Evidence for a clumpy disc-wind in the star forming Seyfert\,2 galaxy MCG--03--58--007
G. A. Matzeu, V. Braito, J. N. Reeves, P. Severgnini, L. Ballo, A. Caccianiga, S. Campana, C. Cicone, R. Della Ceca, M. L. Parker, M. Santos-Lleó, N. Schartel
MMNRAS , 1– ?? (?) Preprint 5 December 2018 Compiled using MNRAS L A TEX style file v3.0
Evidence for a clumpy disc-wind in the star forming Seyfert 2 galaxyMCG–03–58–007
G. A. Matzeu , (cid:63) , V. Braito , , J. N. Reeves , P. Severgnini , L. Ballo , A. Caccianiga ,S. Campana , C. Cicone , R. Della Ceca , M. L. Parker , M. Santos-Lleó , and N. Schartel European Space Agency (ESA), European Space Astronomy Centre (ESAC), E-28691 Villanueva de la Cañada, Madrid, Spain INAF – Osservatorio Astronomico di Brera, Via Bianchi 46, I-23807 Merate (LC), Italy Center for Space Science and Technology, University of Maryland Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, USA INAF – Osservatorio Astronomico di Brera, Via Brera 28, I-20121 Milano, Italy
ABSTRACT
We report the results of a detailed analysis of a deep simultaneous
130 ks
XMM-Newton & NuSTAR observation of the nearby ( z = 0 . ) and bright ( L bol ∼ × erg s − )starburst-AGN Seyfert 2 system: MCG–03–58–007. From the broadband fitting we show thatmost of the obscuration needs to be modeled with a toroidal type reprocessor such as MYTorus (Murphy & Yaqoob 2009). Nonetheless the signature of a powerful disc-wind is still apparentat higher energies and the observed rapid short-term X-ray spectral variability is more likelycaused by a variable zone of highly ionized fast wind rather than by a neutral clumpy medium.We also detect X-ray emission from larger scale gas as seen from the presence of several softnarrow emission lines in the RGS, originating from a contribution of a weak star formingactivity together with a dominant photoionized component from the AGN.
Key words:
Subject headings: Black hole physics – galaxies: active – galaxies: nuclei –Seyfert 2: individual (MCG–03–58–007) – X-rays: galaxies
It is widely accepted that the central engine of active galactic nu-clei (AGN) is powered by accretion of matter onto a super-massiveblack hole (SMBH). They are also thought to harbor circumnuclearmaterial in a toroidal structure (i.e., dusty torus) that absorbs andreprocesses the high-energy radiation emitted from these centralregions. Although the dusty torus is predicted to be ubiquitous inAGN by the unification scheme (Antonucci 1993), the exact geom-etry, location and composition are poorly understood. Nonetheless,important advances have been made in recent years through theobserved long-term variability (i.e., months–years) of the X-ray ab-sorber’s column density ( N H ). This ruled out the standard view ofan homogeneous ‘doughnut’ shaped absorber in favour of a more‘clumpy’ torus (e.g., Risaliti, Elvis & Nicastro 2002; Markowitz,Krumpe & Nikutta 2014) characterized by a distribution of a largenumber of individual clouds.Since the first detection of resonance iron K shell absorptionlines blueshifted to rest-frame energies of E > in luminousAGN (Chartas, Brandt & Gallagher 2003; Reeves, O’Brien & Ward2003; Pounds et al. 2003), high velocity outflows have becomean essential component in the overall understanding of AGN. The (cid:63)
Correspondence to: [email protected] faster components of these winds are thought to occur as a resultof the accretion process and originates in the inner regions near theblack hole whereas the warm absorbers at lower velocities probablyoriginate from much further out e.g., >
10 pc and greater (e.g.,Blustin et al. 2005; Reeves et al. 2013). Fast disc-winds are presentin ∼ of the bright AGN (Tombesi et al. 2010; Gofford et al.2013), suggesting that their geometry is characterized by a wideopening angle, as confirmed in the luminous quasar PDS 456 byNardini et al. (2015). These winds are considered the key ingredientof AGN feedback models and might represent the missing link in theobserved galactic feedback process, by driving massive molecularoutflows out to large ( ∼ kpc) scales in galaxies (Cicone et al. 2014;Tombesi et al. 2015; Feruglio et al. 2015; Cicone et al. 2018).Disc-winds are characterized by a high column density ( N H ∼ cm − ) and a mean velocity (cid:104) v w (cid:105) ∼ . c (Tombesi et al. 2010).These high velocities can result in a large amount of mechanicalpower, possibly exceeding the . – of the bolometric luminosity( L bol ), suggesting a significant AGN feedback contribution to theevolution of the host galaxy (King 2003; King & Pounds 2003;Di Matteo, Springel & Hernquist 2005; Hopkins & Elvis 2010;Tombesi et al. 2015). However, the velocity of the Fe K absorbers canspan over a wide range, from as low as a few × − km s − (more typical of what is seen in the soft X-ray warm absorbers; e.g.,Kaastra et al. 2000; Blustin et al. 2005; Reeves et al. 2013) up to © ? The Authors a r X i v : . [ a s t r o - ph . H E ] D ec Matzeu et al. mildly relativistic values of ∼ . − . c in the most extreme cases(e.g., Chartas et al. 2002; Reeves et al. 2009; Parker et al. 2017,2018. A recent simultaneous XMM-Newton & NuSTAR observationof the luminous quasar PDS 456 carried out in March 2017 revealeda new relativistic component of the fast wind while observed in thelow-flux state (Reeves et al. 2018), which provided the evidenceof the multi-phase structure disc-wind. In recent studies a positivecorrelation was also found between the outflow velocity of the disc-wind and the X-ray luminosity in PDS 456 (Matzeu et al. 2017) andIRAS 13224–3809 (Pinto et al. 2018), which is what is expected ina radiatively driven wind scenario.MCG–03–58–007 is a nearby ( z = 0 . ) bright Seyfert 2galaxy. It was first selected as a Compton-thick AGN (i.e., N H > cm − ) candidate due to its faint X-ray flux (Severgnini, Cac-cianiga & Della Ceca 2012) as measured in a
10 ks observationwith
XMM-Newton in 2005. A longer
80 ks
Suzaku observation in2010 revealed an obscured AGN but in the Compton-thin (i.e., N H ∼ cm − ) regime with two deep blue-shifted absorp-tion troughs between . – . (Braito et al. 2018, B18 here-after). These are associated with two zones of a highly ionized( log( ξ/ erg cm s − ) ∼ . ) disc-wind with column density of log( N H / cm − ) ∼ . and log( N H / cm − ) ∼ . , and out-flowing at a mildly relativistic speed of v w ∼ . – . c . A morerecent
130 ks simultaneous
XMM-Newton & NuSTAR observationconfirmed the presence of the wind and showed a rapid N H varia-tion on a timescale of ∼ (B18). Such short-term variabilitysuggests the presence of inhomogeneous material as part of thedisc-wind, rapidly crossing the line-of-sight. The work presentedin B18 was focused on the analysis of the hard X-ray band andon the detection of the blueshifted iron K absorption features inMCG–03–58–007.The main motivation of this follow up paper is to undertake afull broadband analysis of MCG–03–58–007 with the simultaneous XMM-Newton & NuSTAR data by parameterizing the observationwith a more physically motivated model for the neutral absorbersuch as
MYTorus (Murphy & Yaqoob 2009). With this compre-hensive broadband analysis we can: (a) investigate whether thespectral variability, as seen through the hardening of the spectra,is to a change in the clumpy neutral absorber or to an inhomo-geneous highly ionized disc-wind; (b) to investigate in detail, forthe first time in this source, the physical properties of the soft X-ray emission. This paper is organized as follows: in Section 2, wesummarize the data reduction, whereas in Section 3 we describethe spectral analysis that focuses on the RGS spectra. In Section 4we focus on the broadband spectral analysis of the simultaneous
XMM-Newton & NuSTAR observation of MCG–03–58–007 whichincludes a time-dependent spectral analysis. In Section 5 we discussthe possible origin of the observed soft X-ray emission lines andwe discuss whether the obscuration event is caused by a transitingneutral absorber or by an inhomogeneous highly ionized disc-wind.In Section 5 we also discuss the energetic and location of the highlyionized absorber and compare it with what was found in B18.In this paper the values of H = 70 km s − Mpc − and Ω Λ =0 . are assumed throughout and errors are quoted at the confidence level ( ∆ χ = 2 . and ∆ C = 2 . ) for one parameterof interest unless otherwise stated. XMM-Newton observed MCG–03–58–007 simultaneously with
NuSTAR between the 6th–9th of December 2015 with an exposure
Figure 1.
