Mon. Not. R. Astron. Soc. 000, 1–6 (2006)

Printed 24 July 2006

(MN LATEX style file v2.2)

A 3D View of The Central Kiloparsec Regions of the Seyfert Galaxy NGC1358 G. Dumas1,2⋆ , E. Emsellem1, P. Ferruit1 1

Universit´ e de Lyon 1, Centre de Recherche Astronomique de Lyon, Observatoire de Lyon, 9 avenue Charles Andr´ e, F-69230 Saint-Genis Laval, France ; CNRS, UMR 5574 ; Ecole Normale Suprieure de Lyon, Lyon, France. 2 ARI Liverpool John Moores University, UK

Accepted ... Received ... in original ...

ABSTRACT

We present a study of the stars and ionized gas in the central kpc of the Seyfert 2 galaxy NGC 1358, using integral-field spectroscopy obtained with OASIS (CFHT). We have derived both stellar and gaseous kinematical maps, including higher order Gauss-Hermite stellar velocity moments. Emission line (Hα, Hβ, [O iii], [N ii], [S ii]) and emission line ratios ([O iii]/Hβ and [N ii]/Hα) maps have also been constructed. These data reveal an inner gaseous spiral structure (. 1 kpc) which clearly shows up as a perturbation in the ionised gas kinematic maps, but is not seen in the stellar velocity maps. The presence of an inner bar could explain this feature. Simple dynamical models are built to constrain the mass-to-light ratio of the galaxy and the properties of the presumed inner bar. Key words: galaxies : active - galaxies : kinematics and dynamics - galaxies : individual (NGC1358).

1

INTRODUCTION

Seyfert galaxies are ideal targets to study the AGN phenomenon in the nearby Universe. At the kiloparsec scale, observed inner bars or spiral perturbations are often presumed to be efficient drivers of gas toward the very central region ( Erwin and Sparke 2002, Pogge and Martini 2002). Such structures are spatially well resolved with ground-based facilities, and can be mapped in detail using integral-field spectroscopy. In this context, we have conducted an observational campaign using the OASIS 3D spectrograph at the CFHT to probe the central kiloparsec of a few Seyfert galaxies. In this short paper, we focus on the central region of the Seyfert 2 barred galaxy NGC 1358. With an inclination of i ∼ 38◦ and a distance of 53.6 Mpc (1′′ is ∼ 260 pc; Ho et al. 1997b), the OASIS data allowed us to reveal and resolve an inner gaseous spiral structure of less than 1 kpc in diameter. In Section 2, we provide the characteristics of the OASIS data, and briefly describe the resulting two-dimensional maps of the stellar and gas distributions and kinematics. These are finally discussed in the context of a simple bar model in Section 3.

2

NGC 1358 was observed on November 11 2000, with OASIS at the Canada-France-Hawaii Telescope (Mauna Kea, Hawaii). A first exposure was obtained using the MR1 spectral configuration which covers the spectral range 48005550 ˚ A, providing the [O iii], Hβ and [N i] emission lines and various stellar absorption lines (Fe, Mgb). The spatial sampling was 0.27 arcsec, yielding a field of view of 10.′′ 4×8.′′ 3 arcsec2 . Two other overlapping pointings were obtained in the MR2 configuration with the same spatial sampling, providing a field of view of 16.′′ 1×13.′′ 7 including the [O i], [N ii], Hα and [S ii] emission lines (spectral range : 6200-6990 ˚ A). The resulting seeing was about 1 arcsec (Full Width at Half Maximum). The two data sets were reduced using the dedicated XOASIS software (Rousset 1992), resulting in fully calibrated datacubes. After merging of the exposures, we obtained two datacubes, one in each spectral configuration. In Figure 1, we present the MR1 recontructed intensity map of NGC 1358 as compared to the larger scale HST/WFPC2 (F606W) (Malkan et al. 1998) and R-band Digital Sky Survey images.

