1 INTRODUCTION
Do interactions or mergers trigger activity in the nucleus of the galaxy? Many theoretical models suggest that the active galactic nucleus (AGN) activity is closely linked to galaxy interactions and mergers (e.g., Kauffmann & Haehnelt Reference Kauffmann and Haehnelt2000; Cattaneo Reference Cattaneo2001; Wyithe & Loeb Reference Wyithe and Loeb2002; Di Matteo et al. Reference Di Matteo, Croft, Springel and Hernquist2003; Kang et al. Reference Kang, Jing, Mo and Börner2005; Bower, Benson, & Malbon Reference Bower2006; Croton et al. Reference Croton2006). But observational studies have yielded contradictory results (e.g., Dahari Reference Dahari1984, Reference Dahari1985; Keel et al. Reference Keel, Kennicutt, Hummel and van der Hulst1985; Kennicutt et al. Reference Kennicutt1987; Barton, Geller, & Kenyon Reference Barton, Geller and Kenyon2000; Virani, De Robertis, & VanDalfsen Reference Virani, Robertis and VanDalfsen2000; Schmitt Reference Schmitt2001; Miller et al. Reference Miller2003; Grogin et al. Reference Grogin2005; Waskett et al. Reference Waskett2005; Koulouridis et al. Reference Koulouridis2006; Serber et al. Reference Serber, Bahcall, Ménard and Richards2006; Alonso et al. Reference Alonso, Lambas, Tissera and Coldwell2007; Woods & Geller Reference Woods and Geller2007; Ellison et al. Reference Ellison, Patton, Simard and McConnachie2008; Li et al. Reference Li2006, Reference Li2008). Dahari (Reference Dahari1984) searched for close companion galaxies in a redshift-limited sample of Seyfert galaxies, and found that there is a definite excess of companions in the Seyfert sample, compared with a control sample of field galaxies. Woods & Geller (Reference Woods and Geller2007) detected a significantly increased AGN fraction in the pair galaxies compared to matched sets of field galaxies. Ellison et al. (Reference Ellison2011) found a clear increase in the AGN fraction in close pairs of galaxies relative to the control sample, and further demonstrated that the increase in AGN fraction is strongest in equal mass galaxy pairings, and weakest in the lower mass component of an unequal mass pairing. However, Li et al. (Reference Li2006) did not observe strong evidence that interactions and mergers are playing a significant role in triggering the AGN activity, and suggested other physical mechanism responsible for explaining the AGN activity: AGN are preferentially located at the centers of dark matter haloes. Ellison et al. (Reference Ellison, Patton, Simard and McConnachie2008) found little evidence for increased AGN activity in their close-pairs sample and concluded that, if AGNs are induced by mergers, then they must occur at stages later than close-pairs typically examine. Li et al. (Reference Li2008) also failed to find any corresponding relation between enhanced AGN activity and interactions.
In this study, we use the Main galaxy sample of the Sloan Digital Sky Survey Data Release 7 (SDSS DR7) (Abazajian et al. Reference Abazajian2009) and a relatively new and publicly available catalogue of the flux and error, and explore influences of galaxy interactions on the AGN activity. Close-paired galaxies often be defined as interacting and merging galaxies, and are used to study the effect of galaxy interactions (e.g., Lambas et al. Reference Lambas, Tissera, Alonso and Coldwell2003; Alonso et al. Reference Alonso, Tissera, Coldwell and Lambas2004a). On the contrary, isolated galaxies are a group of galaxies that may have experienced no major interactions in billions of years. Undoubtedly, the comparison between the properties of galaxies in pairs and isolated is a useful method to unveil the effects of interactions on the AGN activity.
Our paper is organised as follows. In Section 2, we describe the data used. In Section 3, we investigate the AGN fraction of galaxies in pairs and isolated. Our main results and conclusions are summarised in Section 4.
In calculating the distance, we used a cosmological model with a matter density Ω0=0.3, cosmological constant ΩΛ=0.7, and Hubble's constant H 0=70 km s−1 Mpc−1.
