BH mass estimation for 1H 0323+342

Independent Black Hole Mass Estimation for the -ray Detected Archetypal Narrow-Line Seyfert 1 Galaxy 1H 0323+342 from X-ray Variability

Hai-Wu Pan11affiliation: Key Laboratory of Space Astronomy and Technology, National Astronomical Observatories, Chinese Academy of Sciences, 20A Datun Road, Chaoyang District, Beijing, China;, 22affiliation: University of Chinese Academy of Sciences, School of Astronomy and Space Science, Beijing 100049, China , Weimin Yuan11affiliation: Key Laboratory of Space Astronomy and Technology, National Astronomical Observatories, Chinese Academy of Sciences, 20A Datun Road, Chaoyang District, Beijing, China;, 22affiliation: University of Chinese Academy of Sciences, School of Astronomy and Space Science, Beijing 100049, China , Su Yao33affiliation: Kavli Institute for Astronomy and Astrophysics, Peking University, Beijing 100871, China 44affiliation: National Astronomical Observatories, Chinese Academy of Sciences, Beijing, China , S. Komossa55affiliation: Max-Planck Institut für Radioastronomie, Auf dem Hügel 69, 53121 Bonn, Germany & Chichuan Jin11affiliation: Key Laboratory of Space Astronomy and Technology, National Astronomical Observatories, Chinese Academy of Sciences, 20A Datun Road, Chaoyang District, Beijing, China;,

As a newly discovered class of radio-loud active galactic nuclei (AGNs), the -ray detected narrow-line Seyfert 1 galaxies (NLS1s) launch powerful jets which are generally found only in blazars and radio galaxies. However, their black hole (BH) masses as estimated from the broad emission lines are one order of magnitude or more lower than those in blazars. This brings new challenges for explaining the radio loudness triggering in AGNs. It is still under debate whether their BH masses from the commonly used virial method are underestimated. Here we present an estimate of the BH mass for the -ray detected NLS1 1H 0323+342, an archetype of this class, from the X-ray variability which is inclination independent. Our results independently confirm that this -ray detected NLS1 harbors a BH similar to those in normal NLS1s rather than those in blazars.

galaxies: active — galaxies: nuclei — X-rays: galaxies

1 Introduction

Narrow-line Seyfert 1 galaxies (NLS1s), which are defined as active galactic nuclei (AGNs) with the broad Balmer lines narrower than 2000 km s (Osterbrock & Pogge, 1985), appear mostly to be radio-quiet with radio loudness111The radio loudness is defined as the ratio between the radio 5 GHz to optical B-band flux. . Only a small subset were noticed to be radio-loud sources (e.g., Zhou et al., 2006; Komossa et al., 2006). Yuan et al. (2008) found there exist a population of genuine radio-loud NLS1s showing observational properties characteristic of blazars with relativistic jets. The finding was confirmed and highlighted by the detection of -ray emission from a small number of NLS1s with the Fermi satellite (e.g. Abdo et al., 2009a, b; D’Ammando et al., 2012; Yao et al., 2015a). Radio-loud NLS1s exhibit some distinct properties compared to those classical radio-loud AGNs, e.g., blazars. Nearby NLS1s generally reside in pseudo-bulges (Mathur et al., 2012), while the common hosts of classical radio-loud AGNs are elliptical galaxies (e.g., León-Tavares et al., 2011). The key difference is the mass of the central black hole (BH).

