The lively accretion disc in NGC 2992-II. The 2019/2021 X-ray monitoring campaigns

MNRAS 514, 2974–2993 (2022) https://doi.org/10.1093/mnras/stac1381 Advance Access publication 2022 May 18 The li v ely accretion disc in NGC 2992 – II. The 2019/2021 X-ray monitoring campaigns R. Middei , 1 , 2 ‹A. Marinucci , 3 V. Braito , 4 , 5 S. Bianchi , 6 B. De Marco, 7 A. Luminari , 1 , 8 G. Matt, 6 E. Nardini , 9 M. Perri, 1 , 2 J. N. Reeves 4 , 5 and F. Vagnetti 8 , 10 1 INAF – Osservatorio Astronomico di Roma, Via Frascati 33, Monte Porzio Catone, I-00078 Roma, Italy 2 Space Science Data Center –ASI, Via del Politecnico s.n.c., I-00133 Roma, Italy 3 ASI – Italian Space Agency, Via del Politecnico snc, I-00133 Rome, Italy 4 Center for Space Science and Technology, University of Maryland Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, USA 5 INAF – Osservatorio Astronomico di Brera, Via Bianchi 46, I-23807 Merate, LC, Italy 6 Dipartimento di Matematica e Fisica, Universit ` a degli Studi Roma Tre, via della Vasca Navale 84, I-00146 Roma, Italy 7 Departament de F ` ısica, EEBE, Universitat Polit ` ecnica de Catalunya, Av. Eduard Maristany 16, E-08019 Barcelona, Spain 8 INAF – Istituto di Astrofisica e Planetologia Spaziali, Via del Fosso del Cavaliere, I-00133 Roma, Italy 9 INAF – Osservatorio Astrofisico di Arcetri, Largo Enrico Fermi 5, I-50125 Firenze, Italy 10 Dipartimento di Fisica, Universit ` a degli Studi di Roma ‘Tor Vergata’, Via della Ricerca Scientifica 1, I-00133 Roma, Italy Accepted 2022 May 12. Received 2022 May 11; in original form 2022 February 28 A B S T R A C T We report on the shortand long-term X-ray properties of the bright nearby Seyfert 2 galaxy NGC 2992, which was e xtensiv ely observ ed with Swift , XMM–Ne wton , and NuSTAR . Swift targeted the source more than 100 times between 2019 and 2021 in the context of two monitoring campaigns. Both time-averaged and time-resolved analyses are performed, and we find that the short-to-long term spectral properties of NGC 2992 are dominated by a highly variable nuclear continuum. The source varied in the 2–10 keV energy band from 0.6 to 12 ×10 −11 erg cm −2 s −1 during the two year long Swift monitoring. The fastest 2–10 keV flux change (by a factor of ∼60 per cent ) occurred on a time-scale of a few hours. The o v erall emission spectrum of the source is consistent with a power law-like continuum (  = 1.69 ±0.01) absorbed by a constant line-of-sight column density N H = (7.8 ±0.1) ×10 21 cm −2 . The reflected emission is likely due to matter with an average column density N H = (9.6 ±2.7) × 10 22 cm −2 , thus NGC 2992 appears to have a globally Compton-thin circumnuclear medium. This scenario is fully supported by an independent analysis of the fractional variability and by XMM–Newton multiyear spectra. K ey words: galaxies: acti ve – galaxies: Seyfert – X-rays: galaxies – X-rays: individuals (NGC 2992). 1 INTRODUCTION Active galactic nuclei (AGNs) are extragalactic sources that emit across the whole electromagnetic spectrum. Such systems are composite and each sub-structure has its own role in shaping the emerging spectrum (see P ado vani et al. 2017 , for a comprehensi ve re vie w). It is ubiquitously accepted that the X-ray emission originates in the very inner regions of AGNs, near the central supermassive black hole (SMBH). Accretion of matter infalling on to the SMBH is responsible for the enormous amount of optical-UV photons, a fraction of which can be further energized via inverse-Compton (Sunyaev & Titarchuk 1980 ) off thermal electrons (the so-called hot corona: Haardt & Maraschi 1991 , 1993 ; Madejski et al. 1995 ; Zdziarski et al. 1995 ) up to the X-rays. The maximum energy gain for these seed photons is mainly set by the hot plasma’s temperature, and, to a lower extent, by its opacity (e.g. Rybicki & Lightman 1979 ; Beloborodov 1999 ; Middei et al. 2019 ). In fact, the X-ray continuum in AGNs is well modelled by a power law with a high energy roll-o v er (e.g. Perola et al. 2002 ; Dadina 2007 ; Molina et al.  E-mail: [email protected] 2009 , 2013 ; Malizia et al. 2014 ; Fabian et al. 2015 , 2017 ; Ricci et al. 2018 ; Tortosa et al. 2018 ). AGN X-ray spectra may show additional features due to reprocessing of the primary X-ray emission by the circumnuclear material. A fluorescence emission line from the Fe K-shell is commonly observed as the most prominent feature (e.g. Bianchi et al. 2009 ) and its analysis carries a wealth of information on the physics of the reflecting material. This emission line has an intrinsically narrow profile that can undergo distortions, such as broadening, due to special and general relativistic effects. In particular, the closer to the SMBH the reflectors, the more distorted (i.e. the broader) the neutral or ionized Fe line profile (e.g. Fabian et al. 1995 ). On the contrary, at larger distance, these effects are negligible, thus the Fe K αshape is consistent with a narrow profile. Additionally, in the case of Compton-thick reflectors (i.e. N H  1.5 ×10 24 cm −2 ) the X-ray spectra show a typical emission excess around 30 keV, the so-called Compton-hump (e.g. Matt, Fabian & Ross 1993 ). The effect of any absorbing matter crossing our line of sight can significantly attenuate the observed number of photons, especially in the soft X-rays (e.g. Cappi et al. 1999 ; Awaki et al. 2000 ; Matt 2002 ; Bianchi et al. 2009 ; Middei et al. 2021 ). The X-ray emission of AGNs is also well known to be variable in spectral shape and amplitude. Nearby Seyfert galaxies as well © 2022 The Author(s) Published by Oxford University Press on behalf of Royal Astronomical Society Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 The 2019/2021 X-ray monitoring campaigns 2975 MNRAS 514, 2974–2993 (2022) Table 1. The observation log for the XMM–Newton and NuSTAR data is presented. The NuSTAR exposure was simultaneously taken during the second orbit of XMM–Newton . Satellite Detector Obs. ID Obs. Net exposure Start-date XMM–Newton pn 0840920201 92.6 ks 2019-05-07 XMM–Newton pn 0840920301 92.8 ks 2019-05-09 NuSTAR FPMA/B 90501623002 57.4 ks 2019-05-10 as distant quasars show a typical softer-when brighter behaviour (e.g. Sobolewska & Papadakis 2009 ; Serafinelli, Vagnetti & Middei 2017 ), where softer spectral states, characterized by a photon index  > 2, generally correspond to higher flux states. Variability is also commonly witnessed in terms of flux changes that occur on different time intervals. Changes from months to decades are common (e.g. Papadakis et al. 2008 ; Vagnetti, Turriziani & Trevese 2011 ; Vagnetti et al. 2016 ; Falocco et al. 2017 ; Paolillo et al. 2017 ) and, in the X-rays, variations are also observed down to kiloseconds time-scales (e.g. Uttley, McHardy & Papadakis 2002 ; Ponti et al. 2012 ). The origin of such a rapid variability cannot be solely ascribed to the X-ray band merely mimicking the variations of the disc optical–UV photons (Nandra 2001 ). Moreo v er, a tight relation between short time-scales variations and SMBH mass (e.g. Papadakis 2004 ; O’Neill et al. 2005 ; McHardy et al. 2006 ; Ponti et al. 2012 ) is well established. Multi-epoch high S/N spectral and timing data provide compelling pieces of information to shed light on to the physics behind X-ray variability and its tight link with the emerging X-ray spectrum. In this context, we report on the X-ray spectral properties of NGC 2992, a nearby highly inclined spiral galaxy ( z = 0.00771, Keel 1996 ) classified as a Seyfert 1.5–1.9 galaxy (Trippe et al. 2008 ). This source was the target of two consecutiv e XMM –Ne wton orbits in 2019, the second of which had a simultaneous but shorter NuSTAR exposure. Multiple transient Fe K emission lines between 5 and 7 keV were found, originating from several flaring sectors of the accretion disc (Marinucci et al. 2020 ). Moreo v er, variable absorption structures abo v e 9 keV were also detected, associated with an intermittent disc wind (Luminari et al., submitted). The general trend of the source (already reported in Yaqoob et al. 2007 ; Shu et al. 2010 ; Marinucci et al. 2018 ), where relativistic emission lines are observed at high flux levels, was therefore confirmed. In this paper, we report on the analysis of all the Neil Gehrels Swift Observatory (hereafter Swift) data, most of which were taken in the context of two monitoring campaigns. Then, we present a detailed timing and spectral analysis of the XMM–Newton and NuSTAR 2019 observations. 2 DATA REDUCTION AND SCIENCE PRODUCTS EXTRACTION This paper focuses on NGC 2992 observations (longer than 100 s) taken in the context of Swift monitoring campaigns. In the first one, Swift-XRT monitored NGC 2992 throughout 2019 (from March 26 to December 14), with the aim of triggering a deep, high flux observation of the source. On 2019 May 6, the triggering flux threshold was met ( F 2–10 = 7.0 ×10 −11 erg cm −2 s −1 ) and XMM–Newton started observing the source on 2019 May 7 for two consecutive orbits (ObsIDs: 0840920201, 0840920301). NuSTAR (Harrison et al. 2013 ) observed NGC 2992 on 2019 May 10 for ∼120 ks, simultaneously with the second XMM–Newton orbit. In this paper, we consider the same data set presented in Marinucci et al. ( 2020 ) and Luminari et al. (submitted) (see also Table 1 ) and we address the reader to these papers for details on the data reduction. Table 2. Extraction regions properties as a function of the XRT observed rate. Region’s shape Radius (inner radius) Rate pixel cts s −1 Circle 20 < 0.6 Annulus (2) > 0.6 Annulus (3) > 1.4 Annulus (4) > 1.7 Annulus (5) > 2.8 Annulus (6) > 3.2 Then, we also consider exposures obtained during a no v el 2021 Swift monitoring aimed at keeping track of the extreme variability of NGC 2992. The extraction of the high level science products of each Swift– XRT exposure resulted from an automatic process that downloads and reduces raw data taken via photon counting acquisition mode. The procedure is based on the standard pipelines xrtpipeline and xrtproducts described in Capalbi et al. (2005). 1 The regions used to extract the source and background spectra and light curves are selected taking into account any pile-up affecting that specific observation. In particular, after computing the source net count rate with xima g e , the procedure selected a circular region or an annulus in the case of a rate < 0.6 or > 0.6 cts s −1 , respectively. Then, the inner radius of the annulus is determined on the basis of the observed count rate. In Table 2 , we list the rates limits for each inner radius of the extracting annulus. The region used to extract the source al w ays has an outer radius of 50 arcsec, regardless of its circular or annular shape. On the other hand, the background is al w ays extracted using an annular region centred on the source. A difference of 25 pixels ( ∼60 arcsec) is set between the inner and outer radii of such a region. The outer radii of the source and background regions are al w ays ∼60 arcsec apart. Spectra were then binned requiring a minimum of 5 counts per bin and were fitted adopting the Cash statistic (Cash 1979 ). Finally, we relied on a similar automatic procedure to extract science products for each UV O T exposure. In particular , we checked that two regions, one circular and centred on the source (radius = 6 arcsec) and a concentric annulus (  radius = 7 arcsec) were free of any other sources or spurious detection. The returned count rates for the UV O T filters are not corrected for Galactic nor intrinsic reddening. We notice that correcting for such effects would not modify our results, instead it would lead to a shift of the rates. Our computations showed that considering a reddening of E ( B − V) = 0.0519 (Schlafly & Finkbeiner 2011 ) would decrease the rates by ∼13 or ∼30 per cent for the V and the UVW 2 filters, respectively. 3 TIMING PROPERTIES The Swift , XMM–Newton , and NuSTAR observations provide a compelling data set to shed light on to the temporal properties of NGC 2992 at different time-scales. 3.1 Daily to yearly variations Swift e xtensiv ely observ ed NGC 2992. From 2006 to present days more than one hundred observations are available, and the bulk of 1 https:// swift.gsfc.nasa.gov/ analysis/xrt swguide v1 2.pdf Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 2976 R. Middei et al. MNRAS 514, 2974–2993 (2022) Figure 1. Swift –XRT and - UVOT light curves from the 123 observ ations. Remarkable v ariability from short to long time-scales can be observed in the X-rays, while the optical–UV light curves (not corrected for the extinction) are characterized by a more constant behaviour. UVOT filters are labelled in the plot while Xs and Xh account for X-rays in the 0.3–2 and 2–10 keV energy ranges. We notice that for visual purposes, each segment of the x -axis has a different length. the exposures belong to two distinct monitoring campaigns, one held during 2019 from which 60–2 ks exposures were derived, and a novel one co v ering 2021. Multi-epoch and multiwavelength light curves are showed in Fig. 1 , where remarkable variations are observed from daily to yearly time-scales. The lowest X-ray state corresponded to the archi v al observations from 2006 and a maximal variation larger than a factor of 10 is found. To the fast X-ray variations correspond fairly flat optical and UV time series. Such a constant behaviour, is clearly observed in the optical filters U (with both V and B showing the same trend) and UVW 2, although in this ultraviolet band we can see a marginal long-term increase. The flat shape of the optical– UV light curves for NGC 2992 can be straightforwardly accounted for by the obscured nature of this source, implying that even if the AGN may contribute, it does not dominate the emission in these bands. The well-sampled X-ray light curves obtained from the monitoring campaigns allow us to compute the corresponding structure functions (SFs). The SF has been widely adopted in different bands of the electromagnetic spectrum (Trevese et al. 1994 ; de Vries et al. 2005 ; Bauer et al. 2009 ) and quantifies the amount of variability