Free Access
Issue
A&A
Volume 597, January 2017
Article Number A100
Number of page(s) 9
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/201629444
Published online 11 January 2017

© ESO, 2017

1. Introduction

X-ray variability is a well-known property of active galactic nuclei (AGN). In recent years, many studies have focused on its characterization, and a class of extremely variable sources was found. Sources showing a transition between Compton-thin (i.e. with an obscuring column density of 1022 < NH < 1024 cm-2) and Compton-thick levels of obscuration (NH > 1024 cm-2) are called “changing-look AGN”, and are important in assessing the relevance and physics of variability. The most famous cases are NGC 1365 (Risaliti et al. 2005; Walton et al. 2014; Rivers et al. 2015b), NGC 6300 (Guainazzi 2002), NGC 2992 (Gilli et al. 2000) , NGC 7674 (Bianchi et al. 2005; Gandhi et al. 2016), and NGC 7582 (Piconcelli et al. 2007; Rivers et al. 2015a). The nature of X-ray variability could be explained in different ways. A drop in flux can either be due to intrinsic fading of the central engine since we do not expect the accretion of matter on supermassive black holes (SMBHs) to be constant in time, or to an eclipsing phenomenon caused by some clumpy material absorbing the radiation along the line of sight (l.o.s.). There could also be other effects, such as instabilities in the accretion flow. A significant change in the spectral shape with a constant flux can also occur.

In order to adequately study this complex property of AGN, monitoring sources on a wide range of timescales from weeks to years is needed, ideally with high sensitivity in the hard X-ray band (>10 keV), which more directly probes the primary emission from the innermost regions of the AGN.

Also known as UGC 4203, Mrk 1210 (z = 0.0135, Fig. 1) hosts a Seyfert 2 AGN which was initially observed by ASCA in 1995 (Awaki et al. 2000). The flat spectrum and high equivalent width (EW) of the iron line at 6.4 keV were interpreted as emerging from reflection off circumnuclear matter of the AGN primary X-ray continuum, severely suppressed by Compton-thick absorption. In 2001, an XMM-Newton observation (Guainazzi et al. 2002) found that Mrk 1210, six years after the first observation, was still obscured but at the Compton-thin level only. Interpreting the change as an intrinsic flux enhancement, they coined the name “Phoenix galaxy” for this new changing-look AGN. Also Ohno et al. (2004), in the same year, observed Mrk 1210 with BeppoSAX and found similar results. Later on, the Phoenix galaxy was observed by other instruments, always showing variability of less than a factor of two in both intrinsic emission and column density, with the latter always in the Compton-thin regime (Matt et al. 2009; Risaliti et al. 2010). The list of all X-ray observations is presented in Table 1.

thumbnail Fig. 1

Sloan Digital Sky Survey (SDSS) optical image of Mrk 1210 (gri composite). The cyan and red circles are the source extraction regions for the NuSTAR (see Sect. 2) and Chandra (see Appendix A) spectra, respectively.

Open with DEXTER

While the source has been extensively studied for the past twenty years, the reason for the changes observed between 1995 and 2001 remains unclear. A definitive answer is not yet available; however, as time went by, a change in NH obscuring the nucleus was progressively addressed as the principal effect causing the change in flux. A better understanding of this source can be achieved thanks to the Nuclear Spectroscopic Telescope Array (NuSTAR). Launched in 2012, NuSTAR is the first telescope able to focus hard X-ray photons, enabling a gain of a factor of ~100 in sensitivity and one order of magnitude in angular resolution with respect to previous facilities in the hard (>10 keV) X-ray band (Harrison et al. 2013). In its operating band (3–79 keV), thanks to a field of view (FoV) at 10 keV of 10× 10, and an 18′′ FWHM with a half-power diameter of 58′′, NuSTAR is suitable for studying the hard X-ray spectra of AGN with high sensitivity. In this paper, we report on the NuSTAR observation of Mrk 1210 in 2012. After describing the data reduction (Sect. 2), we present the spectral analysis in Sect. 3. A discussion of our results in the context of the previous findings in the literature is provided in Sect. 4, and we draw our conclusions in Sect. 5.