Time-averaged background-subtracted source and the time-averaged background spectra of the simultaneous
XMM-Newton & NuSTAR observation of MCG–03–58–007. The net spectra shown here are the pn(black), MOS 1 +
50 keV . This data have been binned to σ for clarity. time of ∼
130 ks (see Table 1). The
NuSTAR observation startedapproximately two hours before the
XMM-Newton one and ended
150 ks afterwards (see Table 1), with a total duration of . .All the XMM-Newton data were reduced by adopting the ScienceAnalysis System (sas v16.0.0). The most recent set of calibrationfiles and the EPIC-pn and MOS 1 + MOS 2 adopted in this workare the same used in B18. The same applies for the
NuSTAR data,reduced according to the
NuSTAR
Data Analysis Software pack-age (nustardas v1.6.0). The
XMM-Newton
RGS data have beenreduced by using the standard sas task rgsproc , where the highbackground time intervals were filtered out by applying a thresholdof . − on the background event files. After checking that theRGS 1 and RGS 2 spectra were in good agreement, to within the level, we combined them by using the sas task rgscombine . The twocombined RGS collected a total of net counts. Due to the lownumber of net counts, the spectra were binned to ∆ λ = 0 . whichunder-samples the RGS FWHM spectra resolution of ∆ λ = 0 . – . . At this binning, the spectra are characterized by < countsper channel and hence we adopted C-statistics (Cash 1979) in theRGS analysis, whereas we adopt χ minimization technique forthe EPIC and FPMA/B spectral analysis. All the models adoptedwere fitted to the data by using the standard software package xspecv12.9.1p (Arnaud 1996).In Fig. 1 we show the time-averaged pn, MOS 1 + σ for clarity) between the . –
70 keV rest energy band.In the soft band the EPIC data was filtered for high backgroundwhich severely affected the observation (as reported in B18) whilein the hard X-ray band the background dominates above
50 keV .After screening, the background level is only ∼ – of the netsource rate for the pn and the combined MOS. In the broadbandfitting we will confront different physical reprocessing scenariosincluding the MYTorus model (Murphy & Yaqoob 2009), which isvalid at
E > . and hence for the rest of the broadband anal-ysis we concentrate on the . –
50 keV band. A Galactic absorptioncolumn of the N H = 2 . × cm − (Dickey & Lockman 1990),as from the H i
21 cm measurements, was adopted in all the fits
MNRAS , 1– ????
MNRAS , 1– ???? (?) lumpy disc-wind in MCG–03–58–007 XMM-Newton NuSTAR
Obs. ID
Instrument PN MOS 1 + MOS 2 RGS 1 + RGS 2 FPMA FPMBStart Date, Time (UT) 2015-12-06, 12:56:06 2015-12-06, 12:33:16 2015-12-06, 13:33:07 2015-12-06, 10:36:08 2015-12-06, 10:36:08End Date, Time (UT) 2015-12-08, 01:25:11 2015-12-08, 01:29:12 2015-12-08, 01:33:25 2015-12-09, 17:21:08 2015-12-09, 17:21:08Duration(ks) 131.3 258.0 266.4 281.8 282.8Exposure(ks) a b (0 . − b (2 − b (3 − – – – 4.93 5.12 Table 1.
Summary of the 2015 simultaneous
XMM-Newton & NuSTAR observation of MCG–03–58–007. a Net Exposure time, after background screening and dead-time correction. b Observed fluxes in the 0.5–2 keV, 2–10 keV and 3–50 keV bands in units × − erg cm − s − . Figure 2.
Data/model ratio for the pn (black) and the combined MOS (red) spectra between between . – . . The adopted model is a simple power-lawand Galactic absorption. We note that although MCG–03–58–007 is faint and absorbed in the soft X-ray band, its emission is largely line-dominated. and we also assume solar abundance of the main abundant elementsthroughout the analysis. As already shown in B18, both the
Suzaku and
XMM-Newton soft(
E < ) spectra are rich in emission lines. This is also evidentin Fig. 2 where we show the . – . residuals for the XMM-Newton
EPIC data (black and red lines are the pn and MOS spectra,respectively). The adopted model is a simple power-law and Galac-tic absorption. Since this line-dominated soft spectral range couldreveal information on the circum-nuclear gas surrounding the AGN,we analyzed the high resolution spectrum collected with the RGS.In particular, we considered the RGS spectrum between . – ( . – .
63 ˚A ) in order to build a template for the soft X-ray emis-sion which we later apply to the broadband and the time-dependentspectral analysis in Section 4.
As a first step, the RGS spectrum was fitted with a simple power-law( zpowerlaw ) and a Galactic absorption model (
Tbabs with cross-section and ISM abundances of Wilms, Allen & McCray 2000).This simple continuum model returned a photon-index of Γ ∼ . which is broadly consistent with that found in the broadband anal-ysis in Section 4. Since the residuals in the spectrum suggests thepresence of multiple unresolved narrow emission lines, we mod-elled them with Gaussian components ( zgauss ) in xspec, with theline-width fixed at σ = 1 eV . We note that, the inclusion of aline requires to provide an improvement of ∆ C ≥ . for freeparameters (i.e., line energy and the normalization) correspond-ing to a ≥ confidence level. Following this criterion we for-mally detected eleven emission lines, ten of which have a secureidentification mostly associated, in their rest frame, with He- andH-like of abundant elements such as N, O and Ne. There is an addi-tional line at +1 − eV (or . ) which does not correspond to aknown transition. Two additional emission lines where marginallydetected ( ∆ C/ ∆ ν < . / ) at the expected rest-frame energies of E = 574 +1 − eV (21 . and E = 1884 +27 − eV (6 . whichcan be associated to the resonance transition of O vii and Si xiiirespectively.The overall fit statistic considerably improved upon the addi-tion of the Gaussian line profiles decreasing from C/ν = 407 / ,without any line emissions (null hypothesis probability . × − )to a more statistically acceptable C/ν = 298 / (null hypothesisprobability . × − ) with the included lines. Fig. 3 shows the restframe – RGS spectrum with the best-fitting model (red lines)overlaid on. The emission line fluxes, equivalent widths ( EW ) andidentifications are listed in Table 2. The centroid energy of the O viiline at the rest energy of E = 562 ± ( . ) suggests an MNRAS , 1– ?? (?) Matzeu et al. − × − × − × − no r m a li z ed c oun t s s − Å − c m − Rest Wavelength (Å) | S i X III H e α | N e X L y β | N e I X ( f ) Rest Energy (keV)
16 18 20 22 × − × − × − × − c oun t s s − Å − c m − Rest Wavelength (Å) | F e XV II | F e XV II | O V III L y α | O V II H e α (r) | O V II H e α ( i ) | O V II H e α ( f ) |?
24 26 28 30 × − × − × − × − c oun t s s − Å − c m − Rest Wavelength (Å) | N V II L y α | N V I H e β | N V I H e α ( f ) Figure 3.
Enlarged view of the combined RGS data of MCG–03–58–007 showing the count rate spectrum normalized to the instrumental effective area. Thebest-fit model is shown by the red line. Despite the low number of counts, it was possible to detect eleven emission lines. Ten out of eleven lines were formallyidentified with the most likely He- H-like transitions from N, O and Ne. From the plot it is evident that the forbidden transitions are dominating the spectrum overthe resonance and intercombination, suggesting the low-density nature of the photoionized distant gas. The emission line significantly detected at rest energy of E ∼
620 eV ( ∼
20 ˚A) could not be identified. Two additional emission lines were marginally detected at the rest-frame energies of E = 574 +1 − eV (21 . and E = 1884 +27 − eV (6 . , which are identified to the resonance transitions of O vii and Si xiii respectively. identification with the forbidden transition, detected at σ confi-dence level. The intercombination transition is also detected at theexpected energy of E = 569 ± . but with a lowersignificance at the ∼ σ level.As for the other He-like complexes such as the N vi and Ne ix triplets, we only observed the forbidden component atthe expected energy of E = 419 ± . and E =906 +2 − eV (13 . . Although the RGS spectrum has a low-numberof counts, we can observe that the soft X-ray emission lines are MNRAS , 1– ????
20 ˚A) could not be identified. Two additional emission lines were marginally detected at the rest-frame energies of E = 574 +1 − eV (21 . and E = 1884 +27 − eV (6 . , which are identified to the resonance transitions of O vii and Si xiii respectively. identification with the forbidden transition, detected at σ confi-dence level. The intercombination transition is also detected at theexpected energy of E = 569 ± . but with a lowersignificance at the ∼ σ level.As for the other He-like complexes such as the N vi and Ne ix triplets, we only observed the forbidden component atthe expected energy of E = 419 ± . and E =906 +2 − eV (13 . . Although the RGS spectrum has a low-numberof counts, we can observe that the soft X-ray emission lines are MNRAS , 1– ???? (?) lumpy disc-wind in MCG–03–58–007 E obs (1) Flux (2) EW (3) Transition (4) E lab (5) ∆ C/ ∆ ν (6) +1 − [29 .
59] 14 . +9 . − . +83 − N vi He α (f)
420 7 . / +1 − [25 .
00] 7 . +4 . − . +19 − N vi He β (f)
495 11 . / +1 − [24 .
65] 7 . +3 . − . +17 − N vii Ly α . . / +1 − [22 .
06] 18 . +10 . − . +43 − O vii He α (f)
561 11 . / +2 − [21 .
79] 5 . +4 . − . +9 − O vii He α (i) . . / +1 − [20 .
10] 12 . +6 . − . +124 − ? ? . / +2 − [18 .
99] 6 . +2 . − . +78 − O viii Ly α . . / +1 − [17 .
01] 4 . +2 . − . +69 − Fe xvii 3s → . . / +2 − [14 .
98] 3 . +1 . − . +87 − Fe xvii 3d → . . / +2 − [13 .
69] 3 . +1 . − . +108 − Ne ix He α (f) . . / +5 − [10 .
17] 4 . +2 . − . +260 − Ne x Ly β . / Table 2.