2.1 ⋆

E-mail : [email protected]

c 2006 RAS

OBSERVATIONS AND MAPS

Stellar kinematics

In order to derive reliable stellar kinematics, we spatially binned our merged datacubes using the Voronoi 2D binning

2

G. Dumas et al.

Figure 1. Images of NGC 1358. Left : DSS red image; middle : HST/WFPC2 image (F606W); right : OASIS MR1 continuum reconstructed image.

scheme of Cappellari and Copin (2003) setting the minimal signal to noise ratio to 40. We derived the stellar kinematics from the binned datacube in the MR1 configuration. This was achieved by using the penalized pixel fitting (pPXF) method (Cappellari and Emsellem 2004) with an optimal stellar template built from a combination of stellar spectra from the Indo-US Library (Valdes et al. 2004). The line-ofsight velocity distribution was parameterized with a GaussHermite expansion up to the fourth order, providing maps of the mean velocity V , the velocity dispersion σ and the third and fourth Gauss-Hermite moments h3 and h4 . The stellar continuum in the MR2 configuration datacube could not easily be fitted in the same way, because of the lack of strong absorption lines in this wavelength range to constrain the stellar kinematics. We therefore used the results of the pPXF fit in the MR1 spectral region to build an optimal template for this spectral configuration. The fieldof-view in the MR2 configuration being larger than the MR1 one, we had to extrapolate the values of the velocity V and velocity dispersion σ obtained from MR1 configuration. The derived MR2 stellar continuum does not significantly depend on the details of this extrapolation, because of the very shallow gradients both in V and σ within the OASIS field-ofview. The obtained stellar continuum was subtracted from the spectra of the merged cubes in the two spectral configurations, resulting in pure emission line datacubes. Figure 2 shows the mean velocity V , and velocity dispersion σ maps (all maps have the classical North-up East-left orientation). The stellar velocity field shows a mild rotation with a major-axis with a PA of 11 ± 1◦ , aligned with the orientation of the galactic disk (15◦ , Gerssen et al. 2003). The velocity dispersion is roughly constant over the field-of-view (σ = 240 ± 40 km/s). 2.2

Ionized gas distribution and kinematics

The parameters of the emission lines (intensity, mean velocity and FWHM) were derived from single Gaussian fitting of the lines using the fit/spec software (Rousset 1992). In the MR1 spectral configuration, the fit was performed on two systems of emission lines : the first consisting of the Hβ and [N i] lines, and the second, of the

[O iii] doublet. Within each system, emission lines were assumed to share the same velocity and FWHM. We fitted the Hβ and [O iii] lines independently in order to detect differences in the kinematics and distribution of these two systems. In the MR2 configuration, a single system consisting of all the emission lines ([O i], Hα, [N ii] and [S ii]) was used. Sky emission lines such as the [O i]λ6364 line were used to derive the effective spectral resolution of the MR2 configuration: we obtained F W HMOASIS = 5.3 ± 0.4 ˚ A. As there is no unresolved emission lines in the MR1 configuration, we assumed the effective resolution to be the same in both configurations. All values of the emission line widths were corrected from the instrumental resolution p 2 2 with F W HMgas = − F W HMOASIS , where F W HMobs 2 F W HMobs is the line widths before correction. The ionized gas distribution, kinematics and emission lines ratios maps are presented in Fig. 3. The gas velocity field shows a curved zero-velocity line and the mean kinematical PA is about 50◦ . An inner spiral of about 1 kpc in diameter is observed in the gas intensity field for all of the emission lines, but not in the stellar continuum maps (Fig. 1). The spiral arms correspond to regions of lower gas velocity dispersions, except at the end of the arms where the velocity field appears perturbed and the dispersion increases (Fig. 3, fourth row). The density sensitive [S ii]λ6717/[S ii]λ6731 ratio reaches its minimum in the centre of the field with a value close to 1 and increases at the end of the spiral arms and in the East region (≃ 1.3-1.4), corresponding to an electron density of about 800 cm−3 and less than 100 cm−3 respectively. The [O iii]/Hβ ratio reaches its highest values (around 10) inside the nuclear spiral and the [N ii]/Hα ratio ranges from 1.9 in the centre and the end of the spiral arms and 1.7 inside the spiral. A comparison with the aperture measurements of Ho et al. (1997a) shows a good agreement with our data.

