The imprint of the thick stellar disc in the mid-plane of three early-type edge-on galaxies
Galactic stellar discs, such as that of the Milky Way consisting of a thin and a thick component, have usually a complex structure. The study of galactic disc substructures and their differences can shed light on the galaxy assembling processes and their evolution. However, due to observational difficulties there is a lack of information about the stellar population of the thick disc component in the external galaxies. Here we investigate three edge-on early-type disc galaxies in the Fornax cluster (IC 335, NGC 1380A, NGC 1381) by using publicly available photometrical data and our new deep long-slit spectroscopy along galactic mid-planes obtained with the 10-m SALT telescope. We report that significant changes of the stellar population properties beyond the radius where photometrical profiles demonstrate a knee are caused by an increasing thick disc contribution. Stellar population properties in the the outermost thick-disc dominated regions demonstrate remarkably old ages and low metallicity. We interpret these findings as a consequence of star formation quenching in the outermost regions of the discs due to the ram pressure gas stripping from the disc periphery at the beginning of the cluster assembly while subsequent star formation occurring in the inner discs being gradually extinguished by the starvation.
keywords:galaxies: evolution – galaxies: structure – galaxies: stellar content
Thick stellar discs identified as distinct large-scale components of disc galaxies were initially discovered in S0 galaxies through surface photometry of edge-on objects (Tsikoudi, 1979; Burstein, 1979). Later Gilmore & Reid (1983) found a similar structure in the Milky Way that is a spiral of rather late morphological type. By studying individual stars belonging to the thick disc of the Milky Way, researchers have recognized that the thick disc is an old, rather metal-poor, and magnesium-overabundant component (Fuhrmann, 2011; Bensby et al., 2007). Despite the fact that thick discs are nearly ubiquitous in the local galaxies (Comerón et al., 2012; Comerón et al., 2018) including low-mass dwarf galaxies (Yoachim & Dalcanton, 2006) the spectral studies of thick stellar discs are still rare due to the observational difficulty caused by their low surface brightness. Only a few recent studies have begun to investigate stellar population properties in the thick discs of various galaxies (Yoachim & Dalcanton, 2005, 2008; Comerón et al., 2015, 2016; Guérou et al., 2016; Kasparova et al., 2016) using long-slit spectroscopy with the slits oriented parallel with respect to the galaxy mid-plane or through integral field spectroscopy of edge-on galaxies.
Photometrical studies show that radial surface brightness profiles of galaxies often have breaks with a down-bending (truncation) or an up-bending (antitruncation) shape or a combination of both (Erwin et al., 2005; Pohlen & Trujillo, 2006; Erwin et al., 2008; Comerón et al., 2012, and references therein). Comerón et al. (2012), by studying the photometry of a large sample of edge-on galaxies in the 3.6m and 4.65m images from the S4G project (Sheth et al., 2010), have concluded that the antitruncated type of galactic disc surface brightness profiles (Erwin et al., 2005) may in some cases be ascribed to radial truncation of thin discs: at the truncation radius the thin disc (or the inner disc, to be more precise) disappears, and even with only the cut along the galaxy mid-plane we detect the more extended thick disc in the outer parts of the galaxy. As one can see, in some cases antitruncations may be caused by the superposition of thin and thick disc components. This is also supported by photometrical measurements which claim that thick galactic discs often have longer radial scale lengths than their thin counterparts (Burstein, 1979; Pohlen et al., 2004; Comerón et al., 2012). Note that photometrical thin/thick disc decompositions for external galaxies are purely geometric, in contrast to those made for the Milky Way which are often based on the age and the metallicity (or the -enhancement). Nevertheless this photometrical point of view helps in the interpretation of long-slit spectral data obtained along the mid-planes of edge-on disc galaxies.
In this work we present our new spectral data for three edge-on lenticular galaxies (IC 335, NGC 1380A, NGC 1381) belonging to the Fornax cluster. These galaxies have been the subject of many spectroscopic studies of their kinematics (D’Onofrio et al., 1995; Chung & Bureau, 2004; Bedregal et al., 2006; Spolaor et al., 2010a; Vanderbeke et al., 2011) as well as stellar populations (Kuntschner, 2000; Terlevich & Forbes, 2002; Bedregal et al., 2008; Mármol-Queraltó et al., 2009; Spolaor et al., 2010b; Koleva et al., 2011; Johnston et al., 2012). Nevertheless, the above papers considered these objects within galaxy samples and not on an individual basis although the detailed description of their radial profiles (velocities, velocity dispersions, age and metallicities) has been presented, for instance, in Koleva et al. (2011) and Spolaor et al. (2010b). Our spectral measurements are rather deep, and we have reached the outermost parts of these galaxies and detected strong changes in the stellar population properties at some crucial radii. Combining spectroscopic evidence with photometrical data we argue that the changes of stellar population properties are caused by the increasing thick disc contribution in the outermost galactic regions along the galactic mid-plane. We discuss how environment-driven mechanisms in the Fornax cluster could lead to the observed properties of these galaxies.
In Section 2, we describe observations and data reduction process including developed framework for taking into account scattered light in the spectra. In Section 3 we analyse our spectroscopic and archival photometric data and present the results. A discussion and our main conclusions are provided in Section 4 and 5, respectively.