Summary of the emission lines in the RGS spectra from the best-fit Gaussian model. Note that all the lines are practically unresolved and hence weassume a σ = 1 eV during fitting.(1) Measured line energy in the quasar rest frame, in units of eV where the corresponding mean wavelength value in units of ˚A is given within the brackets,(2) line photon flux, in units of − cm − s − ,(3) the equivalent width in the AGN rest frame in units of eV,(4) list of the most plausible identification where (f), (i), and (r) refer to forbidden, intercombination, and resonance transitions for He-like species,(5) corresponding laboratory energy for the detected lines, in units of eV,(6) improvement of the fit in C–statistics upon adding the line to the model. largely dominated by forbidden transitions rather than resonanceand/or intercombination emission (see Fig. 3).This result suggests that such emission line mainly origi-nates from a distant low-density photoionized plasma as seen inmany Seyfert 2s (e.g., Sako et al. 2000; Kinkhabwala et al. 2002;Kallman et al. 2014). In particular, a strong emission line wellmatch with the O viii Ly α emission at the expected rest frameenergy of E = 653 +2 − eV (18 . . The line width estimate( σ < . ) corresponds to an upper limit on the velocity of σ v (cid:46) − (FWHM (cid:46) − ) and hence broadlyconsistent with emitting gas possibly located in the BLR/NLR.A possible diagnostic for the density of the emitting gas canbe carried out by quantifying the ratio between the intensity of theforbidden and intercombination components detected in He-liketriplets (Porquet & Dubau 2000). However in this work, the onlypossible diagnostic can be carried out on the O vii He α lines asthis is the only triplet detected. The resulting emission line ratio is R = z/ ( x + y ) = 3 . +1 . − . (where z and x + y are the forbidden andintercombination transition respectively). This ratio places an upperlimit on the electron density of the order of n e < × cm − (seeFig.8 Porquet & Dubau 2000). Nevertheless, the spectral resolutionin the RGS does not allow to put accurate constraints on the widthsand location of these lines. The above estimates suggests that thelines could be, in principle, broad and emitted from a gas with adensity lower than × cm − and hence located at R gas (cid:38) BLR.Furthermore, all the detected soft emission lines are consistentto their expected rest frame energies, except N vii Ly α line whichlies slightly above it (see Table 2). The possible emission line de-tected at E = 729 +1 − eV (17 . , has an uncertain identificationbut it could be associated with an L-shell transition from iron (e.g., Fe xvii 3s → E = 828 +2 − eV (14 . ,which lies very close to the expected rest-frame energy of Fe xvii3d →
2p transition. The iron L transitions are associated with col-lisionally ionized plasma, likely within the star forming regions inthe host galaxy.Having reached a good parameterization of the narrow emis-sions with the Gaussian profiles, we subsequently replaced the Gaus-sian emission line fit with a publicly available xstar photoionizationemission grid (Bautista & Kallman 2001). Despite the low-numberof counts in the RGS, the inclusion of the xstar grid can providea first order estimate on the physical condition of the gas such asionization, column and electron density. As the emission profilesappear to be narrow, we chose a suitable grid with velocity broad-ening of v turb = 100 km s − . Moreover, we find that in order tomodel it with xstar, it is preferable to adopt a larger binning (i.e., ∆ λ = 0 . ) to the spectrum.From the photoionization modelling, the xstar emission flux(or normalization) is expressed as: κ xstar = f cov L D (1)where L is the ionizing luminosity ( L ion ) in units of erg s − , D kpc is the luminosity distance of the ionizing source (AGN) inkiloparsecs, and f cov = Ω / π is the covering fraction of the gas.Thus by keeping the N H fixed to the default value given in the xstar ftp://legacy.gsfc.nasa.gov/software/plasma_codes/xstar/xspectables/MNRAS , 1– ?? (?) Matzeu et al. table of log( N H / cm − ) = 21 . , which is more typical of a dif-fused warm emission component found in Seyfert 1 and Seyfert 2sgalaxies (e.g., Kinkhabwala et al. 2002; Blustin et al. 2005), we geta normalization of κ xstar = 1 . +0 . − . × − . By assuming a bolo-metric luminosity in MCG–03–58–007 of L bol ∼ × erg s − (B18) and that L ion ∼ L bol / ∼ erg s − , at a luminositydistance of D L = 138 . , we obtain a corresponding gas cov-ering fraction of the order of ∼ . . This suggests that the distantphotoionized gas is covering a small fraction of the sky and possi-bly inhomogeneous, however this result is strongly dependent on agiven column density. If we assume a covering factor, whichis broadly expected in the NLR (e.g., Netzer & Laor 1993), it wouldrequire a normalization of κ xstar = 5 . × − corresponding to acolumn density of log( N H / cm − ) = 19 . +0 . − . .The ionization state of the plasma was measured to be log( ξ/ erg cm s − ) = 1 . +0 . − . and the addition of the photoion-ized gas emission component results in a considerable improvementto the overall fit to ∆ C/ ∆ ν = 34 . / which is at > . con-fidence level. This is somewhat expected considering that the softX-ray emission is dominated by forbidden transitions. The inclusionof a thermal component via the xspec ( mekal ; Mewe, Gronenschild& van den Oord 1985) model, led the a modest improvement in thefit statistic of ∆ C/ ∆ ν = 8 . / . σ ) . Here we find a poorlyconstrained temperature of kT = 0 . +0 . − . keV and a normal-ization of . +0 . − . × − − π [ D A (1+ z )] (cid:82) n e n H dV , where n e and n H are the electron and hydrogen densities (measured in cm − )respectively and D A is the angular diameter of the source in cm .Lastly, we kept the photon-index of the soft power-law componentfixed at the best value, found in B18 and later in the broadbandanalysis of Γ = 2 . . Note that with such a complex model, thenormalization of the scattered power-law component is no longerwell constrained. In Fig. 4 we show the best-fit model (red) to thedata with the corresponding contribution of both thermal (green)and photoionized (blue) components. The latter clearly dominateacross the RGS spectrum. XMM-Newton & NuSTAR
ANALYSIS4.1 Description of the
XMM-Newton & NuSTAR lightcurves
In Fig. 5 we show the
XMM-Newton (blue) and
NuSTAR (red)lightcurves in different energy bands, as well as the hardness ratiobetween the –
10 keV and –
40 keV bands observed by
NuSTAR .The lightcurve in the . – band shows no variability whilethe flux in the – band in particular drops strongly at around125 ks, from ∼
11 cts s − to ∼ .
04 cts s − , after the start of the NuSTAR observations. In comparison the –
40 keV hard X-rayband shows little variation throughout the observations. Indeed theratio between the –
40 keV and – bands shows a clear in-crease in hardness at this point, which B18 earlier interpreted as anobscuration event which occurs within an approximate timescale ofone day. Then at the end of the observation at 250 ks, the sourceflux gradually starts to recover in the remaining 30 ks of the NuSTAR observation.Accordingly to this behavior and following B18, the observa-tions were split into three time intervals; from –
125 ks (slice A), –
250 ks (slice B) and from –
280 ks (slice C), as marked byvertical dotted lines in Fig. 5. Slice A has a high count rate and islargely coincident with the whole of the
XMM-Newton observation.On the other hand, in Slice B captures the increase in hardness ofthe source, where in particular the – band count rate drops compared to the –
40 keV lightcurve. Note that slice B includesless than
10 ks of XMM-Newton data, which is only included to con-strain the soft band spectrum, below . In slice C the hardnessratio drops again whereas the source brightness appears to recover.Slice C only consists of
NuSTAR data between –
280 ks and thespectral properties are intermediate between slice A and B. As thequality of spectrum for Slice C is low (with only ∼
30 ks of netexposure for both
NuSTAR
FPM detectors combined), we do notinclude this in our subsequent quantitative analysis and thereafterconcentrate on Slice A and Slice B. Nonetheless, the overall vari-ability provides important physical constraints, which as suggestedby B18, indicates that the AGN went through an absorption eventin order to account for the rapid change in hardness ratio.
For the initial broadband analysis we consider slice A rather than thetime-averaged spectra as the former includes the only time-intervalwhere both telescopes are effectively observing simultaneously (seeFig. 5). Since we obtained a reasonable fit with xstar and mekal in the RGS (see 3.1), we adopt the same best-fit values in theEPIC broadband spectra. This was achieved by fixing the ionizationparameter and by letting the corresponding normalizations readjustaccordingly including the temperature of the mekal component.In this work we adopt two main models. Model A representsa more classical model where the primary continuum is modelledas power-law component transmitted through a neutral absorber.The Compton reflected component is produced by a distant neutralreflector which is geometrically approximated as a slab of neutralmaterial with pexmon (Nandra et al. 2007), whilst a distant scatteredcontinuum component is also included throughout. Model B as-sumes a geometrically toroidal reproccesor modelled with MYTorus (Murphy & Yaqoob 2009), assuming a standard (coupled) config-uration, which takes into account the physical properties of theabsorbing medium, the Compton-down scattering effect and it in-cludes self-calculated reflected components (continuum plus Fe Kemission lines). This model also assumes a fixed geometry of thetoroidal X-ray reprocessor, a single value for the covering factorof the torus (corresponding to a half–opening angle of 60 ◦ ) and auniform composition of the torus itself. Model B will be describedin Section 4.2.1. In the following we adopt model A , defined as: Model A = Tbabs × [ zpowerlw scatt + xstar em + mekal + pexmon + ( xstar FeK , × xstar FeK , ) × zpowerlw intr × zphabs ] , where Tbabs accounts for the Galactic absorption, the scat-tered power-law component is parameterized with zpowerlw scatt . zpowerlw intr models the primary continuum which is absorbed byfully covering neutral material ( zphabs ) with a column density of log( N H / cm − ) = 23 . ± . . The Compton-reflection compo-nent is modelled with pexmon (Nandra et al. 2007) which includesthe power-law continuum reflected from distant neutral material andthe emission from Fe K α , Fe K β , Ni K α and Fe K Compton shoul-der. The photon-indexes of the reflected and primary continuumcomponents are assumed to be the same, the inclination angle wasfixed as ◦ ( ◦ corresponds to face-on), and the cutoff energy was
300 keV . The normalization was allowed to vary and the scalingreflection factor parameter was fixed at R = Ω / π = 1 .B18 detected in the XMM-Newton & NuSTAR spectra twostrong absorption features at rest frame energies of . ± . MNRAS , 1– ????