NLS1s are thought to have lower BH masses than classical blazars and Seyfert galaxies, and hence higher Eddington ratios (see Komossa, 2008; Yuan et al., 2008 for a review). However, this assertion is still under active debate in the forefront research today. There have been studies suggesting that the BH masses of NLS1 may possibly be underestimated from the commonly used virial method involving broad optical emission lines because of the inclination effect, if their broad line regions (BLRs) are planar and seen face on (Collin & Kawaguchi, 2004; Jarvis & McLure, 2006). Early studies have found a positive correlation between the radio loudness and BH mass in AGNs (Laor, 2000). It indicates that a threshold BH mass is required for AGNs that launch with powerful relativistic jets. Thus, radio-loud NLS1s have also been suggested in some studies to have supermassive BHs similar to those in blazars (e.g., Calderone et al., 2013; León Tavares et al., 2014). Moreover, observationally NLS1s are found to locate at one extreme end in the so-called eigenvector 1 parameter space (Boroson, 2002; Sulentic et al., 2008) as revealed by a set of correlations presumably driven primarily by the Eddington ratio (e.g., Boroson, 2002; Grupe et al., 2010; Xu et al., 2012). These include the strong permitted Fe emission lines (Boroson & Green, 1992), steep soft X-ray spectra (Wang et al., 1996) and rapid X-ray variability (Leighly, 1999), etc. In contrast, most blazars lie at the other end of the eigenvector 1 parameter space (Boroson, 2002). Thus, NLS1s were once thought to be radio-quiet, which turns out to be a consequence of their low radio-loud fraction (Zhou et al., 2006; Komossa et al., 2006). This makes the detection of -ray emitting NLS1s to be a new challenge for explaining the radio loudness in AGNs. There are other lines of evidence hinting that radio-loud NLS1s have lower BH masses than blazars, and hence higher Eddington ratios (Komossa et al., 2006; Yuan et al., 2008; Yao et al., 2015b). Therefore, an inclination-independent estimator is essential for resolving the controversy over the BH masses of NLS1s launching relativistic jets.

Only a handful of NLS1s have been detected in -rays so far (Abdo et al., 2009a, b; Foschini, 2011; D’Ammando et al., 2012, 2015; Yao et al., 2015a; Paliya et al., 2018; Yang et al., 2018). 1H 0323+342 is the nearest one (redshift ) among them, allowing detailed observational studies, e.g., on its host morphology (Zhou et al., 2007; León Tavares et al., 2014), the variability of multiwavelength emission (e.g., Paliya et al., 2014; Yao et al., 2015b) and jet structure on pc scale (Fuhrmann et al., 2016). The single-epoch spectra method using several broad emission lines has been employed, resulting in a relative low mass of about (Zhou et al., 2007; Landt et al., 2017), and the Eddington ratio of about (Landt et al., 2017). Recently, Wang et al. (2016) used the reverberation mapping method and obtained a similar result. However, a BH mass of the order of magnitude of has been also suggested based on the bulge luminosity—BH mass relation (León Tavares et al., 2014). This discrepancy can be explained in two ways: either 1H 0323+342 (or perhaps other radio-loud NLS1s too) resides in a luminous bulge (often a pseudo-bulge) while harbors a much less massive BH than what is normally predicted from the bulge luminosity, or the BH mass is underestimated from the virial method using the broad lines due to the inclination effect.

In this paper, we estimate the BH mass of 1H 0323+342 from the variability properties of its X-ray emission. Rapid X-ray variability is one of the basic observational characteristics of AGN (McHardy, 1985). Since the X-ray emission is thought to originate from the innermost region of an accretion flow around the BH, its variability provides a powerful tool to study the dynamics of matter closely around the BH, which is likely encoded with some of the BH parameters. In fact, some of the characteristics of X-ray variability have been used to estimate the BH mass of AGN, including the break frequency of the power spectrum density (PSD) (McHardy et al., 2006; González-Martín & Vaughan, 2012), the normalized excess variance (Zhou et al., 2010; Ponti et al., 2012), and the quasi-periodic oscillation frequency (Pasham et al., 2014; Zhou et al., 2015; Pan et al., 2016). Unlike the commonly used virial method involving the line-of-sight velocities of the orbiting gas which is susceptible to the orientation effect of AGN (Collin & Kawaguchi, 2004), X-ray variability can be considered as essentially inclination-independent. In this work, we perform X-ray timing analysis using data from a high quality X-ray observation made with the XMM-Newton, based on which the BH mass of 1H 0323+342 is derived from the break frequency of the PSD and the excess variance.