giving a measure of the mean change between two observations separated by a time-lag τ. It has been used for ensemble studies (see Vagnetti et al. 2011 , 2016 , for details) as well as for single AGN (e.g. Gallo et al. 2018 ; Laurenti et al. 2020 ). Different mathematical formulations for such an estimator have been proposed (e.g. Simonetti, Cordes & Heeschen 1985 ; di Clemente et al. 1996 ), and we here use the one described in section 3 of Middei et al. ( 2017 ). In Fig. 2 , we show the soft (0.3–2 keV) and hard (2–10 keV) SFs for the two Swift– XRT campaigns. Flux changes in both bands increase with the restframe time-lags, SF hard ∼τ0.12 ±0.02 and SF soft ∼τ0.10 ±0.02 , with these slopes being compatible with the one found from ensemble studies focusing on the average variability of AGNs (SF ensemble ∼ Figure 2. Structure functions for the soft and hard X-rays computed for the two monitoring campaigns. Dashed lines account for the weighted linear regression describing the SF. Light curves in 2021 were more variable than in 2019 and both the hard and soft X-rays varied of the same amount within each monitoring. τ0.121 ±0.004 , Vagnetti et al. 2016 ). The normalizations of the SFs account for two different variability levels, the larger the variability, the higher the SF. Interestingly, larger variability is measured during 2021 corresponding to a lower flux state of the source. Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 The 2019/2021 X-ray monitoring campaigns 2977 MNRAS 514, 2974–2993 (2022) Figure 3. NGC 2992 hardness ratios as a function of the total collected counts in the 0.3–10 keV band. Different colours identify data from the 2019 (blue), 2021 (red), and those already in the archive (green). We then searched for correlations possibly connected with variations in the column density of the obscurer. For this reason, we computed the ratio between the hard and soft X-rays and we studied it as a function of the total counts. In accordance with the plot in Fig. 3 no correlation holds between the hard/soft X-ray ratio and full band rate, and a marginal trend can be only observed for a full band rate below 0.5 cts s −1 . The lack of a noticeable trend is consistent with a fairly constant column density of the obscurer. On the other hand, the soft and hard X-rays are strongly correlated ( P cc = 0.99, P ( < r ) < 0.01 per cent), see Fig. 4 , suggesting that both the soft and hard X-rays are produced by the very same spectral component. Finally, we further stress that neither the soft nor the hard X-ray bands are correlated with the ultraviolet emission, see Fig. 5 . Such a correlation has been commonly observed in samples of unobscured AGN (e.g. Edelson et al. 2002 ; Lusso et al. 2010 ; Edelson et al. 2015 ; Lusso & Risaliti 2017 ) and the lack of correlation has been associated either to an incumbent changing look process (e.g. Ricci et al. 2020 , 2021 ; Laha et al. 2022 ) or to intervening absorbing matter. Due to the Seyfert 2 nature of NGC 2992, the lack of a correlation between the disc and coronal emissions can be ascribed to the pre-dominantly non-nuclear nature of the UV emission observed with Swift–UVOT . We further quantified the variability properties of NGC 2992 computing the so-called fractional variability ( F var ). This estimator (e.g. Edelson et al. 2002 ; Vaughan et al. 2003 ; Ponti et al. 2004 , 2006 ) provides a direct measure of a light-curve variability and is defined as the square root of the normalized excess variance (e.g. Vaughan et al. 2004 ; Ponti et al. 2006 ; Matzeu et al. 2016 , 2017 ; Alston et al. 2019 ; De Marco et al. 2020 ; Igo et al. 2020 ; Parker et al. 2020 ). To derive compatible excess variance spectra, we only considered data taken during the 2019 and 2021 monitoring campaigns. In particular, for a meaningful comparison, we need to compute the F var spectra using light curves of similar length. The 2019 campaign spans a time interval of about 6 months although not performed continuously due to visibility issues. We thus divided each set of observations into subsets roughly co v ering an about 3 months long-time interval. We Figure 4. Correlation between soft and hard count rates (0.3–2 and 2–10 keV, respectively). This suggests that the power-law component dominates the ∼1–10 keV spectrum of NGC 2992. Figure 5. Lack of correlation between the X-rays and ultraviolet UVW 2 filter. Cyan triangles account for the soft X-rays while blue circles are used for the hard band. Pearson cross-correlation coefficients of P cc = −0.21 P( < r ) = 2 per cent and P cc = −0.19 P ( < r ) = 3 per cent are found for the Xs versus UVW 2 and Xh versus UVW 2, respectively. thus ended up with five different subsamples: sample a (2019 March– June), sample b (2019 October–December), sample c (2021 January– March) and sample d (2001 April–July) and sample e (2021 October– December). We then computed the light curves for the samples in dif ferent energy interv als and deri ved the corresponding fractional variability. In Fig. 6 , we show the resulting F var spectra. Aside from some fluctuations abo v e ∼7 k eV, lik ely due to background issues Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 2978 R. Middei et al. MNRAS 514, 2974–2993 (2022) Figure 6. Fractional variability spectra derived for the five Swift’s subsamples described in Section 3 . Excess variance spectra clearly have different normalizations suggesting that the source varied by different amounts o v er monthly time-scales. Error bars only account for the Poissonian noise. and/or low S/N, and the first energy bin that is mostly due to distant scattering, all the spectra have a fairly constant behaviour. This suggests that the variability is driven by a single variable component, which in our case is the primary continuum emission. Concerning the normalization of the spectra, we notice that the higher the F var spectrum the lower the flux, which is in agreement with what is commonly observed in other AGNs (e.g. Barr & Mushotzky 1986 ; Green, McHardy & Lehto 1993 ; Lawrence & Papadakis 1993 ). 