 Name Date Exposure, Seeing, PA, s arcsec deg 2700 2700 1800 6000 5400 2700 7200 6600 5400 3600 8400 4500 2.5 2.5 2.0 1.6 1.2 3.0 1.0 1.3 2.5 2.0 2.0 2.0 -30 52 16 -73 -90 -45 -15 0 -30 0 -67 -105

IC 875 Apr 23, 2012 IC 1502 Nov 19, 2011 NGC 16 Nov 20, 2011

4 NGC 2350 Dec 13, 2012 5 NGC 3098 Apr 18, 2012 6 NGC 3248 Apr 22, 2012 7 NGC 6615 Sep 19, 2012 8 NGC 6654 Sep 20, 2012 9 NGC 6798 Nov 20, 2011 10 NGC 7351 Nov 19, 2011 11 UGC 4551 Dec 12, 2012 12 UGC 9519 Apr 24, 2012

3.3. Subtraction of Night Sky Contribution
In the analysis of objects of low surface brightness, special attention should be given to the careful subtraction of the contribution of the night sky, underestimation of which may lead to the systematic errors in determining the parameters of stellar populations of the studied objects [29]. In this paper we determine the properties of stellar populations of the structural components of the galaxy, including the stellar discs of low surface brightness. Therefore, we believe it is necessary to describe in detail the procedure of the night sky spectrum subtraction. We have previously proposed a refined technique for subtracting the spectrum of the night sky for the long-slit spectroscopy of the low surface brightness objects in case of strong variations of the instrumental
ASTROPHYSICAL BULLETIN Vol. 69 No. 2 2014

profile along the slit [29]. In that procedure the sky model is constructed by recalculating the spectrum of the sky from the edge regions on the slit to the region containing the spectrum of the object, applying the deconvolution procedure to the reference spectrum. The spectrum of the dawn/twilight sky, which carries the data on the variations of the instrumental profile, is supposed to be used as a reference spectrum. Unfortunately, during the observations the spectra of the dawn/twilight sky were not always observed due to a sudden worsening of weather conditions. The standard method, which does not require the use of a reference spectrum and accounts in the zero approximation for the profile variations along the slit, consists in the approximation of the sky spectrum by the low-power polynomial (2­4) in each column of the spectrum image and its recomputation for the area of the object. This method works well for the spectra of non-extended objects. However, this condition is not always fulfilled in the long-slit observations of galaxies. Therefore, we propose another way to build the spectrum of the night sky, which is based on the extrapolation procedure in the frequency domain. The spectrum of the night sky at a given position on the slit y ­S (, y ) can be written as a convolution of the "true" spectrum of the night sky S0 () with the instrumental profile L(, y ): S (, y ) = S0 () L(, y ). (1)


128

KATKOV et al.

In the frequency domain the convolution procedure becomes the multiplication procedure. Hence, if we perform a one-dimensional Fourier transform along the direction of dispersion, then FFT[S (, y )] = FFT[S0 ()] FFT[L(, y )]. (2) As the analysis of the spectra of the dawn and twilight sky has shown, the shape of the instrumental profile of the SCORPIO/SCORPIO-2 spectrographs varies rather monotonously along the slit, therefore its Fourier transform FFT[L(, y )] varies along the slit as monotonously. The first factor in the Fourier transform of the night sky spectrum is a constant function, hence in general the Fourier transform of the night sky along the slit varies monotonously. Using this fact and applying the standard polynomial extrapolation procedure to the spectrum of the night sky in the frequency domain, i.e., to its Fourier transform, and performing the subsequent inverse Fourier transform, we can construct the model of the spectrum of the night sky. The Fourier transform of the spectral image is a complex value, this is why the extrapolation should be performed separately for the real and imaginary parts of the Fourier transform. In this version of the model of the night sky, the reference spectrum of the dawn/twilight sky is not required. At the same time the quality of the model is comparable to the method based on the deconvolution procedure [29].

3.4. Data Analysis
Before the analysis of the spectra of galaxies, we analyzed the spectrum of the dawn sky in order to determine the variation of the parameters of the spectrograph instrumental profile (LSF--Line Spread Function) along and across the direction of dispersion which is required to correctly determine the internal kinematics of stars and gas in the galaxies. To do this, we have split the image of the dawn sky spectrum into many areas: 32 intervals along the slit and 7 segments along dispersion; in each section the dawn sky spectrum was summed to achieve the typical signal-to-noise relationship S/N = 100 per pixel. Next, we approximated the spectra from each area by the high-resolution solar spectrum, taken from the ELODIE 3.1 stellar spectral library [30], using the ppxf procedure of the per-pixel approximation of spectra [31]. In the approximation of the dawn sky spectrum, the instrumental profile has been parameterized by the Gauss­Hermite series of orthogonal functions [32]. As a result of approximation, we have obtained the instrumental profile parameters for different spectral regions. The mean characteristic width of the instrumental profile (in terms of velocity dispersion) for the observational mode with the VPHG 2300G and VPHG 1200@540