2 Observations and data reduction
We have performed long-slit spectroscopy for the following Fornax cluster members – all of them lenticular galaxies IC 335, NGC 1380A, and NGC 1381, in the frame of our project on the stellar populations properties in large-scale discs of S0s in clusters. Observations were performed with the Robert Stobie Spectrograph (RSS; Burgh et al., 2003; Kobulnicky et al., 2003) at the Southern African Large Telescope (SALT) (Buckley et al., 2006; O’Donoghue et al., 2006). We have used the long-slit mode with a slit width of 1.25 arcsec and the volume-phase grating GR900 providing a spectral resolution of 4.8 Å (FWHM) in the Å spectral range. All observational details are given in Table 1. The slit was aligned with the galaxies major axes going through their nuclei and mid-plane. The total exposure time per object is about 120. The seeing during the observations was in the arcsec range. The RSS pixel scale is 0.1267 arcsec, and the length of the slit is 8 arcmin. We used a binning factor of 4 to get a final spatial sampling of 0.507 arcsec pixel. An Ar comparison arc spectrum was exposed to calibrate the wavelength scale after each observation as well as the spectral flats were taken regularly to correct the spectra for pixel-to-pixel variations. Spectrophotometric standard stars were observed during twilights, after observing the objects, for a relative flux calibration.
2.2 Data reduction
The primary data reduction was done with the SALT science pipeline (Crawford et al., 2010). After that, bias and gain corrected long-slit data were reduced as described by Kniazev et al. (2008). The accuracy of the wavelength calibration was checked by measuring the sky line [Oı] 5577; the rms scatter of its wavelength measured along the slit was about 0.04 Å. The observed galaxies are much less than 8 arcmin in diameter, so we used the pure night-sky spectra from the slit edges to subtract the sky background.
2.3 Scattered light
When analysing the reduced spectra we found that our velocity dispersion measurement are overestimated in comparison with those obtained with better spectral resolution (Bedregal et al., 2006; Koleva et al., 2011). We suspected that the reason for such a behavior is the presence of the diffuse scattered light in the spectrograph. It redistributes light coming from the bright galaxy center to the galaxy outskirts and contributes to the galactic continuum affecting the contrast of the absorption lines. Neglecting this effect could lead to the systematic biases in the stellar population properties. In this section we describe a framework to calculate the scattered light contribution in the spectra along the slit.
We have used the spectrum of a standard star observed with the same spectral setup as the galaxies. The stellar light profiles along the slit can be expressed as:
where is a point spread function caused by atmospheric perturbations, – a full PSF of the light scattering in the instrument, is a total stellar flux at a given wavelength, is a scaling factor, denotes convolution and is a Dirac delta function. In the second approximate equality we assume an additive representation for the full scattering function . This is motivated by necessity to estimate extended part of the scattering function. The physical meaning of is a fraction of light which is redistributed within the instrument in accordance to the scattering function . Note that one can use a function since the diffraction PSF is much narrower than the atmospheric one.
This additive representation allows us to propose a procedure for the calculation of in any spectrum. The top panel of Figure 1 demonstrates the procedure.
For further calculations we approximated the shape of by Gaussian and Moffat profiles fitting the upper part () of the observed profile and scaled to contain the full stellar flux along the slit. We cannot be able securely recover profile because the profile wings are affected by the instrumental light scattering. So we used these two extreme parametrizations (Gaussian and Moffat) having completely different wings and considered how they affect the final parameters.
To avoid a degeneracy in the main fitting procedure we fixed the scaling coefficient by requiring equality between the maximum values of the Gaussian/Moffat representation (green line) and the observed stellar profile (black line). In the main minimization loop we approximate the function as the sum of three exponential functions and three Gaussians (blue line) which results in a good modeling of the observed stellar profile. To take into account variations along the wavelengths we determined component in 6 bins along the whole spectral range from 3800 till 6800 Å. The shapes of for different wavelength bins turned out to be identical. The total fraction of the scattered light not accounted by the atmospheric is about 15 (25) per cent for Moffat (Gaussian) parametrization.
One can apply the same additive parametrization of light scattering for a galaxy profile to estimate contribution of scattered light in the observed spectrum:
where is a galaxy light profile at some wavelength affected by atmospheric seeing only. Consecutively convolving with one can obtain:
The term in the square brackets can be considered as an additive component caused by the light scattering within the instrument.
Under this framework we calculated an additive component of scattered light at each wavelength. Nevertheless, photons are also scattered along the dispersion direction resulting in a less deep spectral absorption features. To take into account that fact we convolved the computed component of scattered light along the dispersion direction with the normalised function considering that the light scattering works in the same way along dispersion direction. To obtain spectra unaffected by instrumental light scattering we subtracted scattered additive component at every wavelength in our observed spectra.
The middle panel of Figure 1 shows an application of this framework to the galaxy NGC 1380A. The bottom panel demonstrates the relative contribution of the scattered light to the galaxy profile for two parametrization of the atmospheric PSF (Gaussian/Moffat). Despite our estimation of a total scattered light fraction of 15 (25) per cent based on a standard star, the relative contribution of the scattered light at a given position on the slit can reach higher values. This happens because the PSF is very broad and covers a large fraction of the galaxy.
3 Data analysis and results
Stellar kinematics and stellar population properties resolved along the slit were derived by the full spectral fitting technique NBursts Chilingarian et al. (2007a, b). This technique implements a pixel-to-pixel minimization fitting algorithm, where an observed spectrum is approximated by a stellar population model broadened with a parametric line-of-sight velocity distribution (LOSVD). We used a grid of PEGASE.HR high-resolution simple stellar population (SSP) models (Le Borgne et al., 2004) based on the ELODIE3.1 empirical stellar library (Prugniel et al., 2007) with the fixed Salpeter initial mass function (IMF) and pre-convolved with the RSS spectrograph instrumental function recovered from the spectrum of a Lick standard star. During minimization loop, SSP model template is interpolated from the models grid at given age and metallicity, then broadened with stellar LOSVD and multiplied by Legendre polynomial continuum to take into account possible internal dust reddening and/or spectral calibration errors both in the data and models.