2 DATA
2.1 Summary of the data
Many of survey properties of the SDSS were discussed in detail in the Early Data Release paper (Stoughton et al. Reference Stoughton2002). In this study, we use the Main galaxy sample (Strauss et al. Reference Strauss2002) of the SDSS DR7 (Abazajian et al. Reference Abazajian2009). The data were downloaded from the Catalogue Archive Server of SDSS DR7 by the SDSS SQL Search (with SDSS flag: bestPrimtarget&64 > 0) with high-confidence redshifts (Zwarning ≠16 and Zstatus≠0, 1 and redshift confidence level: zconf > 0.95) (http://www.sdss.org/dr7/).
The SDSS Main galaxy sample is an apparent-magnitude-limited sample, which seriously suffers from the Malmquist bias (Malmquist Reference Malmquist1920; Teerikorpi Reference Teerikorpi1997). Because faint galaxies at large distances will not be detected, the average luminosity of galaxies in such a sample increases with increasing distance. In order to decrease this bias, one often used the volume-limited galaxy sample. Our Main galaxy sample of the SDSS DR7 contains 565029 Main galaxies with the redshift 0.02 ≤ z ≤ 0.2. From this apparent-magnitude-limited Main galaxy sample, Deng (Reference Deng2010) constructed a luminous volume-limited Main galaxy sample that contains 120362 galaxies at 0.05 ≤ z ≤ 0.102 with −22.5 ≤ Mr ≤ −20.5 and a faint volume-limited sample that contains 33249 galaxies at 0.02 ≤ z ≤ 0.0436 with −20.5 ≤ Mr ≤ −18.5. In this work, we still use these two volume-limited samples.
2.2 Galaxy pairs
For the identification of galaxy pairs, many authors developed different criteria (e.g., Karachentsev Reference Karachentsev1972; Barton et al. Reference Barton, Geller and Kenyon2000; Lambas et al. Reference Lambas, Tissera, Alonso and Coldwell2003; Patton et al. Reference Patton2005; Focardi et al. Reference Focardi2006; Kewley, Geller, & Barton 2006; Deng et al. 2008a, 2008b). It is important to recognise that up to now, there still is not a widely accepted criterion. Any criterion has its own drawbacks. We noted that for many issues of galaxy pairs, the use of different criteria often can reach the same conclusions. Thus, we believe that the selection of criteria is less important in such issues. In this study, we use a typical criterion developed by Lambas et al. (Reference Lambas, Tissera, Alonso and Coldwell2003). Lambas et al. (Reference Lambas, Tissera, Alonso and Coldwell2003) selected galaxy pairs in the field by radial velocity (ΔV ≤ 350 kms−1) and projected separation (r p ≤ 100 kpc) criteria. r p ≤ 100 kpc and ΔV ≤ 350 km s−1 can be defined as reliable upper limits for the relative radial velocity and projected distance criteria to select galaxy pairs with stronger-specific star formation than the averaged galaxies in the SDSS and 2dF galaxy redshift survey (Lambas et al. Reference Lambas, Tissera, Alonso and Coldwell2003; Alonso et al. Reference Alonso, Tissera, Coldwell and Lambas2004b, Reference Alonso, Lambas, Tissera and Coldwell2006). By applying the same selection criteria, we identify 1654 pairs in the luminous volume-limited sample and 1133 pairs in the faint volume-limited sample.
It has been known for a long time that the fiber collisions are main sources of incompleteness in SDSS pair catalogues. If one attempts to study the large-scale distribution of pairs, this incompleteness of the pair sample will be a large drawback. But in this study, the influence of this incompleteness is not crucial. In addition, correcting of some incompleteness likely results in new bias. For example, Berlind, Frieman, & Weinberg (Reference Berlind2006) corrected for fiber collisions by giving each collided galaxy the redshift of its nearest neighbor on the sky (usually the galaxy it collided with). Putting collided galaxies at the redshifts of their nearest neighbors will cause some nearby galaxies to be placed at high redshift, which artificially makes their estimated luminosities very high. Therefore, we do not make efforts to correct fiber collisions.