Table 1

History of X-ray observations of Mrk 1210.

2. Data reduction

NuSTAR observed Mrk 1210 on October 5, 2012 for 15.4 ks. The raw event file was processed using the NuSTAR Data Analysis Software package v. 1.5.1 (NuSTARDAS)1. Calibrated and cleaned event files were produced using the calibration files in the NuSTAR CALDB (20150316) and standard filtering criteria with the nupipeline task. We used the nuproducts task included in the NuSTARDAS package to extract the NuSTAR source and background spectra using the appropriate response and ancillary files. We extracted the source spectrum and light curve in both focal plane modules, FPMA and FPMB, using 87′′-radius circular apertures, while the background was extracted using three source-free circular regions on the same chip as the source. All spectra were binned to a minimum of 20 photons per bin using the HEAsoft task grppha.

thumbnail Fig. 2

NuSTAR background-subtracted spectrum of Mrk 1210 (FPMA in black, FPMB in red). The background is shown by diamonds, and is lower than the source signal up to ~50 keV. Both have been rebinned for plotting clarity.

Open with DEXTER

thumbnail Fig. 3

3–79 keV lightcurve of Mrk 1210. The background-subtracted FPMA and FPMB lightcurves were summed.

Open with DEXTER

3. Spectral analysis

We used the XSPEC software version 12.9.0 (Arnaud 1996) to carry out the spectral analysis. The source is clearly detected by NuSTAR up to ~60 keV (Fig. 2), with a net count rate of 0.3894 ± 0.0052 cts s-1 and 0.3762 ± 0.0052 cts s-1 for FPMA and FPMB, respectively. During the observation, the source kept a nearly constant flux, with amplitude variations of less than 30% and no systematic trend, as shown in Fig. 3.

We included in all our fits a Galactic column density NH,gal = 3.45 × 1020 cm-2(Kalberla et al. 2005).

3.1. Phenomenological models

Since we know from previous work that a simple Galactic-absorbed power law is not able to reproduce the spectral complexity of Mrk 1210, we began fitting its spectrum with a plcabs model; a power law with an exponential cut-off, taking into account Compton scattering and absorption at the source (Yaqoob 1997). We also added a narrow (σ = 10 eV, fixed) Gaussian line component to fit the clearly visible feature at ~6 keV. The fit was good (χ2/ν = 537/480), with a photon index Γ = 1.40 ± 0.05 and an obscuring column of gas along the l.o.s. of NH = (1.7 ± 0.2) × 1023 cm-2. From previous studies, an additional soft component in the X-ray spectrum of the Phoenix galaxy is known (e.g., Guainazzi et al. 2002). Since this component can contribute between 3 and ~5 keV in the NuSTAR band, we modeled it adding another power law, and linking all the parameters (photon index, redshift, normalization) to those of the plcabs component. We then allowed for a constant multiplicative value (referred to as fs) to vary in the fit, which represents the fraction of the X-ray continuum Thomson-scattered into the soft X-rays. This procedure is often employed in the literature, and fractions of a few percent are typical (Bianchi & Guainazzi 2007). Adding such a scattered power law (SPL) improved the fit (χ2/ν = 521/479) at more than a 99% confidence limit, based on an F-test.

Although the fit was already acceptable, we wanted to test if any reflection component was required by the data. We therefore added a pexrav model (Magdziarz & Zdziarski 1995), which includes Compton reflection on a slab of neutral material with infinite optical depth. The final model in XSPEC notation is the following: (1)We will refer to this as the “baseline” model. The fit improved dramatically (χ2/ν = 478/478, Δχ2 = 43, see Fig. 4): the source had a photon index Γ = 1.9 ± 0.1, a column density cm-2, a reflection parameter (intended as the ratio of the pexrav normalization and the plcabs one) and a fraction of the primary power law scattered in the soft part of the spectrum of , which is higher than the average of Seyfert 2 galaxies, but not unusual (Bianchi & Guainazzi 2007). This latter component, however, is now only significant at the 97% confidence limit, due to the high level of reflection. Upon removing it, the fit becomes slightly worse (χ2/ν = 483/479), but the parameters are the same within the uncertainties.

thumbnail Fig. 4

Fit with the baseline model (1)to the NuSTAR spectrum of Mrk 1210 (FPMA in black, FPMB in red).