The used stellar population models are computed for the solar element abundance ratio only because they are based on an empirical library of stars from the solar vicinity. To check the relative -element abundance we calculated the Lick indices Mgb and (Worthey et al., 1994; Worthey & Ottaviani, 1997) and compared them to the SSP evolutionary synthesis models by Thomas et al. (2003).
To obtain reliable radial profiles we made spatial binning of the spectra along the slit. We used linearly increasing bins from 2 to 20 pixels ( arcsec) adjusting the signal-to-noise ratio within bin to be greater than 15.
To determine the full spectral fitting parameter uncertainties we carried out Monte Carlo simulation for each spatial bin. We generated a hundred realization of synthetic spectra by adding a random noise to the best-fitting model corresponding to the signal-to-noise ratio in the bin. Then we fitted each synthetic spectrum and estimated the errors as the standard deviation of the output model parameters. The uncertainties for the Mgb and indices were computed by using photon Poisson errors propagated through all the data reduction steps.
Having solved the scattered light problem we compared our velocity dispersion measurements with the data obtained at the ESO FORS2 spectrograph with significantly higher resolution ( km s) by Bedregal et al. (2006). We found that the use of Gaussian shape for the atmospheric provides the velocity dispersion measurements in good agreement with the higher resolution data down to km s111Instrumental resolution km saround Mgb band at 5100Å. This is quite sufficient for measuring the velocities, however, the measurements of the velocity dispersions in the cold thin component of stellar discs ( as low as km s) could be affected by systematic bias to higher values. In the same time, subtraction of the scattered light component computed with the Moffat parametrization of provides velocity dispersions overestimated by 20 km s as well as underestimation of stellar metallicities by 0.1 dex at the 60 km s level of velocity dispersion. For higher velocity dispersions the biases become negligible. So through the further stellar population analysis we used only spectra with removed scattered light component computed with the Gaussian shape of .
We present the profiles of the SSP-equivalent stellar ages and metallicities and stellar kinematics obtained by full spectrum fitting in Fig. 2. The main feature of all studied galaxies is a strong change in the properties of the stellar populations, in particular in the stellar metallicities, beyond a certain radius.
Our photometrical analysis (see Section 3.2 for details) revealed that the studied galaxies have complex disc structure. We denoted the knee radii where a break (truncation) in the surface brightness profile as well as significant changes in the stellar population parameters have appeared. To compare the properties of the two regions separated by the knee radius we binned their spectra correcting for the LOS velocity variations and analyzed these in the same manner as the original radial bins. We used luminosity weighted integration within the bins and for the inner and outer discs correspondingly (see Table 2). We exclude the central disc regions () to avoid possible contribution of the bulge and/or the bar. We defined the maximum radii of outermost radial bin as radii where per spatial bin. The resulting parameters are presented in Table 3 and are shown in Fig. 2 via green and red lines.
|IC 335||NGC 1380A||NGC 1381|
IC 335 (IC 1963, FCC 153): This galaxy demonstrates quite a flat velocity dispersion profile ( km s) which is in agreement with the photometrical decompositions by Salo et al. (2015) and Comerón et al. (2018) where bulge was fitted by a point-like source and contains only % of galaxy light.
The stellar age profile demonstrates a steady increase of age from Gyr at the central region of the disc to Gyr at the outskirts. The metallicity profile has smooth variations with local maxima at arcsec and gradually decreases to a value of [Fe/H] dex at arcsec beyond which it drops further down to [Fe/H] dex.
Generally, the -elements ratio is slightly positive ([Mg/Fe] dex) over the inner disc with larger values at radii larger than 25 arcsec. The analysis of the spectra binned over the whole inner and outer discs does not show significant differences in the -elements abundances.
NGC 1380A (FCC 177): This galaxy also has a small bulge component that is indicated by the flat velocity dispersion profile. The central region of the galaxy (inner arcsec) shows a signature of rejuvenation of the stellar population due to recent star formation event: while the inner disc has average ages of Gyr and metallicities of dex, the central region has Gyr and [Fe/H] dex. Again we detected a significant decrease of the stellar metallicity at the knee radius () down to a value of [Fe/H] dex. The averaged ages and metallicities for the outer disc are not as different from those of the main disc and one could suppose just by looking at the individual data points. This is because the luminosity-averaging causes the innermost points to dominate. The [Mg/Fe] profile has a large scatter but it seems that the last detected points have slightly higher values than the regions at arcsec.
NGC 1381 (FCC 170): This is the only galaxy in our sample with a prominent bulge that can be clearly seen on the galaxy image (top right panel in Fig. 2) as well as on the stellar population profiles. Stellar kinematics show double-humped features on the line-of-sight velocity profile and a shoulder in the velocity dispersion profile that are in agreement with the measurements by Chung & Bureau (2004) and Bedregal et al. (2006) and indicate the presence of a bar (Bureau & Athanassoula, 2005). The metallicity profile looks flat within the inner disc. Beyond arcsec the metallicity decreases down to [Fe/H] dex.
The age profile stays constant within the inner disc at level of Gyr and increases to Gyr in the outskirts. The [Mg/Fe] element ratio indicates that stellar populations are enriched by -elements throughout the entire disc ([Mg/Fe] dex).