2.3 Control sample
A control sample is constructed by randomly selecting galaxies without close companions within r p < 100 kpc and ΔV < 350 km s−1. In order to investigate the effects of galaxy interactions, one often compared galaxies in pairs with isolated galaxies. Perez, Tissera, & Blaizot (Reference Perez, Tissera and Blaizot2009) explored how the way of building a control sample introduces biases that could affect the interpretation of results, and claimed that a suitable control sample for isolating the effects of interactions should be built by imposing constraints on redshift, stellar mass, local environment, morphology, and halo mass. But if considering the correlations among galaxy properties, one should realise that imposing too many constraints also washes out the difference of galaxy properties between the control and the pair samples introduced by any physical mechanism. Because the redshift is the most fundamental quantity in selection effects, in this work, the control sample is required to have the same galaxy number and the same redshift distribution as the pair sample.
3 AGN FRACTION OF GALAXIES IN PAIRS AND ISOLATED
3.1 Identification of AGNs
By considering the classical diagnostic ratios of two pairs of relatively strong emission lines, Baldwin, Phillips & Terlevich (Reference Baldwin, Phillips and Terlevich1981, hereafter BPT) demonstrated that it is possible to distinguish AGNs from normal star-forming galaxies. We download the flux and error in the flux of four lines from http://www.mpa-garching.mpg.de/SDSS/DR7/. Like Li et al. (Reference Li2006) did, AGNs are selected from the subset of galaxies with signal-to-noise ratio S/N > 3 on the four emission lines [OIII]λ5007, Hβ, [NII]λ6584, Hα. In our apparent-magnitude-limited Main galaxy sample, 253594 galaxies have S/N > 3 on the four lines. Following Kauffmann et al. (Reference Kauffmann2003), a galaxy is defined to be an AGN if
\begin{eqnarray*}
&&\log ([{\rm OIII}]\,\lambda \,5007/{\rm H}\,\beta ) > 0.61/\{ \log ([{\rm NII}]\,\lambda \,6584/H\alpha )\\
&&\quad -\, 0.05\} + 1.3.
\end{eqnarray*}
Our apparent-magnitude-limited Main galaxy sample contains 89716 AGNs. Figure 1 shows redshift distribution of Main galaxies and AGNs for the apparent-magnitude-limited Main galaxy sample of the SDSS DR7. As seen from this figure, there is a quite large difference of AGN fraction in different redshift bins, but the trend of change for AGNs is nearly the same as the one for Main galaxies. This shows that there are serious selection effects in the apparent-magnitude-limited Main galaxy sample. In this work, we use the volume-limited galaxy samples in which the radial selection function is approximately uniform. In addition, when performing comparative studies, the control sample is required to have the same galaxy number and the same redshift distribution as the pair sample. So, selection effects in this work are less important.
Redshift distribution of Main galaxies (left panel) and AGNs (right panel) for the apparent-magnitude-limited Main galaxy sample of the SDSS DR7.
3.2 AGN fraction of galaxies in pairs and isolated
The luminous volume-limited sample contains 28674 AGNs, the faint volume-limited sample includes 5021 AGNs. We compute the AGN fraction of galaxies in pairs and isolated: in the luminous volume-limited sample, 0.2597 ± 0.0089 for isolated galaxies and 0.2358 ± 0.0084 for paired galaxies; in the faint volume-limited sample, 0.1545 ± 0.0083 for isolated galaxies and 0.1664 ± 0.0086 for paired galaxies. Here, the Poissonian error is taken into account. In the faint volume-limited sample, paired galaxies have a slightly higher AGN fraction than isolated galaxies, which seemingly shows that interactions or mergers likely trigger the AGN activity. But in the luminous volume-limited sample, an opposite trend can be observed. The significance is <1σ. So, it is difficult to conclude whether interactions or mergers likely trigger the AGN activity.
Considering the variation in AGN fraction with redshift, we divide the whole redshift region of two volume-limited samples into redshift bins with width 0.01 (The last redshift bin is 0.100–0.102 for the luminous volume-limited Main galaxy sample, and 0.04–0.0436 for the faint volume-limited Main galaxy sample.), and focus the analysis on the statistical differences of the AGN fraction between paired galaxies and isolated ones in each redshift bin. Figure 2 shows the fraction of AGNs as a function of redshift z for paired galaxies (red triangle) and isolated galaxies (blue dot) in the luminous (on the right-hand side of the green vertical line) and faint (on the left-hand side of the green vertical line) volume-limited samples. As can be seen from Figure 2, in the faint volume-limited sample (low redshift range), paired galaxies have a slightly higher AGN fraction than isolated galaxies, whereas in the luminous volume-limited sample (high redshift range), the AGN fraction of isolated galaxies is slightly higher. This finding further confirms the above-mentioned conclusion.