Open with DEXTER

We also note that the best fit line energy is lower than the iron Kα 6.4 keV centroid, although consistent with it at a 99% confidence limit (Fig. 5). Using previous versions of the NuSTARDAS software (v 1.2.1) and NuSTAR CALDB (20130909), and applying the same baseline model, we find that all the fit parameters are the same within the uncertainties, with a centroid line energy of 6.35 ± 0.06 keV, which is now consistent with the expected value of 6.4 keV at a 90% confidence limit. Moreover, as we shall see in the following subsections (in particular in Sect. 3.3), we get good fits for the line feature adopting both toroidal models, in which the energy line is fixed to 6.4 keV, and a Compton shoulder is self-consistently calculated. Finally, since none of the previous observations found such a significant line energy shift, data with an improved spectral resolution would be needed to assess the relevance of the one found here.

thumbnail Fig. 5

68% (black), 90% (red), 99% (green) confidence contours. The two-dimensional confidence contours allow a marginal consistency with the iron Kα line at 6.4 keV, indicated by the dashed horizontal line.

Open with DEXTER

3.2. Toroidal models

The high reflection value obtained by applying the baseline model implies a more complicated geometry or some time delay between the components; at the same time, a reflection component alone is not able to successfully fit the spectrum (χ2/ν = 609/481). Therefore, we explored more physically motivated models which can take into account both absorption and reflection in a self-consistent way.

To this aim, we tried the BNTorus model (Brightman & Nandra 2011) and the MYTorus model (Murphy & Yaqoob 2009). These were developed based on Monte Carlo simulations, adopting a toroidal geometry for the material responsible for obscuration, scattering, and line fluorescence, all self-consistently treated.

We first applied the BNTorus model to the data, fixing the inclination angle of the torus (i.e., the parameter θinc; θinc = 0 describes a face-on view) to 87°. The fit was acceptable (χ2/ν = 517/481), with , cm-2, and half-opening angle of the torus θtor< 65°, from which a covering factor of >0.4 is inferred. A very high fraction, , of the primary power law is scattered below 5 keV.

Similarly, the MYTorus model in its “coupled”, default mode (i.e., with a donut-shaped, smooth reprocessor with a covering factor of 0.5) gives an acceptable fit: χ2/ν = 512/480, Γ = 1.63 ± 0.08, NH = (3.5 ± 0.6) × 1023 cm-2, with a high fraction, (12 ± 3)%, of the primary power law scattered below 5 keV. However, using either torus model, the flattening in the spectrum above 20 keV is not properly fitted (see Fig. 6), and we conclude that they are not able to capture the spectral shape of Mrk 1210 while the phenomenological baseline model does. This suggests that the Mrk 1210 torus contains Compton-thick material producing the pronounced Compton hump, but our line of sight does not pass the Compton-thick part in the NuSTAR observation.

3.3. A physical picture

In the previous section we found that the reflection component, modeled with a pexrav model, has a normalization a factor of ~2.5 larger than the primary continuum from the AGN. Indeed, applying other self-consistent models based on Monte Carlo simulations with a classic toroidal geometry, we found worse fits with respect to that obtained with the baseline model, due to the presence of the reflection excess.

thumbnail Fig. 6

Fit with the BNTorus model (top) and the coupled MYTorus model (bottom) to the NuSTAR spectrum of Mrk 1210 (FPMA in black, FPMB in red).