The detailed stellar population profiles for considered galaxies IC 335, NGC 1380A, NGC 1381 were previously obtained from ESO/FORS2 data in Bedregal et al. (2006, 2008) and Gemini/GMOS data in Spolaor et al. (2010b) and later re-analysed by Koleva et al. (2011) by using a full spectral fitting technique. We compare our profiles with those by Koleva et al. (2011) in Fig. 2 (blue and orange lines with dots) and found that our measurements are in good agreement with theirs.
|(Thin disc)||(Thick disc)|
|Binned regions, arcsec||10…47||53…64|
|Binned regions, arcsec||12…32||38…60|
|Binned regions, arcsec||20…47||53…78|
Since our spectral analysis has shown sharp changes of the stellar population parameters at the knee radius, we would like to understand whether this happens due to the intrinsic features (internal gradients) of the thin stellar disc or due to changes in the relative contribution of the thin and thick disc subsystems in the radial direction. Are there any reasons from the photometric point of view to assume that thick disc stars in the studied galaxies have a considerable contribution to the light in the mid-plane at radii arcsec?
3.2.1 Radial structure
We have used HST images of NGC 1380A and NGC 1381 obtained with ACS/WFC in the F850LP band. In Fig. 3 we present the radial photometric profiles at different -distances below and above the main plane of the galaxies. Radial profiles have a stepped structure, and the knees, separating the different exponential sections, separate also the regions with different stellar populations (see Fig. 2). The knee radii of each galaxy are listed in Table 2. It is well seen (Fig. 3) that the steps at higher galactic altitudes are less prominent, therefore the radial profiles outside the main plane ( arcsec) can be described by a single exponential law. Martinez-Lombilla et al. (2018) found a similar behavior for two highly inclined nearby galaxies NGC 4565 and NGC 5907.
It is difficult to disentangle correctly the individual contributions of the two discs due to parameter degeneracy and the unknown truncation law. For this reason, it makes sense to fit each segment of the profile using one exponential component. In this way we get a radial scale of some superposition of a thick and a thin discs. We have estimated the radial scale length inside and outside using the expression (van der Kruit & Searle, 1981)
where is the exponential scale length at a given , and is the modified Bessel function. In order to analyse only the data concerning the disc components, we exclude the inner regions of the galaxies , where the influence of spherical subsystems or bars is possible.
In Fig. 4, on the top, we present the radial scale length of the outer and inner segments (red and green signs) as a function of the -distance from the mid-plane. We see a clear trend for the radial scale length to grow with increasing , which can be interpreted as an evidence of the disc heterogeneity. For two of the three galaxies – NGC 1380A and NGC 1381 – we see that there are significant differences in the radial scale lengths of both segments only within the inner layer with arcsec. This corresponds to the fact that the stepped profile becomes a single exponential law far away from the mid-plane of the disc.
Note that the scale length estimates of the outer segment are very sensitive to the quality of background subtraction. Moreover the truncation law of inner disc is unknown and may depend on the ram pressure stripping process which in turn depends on the trajectory of the galaxy and on the disc orientation with respect to the incoming flow. Therefore, one should not expect that the outer discs represent only a thick disc, and the measured radial scale corresponds to the true value of thick disc radial scale.
3.2.2 Vertical structure
Next, we investigated the vertical structure of the edge-on discs in the studied galaxies. We pursued two aims: i) to check whether vertical structure of the discs changes with radius particularly around and ii) to demonstrate how contribution of the embedded thinnest disc component changes with . We assessed parameters of the vertical profiles for a given galactocentric distance by using multi-component model (Spitzer, 1942):
To successfully describe vertical profiles we used a three-component model. Unfortunately, a model consisting of only two components does not well fit extended wings of the vertical profiles. This problem was recently addressed in the paper by Comerón et al. (2018) where the authors demonstrated that PSF effects could be responsible for the extended wings at least for the Spitzer data. We utilized Tiny Tim PSF modeling tool (Krist et al., 2011) and tested PSF effects in the similar manner as Comerón et al. (2018). We concluded that extended wings of vertical profiles in the studied galaxies cannot be described by PSF effects in the used HST data222We also tested the effects of increasing thickness with radius and deviation from the precise edge-on orientation by means of integration of three-dimensional model of galaxy light along line-of-sight. None of the effects make it possible to describe the extended wings in the profiles. Detailed description of our experiments will be given in forthcoming paper.. This has motivated us to apply a three-component model consisted of thin, mid, and thick components. Note, that we do not focus on physical interpretation of each disc component because we have only long-slit data along the mid-plane of the studied galaxies. This question might be naturally considered basing on the long-slit data taken in perpendicular orientation or on the IFU data. Thus, recently announced Fornax 3D project (Sarzi et al., 2018), including MUSE observations of our studied galaxies, would be particular useful for this aim.
We computed vertical profiles as a median average of the galaxy cuts in the small radial bins of arcsec size covering radial distances from to arcsec where surface brightness drops down to mag arcsec. Since vertical profile decomposition into three components is not unambiguous and is affected by degeneracy between the model parameters we applied the following trick. Firstly we fitted vertical profile which is closest to the by using hand-tuned initial parameter guesses. Then, for the next profile we used output parameters from previous step as an initial guess and required that vertical scales of all components can vary in some range around the initial guess. The range was calculated from the condition that gradient of vertical scale should be less than 0.5 arcsec per arcsec in radial distance. We decomposed vertical profiles by means of non-linear minimization method using lmfit package (Newville et al., 2016).