Fraction of AGNs as a function of redshift z for paired galaxies (red triangle) and isolated galaxies (blue dot) in the luminous (on the right-hand side of the green vertical line) and faint (on the left-hand side of the green vertical line) volume-limited samples. The error bars are 1σ Poissonian errors.
In dense systems of a galaxy sample, interactions and mergers often occur in a large fraction of galaxies (e.g., Rubin, Hunter, & Ford 1991; Mendes de Oliveira & Hickson 1994; Lee et al. Reference Lee2004). For example, Lee et al. (Reference Lee2004) showed that there is strong evidence of interactions and mergers within a significant fraction of SDSS CGs (compact groups of galaxies). Paired galaxies also often be located in dense systems such as groups and clusters. Many authors showed that there is no evidence for the environmental dependence of the AGN fraction (e.g., Monaco et al. Reference Monaco, Giuricin, Mardirossian and Mezzetti1994; Coziol et al. Reference Coziol, de Carvalho, Capelato and Ribeiro1998; Shimada et al. Reference Shimada2000; Carter, Fabricant, & Geller 2001; Schmitt Reference Schmitt2001; Miller et al. Reference Miller2003). For example, Carter et al. (Reference Carter2001) showed that the AGN fraction is insensitive to the local environment. Miller et al. (Reference Miller2003) also observed that this fraction is constant from the cores of galaxy clusters to the rarefied field population. There also have been a number of dissenting papers. For example, Dressler, Thompson, & Shectman (Reference Dressler, Thompson and Shectman1985) found five times as many AGNs in the field as in clusters. Popesso & Biviano (Reference Popesso and Biviano2006) also reported a lower fraction of (weak and strong) optical AGN in clusters than in the field and smaller systems. According to these two standpoints, it is difficult to reach the conclusion: paired galaxies have a higher AGN fraction than isolated galaxies.
4 SUMMARY
Using two volume-limited Main galaxy samples of the SDSS DR7, we explore influences of galaxy interactions on AGN activity. In each sample, we construct a paired sample and a control sample, and compared the AGN fraction of galaxies in pairs and isolated. The control sample is required to have the same galaxy number and the same redshift distribution as the pair sample. It is found that in the faint volume-limited sample, paired galaxies have a slightly higher AGN fraction than isolated galaxies, whereas in the luminous volume-limited sample, an opposite trend can be observed. The significance is <1σ. Thus, we do not observe strong evidence that interactions or mergers likely trigger the AGN activity.
ACKNOWLEDGMENTS
We thank the anonymous referee for many useful comments and suggestions. Our study was supported by the National Natural Science Foundation of China (NSFC, Grant 11263005).
Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation, the US Department of Energy, the National Aeronautics and Space Administration, the Japanese Monbukagakusho, the Max Planck Society, and the Higher Education Funding Council for England. The SDSS Web site is http://www.sdss.org.
The SDSS is managed by the Astrophysical Research Consortium for the Participating Institutions. The Participating Institutions are the American Museum of Natural History, Astrophysical Institute Potsdam, University of Basel, University of Cambridge, Case Western Reserve University, University of Chicago, Drexel University, Fermilab, the Institute for Advanced Study, the Japan Participation Group, Johns Hopkins University, the Joint Institute for Nuclear Astrophysics, the Kavli Institute for Particle Astrophysics and Cosmology, the Korean Scientist Group, the Chinese Academy of Sciences (LAMOST), Los Alamos National Laboratory, the Max Planck Institute for Astronomy (MPIA), the Max Planck Institute for Astrophysics (MPA), New Mexico State University, Ohio State University, University of Pittsburgh, University of Portsmouth, Princeton University, the US Naval Observatory, and the University of Washington.