Open with DEXTER

While the BNTorus model allows the user to fit for the torus opening angle and hence to get an estimate of the covering factor of the source, the three components of the MYTorus model can be decoupled (transmitted flux, reflected flux, and fluorescence lines) to simulate different and more complex geometries (we refer to Yaqoob 2012 for an extensive explanation of the decoupled version of the model, and also LaMassa et al. 2014, for a systematic application of the model on a sample of sources). Briefly, together with the canonical components with a l.o.s. angle fixed at 90° (called front-scattered, with the subscript 90), the decoupled version includes an additional couple of scattered+line components seen face on, that is, with the l.o.s. angle fixed to 0° (called back-scattered, with the subscript 00). The relative normalizations between the 90° and 0° components (called AS90 and AS00, respectively) can be untied and left free in the fit. Additionally, one can also untie the column densities of the two scattered/reflected components. In XSPEC notation, this model is described as (2)If we decouple the model (i.e., we leave free the constants AS90 and AS00), and allow the obscuring columns to be different (i.e., NH90NH00), we get a good fit (χ2/ν = 489/479, Fig. 7). The front-scattered component vanishes (i.e., AS90 → 0), while the back-scattered component converges to the same normalization as the primary continuum (i.e., AS00 → 1, preserving the internal self-consistency of the model), and the clouds responsible for the reflection component are Compton-thick, with an optical depth τ ~ 2.7 (NH ~ 4 × 1024 cm-2). Moreover, the upper limit on the column density is unconstrained by the fit. In other words, the column density obscuring the l.o.s. is different from the column density of the clouds responsible for the back-scattered reflection. Fixing the normalizations of the back-scattered component and the primary power law to be equal as suggested by the fit (i.e., AS00 = 1) does not change the result regarding the different column densities for the two components. On the other hand, the fit becomes worse (χ2/ν = 505/479, Δχ2 = 16 for the same number of degrees of freedom) if the reflection component is front-scattered, that is, if we simply decouple the column densities from the default configuration of MYTorus. Also in this case, the data require two different absorbing columns: a Compton-thin one obscuring the primary emission, and a thick one producing the reflection. The parameters of the models used for fitting (baseline, toroidal, MYTorus decoupled) are shown in Table 2.

thumbnail Fig. 7

Fit with the MYTorus decoupled model (2), with normalizations and column densities decoupled (FPMA in black, FPMB in red). Fitting parameters are reported in Table 2.

Open with DEXTER

Table 2

Summary of spectral analysis.

3.4. Intrinsic luminosity

thumbnail Fig. 8

Observed flux in the 210 keV band of Mrk 1210 (top panel), Fe Kα line equivalent width (middle panel) and column density (bottom panel) as a function of time. The inset panels show the zoomed-in region of the five Chandra observations during 2008. (*) We report here the equivalent width of model 1 instead of that of model 2 in Ohno et al. (2004). The two values are consistent within the uncertainties, and we adopt the more constrained one for clarity.

Open with DEXTER

Our models find an intrinsic 2–10 keV luminosity in the range 0.6−1 × 1043 erg s-1. It is interesting to compare these results with other commonly used proxies for the X-ray luminosity. Using the mid-infrared luminosity, the WISE all-sky catalog (Wright et al. 2010; Cutri et al. 2013) reports a W3 (12 μm) magnitude of 4.634 ± 0.015, which translates to a 12 micron luminosity L12 μm = 4.6 × 1043 erg s-1, with a ~1.5% statistical error, using the standard WISE zeropoints. The W1−W2 color is 1.392 ± 0.030, suggesting that the mid-infrared luminosity is AGN-dominated, being the source above the color threshold of W1−W2 > 0.8 identified by Stern et al. (2012). Then, from the mid-IR/X-ray relation (Gandhi et al. 2009; Asmus et al. 2015), one would predict log (L2−10 [erg/s] ) ~ 43.3, which is consistent with our MYTorus luminosity.