Variations of the vertical scale length of the different components obtained from the photometric cuts at different are shown in blue, orange, and green lines in Fig. 4 (middle row) while the surface brightness along the mid-plane for the same model components and the observed profiles (black symbols) are presented in the bottom row panel. Fig. 5 demonstrates a few examples of vertical profiles with overplotted best-fit models and separate components at different radial distances. From our analysis we concluded that i) contribution of the thin disc component (that is associated with the thinnest and probably with some additional mixture of the mid component) significantly decreases around , and the most vertically extended components (mid and thick ones) dominate in the outer disc; ii) all disc components demonstrate moderate or significant growth of their thickness with the radius (except the regions where the component contribution is negligible).
In the work by Comerón et al. (2018) there is photometric analysis of our galaxies based on Spitzer data but their image approximation by Comerón et al. (2018) implies an absence of any change in scale heights of the discs along the radius, and this is why it is difficult to compare their results with ours. Moreover, for each galaxy, they considered the mean scale height values in four segments ( and ), which can include our . Nevertheless, comparing both approaches we found that our estimation of vertical scale for the thinnest components is compatible with their values for thin discs while our mid and thick components have in general higher scales than their thick discs. For the internal segments (), Comerón et al. (2018) obtained 1.4 arcsec and 7.1 arcsec for vertical scales of the thin and thick disc components of IC 335 ( arcsec); 2.1 arcsec and 10.5 arcsec – for NGC 1380A ( arcsec); 2.2 arcsec and 10.7 arcsec – for NGC 1381 ( arcsec). For the external segments () in most cases they did not get good fits.
3.2.3 Isophote analysis
We applied the fitting formalism Isofit recently developed by Ciambur (2015) to the images. This formalism provides an appropriate description of deviations from ellipticity and, therefore, is useful for isophote analysis in edge-on galaxies (see examples in Ciambur & Graham, 2016). The resulting ellipticity and the , coefficients of the Fourier harmonics are presented in Fig. 6. We used eight harmonics in the Isofit tool. Fig. 6 clearly demonstrates that all three objects have disky isophotes in their inner parts. The parameter (negative values indicate diskiness) increases beyond the knee radius for the three galaxies.
3.2.4 Photometry analysis summary
To sum up, our photometric analysis supports the fact that the considered galaxies have more than one disc component, since i) their vertical profiles have complex structure and are not fitted by a single and even two disc components within ; ii) radial scale lengths grow with again indicating complex vertical disc structure and iii) isophotes change their shape sharply, reducing the diskyness beyond the .
We interpret that the thick disc components increase their contribution to the total light around and dominate the disc peripheries. It results in significant variations of the stellar population properties as a function of the radius.
This is in good agreement with the recent paper by Comerón et al. (2018) where thick discs of edge-on galaxies are studied in the SG Survey. The authors have demonstrated that thick discs are nearly ubiquitous, and their contribution to the surface brightness in the mid-plane can increase as radius grows (see their Fig. 22 and similar figures in their appendices for the galaxies studied here).
We have studied three edge-on galaxies (IC 335, NGC 1380A and NGC 1381) belonging to the Fornax cluster. Our study reveals important information for the stellar disc formation theory complementing the investigation of these galaxies by other authors (for instance, Bedregal et al., 2008; Spolaor et al., 2010b; Koleva et al., 2011) .
The galaxies under consideration naturally fit into a two-phase model of galaxy assembly. First, thick stellar discs formed rapidly at high redshifts in dense turbulent unstable gas-rich discs (Elmegreen & Elmegreen, 2006; Bournaud et al., 2009; Comerón et al., 2014; Elmegreen et al., 2017). After that the thin disc components grew for a long time from gas freshly accreted through cosmological filaments (Sancisi et al., 2008; Combes, 2014), minor gas-rich mergers (Robertson et al., 2006; Sancisi et al., 2008), or by accretion of cooled left-over gas (Burkert et al., 1992) or coronal gas cooled by the fountain mechanism. (Fraternali, 2009; Fraternali et al., 2013).
Our galaxies belong to the Fornax cluster which has a complex structure and where the mass assembly processes are still going on (Drinkwater et al., 2001; Iodice et al., 2017; Spiniello et al., 2018). Our galaxies are located near the main cluster core associated with NGC 1399 where the majority of galaxies are early-type ones (Ferguson, 1989). Fig. 7 demonstrates that these galaxies are strongly dynamically bound to the main core of the Fornax cluster. It is natural to assume that they could have experienced dense-environment effects (Boselli & Gavazzi, 2006) in the past such as ram pressure stripping (Gunn & Gott, 1972; Abadi et al., 1999; Quilis et al., 2000), tidal interactions with the cluster gravitational potential and high-speed galaxy-galaxy encounters (Moore et al., 1996; Moore et al., 1998), which could affect the galaxy evolutionary phase while the thin disc grew.
All our galaxies demonstrate a significant increase of the SSP equivalent age and decline of the stellar metallicity towards the galaxy peripheries (see Fig. 2), where the thick disc components dominate. Recently Kasparova et al. (2016) studied the edge-on galaxy NGC 4710 belonging to the Virgo cluster, which has a similar behavior. They proposed that the Hı gas layer has been stripped by ram pressure and as a result its thin disc is “unfinished” and we can observe a stellar population of the thick disc at the outskirts of NGC 4710. The same scenario takes place for the galaxies studied in this paper. Note that ram pressure can effectively remove the gas starting from some particular radius where the ram pressure overcomes the gravitational pressure of the disc (Boselli & Gavazzi, 2006). We suggest that could be this radius.
