The Changing-look Quasar Mrk 590 Is Awakening

Active galactic nuclei (AGNs) are characterized by continuum emission across the electromagnetic spectrum and strong emission lines in the UV, optical, and near-IR. AGNs with both broad and narrow emission lines are classified as "Type 1" and those with only narrow emission lines as "Type 2," and this difference is usually attributed to our viewing angle relative to an obscuring midplane. AGNs are also known to be variable sources, showing variability in the continuum as well as in emission lines. The continuum variability is observed on both short and long timescales, but the variability amplitude is usually small, of the order of ∼20% over a year for low-luminosity AGNs (i.e., with Seyfert-like luminosity of about 10⁴¹–10⁴⁴ erg s⁻¹). The emission-line variations usually track the continuum variations, but see Goad et al. (2016).

A special class of AGN that has recently been gaining recognition has been termed "changing-look quasars."¹³ Their broad emission lines are observed to appear or disappear together with large changes in continuum luminosity. The changes in the emission-line spectra are such that they change type, e.g., from "Type 1" to "Type 2," and this is clearly an intrinsic change, implying that "type" is not always associated with viewing angle. This phenomenon has become of recent interest due to a combination of several serendipitous discoveries of changing-look AGNs (e.g., Shappee et al. 2014; Denney et al. 2014; LaMassa et al. 2015) and the growing availability of large spectroscopic quasar databases with long time-baselines of multi-epoch photometry and spectroscopy. These rare occurrences have the potential to contribute to our limited understanding of the central engines of active galaxies. Systematic searches of databases such as the SDSS/BOSS/TDSS have begun to uncover more changing-look quasars at larger redshifts and have led to predictions of the possible occurrence rates for quasars to change type (MacLeod et al. 2016; Runnoe et al. 2016).

Changing-look quasars are of particular interest for their potential to shed light on the physical origin of quasar variability and the details of accretion, since they have undergone such an extreme apparent change. Additionally, what is the role these objects play in the accretion history of the black hole (BH)—are we observing the beginning/end of a quasar phase? What role do these objects play in the larger context of galaxy and BH coevolution? Recent investigations of changing-look quasars have tried to use the observations before and after the change and any additional data available on the few known sources to test whether or not the changes are due to changes in accretion rate or obscuration, and to determine whether the quasar is turning on for the first time, reviving after a period of quiescence, or something else. In all cases so far, obscuration has not been identified as the major cause of changing look. Instead, a significant increase (decrease) in the mass accretion rate is the most favored explanation for the appearance (disappearance) of the broad emission lines.

The nearby, low-luminosity AGN Markarian 590 (Mrk 590 hereafter) remains an interesting case of a changing-look quasar in the local universe. Denney et al. (2014; D14 hereafter) have described the initial set of observations that catalog this object's observed history and "change," where it went from being a strong broad-line emitter three decades ago, allowing a reverberation mapping-based direct BH mass measurement to be made ((4.75 ± 0.74) × 10⁷ M_⊙, Peterson et al. 2004), to being broad-line weak—the optical broad lines all but disappeared—in the past decade. Mrk 590 also shows the presence of ultrafast outflows in the X-ray band (Gupta et al. 2015). We have obtained additional UV and X-ray spectra since those presented by D14 that we describe in Sections 2 and 3. Implications of our results are discussed in Section 4 and we show that the "soft X-ray excess," observed in a large fraction of AGNs, is not due to the reprocessing of hard X-ray continuum, but is instead a result of thermal Comptonization of UV photons. We also argue that the changing-look phenomenon is a normal event in the duty cycle of "normal" quasars.