2 Observation and Data Reduction

1H 0323+342 was observed with XMM-Newton with an exposure of  ks on August 24 2015 in the large window imaging mode (ObsID: 0764670101). To improve the signal-to-ratio, the data taken from the EPIC PN and MOS cameras are utilized. We follow the standard procedure to reduce the data and extract science products from the observation data files (ODFs) using the XMM-Newton Science Analysis System (SAS, version 15.0.0). Only good events (single and double pixel events, i.e., PATTERN 4 for PN or 12 for MOS) are used. Source events are extracted from a 40-arcsec circular region, and background events from a source-free circle with the same radius on the same CCD chip. No pile-up effect is found for each of the detectors, by applying the EPATPLOT task. X-ray light curves are constructed for all the three EPIC cameras in the 0.2-10 keV energy band with a binsize of 10 s. Final source light curves are obtained by subtracting corresponding background light curves extracted from the background regions, and corrected for instrumental factors using EPICLCCORR.

Figure 1: The PN (up panel) and combined MOS (MOS1+MOS2; middle panel) light curves of 1H 0323+342 in the energy band of 0.2-10 keV. The black lines represent the light curves with a binsize of 100 s, while the light curves with a binsize of 10 s are plotted in grey. The ratios between the PN and combined MOS (MOS1+MOS2) light curve are plotted in the low panel.

For count rate light curves of a source measured simultaneously by independent detectors, the light curves can be co-added in order to increase the signal-to-noise ratio, provided that the detectors have the same responses (effective areas) or the source’s spectral shape does not vary with time during the observation. Since the two MOS detectors are essentially identical and have very similar responses, the two MOS light curves are co-added into a combined MOS light curve. Figure 1 shows the PN and the combined MOS light curves. Although the PN and MOS detectors have different responses, the two light curves follow each other closely, indicating weak or no changes in the spectral shape. This can be demonstrated by the ratio of the two light curves (bottom panel in Figure 1), which appears to be constant. We thus also co-added the PN and MOS light curves (CR=CR+CR+CR) to form a combined single PN+MOS light curve of the source, assuming no significant spectral variations during the observation. The combined light curve for all the detectors are shown in Figure 2. In the following analysis, both the PN and MOS light curves are used individually, as well as the combined single source light curve with the best signal-to-noise ratio for comparison. As shown in Figure 1 and Figure 2, the X-ray emission of 1H 0323+342 exhibits significant variability on various timescales, typical of NLS1 galaxies.

Figure 2: X-ray light curve of 1H 0323+342 obtained with XMM-Newton, which is a combination of background-subtracted light curves extracted from the PN, MOS1 and MOS2 detectors, respectively, in 0.2-10 keV. The black line represents the light curve with a binsize of 100 s, while the light curve with a binsize of 10 s is plotted in grey. The background (red dashed line) is significantly fainter than the source.

During the whole observation the particle background is at a normal level except for the last 15 ks when an enhanced background flare occurred. We have carefully inspected the data set for the possible influence of the background on the measured light curves, and especially, during the intervals of the flaring background. We find that the background is significantly fainter (mean count rate=0.05 counts s; Figure 1) than the source, which is rather bright (mean count rate=7.80 counts s). Even during the enhanced background period toward the end of the observation, the background is more than 5 times fainter than the source for each time bin. We therefore consider the effect of background fluctuations on the measured light curves to be negligible, and thus make use of the full, uninterrupted source light curve for timing analysis.

3 Timing Analysis and Bh Mass Estimation

3.1 Power Spectrum Density

Early studies showed that the PSD of AGN could be described by a simple power-law (Lawrence & Papadakis, 1993), whereas later investigations with high quality PSD revealed a flattening toward the low frequency end (e.g., Papadakis & McHardy, 1995). It has been established that the AGN PSD can be best described by a bending (or broken) power-law, two slopes with universal values of in the high frequency and in the low frequency parts, respectively (Markowitz et al., 2003; Vaughan et al., 2003a). McHardy et al. (2006) first suggested a tight correlation between the bend/break frequency and the BH mass as well as the bolometric luminosity over a wide range of BH masses from black hole X-ray binary (BHXB) to AGN. Using a larger sample of AGNs, González-Martín & Vaughan (2012) found a similar result, but with a weaker dependence on the bolometric luminosity. This indicates that the bend/break frequency depends mostly on the BH mass.