Intermediate-age stellar populations in the inner discs and the absence of any emission lines in the spectra point out the quenching of active star formation some time ago. The most obvious explanation for this is a process of starvation (Larson et al., 1980; Bekki et al., 2002; Bekki, 2009; Zinger et al., 2018), which consists in removing the extended gas reservoir from a galaxy halo. This results in the quenching of further star formation activity after a few Gyrs. Due to the fact that SSP-equivalent ages are strongly biased towards the age of the younger population (Serra & Trager, 2007), we can consider the stellar ages in the thin-disc dominated regions as a time stamp for the quenching of active star formation. Hence, the star formation in the main discs of IC 335 and NGC 1380A had stopped approximately Gyr ago due to the dense cluster environment. The inner disc of NGC 1381 demonstrates an older age and a higher -element enhancement of its stellar population ( Gyr, [Mg/Fe) with respect to other galaxies. This indicates that the star formation in this galaxy was rapidly quenched in earlier epoch and can be explained if it entered earlier into the dense environment and if it was already in place Gyr ago (). Another feature of this object is the presence of a prominent bulge, therefore, bulge-driven processes, for instance, morphological quenching (Martig et al., 2009) or active galactic nucleus feedback (Di Matteo et al., 2005; Croton et al., 2006), could be alternatively responsible for the early star formation quenching, without any relation to the cluster environment.
Recent investigations of lenticular galaxies in different environments (Sil’chenko et al., 2012; Katkov et al., 2014; Katkov et al., 2015) have led to the scenario of general evolution of disc galaxies formulated by Sil’chenko et al. (2012). The main idea is that lenticulars are primordial disc galaxies which formed at high redshift () as a thick disc component (Elmegreen & Elmegreen, 2006; Bournaud et al., 2009). The further fate of the galaxy strongly depends on the mode of gas accretion into their discs. If there is persistent gas accretion and dynamical gas cooling, spiral arms can develop and star formation re-ignites: the galaxy is transformed into a typical spiral. In absence of a gas-accretion source, which most commonly happens in dense environments, the galaxies preserve their lenticular morphology during all their life; therefore S0s are the dominant galaxy population in galaxy clusters at . Similar idea have been also suggested by Comerón et al. (2016) for the ESO243-49 evolution.
In the galaxies investigated in this paper we have found imprints of all galaxy formation stages discussed above: the primordial formation of the thick discs in the outermost regions and the subsequent development of a thin disc component in their internal parts that has been stopped by environmental effects within the Fornax cluster.
In this paper we have performed a detailed study of three edge-on galaxies (IC 335, NGC 1380A, NGC 1381) belonging to the Fornax cluster. We explored publicly available photometrical HST data as well as new deep spectroscopic observational data obtained at the 10-m SALT telescope.
We have demonstrated that the long-slit spectra obtained with the RSS spectrograph have sufficient scattered light to bias the measurements of the stellar population properties in the outer parts of galactic discs. We have developed a framework to take into account the scattered light which can be used for any kind of long-slit data analysis.
The stellar population properties of the outer disc regions in all three galaxies demonstrate a significantly older ages and lower metallicity than the inner ones. Combining these data with a photometric analysis we have concluded that the changes of the stellar population properties are caused by an increase of the thick-disc contribution in the outermost galactic regions. We interpreted this in the frame of a two-phase process of disc galaxy assembly where the thick disc component formed at high redshift while the thin disc developed later from the gas accreted from outside. We suggest that the star formation in the outer disc has been quenched due to ram pressure stripping at the beginning of the cluster assembly while the rest of star formation in the discs was gradually extinguished by starvation.
We are very grateful to the referee Sébastien Comerón for comments and suggestions that improved this manuscript. We also thank Prof. Anatoly Zasov and Dr. Igor Chilingarian for fruitful discussions. The spectroscopical observations reported in this paper were obtained with the Southern African Large Telescope (SALT), under programmes 2014-2-MLT-001 and 2015-2-MLT-002 (PI: Alexei Kniazev). AYK acknowledges the support from the National Research Foundation (NRF) of South Africa. IYK, AVK are grateful to the Russian Science Foundation grant 17-72-20119 which supported the photometrical analysis as well as the manuscript preparation. IYK is also thankful to the RFBR grant number 16-02-00649. The authors acknowledge partial support from the M.V. Lomonosov Moscow State University Program of Development. Based on observations made with the NASA/ESA Hubble Space Telescope, and obtained from the Hubble Legacy Archive, which is a collaboration between the Space Telescope Science Institute (STScI/NASA), the Space Telescope European Coordinating Facility (ST-ECF/ESA) and the Canadian Astronomy Data Centre (CADC/NRC/CSA). This research made use of Astropy, a community-developed core Python package for Astronomy (The Astropy Collaboration et al., 2018); The Atlassian JIRA issue tracking system and Bitbucket source code hosting service.