We can also use the optical [OIII] emission line to provide another independent estimate of the X-ray luminosity. Koss et al. (in prep.) report a dust reddening-corrected [OIII]λ5007 flux of erg cm-2 s-1, from which we derive a luminosity of L[OIII] ~ 3.0 × 1041 erg s-1. Using the relationship for Seyfert galaxies between L2−10 and L[OIII] with a scatter of 0.5 dex presented in Berney et al. (2015), we predict log (L2−10 [erg/s] ) ~ 43.3, consistent with the luminosity derived from the infrared and with that obtained by our spectral analysis. These values suggest that the luminosity derived by the baseline model is likely underestimated by a factor of ~2, due to the phenomenological combination of the absorbed power law and infinite slab reflection models. The physically motivated, decoupled MYTorus model alleviates this problem and allows for a more reliable estimate of the intrinsic luminosity of the source.

4. Long-term behavior and discussion

In order to understand the time dependent variation of the X-ray spectrum, we compiled the observed 210 keV fluxes, iron line equivalent widths, and column densities from the literature together with results of this work to gain a global picture of the behavior of the source. We choose to use our baseline model parameters (shown in Table 2) in order to compare directly with the results from previous papers based on the same phenomenological models. To do so, we reduced the data of the six Chandra observations with standard procedures. See the Appendix for the analysis of the 2004 observation (hereafter “C0”), while for the other five observations during 2008 (namely, “C1-5”) the best-fitting models of Risaliti et al. (2010), analogous to our baseline model, were applied. We also note that Mrk 1210 is bright enough to be detected by the Burst Alert Telescope (BAT) onboard Swift, and it is indeed present in the Swift/BAT 70 month catalog (Baumgartner et al. 2013). As can be seen easily from the top panel of Fig. 8, the source was in a low-flux state during the ASCA observation. It has been seen in a high-flux state since then, with different observatories. The equivalent widths of the iron line (Fig. 8, middle panel) reflect this trend. From the bottom panel of Fig. 8 it is clear that the column density of the source has been around 2−4 × 1023 cm-2 since 2001, while the ASCA data do not help in shedding light on the column density in 1995. Both Awaki et al. (2000) and Guainazzi et al. (2002) found a Compton-thin obscuration applying a pure transmission model to the ASCA data, despite obtaining a worse fit with respect to a reflection-dominated scenario. The absorption-corrected 210 keV luminosity reported by Awaki et al. (2000), on the other hand, is a factor of approximately five lower than the later, intrinsic luminosities, but the quality of the data does not allow for a robust estimate of the intrinsic luminosity, nor a robust detection of the transmitted continuum, leaving total degeneracy between heavily absorbed and intrinsically faint primary emission scenarios. In the following, both the flux change (Sect. 4.1) and the eclipsing scenario (Sect. 4.2) are discussed.

4.1. Change in intrinsic luminosity

If we suppose that Mrk 1210 was Compton-thin also during the ASCA observation, but with the intrinsic emission shut off, the reflection-dominated spectrum was entirely due to an echo of a previous high-flux state.

Thanks to the high-quality NuSTAR spectrum, we are able to disentangle the intrinsic power law and the reflection component. If we shut off the intrinsic component in our NuSTAR data, leaving only the pexrav model and the line component (i.e., a reflection dominated model), we find an observed flux and line equivalent width consistent with the values reported by Awaki et al. (2000), albeit that uncertainties on these values cannot be reliably computed, being a reflection dominated scenario highly disfavored by the NuSTAR data, as already discussed in Sect. 3 (see Table 3).

Moreover, from the NuSTAR best fit baseline model, we can easily adjust the fit parameters, and vary only the normalization of the intrinsic continuum, to recover all the subsequent states of Mrk 1210, which are similar to the NuSTAR one. This means that keeping a reflection component constant and varying only the intrinsic continuum, together with minor (factor ~2) but required column density variations on shorter timescales (as noted by Risaliti et al. 2010), all the observations of the Phoenix galaxy during the last ~20 yr can be explained.

As a further check, the 10–40 keV observed flux and line flux are plotted in Fig. 9. In the top panel, we computed the 10–40 keV flux, which is the band where the reflection component is thought to dominate, from BeppoSAX/PDS2, Swift/BAT 70 month catalog3, and NuSTAR. We also extrapolated the ASCA 1040 keV flux, using the same procedure of Swift/BAT. The bottom panel shows the line flux for the iron line component, measured using different instruments. Both panels appear to confirm the constancy of the reflection through the years, with a hint of a slightly enhanced line during the Suzaku observation.