300 keV . The normalization was allowed to vary and the scalingreflection factor parameter was fixed at R = Ω / π = 1 .B18 detected in the XMM-Newton & NuSTAR spectra twostrong absorption features at rest frame energies of . ± . MNRAS , 1– ???? (?) lumpy disc-wind in MCG–03–58–007
10 15 20 25 30 × − − ν F ν ( e r g s c m − s − ) Wavelength (Å) | F e XV II | O V III L y α | O V II H e α | N V I H e β ( f ) | Thermal emission | Photoionized emission | F e XV II | Total model | N V I H e α ( f ) Figure 4.
Best-fit model (red) overlaid on the RGS spectrum with group ∆ λ = 0 . for extra clarity. The corresponding contribution of both thermal (green)and photoionized (blue) components are shown, where the former is responsible for the collisionally ionized Fe L emission between – and the lattermainly responsible to all the forbidden transition emission e.g., Nvi and Ovii He α . The RGS spectrum is clearly dominated by emission from photoionizedelements from distant gas possibly located in the NLR. and . ± . associated to blueshifted s → p transitionof Fe xxv and Fe xxvi. These absorption profiles likely correspondto two zones of highly ionized outflowing absorbers with velocitiesof v w ∼ − . c and v w ∼ − . c . Since the parameterization withthe Gaussian absorption profiles has been already explored in B18,here we simply model these features with two multiplicative gridsof photoionized absorbers generated with xstar photoionizationcode (v2.21bn13, Kallman et al. 2004), defined as xstar FeK , and xstar FeK , respectively. Since these lines are broad, we adopt forall the subsequent fits the original grid used for PDS 456 (Nardiniet al. 2015; Matzeu et al. 2016) with a high velocity broadeningof v turb = 10000 km s − . In generating this grid, it was also as-sumed an intrinsically steep ionizing continuum of Γ = 2 . , whichis consistent with what found in MCG–03–58–007. We find thatthe addition of two absorption zones with outflow velocities of v w /c = − . ± . and v w /c = − . ± . and columndensity of log( N H / cm − ) = 23 . ± . and log( N H / cm − ) =24 . +0 . − . improved the fit by ∆ χ / ∆ ν = 28 . / ( . σ ) and ∆ χ / ∆ ν = 19 . / ( . σ ) for zone 1 and zone 2 respectively.The slower zone prefers a lower ionization ( log( ξ/ erg cm s − ) =5 . ± . ) than the faster zone ( log( ξ/ erg cm s − ) = 6 . +0 . − . ).For the latter zone, the ionization and column density of the ion-ized absorber are highly degenerate which explains the large er-ror. Here we note that the XMM-Newton & NuSTAR spectra aredominated by the transmitted primary absorbed component above ∼ . The intensities of the reflected and primary component are norm pexmon = 1 . ± . × − ph cm − s − keV − and norm zpow = 2 . +0 . − . × − ph cm − s − keV − respectivelywhich corresponds to a reflection fraction R = 0 . ± . . Overallmodel A provided a good fit to slice A of χ /ν = 714 / . Thedetails of the best-fit model are listed in Table 3.In terms of the soft excess, the soft photoionized emis-sion component is modelled with the same xstar emission grid( xstar em ) and the corresponding N H and ξ values used in theRGS. Its normalization is readjusted to a slightly higher mea-surement (but consistent within the errors found in the RGS) of κ xstar = 1 . +0 . − . × − . If we let the ionization parameter to befree to vary, the resultant fit would lead to a considerable decreaseof ξ from log( ξ/ erg cm s − ) ∼ . to log( ξ/ erg cm s − ) ∼ . .However if we assume this value in the RGS model it would wouldpractically account only for the steep continuum and not modelthan the individual emission lines. The other contribution to thesoft excess is associated with starburst emission which is modelledwith a thermal component emission ( mekal ) with a temperature of kT = 0 . +0 . − . keV . We find that after including both the thermaland the photoionized components, the photo-index of the scatteredcontinuum is still steep at Γ = 3 . +0 . − . . Note that such steep valuecannot be decreased even by adding a second mekal component.This might indicate that the scattered soft X-ray power-law is steeperthan what observed in the hard X-rays. This could be caused by thepresence of an intrinsic soft excess as seen in many type 1 AGN (e.g., MNRAS , 1– ?? (?) Matzeu et al.
Figure 5.
Lightcurve in different energy bands (i.e., . – – –
10 keV and –
40 keV ) corresponding to EPIC-pn (blue) and FPMA + B (red)both binned at 5814 seconds (one
NuSTAR orbit) respectively. The hardness ratio between –
40 keV and –
10 keV corresponding to the FPMA + B isplotted on the bottom panel. MCG–03–58–007 was observed simultaneously with
XMM-Newton & NuSTAR , the duration of the latter extended
150 ks beyond
XMM-Newton to . . According to the behavior of the lightcurves, the XMM-Newton and
NuSTAR spectra were sliced in three segments named slice A–Cwhich are visually marked on the plot by the vertical black dotted lines. Thus slice A and slice B are characterized by the
XMM-Newton (pn and MOS 1 +
2) and
NuSTAR (FPMA + B) spectra separated at –
125 ks and –
250 ks respectively. However, in slice B we are only left with less than
10 ks of XMM-Newton data which implies that the
XMM-Newton contribution will be only from a short EPIC-pn spectrum. Furthermore, slice C is entirely consisting of
NuSTAR databetween –
280 ks . Singh, Garmire & Nousek 1985; Turner & Pounds 1988; Nardiniet al. 2011).
In the above fitting, the pexmon model assumes geometrically asimple slab reflector. In the following we replace both the pexmon and the simple neutral absorber ( zphabs ) with the
MYTorus model.It is essentially composed by three tables, developed for xspec, ofreprocessed spectra (for more details see Murphy & Yaqoob 2009)assuming a primary power-law input spectrum that interacts witha reprocessor with toroidal geometry. We therefore analyze slice Awith model B which is constructed mathematically as follows: Model B = Tbabs × [ zpowerlw scatt + xstar em + mekal + A S × MYTorusS + MYTorusZ × ( xstar FeK , × xstar FeK , ) × zpowerlw intr + A L × gsmooth × MYTorusL ] , where MYTorusS and
MYTorusL are publicly available emissiongrids that reproduce respectively to the reflected continuum and theFe K α , Fe K β emission line spectrum. MYTorusZ is a multiplicativetable corresponding to the zeroth-order transmitted continuum, con-taining pre-calculated transmission factors that affect the incidentcontinuum due to photoelectric absorption. The gsmooth compo-nent is a convolution model, available on xspec, which takes intoaccount the broadening of the Fe K emission lines. The soft ex-cess and the iron K absorption features have been modelled as inmodel A returning consistent values as listed in Table 3.Here we constructed MYTorus assuming the standard ‘cou-pled’ configuration (Yaqoob 2012) which assumes that the angle atwhich the line-of-sight directly intercepts the torus is the same (i.e.,coupled) as the scattered one. Here the relative reflected and fluores-cence line emitting normalizations are set to be equal to the primarypower-law normalization i.e., norm S = norm L = norm intr zpowl ,while all the constant factors are set to unity i.e., A S = A L = 1 . Theline-of-sight inclination angle in respect to the polar axis was fixedat its best-fit value of θ obs = 70 . ◦ , consistent with the type 2 classi- MNRAS , 1– ????
MYTorusL are publicly available emissiongrids that reproduce respectively to the reflected continuum and theFe K α , Fe K β emission line spectrum. MYTorusZ is a multiplicativetable corresponding to the zeroth-order transmitted continuum, con-taining pre-calculated transmission factors that affect the incidentcontinuum due to photoelectric absorption. The gsmooth compo-nent is a convolution model, available on xspec, which takes intoaccount the broadening of the Fe K emission lines. The soft ex-cess and the iron K absorption features have been modelled as inmodel A returning consistent values as listed in Table 3.Here we constructed MYTorus assuming the standard ‘cou-pled’ configuration (Yaqoob 2012) which assumes that the angle atwhich the line-of-sight directly intercepts the torus is the same (i.e.,coupled) as the scattered one. Here the relative reflected and fluores-cence line emitting normalizations are set to be equal to the primarypower-law normalization i.e., norm S = norm L = norm intr zpowl ,while all the constant factors are set to unity i.e., A S = A L = 1 . Theline-of-sight inclination angle in respect to the polar axis was fixedat its best-fit value of θ obs = 70 . ◦ , consistent with the type 2 classi- MNRAS , 1– ???? (?) lumpy disc-wind in MCG–03–58–007 Figure 6.