3

MODELING AND DISCUSSION

The two-arm gaseous spiral and the perturbations observed in the OASIS kinematic maps of NGC 1358 (Sect. 2.2) are reminiscent of gas flows in a barred potential (Athanassoula c 2006 RAS, MNRAS 000, 1–6

3

Arcsec

central kiloparsec regions of NGC1358

Arcsec

Arcsec

Figure 2. NGC 1358 OASIS stellar kinematics. Left : mean velocity V (in km/s); right : velocity dispersion σ (in km/s). The ranges are indicated in the bottom right corner of each map. Both maps are North-up, East-left.

1992 ; Maciejewski et al. 2002). In the present Section, we wish to examine this hypothesis by assuming the ionized gas is responding to a barred gravitational potential. We therefore build a simple dynamical model of the central kiloparsec of NGC 1358 following the formalism described in Emsellem et al. (2003). We first create an axisymmetric luminosity model using the Multi-Gaussian-Expansion formalism (MGE hereafter)( Emsellem et al. (1994), Cappellari (2002)), using the HST/WFPC2 image (F606W, Fig. 1) as a reference. This HST image is deprojected assuming a position angle of P A = 195◦ for the line of nodes and a disk inclination of i = 38◦ . The resulting deprojected image is fitted with a two-dimensional MGE model , which is then artificially thickened with a constant axis ratio of 0.36 which minimized the differences between our axisymmetric 3 dimensional luminosity model and the HST image. Figure 4 compares this model (projected with i = 38◦ ) to the initial HST image. The MGE model does obviously not reproduce the large-scale bar, but we only aim here at constraining the potential in the central 15′′ . The gravitational potential is then derived by using Jeans equations assuming isotropy, a constant mass-to-light ratio (M/L) and an inclination of i = 38◦ . A best-fit model was obtained with M/L = 6.1, a value used for the rest of the modeling. We then add a bar-like perturbation to this axisymmetric gravitational potential, as described in Emsellem et al. (2003) : δΦ(r, θ) = Φ2 (r) cos (2θ). Gaseous orbit are derived using the linear epicycle approximation, including an artificially induced damping (simulating the dissipative nature of the gas) described by the parameter Λ (km2 .s−2 ). The free parameters of this model are : the characteristic radius Rbar (arcsec), force Qbar (km2 s−2 ), the pattern speed Ωp (km s−1 kpc−1 ) and the orientation φ of the bar, and the damping coefficient Λ. In order to constrains some of these parameters, we assume that the observed spiral arms are at least partly outside the Inner Lindblad Resonance (ILR here after), and that the shocks lie inside the radius of the Ultra Harmonic Resonace 4:1 (UHR). This implies upper and lower limits on the pattern speed of the bar : 230 km s−1 kpc−1 6 Ωp 6 475 km s−1 kpc−1 . Moreover, since the bar ends before its c 2006 RAS, MNRAS 000, 1–6

corotation radius, we have limits on the bar radius : Rbar 6 8 ′′ . We then derived models for pattern speeds and bar radius within these ranges, for different values of Λ, Qbar and φ. For each model, we compute the density distribution, the velocity field and the velocity dispersion maps. These maps were constructed taking into account the resolution of our observations with OASIS (seeing of 0.9 ′′ ) and we compared them to the observed data : [OIII] distribution and ionised gas velocity fields. We are able to roughly reproduced the gaseous spiral feature (length and shape) with parameters taking values in ranges : • • • •

5 ′′ 6 Rbar 67 ′′ , 260 km s−1 kpc−1 6 Ωp 6 290 km s−1 kpc−1 , 200000 km2 s−2 6 Qbar 6 350000 km2 s−2 100 km2 .s−2 6 Λ 6 120 km2 .s−2