3.2 Hourly to daily changes The two consecutiv e XMM–Ne wton orbits, one of which also has a simultaneous NuSTAR exposure, are extremely suitable for quantifying NGC 2992 variations on very short term. We started deriving the NGC 2992 light curves, see Fig. 7 . The different panels, from top to bottom, account for the 0.3–1, 1–3 and 3–10, 10–79 keV bands, while the last ro w sho ws on the ratios between the soft and hard X-rays (1–3 k eV/3–10 k eV). The constancy of the 0.3–1 light curves is consistent with extra-nuclear, ionized emission. In orbit 1, XMM–Newton caught the source in a higher flux state than in orbit 2, for which data up to 79 keV are also showed. Variations during the first orbit are about a factor of 60 per cent in both the 1–3 and 3–10 keV energy bands. Such fast flux changes occur on a time-scale of ∼40 ks. On the other hand, the amount of variability in orbit 2 is reduced to a few per cent for both the soft and the hard X-ray bands. NuSTAR ’s light curve is consistent with the 3–10 keV XMM –Newton time series and shows a similar amount of variability suggesting for the presence of a weak reflected component and a primary continuum still dominating this high-energy band. The hardness ratios reported in the last row of Fig. 7 are suggestive of a smooth spectral softening o v er the observ ed temporal window. In fact, ratios between the 1– 3 and 3–10 keV bands increases from 0.65 at the beginning of the first orbit to ∼0.75 at the end of the second XMM–Newton exposure. Thanks to the high S/N of this data set, we derived the F var spectra of the two XMM–Newton exposures. We used the background subtracted light curves binned every 1000 s. The resulting variability spectra are shown in Fig. 8 . The spectrum of orbit 1 has a larger normalization than orbit 2, in agreement with the light curves in Fig. 7 . During orbit 1, the 1–3 keV X-rays are more variable than the hard X-rays. A moderate drop in the variability spectrum is observed around 6.4 keV, as expected for a constant Fe K αemission. A similar drop is also observed in orbit 2, despite the ∼3 times less variable spectrum. Both spectra show a drop below ∼1 keV, as this energy range is dominated by a constant component. Finally, both spectra show an interesting variable feature around 5 keV that can possibly be associated with the transient emission line component discussed in Marinucci et al. ( 2020 ). In a recent work by Parker et al. ( 2020 ), the authors computed different tables to model F var spectra with standard spectral fitting packages, such as Xspec (Arnaud 1996 ). Following Parker et al. ( 2020 ) and the prescriptions in the web page, 2 we tried to model our excess variance spectra. In particular, we used the following model: Fvar pidamp 1 . fits ×Fvar pow . fits ×Fvar xildamp . fits . The first table is based on the Spex photoionization model Pion (Miller et al. 2015 ; Mehdipour, Kaastra & Kallman 2016 ) and accounts for the drop in variance due to a constant photoionized emission. The model’s parameters are frac and ξ, the ratio of the 0.5–10 keV flux of the reflection to the average log primary flux and the disc’s ionization in erg cm s −1 . The second table reproduces the variance of a power law-like continuum changing in log(flux). Its parameters, var and corr define the variance of the logarithmic flux for the primary continuum in the 0.5–10 keV energy range and the correlation between the photon index of the power law and the variable flux. Finally, the last table is needed to account for the reduction of F var due to unblurred reflection. In this case, the frac xill is the ratio of the 0.5–10 keV flux of the reflection to the average power-law flux whose flux is computed in logarithm. We fitted this very same model to the two spectra finding that the first high flux orbit is well described by the model with χ2 = 15 for 13 d.o.f. and that the model works fairly well for the continuum variability of orbit 2, despite the χ2 /d.o.f. = 38/12. This poor statistics is indeed mainly due to the unaccounted excess around 5 keV and, more marginally, by scattered data abo v e 7 keV. In Table 3 the inferred parameters for both the orbits are reported. 4 SPECTRAL PROPERTIES As for Section 3 , the multi-epoch broad-band data presented in this paper provide a compelling collection of observations suitable to perform both time-average and -resolved spectral analyses. 4.1 Mid-to-long term spectral properties The high flux of NGC 2992 allows us to extract XRT spectra for each of the 123 observations presented in this paper. We thus adopted a simple model to determine the basic properties of the primary continuum and the neutral absorbing column of NGC 2992 across the years, by fitting within Xspec the following model: tbabs ×ztbabs ×powerlaw . 2 https:// www.michaelparker.space/ variance-models Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 The 2019/2021 X-ray monitoring campaigns 2979 MNRAS 514, 2974–2993 (2022) Figure 7. Background subtracted light curves in units of counts/sec are shown for XMM–Newton and NuSTAR data. Aside from the 0.3–1 keV energy band, light curves below 10 keV show remarkable changes in the first orbit and moderate flux variability in the second one. The NuSTAR light curve (10–79 keV) is not completely simultaneous with the second XMM–Newton orbit. The 3–10 keV XMM–Newton (orbit 2) and the 10–79 keV NuSTAR light curves are characterized by a similar amount of variations abo v e 1, which suggests the primary continuum to dominate also the NuSTAR energy band. The power-law models the nuclear X-ray emission and both the local and Galactic absorptions are accounted for. For each observation, we fitted the column density of the local absorber, as well as the continuum photon index and its normalization. Via this procedure we determined the best-fitting parameters shown in Fig. 9 and quoted in Table A1 . Although the adopted model does not include the soft scattered component, all the spectra are well accounted for, as they all have a Cstat/d.o.f. ratio close to unity. Hard and soft X-ray fluxes show remarkable variations also down to daily time-scales, see Fig. 9 . It is hard to assess whether there are spectral changes or not. As expected, in fact, the power-law photon index and the obscurer column density are strongly correlated, diluting any intrinsic spectral variability. We thus refitted all the observations keeping N H fixed to its average value of N H = 7.8 ×10 21 cm −2 (determined from the 123 Swift observations). This new attempt led to steeper values of , found to have an average value of  = 1.59 ±0.04 and covering the range 1.37–1.97, see blue points in Fig. 10 . Interestingly, no correlation holds between these photon indices and the total flux, so that the source does not obviously show the typical softer when brighter behaviour commonly observed in AGNs (e.g. Sobolewska & Papadakis 2009 , see example in the inset of Fig. 10 ). 4.2 Short term spectral properties: the Fe K αcomplex We started focusing on the Fe complex of NGC 2992 testing a simple power law to the XMM–Newton spectra. We w ork ed on Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 2980 R. Middei et al. MNRAS 514, 2974–2993 (2022) Figure 8. Fractional variability spectra of Orbits 1 and 2 derived from the background subtracted light curves. The adopted temporal binning is set to 1000 s. Interestingly, these spectra are characterized by two different amounts of variability. In particular, the higher the flux the larger the amount of variability. We notice that the second orbit spectrum lacks the 0.3–0.5 keV energy bin as it was found consistent with zero. Fitted F var spectra are showed. In both orbits, substantial residuals can be observed around 5 keV. the 3–8 keV energy range and we fitted the photon index and the normalization of the continuum for both orbits 1 and 2. In Fig. 11 , we show zoomed spectra where modelling the sole continuum leaves prominent residuals between 6 and 7 keV. Then we added two Gaussian components with zero width