Top panel: The simultaneous
XMM-Newton & NuSTAR data ofMCG–03–58–007 corresponding to slice A. The fluxed spectra was unfoldedagainst a simple power-law with
Γ = 2 with the best-fit model B superim-posed shown in red. The pn, MOS 1 + + B spectra are shown inblack, green and blue respectively where the latter have been only co-addedfor plotting purposes. The contribution of model B are: distant scatteredcomponent (gray), photoionized emission (dark cyan), hot thermal emission(dark magenta), whereas the corresponding MYTorus model contributionsare the (zeroth-order) transmitted absorbed primary component (cyan), thereflected component (blue) and the Fe K α and Fe K β fluorescent line emis-sion lines (magenta). Bottom panel: The corresponding data/model ratiowhich shows some residual left >
10 keV . fication. The overall column density measured in this (coupled) con-figuration indicates that we are viewing the source through Comptonthin material i.e., log( N H / cm − ) = 23 . ± . , which mightsuggest that the upper regions of the torus are indeed less dense thanthe more central one. Overall in model B the transmitted compo-nent also dominates over the reflected component above . Thephoto-index was found to be Γ = 2 . ± . and the normalizationof the primary is . ± . × − ph cm − s − keV − . Model B also provided a very good fit to slice A at χ /ν = 719 . / whichis comparable to model A .Fig. 6 shows the best-fit model B to slice A and it seems toreproduce quite well the overall spectra. In the following analysis inSection 4.3 we will only focus on the MYTorus model as it assumesa better and a more realistic physical geometry of the reprocessor. Inboth model A and model B adopted here in slice A, the two distinctzones of the wind are ubiquitous, with consistent column densitiesand outflow velocities (see Table 3). From the above broadband spectral analysis, we show that bothmodel A and B successfully fitted slice A. However, B18 detected arapid spectral variability in MCG–03–58–007 between slice A andslice B which was likely caused by an obscuration event during the XMM-Newton & NuSTAR observation. In the following we verifywhether, by adopting a more self-consistent model for the neutralabsorber such as
MYTorus (model B ), rather than a slab reprocessor (model A ), the result discussed in B18 is still valid. In particular, weexplore whether the obscuration event is caused by: (i) a transitingneutral absorber within a clumpy torus or rather is (ii) the resultof the inhomogeneous nature of the highly ionized and fast disc-wind where a filament or clump rapidly crosses the line-of-sight(as also seen in PDS 456 Matzeu et al. 2016). We first attemptto test scenario (i) with model B by only varying the line-of-sightcolumn density and the primary continuum normalization, whilstkeeping the wind parameters tied between the two slices. Howeversuch model simply fails to fully reproduce the curvature in slice B,leaving a strong excess particularly between –
40 keV and yields apoor fit ( χ /ν = 893 . / ) as shown in Fig. 7. The harder shapeof the spectrum in slice B due to a higher obscuration compared toslice A.We then explored a more complex geometry of the reproces-sor by adopting the ‘decoupled’ configuration of MYTorus definedas model C . By following the methodology presented in details inYaqoob (2012), we want to represent a physical scenario where theneutral absorber is inhomogeneous in nature and hence character-ized by a more ‘patchy’ distribution of reprocessing clouds. This canbe achieved by decoupling the inclination angle parameters of thezeroth-order (line-of-sight) and reflected continua and allowing thecorresponding column density defined as N H , Z (line-of-sight N H )and N H , S (global N H ) respectively. The inclination of the zeroth-order component is fixed at ◦ whereas the reflected component at ◦ . The fluorescence emission lines, and the relative normalizationsare all tied to the primary power-law i.e., norm S00 = norm
L00 =norm
S90 = norm intr zpowl . As in model B , all the correspondingconstant factors A S00 = A
L00 = A
S90 are set to unity.In this model we find that the global N H is much larger thanthe zeroth-order at log( N H , S / cm − ) = 24 . +0 . − . which imprintsa much stronger reflection component in the spectrum. Althoughbetter than the coupled configuration, model C still forces a ∼ decrease of the primary continuum which still produces some excessresiduals in slice B between –
40 keV (see bottom-left panel inFig. 8). Although the transmitted N H increases and hence explainwell the spectral curvature in slice B, is still not enough to fullyaccount for it (see top-left panels in Fig. 8). Instead, this seems tosuggest that an additional variable absorber, such as the wind, isrequired to explain the increase in hardness of slice B. In summaryby investigating case (i) with model C , resulted in an improved fitcompared to model B at χ /ν = 860 . / (corresponding toa > . improvement). Even if the fit statistically improves,model C is not able to fully reproduce the spectral shape of slice B,leaving strong residuals in the –
40 keV range. Subsequently wetested case (ii) with model C by keeping the transmitted N H , Z of the MYTorus model tied between slices, but allowing the column densityof the slower ( v w /c = 0 . ) zone to vary. In Fig. 8 (right panels) weshow the best-fit model C for case (ii) overlaid on the fluxed slice Aand B spectra (top) and the corresponding residuals (bottom). Thistime the spectral curvature in slice B is well reproduced. This isexplained by a drastic increase in column density of the slowerionized absorber (zone 1) by almost one order of magnitude i.e.,from log( N H / cm − ) = 23 . ± . to log( N H / cm − ) = 24 . ± . between slice A and B.In summary, the observed variability of spectral hardness andline-of-sight column density is best modelled by a change in N H ofthe ionized absorber rather than by neutral absorption. Moreover,as in B18, we also found that there is no improvement in the fit byletting the N H of the fast zone 2 to vary, hence suggesting that thedensity change in the v w /c = 0 . zone is what drives the observedvariability. We also found that the overall power-law normalization MNRAS , 1– ?? (?) Matzeu et al.
Component Parameter Model A Model B (Slab reprocessor) (Toroidal reprocessor) Continuum
Primary power-law
Γ 2 . +0 . − . . +0 . − . norm a . +0 . − . . +0 . − . Scattered continuum
Γ 3 . +0 . − . . t norm b . +0 . − . . +0 . − . Soft X-ray emission
Photoionized c log( N H / cm − ) 21 . ∗ . ∗ log( ξ/ erg cm s − ) d . ∗ . ∗ κ e xstar . +0 . − . . +0 . − . Thermal kT (keV) . +0 . − . . +0 . − . norm ( × − ) f . +0 . − . . +0 . − . Distant reprocessor and neutral absorber
MYTorus Γ – . t log( N H / cm − ) g – . +0 . − . log( N H , S / cm − ) – – log( N H , Z / cm − ) – – norm S = norm L – . t norm S00 = norm
L00 = norm
S90 – – pexmon
Γ 2 . t –norm h . +0 . − . – zphabs log( N H / cm − ) 23 . +0 . − . – Highly ionized absorber
Zone 1 k log( N H / cm − ) 23 . +0 . − . . +0 . − . log( ξ/ erg cm s − ) 5 . +0 . − . . +0 . − . v w /c − . +0 . − . − . +0 . − . Zone 2 k log( N H / cm − ) 24 . +0 . − . . +0 . − . log( ξ/ erg cm s − ) 6 . +0 . − . . +0 . − . v w /c − . +0 . − . − . +0 . − . cross-normalization MOS . ± .
02 0 . ± . FPMA . ± .
04 1 . ± . FPMB . ± .
05 1 . ± . Fit statistic χ /ν . /
660 719 . / Table 3.
Summary of the broadband best fits parameters applied for slice A of
XMM-Newton & NuSTAR (see text for details). t and ∗ denote tied and frozenparameters respectively during fitting. a primary power-law normalization in unit of × − photons cm − s − keV − , b scattered power-law component in units of × − ph cm − s − keV − , c xstar emission grid with v turb = 100 km s − , d ionization parameter with fixed value obtained from the RGS fit, e normalization of the xstar emission component, in units of × − in terms of f cov L/ D , f normalization in units of − − π [ D A (1+ z )] (cid:82) n e n H dV , where n e and n H are the electron and hydrogen densities (measured in cm − ) respectivelyand D A is the angular diameter of the source in cm as defined in the mekal model, g mean line-of-sight column density, integrated over all lines of sight through the torus calculated as ( π/ N H , h ( × − ph cm − s − keV − ) , k xstar absorption grid with v turb = 10000 km s − . MNRAS , 1– ????