In the analysis below, the combined PN+MOS light curve is mainly used given its highest signal-to-noise ratio, while the PN and the combined MOS light curves are also analysed for the purpose of comparison of results. The PSDs of the light curves of 1H 0323+342 are computed as the modulus-squared of the discrete Fourier transform, and the (rms/mean) normalisation is chosen (Vaughan et al., 2003b). The PSD derived from the combined PN+MOS light curve is shown in Figure 3. It shows a typical red-noise spectrum in the frequency range from  Hz to  Hz, while above  Hz it is dominated by Poisson noise. As can be seen, a flattening toward the low frequency end is apparent in the red noise of the PSD.

Figure 3: Power spectrum density of the combined PN+MOS X-ray light curve of 1H 0323+342. The red solid line represents the best-fit power-law, while the blue solid line donates the best-fit bending power-law. The data/model residuals for power-law and bending power-law model are shown in the middle and bottom panel, respectively.

We first fit the PSD with a model of a simple power-law (red noise) plus a constant (Poisson noise): (González-Martín & Vaughan, 2012, see also the model in Vaughan, 2010), where the normalization , power-law index and additive constant are all set to be free parameters. The maximum likelihood method is used to find the best-fitting model parameters by minimizing the fit statistic which is (twice) the minus log-likelihood (González-Martín & Vaughan, 2012; see also Vaughan, 2010), yielding a slope of -2.47 (a red line in Figure 3). The uncertainty of each parameter can be estimated by finding the range of parameter value in which , corresponding to a significance level of 68.3% (González-Martín & Vaughan, 2012). However, a single power-law does not describe the PSD well in the low frequency regime. Judging from the data/model residuals (see the middle panel of Figure 3), the slope of the PSD appears to flatten at around frequency  Hz. Thus we adopt a bending power-law () (González-Martín & Vaughan, 2012, see also the model in Vaughan, 2010) instead of a simple power-law to fit the PSD. The low-frequency slope is fixed to be the canonical value -1 for AGN found from long-term X-ray monitoring studies (e.g., Markowitz et al., 2003; McHardy et al., 2006), given the relatively poor data quality in the low-frequency end (González-Martín & Vaughan, 2012). The high-frequency slope is fitted to be 3.66, with a bend frequency of  Hz. Hence, the likelihood ratio test (LRT) suggested in Vaughan (2010) (see also González-Martín & Vaughan, 2012) is used to select between the simple power-law and bending power-law models, giving a posterior predictive -value from the Markov Chain Monte Carlo (MCMC) simulations. It indicates that the PSD for 1H 0323+342 is well described by a bending power-law, as commonly found in Seyfert AGNs.

We also analyze the PSDs of the PN and combined MOS light curve, respectively. Both PSDs can be fitted by a bending power-law better, with a -value of (PN) or (MOS) based on LRT method, respectively, as shown in the Figure 4. The fitted bend frequencies are  Hz (PN) and  Hz (MOS), respectively, which are consistent with each other as well as that obtained from the combined PN+MOS light curve. Thus, the bend frequency of  Hz obtained from the combined light curve is used to estimate the BH mass of 1H 0323+342 hereinafter. Besides, we also fit the PSD of 1H 0323+342 with a broken power-law model, given that the break frequency may be slightly different from the bend frequency, as discussed in Vaughan (2010). The fitted break frequency is  Hz, which is consistent with the bend frequency derived above.

Figure 4: Power spectrum densities of the PN and combined MOS light curves of 1H 0323+342. The red solid lines represent the best-fit power-law, while the blue solid lines donate the best-fit bending power-law. The data/model residuals for power-law and bending power-law model are shown in the middle and bottom panels, respectively.

The relation between the bend/break frequency and the BH mass as well as the bolometric luminosity was suggested in McHardy et al. (2006) first. Later González-Martín & Vaughan (2012) have found that the relation depends weakly on the bolometric luminosity, thus a relation only between the bend/break frequency and the BH mass is also suggested. All these relations are considered for the BH estimations of 1H 0323+342, and the results are listed in Table 1. First, we derive the BH mass for 1H 0323+342 from the bend frequency of  Hz using the relation with BH mass only:


where is the bend timescale () (González-Martín & Vaughan, 2012)222 is in units of days, in units of , and in units of  ergs s for the equations from McHardy et al. (2006) and González-Martín & Vaughan (2012). These equations (including the equation suggested in Ponti et al., 2012) are also used for uncertainty calculation of BH mass.. The BH mass can be calculated quickly via the simplified formalism:


from the bend frequency directly333The simplified formalisms are derived from the equations in the literature (McHardy et al., 2006; González-Martín & Vaughan, 2012; Ponti et al., 2012) exactly. In all these simplified formalisms, is in units of Hz, in units of , and in units of ergs s.. The BH mass is estimated to be . The uncertainty range is calculated from Monte Carlo simulations, in which the uncertainties of the parameters in the correlations (McHardy et al., 2006; González-Martín & Vaughan, 2012) as well as , are assumed to follow a Gaussian distribution444The same method is also used for the following BH mass estimations.. For , the uncertainty is computed via the propagation of error from bend frequency.

Next we estimate the BH mass for 1H 0323+342 using the relation which depends on both the BH mass and bolometric luminosity. Using the bolometric correction factor (Elvis et al., 1994), Zhou et al. (2007) obtained a bolometric luminosity of  ergs s for 1H 0323+342 from the 5100 Å luminosity. Wang et al. (2016) monitored 1H 0323+342 for more than two months using the Lijiang 2.4 m Telescope, and obtained a mean of  ergs s. The bolometric luminosity can be estimated as by making use of the bolometric correction factor suggested in Kaspi et al. (2000). The value  ergs s is nearly consistent with the result in Zhou et al. (2007), indicating that the optical variability of the source is weak in a timescale of about ten years. We make use of the relationship suggested in McHardy et al. (2006):


with a simplified formalism:


as well as the relationship in González-Martín & Vaughan (2012):


with a simplified formalism:


The BH mass of 1H 0323+342 is estimated to be (Equation 3) or (Equation 5), respectively, using the bolometric luminosity of  ergs s (Wang et al., 2016). The BH mass will be and if the bolometric luminosity of  ergs s (Zhou et al., 2007) is used, which are consistent with the former results.

Finally, we calculate the PSDs of 1H 0323+342 separately for the soft-band light curve (0.2-1.0 keV) and the hard-band light curve (2.0-10.0 keV). We find that both PSDs show similar results as presented above for the full band, albeit with lower significance of the existence of the bend/break frequency.

Measurements Reference
From the PSD bend frequency
 Hz 1
 Hz 1
 erg s
 Hz 2
 erg s
From the normalized excess variance of variability

Note. – Column 1: the BH mass estimators with the uncertainties at the 1  level; Column 2: BH mass derived from different methods for 1H 0323+342, with the uncertainties at the 1  level; Column 3: corresponding references, 1-González-Martín & Vaughan (2012), 2-McHardy et al. (2006), 3-Ponti et al. (2012).

Table 1: Estimated BH mass for 1H 0323+342

3.2 Excess Variance

Besides PSD, the X-ray normalized excess variance () of X-ray variations is also a useful and convenient tool to describe the variability amplitude. Following Nandra et al. (1997), the normalized excess variance can be calculated using the definition:


where is the number of good time bins of an X-ray light curve, the unweighted arithmetic mean of the count rates, and the count rates and their uncertainties, respectively, in each bin.

It has been shown that the tight inverse correlation between the X-ray excess variance and the BH mass of AGNs can be used to estimate the BH mass for AGNs (Lu & Yu, 2001; Papadakis, 2004; Zhou et al., 2010). Ponti et al. (2012) reaffirmed this relationship by making use of a large sample of 161 AGNs observed with XMM-Newton, and found a small scatter of 0.4 dex from the reverberation mapping sample. In fact, the relation is actually a manifestation of the inverse scaling of the break frequency of the AGN PSD with the BH mass, since the excess variance is the integral of the PSD over the frequency domain (Papadakis, 2004; Pan et al., 2015).