to model the Fe K αand its accompanying Fe K β. We assumed the lines not to change between the two orbits so that we fitted the Gaussian energy centroid and normalization for the Fe K α. The Fe K β, had its energy fixed to 7.06 keV and its normalization was free to vary up to 14 per cent of the Fe K αflux (Molendi, Bianchi & Matt 2003 ). Although this leads to a significant reduction in the fit statistic ( χ2 /  d.o.f. = 264/160), the data are far from being well reproduced. First of all, data between 6.4 and 7.1 keV are not yet accounted for, suggesting that the Fe K α profile might be the superposition of different components. We then allowed the Fe K αwidth to vary finding a better fit ( χ2 = −19). The residuals around 6.4 keV are now accounted for. Then, we added two additional Gaussian components for the additional residuals at ∼6.7 and ∼7 keV. As for the 6.4 keV Fe K αline, we fitted the central energy and normalization of these two Gaussians (with null width) tying the values between the orbits. In particular, one line models the Fe Ly αemission at 6.96 keV ( χ2 /  d.o.f. = −42/ −2) while the second accounts for the Fe He αat 6.7 keV ( χ2 /  d.o.f. = −47/ −2). These steps led us to a best fit of χ2 = 156 for 155 d.o.f. and the inferred parameters are quoted in Table 4 . These tests are consistent with a weakly broad Fe K αthat may be the superposition of two different components. 4.3 The XMM–Newton/NuSTAR 2019 obser v ations: Time averaged spectral properties We here derive the NGC 2992 time-average properties in the 0.5– 79 keV energy range by fitting the XMM–Newton and the NuSTAR spectra. Relying on our findings in previous Section 4.2 and the phenomenological model by Marinucci et al. ( 2018 ), we built the following model in Xspec : tbabs ×(apec + Cloudy + (ztbabs ×po) + + zGauss + MyTorusL + MyTorusS + zGauss + zGauss) . (1) 4.3.1 Soft X-rays The neutral Galactic and intrinsic absorption is taken into account using the tbabs model. The soft X-rays of type 2 AGNs are generally dominated by emission lines whose origin is a photonionized gas consistent with the narrow line region (NLR; e.g. Awaki et al. 1991 ; Turner et al. 1997a , b ; Bianchi, Guainazzi & Chiaberge 2006 ; Guainazzi & Bianchi 2007 ; Laha et al. 2020 ), thus we accounted for this emission component using a grid model for Xspec computed with CLOUDY 17 (Ferland et al. 2017 ). This table, already presented in past studies (Bianchi et al. 2010 ; Marinucci et al. 2011 , 2017 ), has two parameters: the ionising flux log U = [ −2.00:4.00], with step of 0.25, and cloud column density log N H = [19.0:23.5], with step of 0.1. Then, the apec model was used to reproduce thermal emission from extra-nuclear material observed with Chandra (Colbert et al. 2005 ). We fitted the temperature and the normalization for the apec component and the column density, the ionization and the normalization of the Cloudy one. Ho we ver, these parameters were tied between the spectra as no variations are expected for this larger scale gas down to the investigated time-scales. 4.3.2 Hard X-rays A power law reproduces the nuclear continuum while MyTorusS and MyTorusL , both additive components, model the reflected emission plus its accompanying Fe K α, Fe K βfluorescent emission lines. MyTorus (Murphy & Yaqoob 2009 ; Yaqoob 2012 ) includes the Compton do wn-scattering ef fect and the self-consistent reflected components assuming a fixed geometry of the toroidal X-ray reprocessor, for which the co v ering factor of the torus corresponds to a fixed halfopening angle of 60 ◦. Here we assumed: the medium absorbing the primary continuum and the one reflecting it to have different column densities. To set this scenario up we fitted independently the column density of ztbabs 3 and MyTorus tables independently i.e. using the so-called decoupled mode to allow the reflector and absorber to have different column densities. Then, we fixed the viewing angle of MyT orusL and MyT orusS to 0 ◦, thus implementing the back scattering scenario. In the fits, the photon index of the power law was tied with those of the two MyTorus tables and computed for both the orbits. The normalization of the nuclear emission was computed for both orbits, similar to the one of the MyTorus model, which we assumed to be the same between MyTorusL and MyTorusS . 4.3.3 Emission lines The 5–7 keV energy range hosts prominent features in emission and we used Gaussian lines to account for all of them except for the Fe K αand the Fe K βlines, which are already included in MyTorus . Ho we ver, in Section 4.2 we found evidence for a weakly broad Fe K α line, thus we added a Gaussian component whose energy centroid 3 The NGC 2992 obscurer has a column density N H < 10 22 cm −2 , too low to adopt the multiplicative table MyTorusZ commonly used to account for absorption in the line of sight. Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 The 2019/2021 X-ray monitoring campaigns 2981 MNRAS 514, 2974–2993 (2022) Figure 9. Best-fitting parameters derived from the analyses of the XRT exposures. The column density is in units of ×10 22 cm −2 and fluxes are in units of erg cm −2 s −1 . The same results are reported in Table A1 . Figure 10. XRT photon indices as a function of the 2–10 keV flux (in units of ×10 −11 erg cm −2 s −1 ). Blue dots account for  v alues deri ved using a fixed absorbing column density N H = 7.8 ×10 21 cm −2 . Grey crosses, instead, represent flatter  values that had been derived with a free to vary N H . The inset (taken from Sobolewska & Papadakis 2009 ) shows MCG-630-15, which in contrast, does show a softer when brighter behaviour more typical of other AGN. Figure 11. Zoom in the 5–7.5 keV energy band of the data to model ratios. The spectrum in black refers to orbit 1 data while the red spectrum accounts for data taken during orbit 2. Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 2982 R. Middei et al. MNRAS 514, 2974–2993 (2022) Table 3. Best-fitting parameters for the two excess variance spectra derived from the XMM–Newton orbits 1 and 2. The fit statistics are χ2 /d.o.f. = 15/13 and χ2 /d.o.f = 38/12 for the two spectra, respectively. Model Parameter obs1 obs2 Fvar pidamp frac 0.26 + 0 . 21 −0 . 09 1.0 + 1 . 9 −0 . 4 xi 0.85 + 0 . 61 −0 . 40 1.6 ±0.3 Fv ar po w var 0 . 059 + 0 . 002 −0 . 004 0.022 + 0 . 005 −0 . 003 corr < 0.24 < 0.63 Fvar xildamp frac 0.26 + 0 . 03 −0 . 