- Abadi et al. (1999) Abadi M. G., Moore B., Bower R. G., 1999, MNRAS, 308, 947
- Bedregal et al. (2006) Bedregal A. G., Aragón-Salamanca A., Merrifield M. R., Milvang-Jensen B., 2006, MNRAS, 371, 1912
- Bedregal et al. (2008) Bedregal A. G., Aragón-Salamanca A., Merrifield M. R., Cardiel N., 2008, MNRAS, 387, 660
- Bekki (2009) Bekki K., 2009, MNRAS, 399, 2221
- Bekki et al. (2002) Bekki K., Couch W. J., Shioya Y., 2002, ApJ, 577, 651
- Bensby et al. (2007) Bensby T., Zenn A. R., Oey M. S., Feltzing S., 2007, ApJ, 663, L13
- Boselli & Gavazzi (2006) Boselli A., Gavazzi G., 2006, PASP, 118, 517
- Bournaud et al. (2009) Bournaud F., Elmegreen B. G., Martig M., 2009, ApJ, 707, L1
- Buckley et al. (2006) Buckley D. A. H., Swart G. P., Meiring J. G., 2006, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series. p. 62670Z, doi:10.1117/12.673750
- Bureau & Athanassoula (2005) Bureau M., Athanassoula E., 2005, ApJ, 626, 159
- Burgh et al. (2003) Burgh E. B., Nordsieck K. H., Kobulnicky H. A., Williams T. B., O’Donoghue D., Smith M. P., Percival J. W., 2003, in Iye M., Moorwood A. F. M., eds, Proc. SPIEVol. 4841, Instrument Design and Performance for Optical/Infrared Ground-based Telescopes. pp 1463–1471, doi:10.1117/12.460312
- Burkert et al. (1992) Burkert A., Truran J. W., Hensler G., 1992, ApJ, 391, 651
- Burstein (1979) Burstein D., 1979, ApJ, 234, 829
- Chilingarian et al. (2007a) Chilingarian I., Prugniel P., Sil’Chenko O., Koleva M., 2007a, in Vazdekis A., Peletier R., eds, IAU Symposium Vol. 241, Stellar Populations as Building Blocks of Galaxies. pp 175–176 (arXiv:0709.3047), doi:10.1017/S1743921307007752
- Chilingarian et al. (2007b) Chilingarian I. V., Prugniel P., Sil’Chenko O. K., Afanasiev V. L., 2007b, MNRAS, 376, 1033
- Chung & Bureau (2004) Chung A., Bureau M., 2004, AJ, 127, 3192
- Ciambur (2015) Ciambur B. C., 2015, ApJ, 810, 120
- Ciambur & Graham (2016) Ciambur B. C., Graham A. W., 2016, MNRAS, 459, 1276
- Combes (2014) Combes F., 2014, in Seigar M. S., Treuthardt P., eds, Astronomical Society of the Pacific Conference Series Vol. 480, Structure and Dynamics of Disk Galaxies. p. 211 (arXiv:1309.1603)
- Comerón et al. (2012) Comerón S., et al., 2012, ApJ, 759, 98
- Comerón et al. (2014) Comerón S., Elmegreen B. G., Salo H., Laurikainen E., Holwerda B. W., Knapen J. H., 2014, A&A, 571, A58
- Comerón et al. (2015) Comerón S., Salo H., Janz J., Laurikainen E., Yoachim P., 2015, A&A, 584, A34
- Comerón et al. (2016) Comerón S., Salo H., Peletier R. F., Mentz J., 2016, A&A, 593, L6
- Comerón et al. (2018) Comerón S., Salo H., Knapen J. H., 2018, A&A, 610, A5
- Crawford et al. (2010) Crawford S. M., et al., 2010, in Observatory Operations: Strategies, Processes, and Systems III. p. 773725, doi:10.1117/12.857000
- Croton et al. (2006) Croton D. J., et al., 2006, MNRAS, 365, 11
- D’Onofrio et al. (1995) D’Onofrio M., Zaggia S. R., Longo G., Caon N., Capaccioli M., 1995, A&A, 296, 319
- Di Matteo et al. (2005) Di Matteo T., Springel V., Hernquist L., 2005, Nature, 433, 604
- Drinkwater et al. (2001) Drinkwater M. J., Gregg M. D., Colless M., 2001, ApJ, 548, L139
- Elmegreen & Elmegreen (2006) Elmegreen B. G., Elmegreen D. M., 2006, ApJ, 650, 644
- Elmegreen et al. (2017) Elmegreen B. G., Elmegreen D. M., Tompkins B., Jenks L. G., 2017, ApJ, 847, 14
- Erwin et al. (2005) Erwin P., Beckman J. E., Pohlen M., 2005, ApJ, 626, L81
- Erwin et al. (2008) Erwin P., Pohlen M., Beckman J. E., 2008, AJ, 135, 20
- Ferguson (1989) Ferguson H. C., 1989, AJ, 98, 367
- Fraternali (2009) Fraternali F., 2009, in Andersen J., Nordströara m B., Bland-Hawthorn J., eds, IAU Symposium Vol. 254, The Galaxy Disk in Cosmological Context. pp 255–262 (arXiv:0807.3365), doi:10.1017/S1743921308027671
- Fraternali et al. (2013) Fraternali F., Marasco A., Marinacci F., Binney J., 2013, ApJ, 764, L21
- Fuhrmann (2011) Fuhrmann K., 2011, MNRAS, 414, 2893
- Gilmore & Reid (1983) Gilmore G., Reid N., 1983, MNRAS, 202, 1025
- Guérou et al. (2016) Guérou A., Emsellem E., Krajnović D., McDermid R. M., Contini T., Weilbacher P. M., 2016, A&A, 591, A143
- Gunn & Gott (1972) Gunn J. E., Gott III J. R., 1972, ApJ, 176, 1
- Iodice et al. (2017) Iodice E., et al., 2017, ApJ, 839, 21
- Johnston et al. (2012) Johnston E. J., Aragón-Salamanca A., Merrifield M. R., Bedregal A. G., 2012, MNRAS, 422, 2590