If this is the correct picture, Mrk 1210 turned on between 1995 and 2001, and stayed more or less constant since then. The reflection component, though, still has to adjust and respond to this change, since it can be considered constant between all the observations. A simple light-travel argument can then put constraints on the distance between the central source and the reflector. For an edge-on view of the system, D = cΔt/ 2, and Δt> 12−17 yr, depending on when the source switched on (between the beginning of 1996 and the beginning of 2001). In this case, the distance between the source and the reflector is D> (1.9−2.6) pc. First a fading, and then an increase in reflection is then expected to occur in the future, and monitoring of the source is the only way to keep track of its changes. Monitoring could also shed new light on the role of the X-ray absorber, which varies on much shorter temporal scales and seems to be associated with broad line region (BLR) clouds (see Risaliti et al. 2010). If this is the case, the physical interpretation of our result applying the decoupled MYTorus model is straightforward: the l.o.s. is intercepting the variable (and presumably compact) Compton-thin absorber, while the reflector is located on larger scales and the photons reflecting back to our l.o.s. are “seeing” a Compton-thick column of gas. Risaliti et al. (2010) suggest that the absorber is a BLR cloud, assuming a central black hole (BH) mass of 5−7 × 107M. In this picture, a second “screen” is required to obscure the broad lines in the optical spectrum, which is instead likely showing an outflow, resulting in the classification of Mrk 1210 as a Seyfert 2 (Mazzalay & Rodríguez-Ardila 2007). However, if the reflection component comes from photons reflecting off the far inner wall of the torus, delayed with respect to the intrinsic emission, a rather distant torus edge is needed with respect to the dust sublimation radius Rsub, which is usually thought to mark the inner wall of the torus, and in this case is ~0.2 pc (Gandhi et al. 2009, and references therein, and adopting a bolometric correction of 20; Lusso et al. 2012).

Table 3

ASCA best fit parameters compared with NuSTARpexrav+zgauss only model.

thumbnail Fig. 9

Observed 1040 keV flux (top panel) and iron line flux (bottom panel) of Mrk 1210, as a function of time.

Open with DEXTER

4.2. Eclipsing event

The ASCA data cannot distinguish between a Compton-thick eclipsing event and an intrinsic low-flux state of the central engine. A random pass of a Compton-thick cloud (or an alignment of clouds along the l.o.s.) during the ASCA observation is consistent with the known clumpiness of the gas surrounding obscured AGN. During the two observations by ASCA in 1995, separated by 25 days, Mrk 1210 was in a low-flux state. In an eclipsing scenario, this means that the putative eclipsing event lasted for at least 25 days. Assuming Keplerian motion of the clump, the linear size eclipsed by a moving Compton-thick cloud in a given time interval can be written as s = 4.5 × 1013(M7/Dpc)1/2Δt25 cm, where Dpc is the distance from the center in parsecs, and Δt25 is the eclipse time interval in units of 25 days. This implies that the cloud is rotating with a Keplerian velocity vK ~ 210(M7/Dpc)1/2 km s-1. Finally, assuming a Compton-thick column density, the average density of the clump is ρ ~ 2 × 1010NH,24(Dpc/M7)1/2/ Δt25 cm-3, where NH,24 is the column density in units of 1024 cm-2. Interestingly, high-density clumps on a parsec-scale (or subparsec-scale) are observed in some systems showing water maser emission at 22 GHz. Mrk 1210 is indeed one of them (Braatz et al. 1994), even if the association of the maser spots to a particular geometry is still unclear and sensitive VLBI observations are needed to further investigate the nuclear environment. The maser activity could indeed be associated with a dusty Keplerian disk orbiting the SMBH, from which the most precise BH mass available to date could be derived (e.g., Kuo et al. 2011). This would allow a direct comparison of this object with other samples of disk masers (Pesce et al. 2015; Masini et al. 2016) and retrieve information about the X-ray absorber. On the other hand, the maser spots could be associated with the outflow seen in the optical spectrum, invalidating any possible BH mass estimate. Moreover, as reported by Storchi-Bergmann et al. (1998) and Mazzalay & Rodríguez-Ardila (2007), Mrk 1210 probably shows a recent circumnuclear starburst. If the masers are associated with the outflow responsible for the broad components of the optical lines, the Phoenix galaxy could be a very interesting and local laboratory to study the interplay between AGN activity and star formation.