12 < 0.34 Table 4. Best-fitting quantities derived from the XMM–Newton exposures fitted in the 3–10 keV energy range. The dagger indicates that the energy centroid of the Fe K βwas fixed. Component Parameter Orbit 1 Orbit 2 Units Tbabs N H 1.0 ±0.2 0.9 ±0.2 ×10 22 cm −2 Pow  1.62 ±0.03 1.66 ±0.02 N pow 1.7 ±0.1 2.1 ±0.1 ×10 −2 ph. keV −1 cm 2 s −1 zGauss Fe K αE 6.38 ±0.05 keV σ46 ±10 eV EW 90 ±5 eV Norm 8.6 ±0.3 ×10 −5 ph. cm −2 s −1 zGauss Fe K β† E 7.06 keV EW < 20 eV Norm < 5.9 ×10 −6 ph. cm −2 s −1 zGauss FeLy αE 6.96 ±0.01 keV EW 25 ±4 eV Norm 2.0 ±0.3 ×10 −5 ph. cm −2 s −1 zGauss Fe He αE 6.71 ±0.04 keV EW 12 ±3 eV Norm 1.2 ±0.3 ×10 −5 ph. cm −2 s −1 was computed tying its value between the spectra and fitting its normalization in both orbits. Moreo v er, we fix ed the centroid energy of two additional ionized Gaussians to E = 6.7 keV, E = 6.96 keV, respectively, and assumed a narrow profile ( σ= 0 eV) for both of them. These steps led to a fit statistic of χ2 /d.o.f. = 1074/731. Residuals between 3 and 5 keV suggest the photon index may not be the same for XMM–Newton orbit 2 and NuSTAR (  ∼0.06). This may be either due to the non-simultaneity of the spectra or due to intercalibration issues among the detectors as also reported for other observations (e.g. Porquet et al. 2018 ; Laha & Ghosh 2021 ). Allo wing for dif ferent  v alues for XMM–Newton and NuSTAR in orbit 2 yields a better fit statistic of χ2 /d.o.f. = 980/730 (see Fig. 12 ), and we report in Table 5 the corresponding best-fitting parameters. In accordance with the light curves, the first orbit NGC 2992 showed a higher flux than in orbit 2, accompanied by a spectral shape characterized by a similar photon index of  = 1.68 ±0.01 and absorbing column density of N H = 7.8 ±0.2 ×10 21 cm −2 . In a similar fashion, the reflected emission has a compatible flux between the two observations and a rather constant column density N H ∼ 9.6 ×10 22 cm −2 for the scattered component out of the line of sight. The presence of Compton-thin matter both along the line of sight and out of the line of sight implies the o v erall emission spectrum of NGC 2992 being globally Compton-thin. Only upper limits were found for the variable Fe K αred tail while the Fe XXV He αand Fe XXVI Ly αare well constrained in both orbits. Despite the fairly acceptable statistics, the model well reproduces the XMM–Newton and NuSTAR spectra as the high χ2 is mainly due to residuals between 1 and 2 keV likely resulting from calibration issues. Figure 12. Best fit for the two XMM–Newton orbits and the accompanying NuSTAR data. The inferred parameters are reported in Table 5 . The model components shown and labelled in the bottom panel are those derived for the simultaneous XMM–Newton–NuSTAR data only. 4.4 High energy cut-off The broad-band co v erage pro vided by a simultaneous XMM– Ne wton–NuSTAR e xposure is extremely suitable to investigate for the high energy roll-o v er of the nuclear continuum emission, which provides direct clues on the physical properties of the hot corona. For this reason, here we focus on data belonging to XMM–Newton orbit 2 and those from its accompanying NuSTAR observation. In Section 4.3 , we found the o v erall emission spectrum of NGC 2992 to be globally Compton thin. We thus further investigate the properties of the X-ray emission in NGC 2992 replacing MyTorus with the Borus model (Balokovi ´ c et al. 2018 ; Balokovi ´ c, Garc ´ ıa & Cabral 2019 ; Balokovi ´ c et al. 2021 ). In this model, in fact, the high energy cut-off of the primary continuum is set as a free parameter and is not fixed to 300 keV. In this model, a homogeneous spherical scattering medium is considered to surround the central X-ray source. Except for Fe, whose relative abundance (A Fe ) can be derived, a solar abundance is considered. We therefore modelled the simultaneous XMM–Newton/NuSTAR orbit 2 replacing the MyTorus tables with the Borus one (borus02 v170323a.fits) ending up with the model: tbabs G ×(apec + Cloudy + (tbabs z ×cutoffpl) + + Borus + zGauss + zGauss + zGauss) . (2) We fit XMM –Newton orbit 2 and NuSTAR data computing the Borus normalization and column density. We assumed the cut-off power law and Borus to have the same primary photon index, high energy Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 The 2019/2021 X-ray monitoring campaigns 2989 MNRAS 514, 2974–2993 (2022) Figure A1. Spectral slices ∼5 ks long best fitted in accordance with Section 5.4. Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 2990 R. Middei et al. MNRAS 514, 2974–2993 (2022) Figure A2. Continues Fig. A1 . Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 The 2019/2021 X-ray monitoring campaigns 2991 MNRAS 514, 2974–2993 (2022) Table A1. Best-fitting parameters derived form the analysis of XRT exposures. Fluxes in the 0.5–2 and 2–10 keV are reported in units of ×10 −12 and ×10 −11 erg cm −2 s −1 , respectively. The column density accounts for 10 22 cm −2 and normalizations are listed in units of photons keV −1 cm −2 s −1 . Date Obs. ID. F 0.5–2 keV F 2–10 keV N H  Norm 2006-06-14 00035344002 1.16 ±0.23 2.30 ±0.23 0.23 ±0.22 0.90 ±0.20 0.0008 ±0.0002 2006-06-25 00035344003 0.68 ±0.09 0.98 ±0.09 0.19 ±0.14 1.05 ±0.17 0.0005 ±0.0001 2006-07-06 00035344004 0.67 ±0.09 0.97 ±0.09 0.16 ±0.13 0.90 ±0.14 0.0004 ±0.0001 2006-07-07 00035344005 0.85 ±0.06 0.59 ±0.06 0.22 ±0.07 1.16 ±0.08 0.0006 ±0.0001 2006-07-11 00035344006 0.78 ±0.16 1.37 ±0.16 0.18 ±0.16 1.19 ±0.24 0.0005 ±0.0002 2006-07-12 00035344007 0.80 ±0.21 1.63 ±0.21 0.34 ±0.28 1.29 ±0.28 0.0007 ±0.0003 2015-12-02 00081055001 6.60 ±0.27 3.14 ±0.27 0.89 ±0.10 1.52 ±0.06 0.0123 ±0.0011 2017-06-02 00035344008 4.19 ±0.29 2.86 ±0.29 0.58 ±0.12 1.31 ±0.09 0.0052 ±0.0006 2017-06-04 00035344009 4.65 ±0.45 4.77 ±0.45 0.80 ±0.22 1.45 ±0.13 0.0077 ±0.0015 2019-03-26 00035344010 11.80 ±0.71 6.87 ±0.71 0.81 ±0.13 1.63 ±0.09 0.0211 ±0.0026 2019-03-30 00035344011 14.24 ±0.91 7.82 ±0.91 0.79 ±0.14 1.70 ±0.09 0.0258 ±0.0033 2019-04-03 00035344012 8.72 ±0.62 6.18 ±0.62 0.84 ±0.18 1.54 ±0.11 0.0155 ±0.0024 2019-04-07 00035344013 10.01 ±0.83 7.48 ±0.83 0.79 ±0.19 1.57 ±0.12 0.017 ±0.003 2019-04-11 00035344014 6.33 ±0.58 5.32 ±0.58 0.44 ±0.15 1.24 ±0.11 0.0064 ±0.0010 2019-04-14 00035344015 10.83 ±0.77 7.95 ±0.77 0.74 ±0.15 1.56 ±0.10 0.0174 ±0.0025 2019-04-19 00035344016 7.75 ±0.55 6.03 ±0.55 0.57 ±0.14 1.29 ±0.10 0.0094 ±0.0013 2019-04-23 00035344017 8.78 ±0.60 6.32 ±0.60 0.63 ±0.13 1.39 ±0.10 0.0118 ±0.0016 2019-05-01 00035344019 10.21 ±0.82 7.67 ±0.82 0.90 ±0.17 1.61 ±0.11 0.0198 ±0.0031 2019-05-04 00035344020 7.28 ±0.63 6.76 ±0.63 0.75 ±0.20 1.42 ±0.13 0.0113 ±0.002 2019-05-06 00035344021 9.73 ±0.68 6.24 ±0.68 0.81 ±0.15 1.58 ±0.1 0.0171 ±0.0024 2019-05-08 00035344022 9.55 ±0.62 6.47 ±0.62 0.81 ±0.15 1.48 ±0.09 0.0161 ±0.0021 2019-05-10 00035344023 9.46 ±1.33 11.36 ±1.33 0.70 ±0.24 1.50 ±0.17 0.0143 ±0.0034 2019-05-12 00035344024 11.91 ±5.47 9.60 ±1.47 0.76 ±0.57 1.53 ±0.39 0.0194 ±0.0109 2019-05-14 00035344025 12.80 ±0.68 7.36 ±0.68 0.80 ±0.13 1.57 ±0.08 0.0221 ±0.0026 2019-05-16 00035344026 15.48 ±0.92 6.97 ±0.92 0.85 ±0.13 1.74 ±0.09 0.0302 ±0.0036 2019-05-18 00035344027 10.39 ±0.68 6.92 ±0.68 0.82 ±0.15 1.54 ±0.10 0.0181 ±0.0025 2019-05-20 00035344028 11.53 ±0.71 6.33 ±0.71 0.93 ±0.15 1.69 ±0.09 0.0238 ±0.0031 2019-05-22 00035344029 12.44 ±0.74 7.10 ±0.74 0.88 ±0.15 1.67 ±0.10 0.0243 ±0.0032 2019-05-24 00035344030 11.33 ±0.74 6.80 ±0.74 0.57 ±0.11 1.47 ±0.09 0.0145 ±0.0017 2019-05-26 00035344031 10.34 ±0.68 6.20 ±0.68 0.66 ±0.12 1.51 ±0.09 0.0151 ±0.0019 2019-05-28 00035344032 8.42 ±0.64 6.09 ±0.64 0.87 ±0.18 1.54 ±0.11 0.0155 ±0.0023 2019-05-30 00035344033 9.43 ±0.67 6.56 ±0.67 0.87 ±0.17 1.56 ±0.11 0.0175 ±0.0026 2019-06-01 