However, it is difficult to further constrain the values of these parameters. Figure 5 shows a comparison between 2 of these models and the observed data, for 2 different sets of the parameters : φ = 80◦ , Ωp = 280 km s−1 kpc−1 , Rbar = 6 ′′ , Λ = 100 km2 s−2 and Qbar = 260000 km2 s−2 for model 1 (left hand side of fig 5) and φ = 130◦ , Ωp = 280 km s−1 kpc−1 , Rbar = 6 ′′ , Λ = 100 km2 s−2 and Qbar = 350000 km2 s−2 for set 2 (right hand side). The corresponding corotation radius is then at ∼ 6.8 ′′ (∼ 1.8 kpc) for both of these models. The model 1 corresponds to one of the sets of parameters wich best reproduce the extent, orientation and shape of the observed gaseous features. The spiral arms end then just outside the ILR, situated at ∼ 1 ′′ (∼ 260 pc). However, the predicted velocity field doesn’t show the structure seen in the gaseous velocity field : it has the same orientation as the stars and the galaxy disk, but the bar perturbation is too weak to reproduce the observed data. We tried to reproduce the gaseous velocity field structure, taking other values for the parameters (e.g. larger φ and Qb ar). Thus the model 2 in fig 5 shows the results of the best model we can reach. We reproduced the change of orientation of the line of nodes but the perturbation of the predicted gas velocity field is not yet sufficient and those of

G. Dumas et al.

Arcsec

Arcsec

Arcsec

Arcsec

Arcsec

4

Arcsec Arcsec

Arcsec

Figure 3. NGC 1358 ionised gas distribution and kinematics. First and second rows : intensity maps (in 10−16 erg s−1 cm−2) : [OIII] and Hβ , and [NII], Hα and [SII] fluxes respectively; third row : [OIII], Hβ and MR2 configuration velocity maps (in km/s); fourth row : [OIII], Hβ and MR2 configuration velocity dispersion maps (in km/s); and fifth row : [OIII]/Hβ, [NII]/Hα and [SII]λ6717/[SII]λ6731 line ratios. The ranges are indicated in the bottom right of each map. In the velocity and velocity dispersion maps, the overlaid contours are respectively the [OIII], Hβ and [NII] intensity. In the line ratio maps, they are the [OIII], [NII] and [SII] intensity.

c 2006 RAS, MNRAS 000, 1–6

5

mag/arcsec

R (Arcsec)

2

central kiloparsec regions of NGC1358

R (Arcsec)

R (Arcsec)

Figure 4. Comparison between the HST image and the MGE axisymmetric luminosity model of NGC 1358. Left : Isophotes of the HST deprojected image (black) and of the MGE model (red). Right : major-axis cuts (black), minor-axis cuts (blue) of the HST image and MGE model cuts (red).

the distribution of the gas is too large compared to the observations. If we try to constrain further the velocity field, then the perturbation become very large : the spiral is destroyed, but the velocity structure is not better reproduced.

4

CONCLUSION

We investigated in detail the central kpc regions of the Seyfert galaxy NGC1358 by means of the integral field spectroscopy. We constructed the stellar and gaseous maps (morphology and velocity fields). The ionised gas distribution revealed a nuclear spiral and the gas velocity field showed a strong perturbation. We interpreted these features as the presence of an inner stellar bar which is also described in Malkan et al. (1998). To test this hypothesis, we constructed a simple dynamical model. However this model failed to reproduce the observed morphology and kinematics of the ionised gas. Moreover, other drivers may be responsible of the formation of such an inner gaseous spiral, as density waves inside the ILR of the major bar as described in Englmaier and Shlosman (2000), and the gasesous response (viscosity, shocks) must certainly be taken into account. The role of the major bar in the formation of the nuclear spiral can be supported by the fact that our best model compared to the gaseous velocity field structure is obtained for an orientation of the inner bar of 130◦ which is the orientation of the major bar. However, our simple linear modelisation can’t reproduce the effects of the main bar. A non-linear analysis is certainly needed. Thus, in this paper, we tried to understand the gaseous morphology and kinematics of the central kpc of NGC1358 in term of a nuclear stellar bar. Such strucutre may be a good mechanisn to fuel the AGN. But more complex and realistic simulations, including hydrodynamical parameters are required to describe with more detailed the kinematics of the ionized gas component of NGC 1358, and to replace it in the AGN feeding issue. c 2006 RAS, MNRAS 000, 1–6