XMM-Newton & NuSTAR (see text for details). t and ∗ denote tied and frozenparameters respectively during fitting. a primary power-law normalization in unit of × − photons cm − s − keV − , b scattered power-law component in units of × − ph cm − s − keV − , c xstar emission grid with v turb = 100 km s − , d ionization parameter with fixed value obtained from the RGS fit, e normalization of the xstar emission component, in units of × − in terms of f cov L/ D , f normalization in units of − − π [ D A (1+ z )] (cid:82) n e n H dV , where n e and n H are the electron and hydrogen densities (measured in cm − ) respectivelyand D A is the angular diameter of the source in cm as defined in the mekal model, g mean line-of-sight column density, integrated over all lines of sight through the torus calculated as ( π/ N H , h ( × − ph cm − s − keV − ) , k xstar absorption grid with v turb = 10000 km s − . MNRAS , 1– ???? (?) lumpy disc-wind in MCG–03–58–007 Component Parameter Model C Model C (decoupled, wind fix) (decoupled, wind vary) Continuum
Slice A Slice B Slice A Slice BPrimary power-law
Γ 2 . +0 . − . . t . +0 . − . . t norm a . +0 . − . . +0 . − . . +0 . − . . +0 . − . Scattered continuum
Γ 2 . t . t . t . t norm b . +0 . − . . t . +0 . − . . t Soft X-ray emission
Photoionized emission κ xstar ( × − ) . +0 . − . . t . +0 . − . . t Thermal emission kT (keV) . +0 . − . . t . +0 . − . . t norm ( × − ) . +0 . − . . t . +0 . − . . t Distant reprocessor and neutral absorber
MYTorus
Γ 2 . t . t . t . t log( N H , S / cm − ) 24 . +0 . − . . t > . > . t log( N H , Z / cm − ) 23 . +0 . − . . +0 . − . . +0 . − . . t norm S00 = norm
L00 = norm
S90 . t . t . t . t Highly ionized absorberZone 1 log( N H / cm − ) 23 . +0 . − . . t +0 . − . +0 . − . log( ξ/ erg cm s − ) 4 . +0 . − . . t . +0 . − . . t v w /c − . +0 . − . − . t − . +0 . − . − . t Zone 2 log( N H / cm − ) 24 . +0 . − . . t . +0 . − . . t log( ξ/ erg cm s − ) 6 . +0 . − . . t . +0 . − . . t v w /c − . +0 . − . − . t − . +0 . − . − . t cross-normalization MOS . ± . – . ± . –FPMA . ± .
04 0 . +0 . − . . ± .
04 0 . +0 . − . FPMB . +0 . − . . +0 . − . . ± .
04 1 . +0 . − . Fit statistic χ /ν . /
761 826 . / Table 4.
Summary of the broadband best fits parameters of model C applied for slice A and B of XMM-Newton & NuSTAR (see text for details). The importantoutcome of this result is that the observed spectral variability between the slices can be explained by a drastic increase in N H by a factor of ∼ × in zone 1of the highly ionized absorber as opposed to a change in the N H of the neutral absorber. This behavior suggests at a high confidence level ( > . ) that theobserved spectral variability is caused by the highly ionized material rather than a neutral inhomogeneous absorber. All the parameters and units are the sameas in Table 3. does not decrease as drastically ( ∼ ) and the model now con-verges at ∼
20 keV , hence all the variations occur at lower energiesdue to absorption variability. Thus this model led to an excellent fitto the overall data at χ /ν = 826 . / , providing a substantialimprovement of ∆ χ / ∆ ν = 34 . / (i.e., ∼ σ ). We presented a detailed broadband spectral analysis of the X-rayspectra emission within . –
50 keV of MCG–03–58–007, wherewe successfully de-convolved, by using
MYTorus , all the layers ofabsorption. We infer a complex structure of absorbers consisting of: (a) a primary power-law component absorbed by a fully coveringneutral medium likely associated with the inhomogeneous toroidalabsorber, which is thicker in the equatorial plane, (b) a reflectedcomponent from a distant reprocessor, and (c) a multi-phase highlyionized fast outflow. In addition to this, we considered the analysisof the RGS data which allowed us to better understand the originof the soft X-ray emission. Our main findings are summarized anddiscussed in the following.
MNRAS , 1– ?? (?) Matzeu et al.
Figure 7.
Top panel: Best-fit model B superimposed (red) between slice Aand B. The corresponding MYTorus model contributions are (zeroth-order)transmitted absorbed primary component (cyan), the reflected continuum(blue) and the Fe K α and Fe K β fluorescent line emission lines (magenta).The distant scattered component is shown in gray. Bottom panel: The corre-sponding data/model ratio which show a clear excess in the residuals above20 keV in slice B, which cannot be accounted for in the coupled MYTorus model. The separate thermal and photoionized components are not includedin the plot for clarity (but are included in the model). Note that for plottingpurposes we adopted the combined
NuSTAR
FPMA + FPMB spectra.
The soft X-ray emission of MCG–03–58–007 can be well describedwith a superposition of a power-law scattered component and severalnarrow emission lines. The latter are associated with optically thin,photoionized gas as well as a weaker collisionally ionized plasmacomponent consistent with starburst activities in the host galaxy.From the luminosity measured in the IR band, Oi, Imanishi &Imase (2010) were able to robustly estimate the star formation rate(SFR) in MCG–03–58–007 to be
SFR = 9 . M (cid:12) yr − . From theSFR and the L x –SFR correlations derived from various sample ofLIRG and ULIRG (e.g., Ranalli, Comastri & Setti 2003; Pereira-Santaella et al. 2011; Mineo et al. 2014), we derived the predictedsoft X-ray luminosity to be L (0 . −
2) keV = 3 × erg s − whichis consistent with what it is measured in MCG–03–58–007 for thestar burst component. In terms of the photoionized component, wemeasured a soft X-ray luminosity of L (0 . −
2) keV ∼ erg s − .We did not observe any variability in the soft band in respect tothe 2010 Suzaku observation (see B18). All the emission lines areunresolved at the RGS spectral resolution ( σ v (cid:46) − ).From the results outlined in Section 3.1, the O vii triplet ratiogave us an upper limit on the gas density of n e < cm − whereas from the photoionization modelling we obtained an ion-ization state of ξ ∼
10 erg cm s − . Given the ionizing luminosityof L ion ∼ erg s − in MCG–03–58–007, this places a lowerlimit on the radial distance of the gas of R > . . On the otherhand a subsequent Chandra imaging observation of MCG–03–58–007 in 2016 reveal that the soft X-ray emitter is largely point-like,within an arc second radius of the nucleus (Braito et al. in prep). For MCG–03–58–007 this corresponds to an emitting gas that isconfined within a scale of few hundreds of parsec and althoughthis distance scale is poorly constrained, it is likely constant withNLR gas on approximately parsec scales or greater as often seen inother Seyfert 2 galaxies (e.g., Sako et al. 2000; Kinkhabwala et al.2002; Braito et al. 2017). As shown in Fig 4, the dominant contrib-utor to the emission lines is mostly a distant gas photoionized bythe AGN often observed in Seyfert 2 galaxies (e.g., Guainazzi &Bianchi 2007). However in all the adopted models we also observeda contribution from a weak thermal emission component with tem-perature of kT ∼ . . Note that if we did not include thephotoionized emission, the luminosity of the emission lines com-ponent (parameterized with mekal ) would be simply too high withrespect to the soft X-ray luminosity expected from the star formationrate. The variability in the
XMM-Newton & NuSTAR spectra in MCG–03–58–007 was caused by a rapid eclipsing event as discussed inB18. In this work we further investigated the cause of this event. Weinitially tested a scenario where the variability between the slices(defined in Section 4.1) could be explained solely with a changein N H of the neutral absorber and the primary continuum nor-malization by adopting the coupled MYTorus configuration whilstkeeping the disc-wind xstar parameters fixed between the slices.As discussed in Section 4.3, in the coupled
MYTorus configurationthe column densities of the transmitted and reflected componentsare coupled together (i.e., have the same N H ). We found that inthis scenario the N H only increased by ∼ , while the primarycontinuum decreased by ∼ . Moreover, as there is no spectralvariability between the slices >
20 keV , the excessive drop in nor-malization fails to account for the sharp hardening in the spectrumin slice B. This results in a very prominent excess residuals in the
NuSTAR spectrum above
20 keV in slice B as shown in Fig. 7.In B18, it was tested whether the spectral variability couldbe explained by the presence of a classical clumpy neutral ab-sorber. Although it produced an acceptable fit, this scenario wasalso ruled out due to its excessively high column density at log( N H / cm − ) > . which required ∼ covering (duringslice B) and the unphysically high luminosity derived when Comp-ton (down)scattering effects were taken to account for such high N H . In this work, we attempted to test the above scenario with themore physical MYTorus model by decoupling the N H of the out ofline-of-sight reflected and zeroth-order (transmitted) components(model C ). The patchy distribution of the reprocessor assumed inmodel C , allowed us to better account for the spectral curvature, inslice B (see Section 4.2.1). Despite this, the overall normalization ofthe primary continuum still excessively decreased and hence signif-icant residuals were still present (cid:38)
20 keV in slice B (see left panelof Fig 8).When fitting the
XMM-Newton & NuSTAR data with the
MYTorus decoupled mode, we were able to measure the columndensities of both the out of the line-of-sight and transmitted repro-cessors. Remarkably, we found that the measured global N H , S wasindeed much larger that the line-of-sight N H , Z by at least an orderof magnitude (see Table 4). Such an extreme difference betweenthese two absorbers can be geometrically explained with an over-all patchy toroidal reprocessor which is broadly Compton-thin witha relatively small equatorial thick layer out of the line-of-sight asshown schematically in in Fig. 9. MNRAS , 1– ????