- Kasparova et al. (2016) Kasparova A. V., Katkov I. Y., Chilingarian I. V., Silchenko O. K., Moiseev A. V., Borisov S. B., 2016, MNRAS, 460, L89
- Katkov et al. (2014) Katkov I. Y., Sil’chenko O. K., Afanasiev V. L., 2014, MNRAS, 438, 2798
- Katkov et al. (2015) Katkov I. Y., Kniazev A. Y., Sil’chenko O. K., 2015, AJ, 150, 24
- Kniazev et al. (2008) Kniazev A. Y., et al., 2008, MNRAS, 388, 1667
- Kobulnicky et al. (2003) Kobulnicky H. A., Nordsieck K. H., Burgh E. B., Smith M. P., Percival J. W., Williams T. B., O’Donoghue D., 2003, in Iye M., Moorwood A. F. M., eds, Proc. SPIEVol. 4841, Instrument Design and Performance for Optical/Infrared Ground-based Telescopes. pp 1634–1644, doi:10.1117/12.460315
- Koleva et al. (2011) Koleva M., Prugniel P., De Rijcke S., Zeilinger W. W., 2011, MNRAS, 417, 1643
- Krist et al. (2011) Krist J. E., Hook R. N., Stoehr F., 2011, in Optical Modeling and Performance Predictions V. p. 81270J, doi:10.1117/12.892762
- Kuntschner (2000) Kuntschner H., 2000, MNRAS, 315, 184
- Larson et al. (1980) Larson R. B., Tinsley B. M., Caldwell C. N., 1980, ApJ, 237, 692
- Le Borgne et al. (2004) Le Borgne D., Rocca-Volmerange B., Prugniel P., Lançon A., Fioc M., Soubiran C., 2004, A&A, 425, 881
- Mármol-Queraltó et al. (2009) Mármol-Queraltó E., et al., 2009, ApJ, 705, L199
- Martig et al. (2009) Martig M., Bournaud F., Teyssier R., Dekel A., 2009, ApJ, 707, 250
- Martinez-Lombilla et al. (2018) Martinez-Lombilla C., Trujillo I., Knapen J. H., 2018, preprint, p. arXiv:1804.01546 (arXiv:1804.01546)
- Moore et al. (1996) Moore B., Katz N., Lake G., Dressler A., Oemler A., 1996, Nature, 379, 613
- Moore et al. (1998) Moore B., Lake G., Katz N., 1998, ApJ, 495, 139
- Newville et al. (2016) Newville M., Stensitzki T., Allen D. B., Rawlik M., Ingargiola A., Nelson A., 2016, Lmfit: Non-Linear Least-Square Minimization and Curve-Fitting for Python, Astrophysics Source Code Library (ascl:1606.014)
- O’Donoghue et al. (2006) O’Donoghue D., et al., 2006, MNRAS, 372, 151
- Pohlen & Trujillo (2006) Pohlen M., Trujillo I., 2006, A&A, 454, 759
- Pohlen et al. (2004) Pohlen M., Balcells M., Lütticke R., Dettmar R.-J., 2004, A&A, 422, 465
- Prugniel et al. (2007) Prugniel P., Soubiran C., Koleva M., Le Borgne D., 2007, ArXiv Astrophysics e-prints
- Quilis et al. (2000) Quilis V., Moore B., Bower R., 2000, Science, 288, 1617
- Robertson et al. (2006) Robertson B., Bullock J. S., Cox T. J., Di Matteo T., Hernquist L., Springel V., Yoshida N., 2006, ApJ, 645, 986
- Salo et al. (2015) Salo H., et al., 2015, ApJS, 219, 4
- Sancisi et al. (2008) Sancisi R., Fraternali F., Oosterloo T., van der Hulst T., 2008, A&ARv, 15, 189
- Sarzi et al. (2018) Sarzi M., et al., 2018, A&A, 616, A121
- Serra & Trager (2007) Serra P., Trager S. C., 2007, MNRAS, 374, 769
- Sheth et al. (2010) Sheth K., et al., 2010, PASP, 122, 1397
- Sil’chenko et al. (2012) Sil’chenko O. K., Proshina I. S., Shulga A. P., Koposov S. E., 2012, MNRAS, 427, 790
- Spiniello et al. (2018) Spiniello C., et al., 2018, preprint, (arXiv:1803.03276)
- Spitzer (1942) Spitzer Lyman J., 1942, ApJ, 95, 329
- Spolaor et al. (2010a) Spolaor M., Hau G. K. T., Forbes D. A., Couch W. J., 2010a, MNRAS, 408, 254
- Spolaor et al. (2010b) Spolaor M., Kobayashi C., Forbes D. A., Couch W. J., Hau G. K. T., 2010b, MNRAS, 408, 272
- Terlevich & Forbes (2002) Terlevich A. I., Forbes D. A., 2002, MNRAS, 330, 547
- The Astropy Collaboration et al. (2018) The Astropy Collaboration et al., 2018, preprint, (arXiv:1801.02634)
- Thomas et al. (2003) Thomas D., Maraston C., Bender R., 2003, MNRAS, 339, 897
- Tsikoudi (1979) Tsikoudi V., 1979, ApJ, 234, 842
- Vanderbeke et al. (2011) Vanderbeke J., Baes M., Romanowsky A. J., Schmidtobreick L., 2011, MNRAS, 412, 2017
- Worthey & Ottaviani (1997) Worthey G., Ottaviani D. L., 1997, ApJS, 111, 377
- Worthey et al. (1994) Worthey G., Faber S. M., Gonzalez J. J., Burstein D., 1994, ApJS, 94, 687
- Yoachim & Dalcanton (2005) Yoachim P., Dalcanton J. J., 2005, ApJ, 624, 701
- Yoachim & Dalcanton (2006) Yoachim P., Dalcanton J. J., 2006, AJ, 131, 226
- Yoachim & Dalcanton (2008) Yoachim P., Dalcanton J. J., 2008, ApJ, 683, 707
- Zinger et al. (2018) Zinger E., Dekel A., Kravtsov A. V., Nagai D., 2018, MNRAS, 475, 3654
- van der Kruit & Searle (1981) van der Kruit P. C., Searle L., 1981, A&A, 95, 105