Mrk 590 was observed by Chandra  for 68.87 ks on 2014 November 15 (ObsID 16109). This is longer than the 30 ks initial Director's Discretionary Time (DDT) observation presented by D14, but the observing parameters were otherwise kept the same. We observed with ACIS-S/HETG; the grating was inserted in front of the detector in order to mitigate pile-up, if any. The data were reduced with CIAO following the normal procedure and analyzed using XSPEC. A Galactic column density of N_H = 2.7 × 10²⁰ cm⁻² was included in all fits and was held fixed. All the quoted errors are for 90% confidence, unless noted otherwise. Figure 1 shows the new, 2014 spectrum, together with the 2013 spectrum presented by D14, and the previous "high state" spectrum observed in 2004. The source continues to be in a low state in 2014, so the quality of the high-resolution grating spectrum is too poor to perform useful science. Hence we present the zeroth-order spectrum here, as was done by D14.

Zoom In Zoom Out Reset image size

A simple absorbed power-law model did not provide a good fit to the 2014 Chandra  spectrum. The residuals to this fit are shown in Figure 2, and clearly show a soft excess, so we added a blackbody component to the model. Clear residuals were also observed around 6 keV, so following Longinotti et al. (2007) we added two Gaussian emission lines at rest-frame 6.4 keV and 6.7 keV to represent neutral and ionized Fe Kα emission lines, respectively. The resulting fit was good (${\chi }_{\nu }^{2}=1.0$ for 104 degrees of freedom); the parameters of the fit are presented in Table 1 and the spectra are shown in Figures 3 and 4. Longinotti et al. (2007) required a third emission line at 7 keV; our data are not good enough in this energy range to require this line. They also detected a line in the soft X-ray band, at 19 Å (0.65 keV), but we see no evidence of that line.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 1.  Chandra  Best-fit Spectral Model^a

Model	Parameter 1	Normalizationb	Parameter 3
Power law	Γ = 1.6 ± 0.1	(2.4 ± 0.3) × 10−4
Blackbody	${kT}={0.11}_{-0.02}^{+0.05}$ keV	${7}_{-4}^{+7}\times {10}^{-6}$
Line 1	E = 6.4 keVc	(5 ± 2) × 10−6	EW = 268 ± 100 eV
Line 2	E = 6.7 keVc	(4 ± 2) × 10−6	EW = 176 ± 100 eV

Notes.

^a ${\chi }_{\nu }^{2}$ for the joint fit is 1.0 for ν = 104 degrees of freedom. ^bIn units of photons s⁻¹ cm⁻² keV⁻¹. ^cRest frame; parameter frozen.

Download table as:  ASCIITypeset image

The flux of the source is 2.2 × 10⁻¹² erg s⁻¹ cm⁻² at 0.3–10 keV (1.9 × 10⁻¹² erg s⁻¹ cm⁻² at 0.5–10 keV and 2.0 × 10⁻¹² erg s⁻¹ cm⁻² at 0.3–8 keV). This is significantly smaller than the 2002 flux (8.4 × 10⁻¹² erg s⁻¹ cm⁻²) reported by Gallo et al. (2006), the 2004 flux (11 × 10⁻¹² erg s⁻¹ cm⁻²) reported by Longinotti et al. (2007), and the Suzaku  flux of 6.8 × 10⁻¹² erg s⁻¹ cm⁻² at 2–10 keV reported by Rivers et al. (2012). D14 reported the low state of Mrk 590 with 0.5–10 keV flux of 1.1 × 10⁻¹² erg s⁻¹ cm⁻², so compared to our DDT Chandra  observation in 2013, the source flux appears to have increased slightly, but as shown in Figure 1, the source continues to be in a low state. The flux in the soft excess component is 3.8 × 10⁻¹³ erg s⁻¹ cm⁻² in the 0.3–8 keV band (though it is concentrated below 2 keV), which is 19% of the total. If we restrict to the 0.3–2 keV soft band, then the total flux is 9.7 × 10⁻¹³ erg s⁻¹ cm⁻² while the soft excess flux remains the same as above, making it 39% of the total soft-band flux.