MYTorus decoupled mode, we were able to measure the columndensities of both the out of the line-of-sight and transmitted repro-cessors. Remarkably, we found that the measured global N H , S wasindeed much larger that the line-of-sight N H , Z by at least an orderof magnitude (see Table 4). Such an extreme difference betweenthese two absorbers can be geometrically explained with an over-all patchy toroidal reprocessor which is broadly Compton-thin witha relatively small equatorial thick layer out of the line-of-sight asshown schematically in in Fig. 9. MNRAS , 1– ???? (?) lumpy disc-wind in MCG–03–58–007 Figure 8.
Best-fit
MYTorus decoupled mode (model C ), applied to slice A and slice B where the left panel shows a variable transmitted N H , Z (and constantwind), whereas the right panel shows a constant transmitted N H , Z (and variable wind). Only the latter is able to fully account for the pronounced spectralcurvature seen up to
40 keV in slice B. Note that for plotting purposes we adopted the combined
NuSTAR
FPMA + FPMB spectra. As the soft X-ray modelcomponents such as mekal and xstar emissions are not variable and do not affect the
MYTorus parameters, we omitted them here for clarity. The distantscattered component is shown in gray and the
MYTorus components are defined as in Fig. 7.
MCG–03–58–007 is not an outlier in this respect, in fact thedifference between the global and zeroth-order N H which sug-gest an inhomogeneous torus has been already observed in otherSeyfert 2 galaxies such as e.g., NGC 4945 (Yaqoob 2012), Markar-ian 3 (Yaqoob et al. 2015). This sort of inhomogeneous nature of thetorus has been widely accepted in the scientific community by de-veloping torus models considering a clumpy gas distribution (e.g.,Elitzur & Shlosman 2006; Nenkova et al. 2008a,b). Furthermorein the most recent models, the overall absorption can be quantifiedmore as a viewing probability dependent on the physical proper-ties of the cloud i.e., size and location, which typically tend to bedistributed towards the equatorial plane.Our modelling with the self-consistent model for a toroidalabsorber confirms that the main driver for the observed variabilityis the wind rather than a clumpy neutral absorber. In all the modelstested, when the wind column is held constant between the slices,the models require the normalization of the primary continuum todrop in slice B to compensate, leaving an excess above
20 keV inthe residuals. However this can be ruled out, as such a drop inthe continuum is not seen in the –
40 keV lightcurve. Instead,if the wind N H is allowed to increase in slice B, then this cannaturally account for the increase in hardness of slice B, while theoverall continuum normalization is now constant, consistent withthe lack of variability above
20 keV . Thus this result suggestedat high confidence level > . that the eclipsing event wasnot caused by a neutral absorber in the clumpy torus but insteaddriven by a transiting clump (or filament) in an inhomogeneous andhighly ionized disc-wind located a few hundreds of R g from theblack hole (B18). Indeed such rapid absorption event caused by the highly ionized disc-wind was also observed in PDS 456 (Goffordet al. 2014; Matzeu et al. 2016), which is considered to be hostingthe prototype disc-wind (Reeves et al. 2018), as well as e.g., inPG 1211 +
143 (Pounds et al. 2003), PG 1126–041 (Giustini et al.2011) and APM 08279 + ∼ . –
10 pc (e.g., Capellupo et al. 2013; McGraw et al.2018), which are increasingly supporting the multi-phase structureof these winds. For example in the narrow-line Seyfert 1 galaxyWPVS 007, Leighly et al. (2015) observed for the first time anoccultation event in the UV associated with broad absorption line(BAL) gas in the torus; whereas in the Seyfert 1 galaxy NGC 5548an ongoing long-term obscuration is observed in both the X-rays andUV bands (Kaastra et al. 2014). Interestingly, Hamann et al. (2018)detected a fast UV counterpart of the X-ray wind in PDS 456 whichwas measured at a comparable outflow velocity of v w ∼ . c . Now we investigate the main properties of the highly ionized fastdisc-wind and compare them with the results obtained in B18.From the observed outflow velocity ( v w ) inferred from the twoblueshifted Fe K absorption features, we can estimate a lower limiton the launching radius. This can be done by equating the v w with its MNRAS , 1– ?? (?) Matzeu et al. escape radius, thus R esc = (2 c /v ) R g where R g = GM BH /c isthe gravitational radius which corresponds to R g ∼ . × cm for a black hole mass of M BH ∼ M (cid:12) in MCG–03–58–007(B18). We therefore derive R in , ∼ R g ( ∼ × cm) and R in , ∼ R g ( ∼ × cm) for zone 1 ( v w /c = − . +0 . − . )and zone 2 ( v w /c = − . ± . ) respectively. These results sug-gest that we are viewing the disc-wind through a line-of-sight thatintercepts two distinct streamlines launched at different radii as alsoobserved in PDS 456 (Reeves et al. 2018). We then utilize thesemeasurements to estimate the disc-wind energetics by quantifyingthe mass outflow rate ( ˙ M w ) expressed as ˙ M w = fµπm p v w N H R in ,derived by Krongold et al. (2007), where a biconical geometry ofthe outflow is assumed. The constant factor for cosmic elementalabundances is set to µ = n H /n e = 1 . and R in is the inner radiusor the starting point of the disc-wind. The function f takes intoaccount the inclination angle with respect to the line-of-sight andthe disc. Unlike in PDS 456, where the disc-wind’s solid angle wasdirectly measured (Nardini et al. 2015), the geometry of the systemin MCG–03–58–007 is currently uncertain and hence we adopt aface value of f = 1 . (see Appendix 2 in Krongold et al. 2007).Thus for the slow zone 1, observed in slice A, we obtain amass outflow rate of ˙ M w , z1 ∼ . × g s − ( ∼ . M (cid:12) yr − ) which implies a kinetic power of ˙ E w , z1 ∼ × erg s − . Thiscorresponds to ∼ of the bolometric luminosity and hencefalls within the, theoretically predicted, minimum requirement i.e., ˙ E w /L bol ∼ . – for providing a feedback mechanisms be-tween the central SMBH and its host galaxy (e.g., Di Matteo,Springel & Hernquist 2005; Hopkins & Elvis 2010). This resultis consistent with B18 with the only difference being that a con-stant v w is assumed across the modelling. On the other hand inslice B, the drastic increase in N H results in a considerably higherkinetic power of ˙ E w , z1 ∼ × erg s − which correspondsto ∼ of L bol and hence exceeding the typical values quotedby AGN feedback models. By assuming that the outflow veloc-ity is comparable to the Keplerian velocity across the source i.e., v K ∼ v w = 0 . c and from the measured timescale of this obscu-ration event ∆ t = 120 ks we derive the radial size of the absorberto be ∆ R c = v K ∆ t ∼ R g ( ∼ × cm) . The hydrogennumber density of the cloud is defined as: n H ∼ ∆ N H ∆ R c ∼ ∆ N H r / √ c ∆ t , (2)where ∆ N H is the observed change in column density betweenslice A and B and r g = R/R g is the radial distance in units of R g .From the definition of the ionization parameter i.e., L ion /n H R (Tarter, Tucker & Salpeter 1969) we have n H ∼ n e = L ion /ξR and hence equating these densities we get: r / = L ion ξ √ t c ∆ N H ( GM BH ) − . (3)From an ionization of ξ ∼ − , an observed changein column density of ∆ N H ∼ . × cm − (obtained inmodel C ) and by assuming an ionizing luminosity of order of L ion ∼ erg s − , the location of the eclipsing cloud is de-rived to be R ∼ R g (5 × cm) , which is consistent to whatwas measured in B18. These results suggest that we are viewingthrough a clumpy disc-wind at a typical distance of few hundreds of R g from the central black hole, launched between tens to hundredsof R g . Regarding the fast zone 2, the energetic that can be derivedare purely speculative as, already noted in B18, the N H can only be constrained by a given best-fit ionization value as is otherwisedegenerate with the ionization(see Section 4.2).A physical picture of the possible geometry of such a system isschematically illustrated in Fig. 9 where the estimated distances ofthe launching radii of both zones and location of the eclipsing clumpand distant circumnuclear gas are illustrated. From the overall infor-mation that it is gathered in this work, we have a scenario in whichthere are three main regions i.e., the highly ionized inhomogeneousfast outflows, the clumpy toroidal neutral absorbers and the diffusecircumnuclear NLR scale gas. The highly ionized outflow is likelyinhomogeneous in structure and it is located closer in to the SMBH,where the regions from high to lower ionization are shown in redto cyan respectively. The clumpy torus is located at intermediatescales where we were able to measure two distinct column densities(in model C ) of the embedded neutral material (light to dark bluerepresent the increasing in N H ).The global N H , S can be associated with a compact region ofCompton-thick material that might be distributed equatorially outof our line-of-sight. On the other hand, the measured column den-sity that intercept the line-of-sight ( N H , Z ) might be associated withCompton-thin material that is located at the edge of the clumpytorus. The diffuse circumnuclear gas is situated at larger scales( > BLR) where the photoionized and thermal emitters are illus-trated in blue and green respectively. The low covering factor thatcharacterizes this gas might suggest that it is also inhomogeneousin structure where the dominant component can be associated withphotoionized emission caused by the AGN. Although MCG–03–58–007 is classified as Seyfert 2 galaxy the fortuitous line-of-sight(black dash line) through Compton-thin material. Thanks to thesynergy between
XMM-Newton & NuSTAR
X-ray observatories, itwas possible to build a complete picture of this complex system.