5. Conclusions

We presented the first NuSTAR observation of Mrk 1210, also known as the Phoenix galaxy. It is a long-studied object, considered to be part of the “changing-look” class of AGN. NuSTAR observed Mrk 1210 obscured by a Compton-thin column of gas similarly to other instruments over the past 16 yr. The data are also showing an enhanced reflection component, which requires two different columns of gas to be taken into account. Many previous studies have suggested the presence of a variable X-ray absorber, inducing changes in the absorbing column of a factor of ~2 on the timescales of days and weeks. We note that these short-term column changes, properly complemented by long-term intrinsic variability of the central engine, are able to explain the line equivalent widths and fluxes observed by different instruments over the last ~20 yr. If the low flux-state observed by ASCA in 1995 is due to an intrinsic fading of the engine, we note evidence that the reflection component has remained constant with time. In this scenario, we infer that the physical distance between the source and the reflector is of the order of at least 2 pc. First a drop, and then an enhancement of the reflection component are then expected to occur in the future, in response to the source “low-flux state” (seen by ASCA) and soft X-ray “awakening” (i.e., the Phoenix effect, Guainazzi et al. 2002) seen between 1996 and 2001. On the other hand, if the low-flux state of Mrk 1210 during the ASCA observation can be ascribed to the presence of a Compton-thick cloud obscuring the l.o.s., the cloud can be identified with a maser-emitting clump on the sub-pc scale. Indeed, if the torus-like structure is clumpy, a random passage of an over-dense cloud along the l.o.s. is expected (Wada 2012). The frequency in which such events are to be expected could be evaluated with precise hydrodynamical simulations of the nuclear environment, but we refer to Markowitz et al. (2014) for a statistical study on this topic.

Monitoring in the X-ray band is needed, either to detect the expected behavior of the reflection component and clarify the relation between the X-ray absorber and reflector, which seem to be on different spatial scales, or to possibly detect another low-flux state and further study the properties of the clumpy (sub)pc-scale absorber. To this extent, radio VLBI monitoring at 22 GHz would also be extremely useful to clarify the nature of the nuclear water maser activity.


3

The 14195 keV flux was converted to a 1040 keV using the dummyrsp command in XSPEC.

Acknowledgments

We thank the anonymous referee for useful suggestions that helped to improve the paper. This work was supported under NASA Contract NNG08FD60C, and it made use of data from the NuSTAR mission, a project led by the California Institute of Technology, managed by the Jet Propulsion Laboratory, and funded by the National Aeronautics and Space Administration. We thank the NuSTAR Operations, Software, and Calibration teams for support with the execution and analysis of these observations. This research made use of the NuSTAR Data Analysis Software (NuSTARDAS) jointly developed by the ASI Science Data Center (ASDC, Italy) and the California Institute of Technology (USA). This research has also made use of data obtained from the Chandra Data Archive and the Chandra Source Catalog, and software provided by the Chandra X-ray Center (CXC). A.M., A.C., and S.P. acknowledge support from the ASI/INAF grant I/037/12/0-011/13. M.B. acknowledges support from NASA Headquarters under the NASA Earth and Space Science Fellowship Program, grant NNX14AQ07H. P.G. and P.B. thank STFC for support (grant ST/J003697/2). S.L.M. is supported by an appointment to the NASA Postdoctoral Program at the NASA Goddard Space Flight Center, administered by Universities Space Research Association under contract with NASA. We acknowledge support from NASA NuSTAR A01 Award NNX15AV27G (F.E.B.), CONICYT-Chile grants Basal-CATA PFB-06/2007 (F.E.B., C.R.), FONDECYT Regular 1141218 (F.E.B., C.R.), “EMBIGGEN” Anillo ACT1101 (F.E.B., C.R.), the China-CONICYT fund (C.R.), and the Ministry of Economy, Development, and Tourism’s Millennium Science Initiative through grant IC120009, awarded to The Millennium Institute of Astrophysics, MAS (F.E.B.).