grating is instr = 65 km s-1 and instr = 90 km s-1 respectively. The further analysis consisted of the approximation of the observed absorption spectra of galaxies by the high-resolution models of stellar population. To do this, we used the NBursts software package [33, 34], which is an extension of the ppxf per-pixel spectrum approximation method [31]. The applied approach allows to retrieve the information about the stellar component from all the available spectral regions simultaneously, in contrast to the analysis of the Lick indices of individual lines--the indicators of the properties of stellar populations. The approach of per-pixel approximation of the spectra of galaxies allows to easily exclude from the analysis the regions around the strong emission lines of ionized gas, it allows to avoid their systematic effect on the estimates of the stellar population parameters which is impossible when analyzing the Lick indices. Chilingarian et al. ([33, 34]) have shown that the per-pixel spectrum approximation method provides a 1.5­2 times higher accuracy of finding the stellar population parameters compared to the approach of the Lick indices. The procedure of determining the parameters of the stellar population consists in the nonlinear minimization of the quadratic difference 2 between the observed and model spectra. We have used synthetic spectra of stellar populations as model spectra. They were computed with the Pegase.HR [35] evolutionary code based on the ELODIE 3.1 highresolution stellar spectral library [30] for the simple star formation history in the form of one short burst (SSP, Simple Stellar Population). The SSP stellar population model is set by the age of the star formation burst T (Gyr) and metallicity [Z/H] (dex), while the Salpeter initial mass function [36] is considered to be fixed in the model. To determine the stellar kinematics of the galaxy, the line-of-sight velocity distribution (LOSVD), which is given as the Gauss­ Hermite quadrature [32], is entered in the model spectrum. Furthermore, the multiplicative continuum is included in the model which allows to consider the effect of the interstellar extinction on the shape of the spectrum of the galaxy as well as the errors of absolute calibration of the fluxes both in the observational data and in the stellar library, based on which the stellar population models were considered. To take into account the effect of the spectrograph on the spectrum of the galaxy, before the approximation of the observed spectra we convoluted the grid of stellar population models with the previously determined instrumental profile. The presence even of the weak emission lines and/or the residues from the subtraction of the strongest lines of the night sky can offset the stellar population parameter estimates. Hence, to eliminate
ASTROPHYSICAL BULLETIN Vol. 69 No. 2 2014


PROPERTIES OF STELLAR POPULATIONS IN ISOLATED LENTICULAR GALAXIES Table 3. Table of averaged parameters of the bulges, discs, lenses/rings Galaxy Averaging range, Number arcsec of measurements Bulge IC 875 IC 1502 NGC 16 NGC 2350 NGC 3098 NGC 3248 NGC 6615 NGC 6654 NGC 6798 NGC 7351 UGC 4551 UGC 9519 4 ­7 4 ­7 2 ­5 4 ­7 4 ­7 4 ­7 4 ­7 2 ­5 4 ­7 4 ­7 4 ­7 4 ­7 10 11 10 8 8 10 8 9 6 10 8 9 Disc IC 875 IC 1502 NGC 16 NGC 2350 NGC 3098 NGC 3248 NGC 6615 NGC 6654 NGC 6798 NGC 7351 UGC 4551 UGC 9519 13­45 7­25 6­30 10­40 25­60 10­39 40­60 35­60 8­55 17­45 38­80 16­30 16 11 18 15 18 31 0 3 18 7 3 4 2.9 1.6 1.3 5.1 3.9 5.8 7.3 4.4 2.9
±0.9 ±1.6

129

T, Gyr
±0.7 ±0.9

[Z/H], dex -0.16 -0.04 -0.04 -0.13 -0.10 -0.11 -0.26 -0.19 -0.20 -0.37 -0.28 -0.12 -0.32 -0.13 -0.19 -0.00 -0.22 -0.21 -0.06 -0.27 -0.57 -0.74 -0.32
±0.05 ±0.06 ±0.05 ±0.08 ±0.02 ±0.05 ±0.05 ±0.07 ±0.05 ±0.08 ±0.08 ±0.06

[Mg/Fe], , dex km s
±0.04

-1

4.3 5.4 1.6 5.4 4.8

0.20 0.3 0.19 0.00 0.00 0.24 0.23 0.13 0.15 0.04

110 168 103 73 77 172

±9

17.6

±0.1 ±0.04

±16 ±6

±0.8 ±0.3 ±0.4 ±0.6 ±1.5 ±1.4

­
±0.02 ±0.05 ±0.03 ±0.04 ±0.04 ±0.06

±15

±6 ±5 ±5 ±5 ±7

10.8 12.2 8.0 2.2 2.5

129 158 115 29 158 76

±1.9