ACKNOWLEDGEMENTS
We want to thank the referee for the detailed and helpful report thatimproved the clarity of this paper. GAM also thank Dr EmanueleNardini for the useful discussions. GAM and MLP are supportedby European Space Agency (ESA) Research Fellowships. Basedon observations obtained with
XMM-Newton , an ESA science mis-sion with instruments and contributions directly funded by ESAMember States and NASA. GAM, VB, PS, AC and RDC acknowl-edge support from the Italian Space Agency (contracts ASI-INAFI/037/12/0 and ASI-INAF n.2017-14-H.0). JR acknowledges finan-cial support through grants NNX17AC38G, NNX17AD56G andHST-GO-14477.001-A. C.C. acknowledges funding from the Eu-ropean Union’s Horizon 2020 research and innovation programmeunder the Marie Skłodowska-Curie grant agreement No. 664931.
REFERENCES
Antonucci R., 1993, ARA&A, 31, 473Arnaud K. A., 1996, in Astronomical Society of the Pacific ConferenceSeries, Vol. 101, Astronomical Data Analysis Software and Systems V,Jacoby G. H., Barnes J., eds., p. 17Bautista M. A., Kallman T. R., 2001, ApJS, 134, 139Blustin A. J., Page M. J., Fuerst S. V., Branduardi-Raymont G., Ashton C. E.,2005, A&A, 431, 111Braito V., Reeves J. N., Bianchi S., Nardini E., Piconcelli E., 2017, A&A,600, A135Braito V. et al., 2018, MNRAS, 479, 3592 MNRAS , 1– ????
Antonucci R., 1993, ARA&A, 31, 473Arnaud K. A., 1996, in Astronomical Society of the Pacific ConferenceSeries, Vol. 101, Astronomical Data Analysis Software and Systems V,Jacoby G. H., Barnes J., eds., p. 17Bautista M. A., Kallman T. R., 2001, ApJS, 134, 139Blustin A. J., Page M. J., Fuerst S. V., Branduardi-Raymont G., Ashton C. E.,2005, A&A, 431, 111Braito V., Reeves J. N., Bianchi S., Nardini E., Piconcelli E., 2017, A&A,600, A135Braito V. et al., 2018, MNRAS, 479, 3592 MNRAS , 1– ???? (?) lumpy disc-wind in MCG–03–58–007 Figure 9.
Schematic representation of the possible geometry for MCG–03–58–007. Our line-of-sight is represented by the dashed line. The different componentsof the system are shown with the corresponding distance scales from the SMBH. Here there are three main regions i.e., the highly ionized outflow, the clumpytorus and the diffuse NLR gas are illustrated. The highly ionized outflow is located closer in, where the regions from high to lower ionization are shownin red to cyan respectively. The launch radii and location of the eclipsing ionized cloud of outflowing material are reported. The clumpy torus is located atintermediate scales where the two distinct column densities (measured in the
MYTorus decoupled mode) of the embedded neutral material are indicated. Theglobal N H , S is possibly associated with Compton-thick material might be distributed equatorially. On the other hand, the measured column density that crossesthe line-of-sight ( N H , Z ) is associated with Compton-thin material that may be located at the edge of the clumpy torus. The diffuse gas is located at larger scaleswhere the photoionized and thermal emitters are illustrated in blue and green respectively.Capellupo D. M., Hamann F., Shields J. C., Halpern J. P., Barlow T. A.,2013, MNRAS, 429, 1872Cash W., 1979, ApJ, 228, 939Chartas G., Brandt W. N., Gallagher S. C., 2003, ApJ, 595, 85Chartas G., Brandt W. N., Gallagher S. C., Garmire G. P., 2002, ApJ, 579,169Cicone C. et al., 2014, A&A, 562, A21Cicone C. et al., 2018, ApJ, 863, 143Di Matteo T., Springel V., Hernquist L., 2005, Nature, 433, 604Dickey J. M., Lockman F. J., 1990, ARA&A, 28, 215Elitzur M., Shlosman I., 2006, ApJ, 648, L101Feruglio C. et al., 2015, A&A, 583, A99Giustini M. et al., 2011, A&A, 536, A49Gofford J. et al., 2014, ApJ, 784, 77Gofford J., Reeves J. N., Tombesi F., Braito V., Turner T. J., Miller L., CappiM., 2013, MNRAS, 430, 60Guainazzi M., Bianchi S., 2007, MNRAS, 374, 1290Hamann F., Chartas G., Reeves J., Nardini E., 2018, MNRAS, 476, 943Hopkins P. F., Elvis M., 2010, MNRAS, 401, 7Kaastra J. S. et al., 2014, Science, 345, 64Kaastra J. S., Mewe R., Liedahl D. A., Komossa S., Brinkman A. C., 2000, A&A, 354, L83Kallman T., Evans D. A., Marshall H., Canizares C., Longinotti A., NowakM., Schulz N., 2014, ApJ, 780, 121Kallman T. R., Palmeri P., Bautista M. A., Mendoza C., Krolik J. H., 2004,ApJS, 155, 675King A., 2003, ApJ, 596, L27King A. R., Pounds K. A., 2003, MNRAS, 345, 657Kinkhabwala A. et al., 2002, ApJ, 575, 732Krongold Y., Nicastro F., Elvis M., Brickhouse N., Binette L., Mathur S.,Jiménez-Bailón E., 2007, ApJ, 659, 1022Leighly K. M., Cooper E., Grupe D., Terndrup D. M., Komossa S., 2015,ApJ, 809, L13Markowitz A. G., Krumpe M., Nikutta R., 2014, MNRAS, 439, 1403Matzeu G. A., Reeves J. N., Braito V., Nardini E., McLaughlin D. E., LobbanA. P., Tombesi F., Costa M. T., 2017, MNRAS, 472, L15Matzeu G. A., Reeves J. N., Nardini E., Braito V., Costa M. T., Tombesi F.,Gofford J., 2016, MNRAS, 458, 1311McGraw S. M., Shields J. C., Hamann F. W., Capellupo D. M., Herbst H.,2018, MNRAS, 475, 585Mewe R., Gronenschild E. H. B. M., van den Oord G. H. J., 1985, A&AS,62, 197MNRAS , 1– ?? (?) Matzeu et al.
Mineo S., Gilfanov M., Lehmer B. D., Morrison G. E., Sunyaev R., 2014,MNRAS, 437, 1698Murphy K. D., Yaqoob T., 2009, MNRAS, 397, 1549Nandra K., O’Neill P. M., George I. M., Reeves J. N., 2007, MNRAS, 382,194Nardini E., Fabian A. C., Reis R. C., Walton D. J., 2011, MNRAS, 410,1251Nardini E. et al., 2015, Science, 347, 860Nenkova M., Sirocky M. M., Ivezić Ž., Elitzur M., 2008a, ApJ, 685, 147Nenkova M., Sirocky M. M., Nikutta R., Ivezić Ž., Elitzur M., 2008b, ApJ,685, 160Netzer H., Laor A., 1993, ApJ, 404, L51Oi N., Imanishi M., Imase K., 2010, PASJ, 62, 1509Parker M. L., Matzeu G. A., Guainazzi M., Kalfountzou E., Miniutti G.,Santos-Lleó M., Schartel N., 2018, MNRAS, 480, 2365Parker M. L. et al., 2017, Nature, 543, 83Pereira-Santaella M. et al., 2011, A&A, 535, A93Pinto C. et al., 2018, MNRAS, 476, 1021Porquet D., Dubau J., 2000, A&AS, 143, 495Pounds K. A., Reeves J. N., King A. R., Page K. L., O’Brien P. T., TurnerM. J. L., 2003, MNRAS, 345, 705Ranalli P., Comastri A., Setti G., 2003, A&A, 399, 39Reeves J. N., Braito V., Nardini E., Lobban A. P., Matzeu G. A., Costa M. T.,2018, ApJ, 854, L8Reeves J. N. et al., 2009, ApJ, 701, 493Reeves J. N., O’Brien P. T., Ward M. J., 2003, ApJ, 593, L65Reeves J. N., Porquet D., Braito V., Gofford J., Nardini E., Turner T. J.,Crenshaw D. M., Kraemer S. B., 2013, ApJ, 776, 99Risaliti G., Elvis M., Nicastro F., 2002, ApJ, 571, 234Sako M., Kahn S. M., Paerels F., Liedahl D. A., 2000, ApJ, 543, L115Severgnini P., Caccianiga A., Della Ceca R., 2012, A&A, 542, A46Singh K. P., Garmire G. P., Nousek J., 1985, ApJ, 297, 633Tarter C. B., Tucker W. H., Salpeter E. E., 1969, ApJ, 156, 943Tombesi F., Cappi M., Reeves J. N., Palumbo G. G. C., Yaqoob T., BraitoV., Dadina M., 2010, A&A, 521, A57Tombesi F., Meléndez M., Veilleux S., Reeves J. N., González-Alfonso E.,Reynolds C. S., 2015, Nature, 519, 436Turner T. J., Pounds K. A., 1988, MNRAS, 232, 463Wilms J., Allen A., McCray R., 2000, ApJ, 542, 914Yaqoob T., 2012, MNRAS, 423, 3360Yaqoob T., Tatum M. M., Scholtes A., Gottlieb A., Turner T. J., 2015,MNRAS, 454, 973 MNRAS , 1– ????