00035344034 10.19 ±0.74 7.04 ±0.74 0.62 ±0.15 1.46 ±0.10 0.0138 ±0.002 2019-06-03 00035344035 7.67 ±0.59 6.19 ±0.59 0.52 ±0.13 1.26 ±0.10 0.0086 ±0.0012 2019-06-05 00035344036 15.33 ±0.92 8.89 ±0.92 0.83 ±0.15 1.64 ±0.10 0.0282 ±0.0037 2019-06-07 00035344037 8.15 ±0.61 6.60 ±0.61 0.62 ±0.13 1.33 ±0.09 0.0106 ±0.0014 2019-06-09 00035344038 5.87 ±0.49 5.00 ±0.49 0.80 ±0.20 1.53 ±0.13 0.01 ±0.0018 2019-06-11 00035344039 10.12 ±0.66 6.27 ±0.66 0.88 ±0.14 1.59 ±0.09 0.0192 ±0.0025 2019-06-16 00035344040 7.40 ±0.58 5.25 ±0.58 0.60 ±0.14 1.38 ±0.10 0.0095 ±0.0013 2019-06-20 00035344041 10.63 ±0.66 6.17 ±0.66 0.83 ±0.14 1.61 ±0.09 0.0192 ±0.0025 2019-06-24 00035344042 6.99 ±0.62 5.94 ±0.62 0.74 ±0.19 1.51 ±0.13 0.011 ±0.002 2019-06-27 00035344043 4.82 ±1.37 5.29 ±1.37 0.13 ±0.23 0.62 ±0.27 0.0026 ±0.001 2019-10-09 00035344044 11.99 ±0.80 7.50 ±0.80 0.77 ±0.13 1.57 ±0.09 0.0201 ±0.0025 2019-10-13 00035344045 4.22 ±0.40 4.00 ±0.40 0.39 ±0.13 1.21 ±0.11 0.004 ±0.0006 2019-10-17 00035344046 5.57 ±0.49 4.71 ±0.49 0.68 ±0.18 1.39 ±0.12 0.0079 ±0.0013 2019-10-21 00035344047 5.71 ±0.67 6.88 ±0.67 0.52 ±0.22 1.31 ±0.16 0.0065 ±0.0015 2019-10-25 00035344048 6.21 ±0.64 5.65 ±0.64 0.80 ±0.23 1.57 ±0.15 0.0107 ±0.0022 2019-10-29 00035344049 5.86 ±0.55 6.52 ±0.55 0.49 ±0.20 1.14 ±0.13 0.0061 ±0.0012 2019-11-02 00035344050 4.05 ±0.45 4.56 ±0.45 0.43 ±0.18 1.19 ±0.14 0.004 ±0.0008 2019-11-06 00035344051 4.57 ±0.61 5.98 ±0.61 0.64 ±0.24 1.42 ±0.17 0.0063 ±0.0015 2019-11-08 00035344052 4.99 ±0.42 3.87 ±0.42 0.60 ±0.15 1.43 ±0.11 0.0066 ±0.001 2019-11-10 00035344053 7.22 ±0.61 5.93 ±0.61 0.77 ±0.17 1.47 ±0.11 0.0115 ±0.0018 2019-11-12 00035344054 7.29 ±0.58 5.55 ±0.58 0.88 ±0.17 1.63 ±0.11 0.0141 ±0.0022 2019-11-14 00035344055 9.14 ±0.71 7.40 ±0.71 0.54 ±0.13 1.27 ±0.10 0.0106 ±0.0014 2019-11-16 00035344056 7.68 ±0.58 6.07 ±0.58 0.82 ±0.17 1.48 ±0.11 0.013 ±0.0019 2019-11-17 00035344057 10.62 ±0.69 6.58 ±0.69 0.85 ±0.13 1.57 ±0.09 0.0193 ±0.0024 2019-11-20 00035344058 8.13 ±0.65 6.70 ±0.65 0.76 ±0.17 1.41 ±0.11 0.0126 ±0.0019 2019-11-22 00035344059 7.80 ±0.98 9.19 ±0.98 0.52 ±0.2 1.31 ±0.15 0.0089 ±0.0018 2019-11-26 00035344061 4.92 ±0.49 4.27 ±0.49 0.63 ±0.18 1.42 ±0.13 0.0067 ±0.0012 2019-11-28 00035344062 10.67 ±0.83 7.15 ±0.83 0.66 ±0.15 1.58 ±0.11 0.0159 ±0.0024 2019-11-30 00035344063 7.84 ±0.73 6.63 ±0.73 0.78 ±0.18 1.53 ±0.12 0.013 ±0.0022 2019-12-02 00035344064 7.08 ±0.63 4.54 ±0.63 0.72 ±0.17 1.77 ±0.12 0.0122 ±0.002 Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 2992 R. Middei et al. MNRAS 514, 2974–2993 (2022) Table A1 –continued Date Obs. ID. F 0.5–2 keV F 2–10 keV N H  Norm 2019-12-04 00035344065 10.82 ±0.75 7.92 ±0.75 0.57 ±0.12 1.37 ±0.09 0.0135 ±0.0017 2019-12-06 00035344066 14.79 ±1.07 8.81 ±1.07 0.83 ±0.15 1.72 ±0.10 0.028 ±0.004 2019-12-08 00035344067 11.25 ±0.78 7.60 ±0.78 0.65 ±0.15 1.47 ±0.10 0.016 ±0.0023 2019-12-10 00035344068 7.88 ±1.00 8.39 ±1.00 0.69 ±0.26 1.54 ±0.18 0.012 ±0.0029 2019-12-12 00035344069 9.70 ±0.67 6.20 ±0.67 0.70 ±0.13 1.55 ±0.09 0.015 ±0.0019 2019-12-14 00035344070 9.36 ±0.74 6.57 ±0.74 0.89 ±0.19 1.66 ±0.12 0.0185 ±0.0031 2019-12-31 00035344071 2.96 ±0.32 2.91 ±0.32 0.53 ±0.18 1.51 ±0.15 0.0037 ±0.0007 2020-01-28 00035344072 11.64 ±0.97 8.21 ±0.97 0.71 ±0.15 1.63 ±0.12 0.0187 ±0.0029 2020-03-25 00035344073 12.08 ±0.79 7.11 ±0.79 0.74 ±0.14 1.55 ±0.09 0.0193 ±0.0025 2020-05-19 00035344074 6.97 ±0.58 5.79 ±0.58 0.83 ±0.19 1.58 ±0.12 0.0125 ±0.0021 2021-01-24 00035344075 3.31 ±0.29 3.11 ±0.29 0.49 ±0.14 1.29 ±0.11 0.0037 ±0.0006 2021-01-29 00035344076 2.95 ±0.41 3.98 ±0.41 0.65 ±0.25 1.41 ±0.18 0.0041 ±0.001 2021-02-03 00035344077 1.54 ±0.84 9.94 ±0.84 0.58 ±0.68 1.11 ±0.47 0.0017 ±0.0012 2021-02-08 00035344078 1.60 ±0.30 2.09 ±0.30 0.83 ±0.36 1.82 ±0.26 0.0032 ±0.0011 2021-02-13 00035344079 2.98 ±0.32 3.07 ±0.32 0.77 ±0.23 1.52 ±0.15 0.0049 ±0.001 2021-02-18 00035344080 2.22 ±0.42 3.81 ±0.42 0.58 ±0.33 1.34 ±0.23 0.0028 ±0.0009 2021-02-22 00035344081 5.27 ±0.49 4.36 ±0.49 0.84 ±0.24 1.71 ±0.15 0.0101 ±0.0021 2021-02-28 00035344082 1.65 ±0.22 2.27 ±0.22 0.32 ±0.17 1.17 ±0.17 0.0014 ±0.0003 2021-03-05 00035344083 1.50 ±0.19 2.32 ±0.19 0.28 ±0.19 0.94 ±0.16 0.0011 ±0.0002 2021-03-10 00035344084 1.04 ±0.19 1.46 ±0.19 0.12 ±0.14 1.14 ±0.20 0.0006 ±0.0002 2021-03-15 00035344085 1.37 ±0.22 1.71 ±0.22 0.42 ±0.20 1.50 ±0.20 0.0015 ±0.0004 2021-03-25 00035344088 2.23 ±0.25 2.78 ±0.25 0.38 ±0.16 1.18 ±0.14 0.0021 ±0.0004 2021-03-30 00035344089 2.00 ±0.30 2.95 ±0.30 0.35 ±0.21 1.09 ±0.17 0.0017 ±0.0004 2021-04-04 00035344090 2.00 ±0.26 2.19 ±0.26 0.41 ±0.17 1.40 ±0.16 0.0021 ±0.0004 2021-04-09 00035344091 3.14 ±0.34 3.27 ±0.34 0.84 ±0.27 1.61 ±0.17 0.0058 ±0.0014 2021-04-14 00035344092 1.52 ±0.22 1.65 ±0.22 0.50 ±0.22 1.58 ±0.20 0.0019 ±0.0005 2021-04-19 00035344093 2.31 ±0.23 2.68 ±0.23 0.28 ±0.14 0.99 ±0.13 0.0017 ±0.0003 2021-04-24 00035344094 1.85 ±0.27 2.57 ±0.27 0.61 ±0.25 1.36 ±0.18 0.0024 ±0.0006 2021-04-30 00035344096 2.76 ±0.30 3.11 ±0.30 0.62 ±0.21 1.38 ±0.15 0.0036 ±0.0008 2021-05-04 00035344097 1.76 ±0.23 2.70 ±0.23 0.15 ±0.12 0.93 ±0.15 0.0011 ±0.0002 2021-05-09 00035344098 1.01 ±0.29 2.40 ±0.29 0.43 ±0.34 1.37 ±0.32 0.0011 ±0.0004 2021-05-12 00035344099 1.34 ±0.26 2.47 ±0.26 0.24 ±0.20 1.08 ±0.21 0.0010 ±0.0003 2021-05-14 00035344100 2.08 ±0.27 2.80 ±0.27 0.46 ±0.17 1.26 ±0.15 0.0022 ±0.0004 2021-05-19 00035344101 2.12 ±0.29 2.85 ±0.29 0.58 ±0.28 1.38 ±0.20 0.0027 ±0.0007 2021-05-24 00035344102 2.32 ±0.25 3.08 ±0.25 0.40 ±0.17 0.96 ±0.13 0.0020 ±0.0004 2021-05-29 00035344103 1.38 ±0.30 3.46 ±0.30 0.20 ±0.16 0.82 ±0.23 0.0008 ±0.0002 2021-06-03 00035344104 0.90 ±0.19 1.74 ±0.19 0.30 ±0.22 1.11 ±0.21 0.0007 ±0.0002 2021-06-08 00035344105 1.88 ±0.41 3.37 ±0.41 0.72 ±0.39 1.68 ±0.29 0.0031 ±0.0012 2021-06-13 00035344106 2.27 ±0.33 2.91 ±0.33 0.70 ±0.31 1.54 ±0.21 0.0035 ±0.001 2021-06-18 00035344107 1.47 ±0.18 2.28 ±0.18 0.25 ±0.15 0.95 ±0.15 0.0011 ±0.0002 2021-06-23 00035344108 3.00 ±0.28 2.78 ±0.28 0.66 ±0.19 1.42 ±0.13 0.0042 ±0.0008 2021-06-27 00035344109 2.43 ±0.27 2.57 ±0.27 0.51 ±0.19 1.35 ±0.15 0.0028 ±0.0006 2021-07-03 00035344110 2.29 ±0.36 3.52 ±0.36 1.10 ±0.36 1.70 ±0.21 0.0056 ±0.0017 2021-07-08 00035344111 2.60 ±0.29 3.03 ±0.29 0.46 ±0.20 1.19 ±0.15 0.0026 ±0.0006 2021-07-13 00035344112 4.03 ±0.39 3.66 ±0.39 0.64 ±0.19 1.46 ±0.14 0.0056 ±0.0011 2021-10-08 00035344113 3.58 ±0.37 3.42 ±0.37 0.55 ±0.17 1.43 ±0.13 0.0044 ±0.0008 2021-10-13 00035344114 3.61 ±0.30 3.21 ±0.30 0.51 ±0.13 1.27 ±0.11 0.0040 ±0.0006 2021-10-18 00035344115 3.55 ±0.45 4.91 ±0.45 0.80 ±0.27 1.43 ±0.17 0.0058 ±0.0014 2021-10-19 00035344116 4.61 ±0.85 8.54 ±0.85 0.77 ±0.38 1.43 ±0.25 0.0073 ±0.0025 2021-10-21 00035344117 2.99 ±0.52 5.37 ±0.52 0.70 ±0.30 1.38 ±0.22 0.0043 ±0.0013 2021-10-23 00035344118 2.59 ±0.37 3.20 ±0.37 0.74 ±0.26 1.55 ±0.19 0.0041 ±0.0011 2021-10-28 00035344119 4.01 ±0.36 3.38 ±0.36 0.65 ±0.18 1.49 ±0.13 0.0057 ±0.001 2021-11-02 00035344120 5.66 ±0.57 6.94 ±0.57 0.66 ±0.24 1.25 ±0.15 0.0075 ±0.0016 2021-11-07 00035344121 3.57 ±0.32 3.31 ±0.32 0.55 ±0.17 1.31 ±0.12 0.0042 ±0.0007 2021-11-12 00035344122 5.43 ±0.51 5.01 ±0.51 