We calculate the excess variances on a timescale of 40 ks, using the light curve of 1H 0323+342 with a binsize of 250 s and energy of 2-10 keV from PN camera only, which is the same as in Ponti et al. (2012). The resulting excess variance is 0.0145. A problem of this method is the large uncertainty. It comes from two sources, one as the measurement uncertainty and the other of the stochastic nature of the variability process, which is non-negligible (Vaughan et al., 2003b; Allevato et al., 2013). The measurement uncertainty is calculated from the equation given in Vaughan et al. (2003b), while the stochastic scatter is computed from the simulation method (see Vaughan et al., 2003b; Pan et al., 2015 for more details), using the best-fit bending power-law PSD model (the Poisson noise is set to be zero) described in Section 3.1. We simulate 1000 light curves using the method of Timmer & Koenig (1995) with a binsize of 250 s and duration of 40 ks. The distribution of the calculated excess variances of all the light curves is obtained, from which the range of stochastic uncertainty at the significance level of 68% is found: and . It is worth noting that the mean excess variance of the simulated light curves is larger than the intrinsic normalized excess variance (integral of the PSD model) with a ratio of 1.31. The total uncertainties then are obtained by combining in quadrature the stochastic and measurement uncertainties, which are at the confidence level. The BH mass for 1H 0323+342 is derived using the relation suggested in Ponti et al. (2012):


with a simplified formalism:


yielding a BH mass of . It has been suggested in Allevato et al. (2013) that the observed calculated from the continuously sampled light curve may overestimate the intrinsic , which is consistent with our simulations. If this is the case, the intrinsic would be 0.0111, indicating a BH mass of .

In fact, it has been suggested that the relation flattens at , below which the excess variances tend to remain constant around 0.01 (Ludlam et al., 2015; Pan et al., 2015). For 1H 0323+342, the value of excess variance is around 0.01, indicating a BH mass being likely the order of or even lower.

4 Summary and Discussion

Thanks to the long, uninterrupted and high quality X-ray observation with XMM-Newton, we are able to estimate the BH mass for the archetypal NLS1-blazar hybrid AGN 1H 0323+342 independently from the X-ray variability, which is free from the possible orientation effect of the BLR. The PSD of the light curve can be better fitted with a bending(broken) power-law model over a simple power-law, with a bend frequency  Hz. The previously established relationships between the PSD bend/break frequency and the BH mass (and possibly bolometric luminosity), as suggested by McHardy et al. (2006) and González-Martín & Vaughan (2012), are employed. The BH mass for 1H 0323+342 is constrained in the range of , hence the Eddington ratio of . As an alternative approach, we also calculate the normalized excess variances of the light curve, giving a BH mass consistent with the results from the PSD break. All these estimates are in reasonable agreement with one another, given the relatively large uncertainties that are calculated from Monte Carlo simulations (see Table 1).

It should be noted that the BH mass—variability relationships used in this paper are based on studies of radio-quiet AGNs. Although 1H 0323+342 is radio-loud, there is strong evidence (based on the X-ray spectral shape and multi-wavelength spectral energy distribution) that its X-ray emission below 10 keV is predominantly from the disk/corona rather than from the relativistic jets, as in radio-quiet AGNs (Yao et al., 2015b). A similar conclusion was also reached by Kynoch et al. (2018), who analysed the same XMM-Newton observation used in this paper. In fact, Yao et al. (2015b) also calculated the X-ray normalized excess variance using a Suzaku observation with somewhat lower data quality and estimated a BH mass . The same practice using Nustar data was applied by Landt et al. (2017) and a similar result was obtained. These results are consistent with our estimates from the excess variance. The BH mass derived independently from the X-ray variability for 1H 0323+342 is thus in agreement with that estimated from the broad optical/infrared emission lines (Zhou et al., 2007; Wang et al., 2016; Landt et al., 2017), which has a typical systematics of  dex.

Laor (2000) found that radio-loud AGNs tend to have BH masses larger than . Furthermore, Ho (2002) demonstrated that radio loudness is strongly inversely correlated with the Eddington ratio. The small BH mass and high Eddington ratio found in 1H 0323+342 are at odds with the above findings. If the BH mass found for 1H 0323+342 is typical of radio-loud NLS1 with relativistic jets, this would imply a new parameter space in which relativistic jets can also be produced, i.e., AGNs with relatively small BH masses and high Eddington ratios (see Komossa et al., 2006; Yuan et al., 2008). An interesting analogy was also suggested to BHXB, which can also produce (transient) jets at the very high state when the Eddington ratios are the highest (e.g. Yuan et al., 2008).