We obtained observations with the Cosmic Origins Spectrograph on board the Hubble Space Telescope (HST) that were coordinated with our approved Chandra  program. The goal of the COS observations was to observe the portion of the UV spectrum not covered by the initial COS G140L DDT observations presented by D14. The new observations were obtained on 2014 November 15 with the G230L grating and covered the observed-frame wavelength ranges ∼1670−2175 Å and ∼2765−3215 Å. Importantly, this coverage was selected to allow a continuum overlap region with the 2013 G140L spectrum and also to cover the Mg ii λ2800 doublet in this object of redshift z = 0.026385. The G230L spectrum was processed with the standard COS pipelines. Figure 5 shows the G230L COS spectrum, which was modeled with a power-law continuum, a Balmer continuum, and an Fe ii template (Vestergaard & Wilkes 2001) as a means to isolate the Mg ii emission, with both a narrow-line component and a broad-line component. The width and velocity shift of the narrow-line components of Mg ii and N iii] 1750 Å were tied, allowing only the strength to differ. We also corrected the spectrum for foreground reddening, assuming E(B − V) = 0.0367 mag determined from the dust maps of Schlegel et al. (1998) and the extinction curve of O'Donnell (1994). As can be seen in the figure, the Mg ii λ2800 emission line clearly shows a broad component. We measured the FWHM = 10028 ± 627 km s⁻¹ and the line dispersion σ = 4000 ± 790 km s⁻¹ for the broad component from the emission-line models.

Zoom In Zoom Out Reset image size

4.1. Soft Excess and the Fe Lines

4.2. The ${Mg}\,{II}$ Emission Line

D14 presented optical, UV, and X-ray observations of the classical Seyfert 1 Mrk 590 that span the past 40+ years. This interesting object brightened by a factor of a few tens between the 1970s and 1990s and then faded by a factor of 100 or more at all continuum wavelengths between the mid-1990s and 2013. Notably, there is no evidence in the current data set that this recent, significant decline in flux is due to obscuration; in particular, the most recent X-ray observations are consistent with zero intrinsic absorption. There were similarly dramatic changes in the emission-line fluxes, the most striking of which is the complete disappearance of the broad component of the Hβ emission line, which had previously been strong (equivalent widths ∼20−60 Å).

In this paper we have presented new Chandra  and HST  observations from 2014 and we find that Mrk 590 is awakening. While the source continues to be in a low state, its soft excess has re-emerged, though not to the historical level. The broad Mg ii emission line is also present. These observations suggest that the soft excess is connected to the UV continuum through warm Comptonization and that the ionizing continuum is also on the rise, but it does not follow the relation between soft excess and UV continuum reported in Mehdipour et al. (2011).

The implications from this long time series of Mrk 590 observations are that (1) Mrk 590 is a direct challenge to the historical paradigm that AGN type is exclusively a geometrical effect, and (2) there may not be a strict, one-way evolution from Type 1 to Type 1.5−1.9 to Type 2 as recently suggested by Elitzur et al. (2014). Instead, for at least some objects, the presence of BLR emission may coincide with episodic accretion events throughout a single active phase of an AGN. If true, such behavior may be more prominent in Seyfert galaxies, where accretion is likely to be a consequence of secular processes (e.g., Mathur et al. 2012; Martin et al. 2018) and therefore likely more episodic than quasar activity, which may be governed more predominantly by major mergers.

Support for this work was provided by the National Aeronautics and Space Administration through Chandra Award Number G04-15114X to S.M. issued by the Chandra X-ray Observatory Center, which is operated by the Smithsonian Astrophysical Observatory for and on behalf of the National Aeronautics Space Administration under contract NAS8-03060. Support for the HST  program number GO-13185 was provided by NASA through a grant from the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS5-26555. K.D.D. was supported by an NSF AAPF fellowship awarded under NSF grant AST-1302093 while she worked on this project. M.V. gratefully acknowledges support from the Danish Council for Independent Research via grant number DFF 4002-00275.