0.62 ±0.17 1.40 ±0.13 0.0073 ±0.0013 2021-11-17 00035344123 4.04 ±0.44 3.65 ±0.44 0.91 ±0.25 1.71 ±0.16 0.0083 ±0.0018 2021-11-22 00035344124 6.69 ±0.62 6.20 ±0.62 0.62 ±0.18 1.31 ±0.12 0.0085 ±0.0015 2021-12-02 00035344125 6.71 ±0.61 6.75 ±0.61 0.42 ±0.14 1.20 ±0.12 0.0065 ±0.001 2021-12-12 00035344126 6.67 ±0.76 7.98 ±0.76 0.55 ±0.20 1.26 ±0.14 0.0078 ±0.0015 2021-12-24 00035344128 3.50 ±0.35 3.85 ±0.35 0.35 ±0.15 1.10 ±0.12 0.0030 ±0.0005 2021-12-28 00035344129 4.30 ±0.30 3.90 ±0.30 0.65 ±0.20 1.35 ±0.20 0.0057 ±0.0003 Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022 The 2019/2021 X-ray monitoring campaigns 2993 MNRAS 514, 2974–2993 (2022) Table A2. Best-fitting parameters of the time-resolv ed XMM–Ne wton –NuSTAR analysis. F or each spectral slice, the corresponding ks are quoted. Fluxes are in units of erg cm −2 s −1 ×10 −11 while Gaussian normalization accounts for 10 −5 ph. cm −2 s −1 . The quoted normalization for power law and MyTorus are in units of ph. keV −1 cm −2 s −1 . Finally, N H and N MyT H are in units of 10 22 and 10 24 cm −2 , respectively. Slice F 0.3–2 keV F 2–10 keV N H  Norm po N MyT H Norm MyT Fe He αFe Ly αRed flare χ2 d.o.f 5 1.51 ±0.03 9.59 ±0.42 0.77 ±0.02 1.76 ±0.03 0.025 ±0.001 < 0.77 0.26 ±0.31 < 2.62 3.22 ±2.16 < 3.3 160 153 10 1.57 ±0.03 10.19 ±0.25 0.78 ±0.02 1.74 ±0.03 0.027 ±0.001 < 0.78 0.13 ±0.23 < 3.78 < 4.13 6.43 ±2.9 164 156 15 1.47 ±0.02 9.88 ±0.26 0.77 ±0.02 1.71 ±0.03 0.025 ±0.001 0.77 ±0.06 0.120 ±0.11 < 3.38 < 1.44 < 5.0 167 154 20 1.35 ±0.03 9.14 ±0.41 0.77 ±0.02 1.69 ±0.04 0.022 ±0.001 < 0.77 0.24 ±0.13 2.96 ±2.05 < 4.12 < 1.26 164 152 25 1.24 ±0.01 8.47 ±0.26 0.78 ±0.02 1.70 ±0.03 0.021 ±0.001 0.78 ±0.17 0.07 ±0.24 < 2.62 < 3.89 < 3.03 136 155 30 1.13 ±0.01 7.69 ±0.22 0.79 ±0.02 1.70 ±0.03 0.019 ±0.001 0.79 ±0.17 0.05 ±0.06 2.76 ±1.91 3.35 ±1.97 < 3.39 210 151 35 1.07 ±0.02 7.48 ±0.26 0.75 ±0.02 1.66 ±0.04 0.017 ±0.001 < 0.75 0.10 ±0.23 2.57 ±1.89 < 2.04 < 1.93 132 150 40 1.03 ±0.01 7.01 ±0.22 0.78 ±0.03 1.71 ±0.03 0.017 ±0.001 < 0.78 0.08 ±0.24 < 2.62 4.01 ±1.96 < 1.24 143 150 45 1.02 ±0.02 7.13 ±0.33 0.74 ±0.02 1.65 ±0.04 0.016 ±0.001 < 0.74 0.22 ±0.26 < 2.14 2.32 ±1.89 < 0.88 169 151 50 1.10 ±0.02 7.45 ±0.61 0.75 ±0.02 1.64 ±0.04 0.016 ±0.001 < 0.75 0.33 ±0.17 < 1.66 < 1.05 < 2.42 163 151 55 1.18 ±0.01 8.13 ±0.26 0.76 ±0.02 1.68 ±0.03 0.019 ±0.001 0.76 ±0.13 0.07 ±0.05 < 1.72 < 2.39 < 1.85 173 151 60 1.30 ±0.02 8.75 ±0.55 0.76 ±0.03 1.68 ±0.03 0.021 ±0.001 < 0.76 0.3 ±0.23 < 2.62 2.75 ±2.05 < 3.86 155 154 65 1.34 ±0.03 9.13 ±0.38 0.79 ±0.02 1.70 ±0.03 0.022 ±0.001 < 0.79 0.2 ±0.26 < 1.86 < 4.03 < 4.56 179 154 70 1.39 ±0.01 9.73 ±0.23 0.78 ±0.02 1.66 ±0.03 0.023 ±0.001 0.78 ±0.75 0.05 ±0.09 < 1.63 < 3.23 < 2.2 123 153 75 1.40 ±0.02 9.62 ±0.32 0.79 ±0.02 1.71 ±0.03 0.024 ±0.001 < 0.79 0.15 ±0.27 2.13 ±2.08 < 2.17 < 2.54 152 151 80 1.36 ±0.01 9.47 ±0.30 0.78 ±0.02 1.69 ±0.03 0.023 ±0.001 0.78 ±0.27 0.05 ±0.07 < 1.95 < 2.15 < 3.86 158 155 85 1.34 ±0.01 9.15 ±0.48 0.78 ±0.02 1.70 ±0.03 0.022 ±0.001 < 0.78 0.03 ±0.01 < 3.98 < 2.07 < 1.8 200 154 90 1.36 ±0.02 9.43 ±0.61 0.80 ±0.03 1.67 ±0.02 0.022 ±0.001 < 0.8 0.54 ±0.17 2.33 ±2.02 < 3.08 < 1.3 181 153 95 1.33 ±0.03 9.12 ±0.47 0.79 ±0.02 1.68 ±0.03 0.022 ±0.001 < 0.79 0.21 ±0.19 < 2.29 < 3.94 < 2.94 130 152 100 1.29 ±0.02 8.99 ±0.27 0.78 ±0.02 1.69 ±0.03 0.022 ±0.001 0.78 ±0.06 0.1 ±0.11 < 3.94 < 1.06 2.78 ±2.64 161 155 105 1.23 ±0.02 8.67 ±0.30 0.77 ±0.02 1.67 ±0.04 0.020 ±0.001 0.77 ±0.12 0.1 ±0.21 < 3.88 < 1.66 < 1.77 189 154 110 1.15 ±0.01 7.98 ±0.30 0.80 ±0.02 1.71 ±0.03 0.020 ±0.001 0.80 ±0.38 0.06 ±0.08 < 3.63 < 2.6 < 3.22 133 152 115 1.09 ±0.03 7.86 ±0.43 0.75 ±0.03 1.62 ±0.04 0.016 ±0.001 < 0.75 0.21 ±0.24 < 2.25 < 3.68 < 1.79 161 152 120 1.07 ±0.02 7.72 ±0.23 0.76 ±0.03 1.63 ±0.04 0.017 ±0.001 < 0.76 0.12 ±0.21 3.48 ±1.94 3.29 ±1.97 < 0.77 158 151 125 1.08 ±0.01 7.76 ±0.46 0.75 ±0.02 1.62 ±0.03 0.017 ±0.001 < 0.75 0.02 ±0.01 < 1.99 3.36 ±2.01 < 3.08 178 151 130 1.14 ±0.02 8.01 ±0.24 0.79 ±0.02 1.67 ±0.03 0.019 ±0.001 0.79 ±0.13 0.06 ±0.14 < 3.27 2.61 ±1.99 < 4.24 162 152 177 0.95 ±0.02 6.78 ±0.32 0.76 ±0.02 1.64 ±0.03 0.015 ±0.001 0.76 ±0.04 0.17 ±0.16 < 1.2 < 4.7 < 2.12 135 147 182 0.99 ±0.02 7.07 ±0.21 0.73 ±0.01 1.60 ±0.02 0.015 ±0.001 0.73 ±0.03 0.07 ±0.05 1.55 ±1.12 < 1.68 < 1.75 165 151 187 1.02 ±0.01 7.39 ±0.30 0.75 ±0.02 1.65 ±0.02 0.016 ±0.001 0.75 ±0.04 0.12 ±0.05 3.65 ±1.18 2.01 ±1.19 < 2.05 189 151 191 1.03 ±0.01 7.29 ±0.25 0.78 ±0.02 1.68 ±0.02 0.017 ±0.001 0.78 ±0.04 0.09 ±0.05 2.38 ±1.63 2.14 ±1.62 < 3.92 447 393 196 0.99 ±0.01 7.15 ±0.28 0.79 ±0.02 1.68 ±0.02 0.016 ±0.001 0.79 ±0.04 0.09 ±0.05 2.49 ±1.61 < 2.77 < 2.25 424 390 202 0.99 ±0.01 7.02 ±0.25 0.79 ±0.02 1.68 ±0.02 0.016 ±0.001 0.79 ±0.04 0.08 ±0.04 < 1.48 < 2.8 < 1.54 469 401 208 1.00 ±0.01 7.16 ±0.25 0.80 ±0.02 1.69 ±0.02 0.017 ±0.001 0.80 ±0.04 0.08 ±0.04 < 1.02 < 3.61 1.97 ±1.76 440 420 214 1.01 ±0.01 7.16 ±0.26 0.78 ±0.02 1.68 ±0.02 0.017 ±0.001 0.78 ±0.04 0.09 ±0.06 < 2.37 2.26 ±1.53 < 1.09 447 429 220 0.97 ±0.01 6.99 ±0.28 0.77 ±0.02 1.67 ±0.02 0.016 ±0.001 0.77 ±0.04 0.10 ±0.05 2.37 ±1.58 < 3.91 3.93 ±1.98 438 415 225 0.97 ±0.01 6.92 ±0.27 0.78 ±0.02 1.67 ±0.02 0.016 ±0.001 0.78 ±0.03 0.09 ±0.05 < 2.09 1.6 ±1.23 2.96 ±1.93 430 420 231 0.97 ±0.01 7.06 ±0.25 0.81 ±0.02 1.67 ±0.02 0.016 ±0.001 0.81 ±0.07 0.07 ±0.06 2.07 ±1.54 < 2.67 < 3.12 456 427 237 0.93 ±0.01 6.74 ±0.27 0.81 ±0.02 1.69 ±0.02 0.015 ±0.001 0.81 ±0.03 0.13 ±0.07 2.03 ±1.5 < 2.52 < 2.53 485 426 243 0.97 ±0.01 6.85 ±0.28 0.77 ±0.02 1.67 ±0.02 0.016 ±0.001 0.77 ±0.04 0.10 ±0.07 1.6 ±1.48 < 2.62 < 2.92 490 428 249 1.04 ±0.01 7.50 ±0.26 0.78 ±0.02 1.66 ±0.02 0.017 ±0.001 0.78 ±0.06 0.07 ±0.08 1.73 ±1.62 < 3.14 < 2.43 433 430 254 1.10 ±0.01 7.67 ±0.27 0.79 ±0.02 1.70 ±0.02 0.018 ±0.001 0.79 ±0.04 0.10 ±0.05 < 2.03 1.88 ±1.58 < 2.72 438 426 260 1.08 ±0.01 7.81 ±0.29 0.82 ±0.02 1.71 ±0.02 0.019 ±0.001 0.82 ±0.04 0.10 ±0.06 < 3.1 < 2.42 ±1.65 < 4.1 482 435 266 1.03 ±0.01 7.42 ±0.28 0.81 ±0.02 1.68 ±0.02 0.017 ±0.001 0.81 ±0.04 0.10 ±0.07 < 2.9 < 2.02 2.62 ±2.05 427 404 272 0.99 ±0.01 7.23 ±0.31 0.78 ±0.02 1.67 ±0.02 0.016 ±0.001 0.78 ±0.04 0.11 ±0.06 < 2.32 4.18 ±1.66 3.38 ±2.03 401 414 278 1.02 ±0.01 7.54 ±0.28 0.77 ±0.02 1.64 ±0.02 0.016 ±0.001 0.77 ±0.04 0.09 ±0.06 < 1.11 < 2.81 3.14 ±2.08 410 403 283 1.06 ±0.01 7.59 ±0.29 0.80 ±0.02 1.69 ±0.02 0.018 ±0.001 0.80 ±0.03 0.12 ±0.07 < 1.17 < 1.65 < 1.76 391 397 289 1.05 ±0.01 7.63 ±0.28 0.79 ±0.02 1.67 ±0.02 0.017 ±0.001 0.79 ±0.04 0.14 ±0.1 < 2.86 2.25 ±1.61 < 3.89 439 413 295 1.04 ±0.01 7.48 ±0.26 0.78 ±0.02 1.66 ±0.02 0.017 ±0.001 0.78 ±0.03 0.13 ±0.08 < 2.82 2.48 ±1.59 2.14 ±2.01 480 423 301 1.09 ±0.01 7.89 ±0.26 0.80 ±0.02 1.68 ±0.02 0.018 ±0.001 0.80 ±0.03 0.10 ±0.05 1.77 ±1.62 1.74 ±1.61 < 2.29 501 439 304 1.06 ±0.03 8.0 ±0.5 0.77 ±0.03 1.62 ±0.05 0.017 ±0.001 < 0.77 0.22 ±0.2 3.77 ±2.53 < 3.36 4.41 ±3.27 133 144 This paper has been typeset from a T E X/L A T E X file prepared by the author. Downloaded from https://academic.oup.com/mnras/article/514/2/2974/6588057 by INAF Brera Milano (Osservatorio Astronomico di Brera) user on 22 December 2022