References

Appendix A: Analysis of the Chandra 2004 (C0) observation

We reduced the Chandra data and produced a grouped spectrum using the CIAO standard chandra_repro and specextract tasks. The source counts were extracted from a 3′′-radius circular region, while the background was extracted from four

10′′ circular regions. Soft X-ray emission below 3 keV is clearly present in the total spectrum (Fig. A.1), and was fitted, following Risaliti et al. (2010), with an apec+zpowerlw model. The hard X-ray part of the spectrum was fitted with our baseline model, Eq. (1). The fitting parameters are reported in Table A.1. We note that the results are consistent with the NuSTAR observation presented in the paper.

thumbnail Fig. A.1

Best fit and data/model ratio to the 2004 Chandra spectrum of Mrk 1210.

Open with DEXTER

Table A.1

C0 observation; best fitting parameters.

All Tables

Table 1

History of X-ray observations of Mrk 1210.

Table 2

Summary of spectral analysis.

Table 3

ASCA best fit parameters compared with NuSTARpexrav+zgauss only model.

Table A.1

C0 observation; best fitting parameters.

All Figures

thumbnail Fig. 1

Sloan Digital Sky Survey (SDSS) optical image of Mrk 1210 (gri composite). The cyan and red circles are the source extraction regions for the NuSTAR (see Sect. 2) and Chandra (see Appendix A) spectra, respectively.

Open with DEXTER
In the text
thumbnail Fig. 2

NuSTAR background-subtracted spectrum of Mrk 1210 (FPMA in black, FPMB in red). The background is shown by diamonds, and is lower than the source signal up to ~50 keV. Both have been rebinned for plotting clarity.

Open with DEXTER
In the text
thumbnail Fig. 3

3–79 keV lightcurve of Mrk 1210. The background-subtracted FPMA and FPMB lightcurves were summed.

Open with DEXTER
In the text
thumbnail Fig. 4

Fit with the baseline model (1)to the NuSTAR spectrum of Mrk 1210 (FPMA in black, FPMB in red).

Open with DEXTER
In the text
thumbnail Fig. 5

68% (black), 90% (red), 99% (green) confidence contours. The two-dimensional confidence contours allow a marginal consistency with the iron Kα line at 6.4 keV, indicated by the dashed horizontal line.

Open with DEXTER
In the text
thumbnail Fig. 6

Fit with the BNTorus model (top) and the coupled MYTorus model (bottom) to the NuSTAR spectrum of Mrk 1210 (FPMA in black, FPMB in red).

Open with DEXTER
In the text
thumbnail Fig. 7

Fit with the MYTorus decoupled model (2), with normalizations and column densities decoupled (FPMA in black, FPMB in red). Fitting parameters are reported in Table 2.

Open with DEXTER
In the text
thumbnail Fig. 8

Observed flux in the 210 keV band of Mrk 1210 (top panel), Fe Kα line equivalent width (middle panel) and column density (bottom panel) as a function of time. The inset panels show the zoomed-in region of the five Chandra observations during 2008. (*) We report here the equivalent width of model 1 instead of that of model 2 in Ohno et al. (2004). The two values are consistent within the uncertainties, and we adopt the more constrained one for clarity.

Open with DEXTER
In the text
thumbnail Fig. 9

Observed 1040 keV flux (top panel) and iron line flux (bottom panel) of Mrk 1210, as a function of time.

Open with DEXTER
In the text
thumbnail Fig. A.1

Best fit and data/model ratio to the 2004 Chandra spectrum of Mrk 1210.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.