Moreover, the consistency of the X-ray variability and virial estimation of the BH mass implies that the BLR in this NLS1 is not heavily flattened, as we have a near pole-on view on 1H 0323+342 as shown by Fuhrmann et al. (2016). Given that 1H 0323+342 is just an individual case, more -ray emitting NLS1s with long, continuous and high quality X-ray observations are further needed to provide independent BH mass estimation.

This work is supported by the National Natural Science Foundation of China (Grant No.11473035). SY thanks the support from the KIAA-CAS fellowship. This work is based on observations obtained with XMM-Newton, an ESA science mission with instruments and contributions directly funded by ESA Member States and NASA. This research has made use of the NASA/IPAC Extragalactic Database (NED).


  • Abdo et al. (2009a) Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009a, ApJ, 699, 976
  • Abdo et al. (2009b) Abdo, A. A., Ackermann, M., Ajello, M., et al. 2009b, ApJ, 707, L142
  • Allevato et al. (2013) Allevato, V., Paolillo, M., Papadakis, I., & Pinto, C. 2013, ApJ, 771, 9
  • Boroson & Green (1992) Boroson, T. A., & Green, R. F. 1992, ApJS, 80, 109
  • Boroson (2002) Boroson, T. A. 2002, ApJ, 565, 78
  • Calderone et al. (2013) Calderone, G., Ghisellini, G., Colpi, M., & Dotti, M. 2013, MNRAS, 431, 210
  • Collin & Kawaguchi (2004) Collin, S., & Kawaguchi, T. 2004, A&A, 426, 797
  • D’Ammando et al. (2012) D’Ammando, F., Orienti, M., Finke, J., et al. 2012, MNRAS, 426, 317
  • D’Ammando et al. (2015) D’Ammando, F., Orienti, M., Larsson, J., & Giroletti, M. 2015, MNRAS, 452, 520
  • Elvis et al. (1994) Elvis, M., Wilkes, B. J., McDowell, J. C., et al. 1994, ApJS, 95, 1
  • Fuhrmann et al. (2016) Fuhrmann, L., Karamanavis, V., Komossa, S., et al. 2016, Research in Astronomy and Astrophysics, 16, 176
  • Foschini (2011) Foschini, L. 2011, Narrow-Line Seyfert 1 Galaxies and their Place in the Universe, 24
  • González-Martín & Vaughan (2012) González-Martín, O., & Vaughan, S. 2012, A&A, 544, A80
  • Grupe et al. (2010) Grupe, D., Komossa, S., Leighly, K. M., & Page, K. L. 2010, ApJS, 187, 64
  • Ho (2002) Ho, L. C. 2002, ApJ, 564, 120
  • Jarvis & McLure (2006) Jarvis, M. J., & McLure, R. J. 2006, MNRAS, 369, 182
  • Kaspi et al. (2000) Kaspi, S., Smith, P. S., Netzer, H., et al. 2000, ApJ, 533, 631
  • Komossa et al. (2006) Komossa, S., Voges, W., Xu, D., et al. 2006, AJ, 132, 531
  • Komossa (2008) Komossa, S. 2008, Revista Mexicana de Astronomia y Astrofisica Conference Series, 32, 86
  • Kynoch et al. (2018) Kynoch, D., Landt, H., Ward, M. J., et al. 2018, MNRAS, 475, 404
  • Landt et al. (2017) Landt, H., Ward, M. J., Baloković, M., et al. 2017, MNRAS, 464, 2565
  • Lawrence & Papadakis (1993) Lawrence, A., & Papadakis, I. 1993, ApJ, 414, L85
  • Laor (2000) Laor, A. 2000, ApJ, 543, L111
  • Leighly (1999) Leighly, K. M. 1999, ApJS, 125, 297
  • León-Tavares et al. (2011) León-Tavares, J., Valtaoja, E., Chavushyan, V. H., et al. 2011, MNRAS, 411, 1127