REFERENCES E. Athanassoula. Morphology of bar orbits. MNRAS, 259: 328–344, Nov. 1992. M. Cappellari. Efficient multi-Gaussian expansion of galaxies. MNRAS, 333:400–410, June 2002. M. Cappellari and Y. Copin. Adaptive spatial binning of integral-field spectroscopic data using Voronoi tessellations. MNRAS, 342:345–354, June 2003. M. Cappellari and E. Emsellem. Parametric Recovery of Line-of-Sight Velocity Distributions from Absorption-Line Spectra of Galaxies via Penalized Likelihood. PASP, 116: 138–147, Feb. 2004. E. Emsellem, G. Monnet, and R. Bacon. The multigaussian expansion method: a tool for building realistic photometric and kinematical models of stellar systems I. The formalism. A & A, 285:723–738, May 1994. E. Emsellem, P. Goudfrooij, and P. Ferruit. A two-arm gaseous spiral in the inner 200 pc of the early-type galaxy NGC 2974: signature of an inner bar. MNRAS, 345:1297– 1312, Nov. 2003. P. Englmaier and I. Shlosman. Density Waves inside the Inner Lindblad Resonance: Nuclear Spirals in Disk Galaxies. ApJ, 528:677–686, Jan. 2000. P. Erwin and L. S. Sparke. Double Bars, Inner Disks, and Nuclear Rings in Early-Type Disk Galaxies. AJ, 124:65– 77, July 2002. J. Gerssen, K. Kuijken, and M. R. Merrifield. Modelindependent measurements of bar pattern speeds. MNRAS, 345:261–268, Oct. 2003. L. C. Ho, A. V. Filippenko, and W. L. W. Sargent. A Search for “Dwarf” Seyfert Nuclei. III. Spectroscopic Parameters and Properties of the Host Galaxies. ApJS, 112:315–+, Oct. 1997a. L. C. Ho, A. V. Filippenko, W. L. W. Sargent, and C. Y. Peng. A Search for “Dwarf” Seyfert Nuclei. IV. Nuclei with Broad H alpha Emission. ApJS, 112:391–+, Oct. 1997b. W. Maciejewski, P. J. Teuben, L. S. Sparke, and J. M.

6

G. Dumas et al.

Observed [OIII] distribution

Modeled gas distribution. Model 2

Arcsec

Modeled gas distribution. Model 1

Observed gas velocity

Modeled gas velocity. Model 2

Arcsec

Arcsec

Arcsec

Arcsec

Modeled gas velocity. Model 1

Figure 5. Comparison between the barred model and the observed data. Left : model 1 (see text), top : intensity map, bottom : velocity map. Middle : observed data, top : [OIII] distribution, bottom : [OIII] velocity field. Right : model 2, top : intensity map, bottom : velocity field.

Stone. Gas inflow in barred galaxies - effects of secondary bars. MNRAS, 329:502–512, Jan. 2002. M. A. Malkan, V. Gorjian, and R. Tam. A Hubble Space Telescope Imaging Survey of Nearby Active Galactic Nuclei. ApJS, 117, July 1998. R. W. Pogge and P. Martini. Hubble Space Telescope Imaging of the Circumnuclear Environments of the CfA Seyfert Galaxies: Nuclear Spirals and Fueling. ApJ, 569:624–640, Apr. 2002. A. Rousset. Contribution des m´ethodes num´eriques au d´epouillement des donn´ees du spectrographe int´egral de champ TIGER. Ph.D. Thesis, 1992. F. Valdes, R. Gupta, J. A. Rose, H. P. Singh, and D. J. Bell. The Indo-US Library of Coud´e Feed Stellar Spectra. ApJS, 152:251–259, June 2004.

c 2006 RAS, MNRAS 000, 1–6

North