  • León Tavares et al. (2014) León Tavares, J., Kotilainen, J., Chavushyan, V., et al. 2014, ApJ, 795, 58
  • Lu & Yu (2001) Lu, Y., & Yu, Q. 2001, MNRAS, 324, 653
  • Ludlam et al. (2015) Ludlam, R. M., Cackett, E. M., Gültekin, K., et al. 2015, MNRAS, 447, 2112
  • Mathur et al. (2012) Mathur, S., Fields, D., Peterson, B. M., & Grupe, D. 2012, ApJ, 754, 146
  • Markowitz et al. (2003) Markowitz, A., Edelson, R., Vaughan, S., et al. 2003, ApJ, 593, 96
  • McHardy (1985) McHardy, I. 1985, Space Sci. Rev., 40, 559
  • McHardy et al. (2006) McHardy, I. M., Koerding, E., Knigge, C., Uttley, P., & Fender, R. P. 2006, Nature, 444, 730
  • Nandra et al. (1997) Nandra, K., George, I. M., Mushotzky, R. F., Turner, T. J., & Yaqoob, T. 1997, ApJ, 476, 70
  • Osterbrock & Pogge (1985) Osterbrock, D. E., & Pogge, R. W. 1985, ApJ, 297, 166
  • Paliya et al. (2014) Paliya, V. S., Sahayanathan, S., Parker, M. L., et al. 2014, ApJ, 789, 143
  • Paliya et al. (2018) Paliya, V. S., Ajello, M., Rakshit, S., et al. 2018, ApJ, 853, L2
  • Pan et al. (2015) Pan, H.-W., Yuan, W., Zhou, X.-L., Dong, X.-B., & Liu, B. 2015, ApJ, 808, 163
  • Pan et al. (2016) Pan, H.-W., Yuan, W., Yao, S., et al. 2016, ApJ, 819, L19
  • Papadakis & Lawrence (1993) Papadakis, I. E., & Lawrence, A. 1993, MNRAS, 261, 612
  • Papadakis & McHardy (1995) Papadakis, I. E., & McHardy, I. M. 1995, MNRAS, 273, 923
  • Papadakis (2004) Papadakis, I. E. 2004, MNRAS, 348, 207
  • Pasham et al. (2014) Pasham, D. R., Strohmayer, T. E., & Mushotzky, R. F. 2014, Nature, 513, 74
  • Ponti et al. (2012) Ponti, G., Papadakis, I., Bianchi, S., et al. 2012, A&A, 542, A83
  • Sulentic et al. (2008) Sulentic, J. W., Zamfir, S., Marziani, P., & Dultzin, D. 2008, Revista Mexicana de Astronomia y Astrofisica Conference Series, 32, 51
  • Timmer & Koenig (1995) Timmer, J., & Koenig, M. 1995, A&A, 300, 707
  • Vaughan et al. (2003a) Vaughan, S., Fabian, A. C., & Nandra, K. 2003a, MNRAS, 339, 1237
  • Vaughan et al. (2003b) Vaughan, S., Edelson, R., Warwick, R. S., & Uttley, P. 2003b, MNRAS, 345, 1271
  • Vaughan (2010) Vaughan, S. 2010, MNRAS, 402, 307
  • Wang et al. (1996) Wang, T., Brinkmann, W., & Bergeron, J. 1996, A&A, 309, 81
  • Wang et al. (2016) Wang, F., Du, P., Hu, C., et al. 2016, ApJ, 824, 149
  • Xu et al. (2012) Xu, D., Komossa, S., Zhou, H., et al. 2012, AJ, 143, 83
  • Yang et al. (2018) Yang, H., Yuan, W., Yao, S., et al. 2018, arXiv:1801.03963
  • Yao et al. (2015a) Yao, S., Yuan, W., Zhou, H., et al. 2015a, MNRAS, 454, L16
  • Yao et al. (2015b) Yao, S., Yuan, W., Komossa, S., et al. 2015b, AJ, 150, 23
  • Yuan et al. (2008) Yuan, W., Zhou, H. Y., Komossa, S., et al. 2008, ApJ, 685, 801-827
  • Zhou et al. (2006) Zhou, H., Wang, T., Yuan, W., et al. 2006, ApJS, 166, 128
  • Zhou et al. (2007) Zhou, H., Wang, T., Yuan, W., et al. 2007, ApJ, 658, L13
  • Zhou et al. (2010) Zhou, X.-L., Zhang, S.-N., Wang, D.-X., & Zhu, L. 2010, ApJ, 710, 16
  • Zhou et al. (2015) Zhou, X.-L., Yuan, W., Pan, H.-W., & Liu, Z. 2015, ApJ, 798, LL5
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description