EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 692 (December 2024) A&A, 692 (2024) A78 Full HTML
Open Access
Issue
A&A
Volume 692, December 2024
Article Number A78
Number of page(s) 11
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/202451688
Published online 03 December 2024

© The Authors 2024

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1. Introduction

Many supermassive black holes (SMBHs) in the center of active galactic nuclei (AGNs) accrete materials through a geometrically thin and optically thick disk (Shakura & Sunyaev 1973). The X-ray spectra of unobscured AGNs have three primary broadband components: a power-law-like continuum originating from the inverse Compton scattering of thermal (UV) disk photons in the central hot plasma (‘hot corona’, Sunyaev & Titarchuk 1980; Haardt & Maraschi 1993), a reflection component from the X-ray reprocessing of coronal emission in the accretion disk (Fabian et al. 1989; George & Fabian 1991; García & Kallman 2010) including a wealth of fluorescent lines (in particular a Fe K line at ∼6.4 keV, George & Fabian 1991; Ross & Fabian 2005) and a Compton hump (> 10 keV), and a blackbody-shape-like soft excess whose nature is still under debate, probably originating from a warm Comptonization (e.g., Done et al. 2012; Petrucci et al. 2018, 2020; Ballantyne et al. 2024) or relativistically blurred reflection (e.g., García et al. 2019; Xu et al. 2021), or the combination of two scenarios (Porquet et al. 2018, 2021; Xiang et al. 2022).

X-ray variability studies of unobscured AGNs have shown that the 0.3–1 keV band (referred to as the soft band), dominated by soft excess, and the Fe K band (5–7 keV), dominated by reflection, lag behind the continuum-dominated 1–4 keV band (referred to as the hard band) in the high-frequency domain (e.g., De Marco et al. 2013; Uttley et al. 2014; Kara et al. 2016). Despite the complicated soft excess origin (perhaps due to the reflection), the Fe K lag explicitly suggests light-travel distances between the corona and the accretion disk at a few gravitational radii, called X-ray reverberation lags (e.g., Fabian et al. 2009; Zoghbi et al. 2010; Kara et al. 2016). On the other hand, the low-frequency variability of unobscured AGNs commonly shows that the hard X-rays lag behind the soft emission with a time delay increasing steadily with energy. This low-frequency hard lag has been observed for several decades (e.g., Papadakis et al. 2001; Kara et al. 2013a,b, 2016). The timescales of the low-frequency X-ray lags are orders of magnitude longer than the viscous timescale of the inner accretion flow. The origin of this lag is still not well understood, though in the prevailing model is the multiplicative mass accretion fluctuation generated at large radii propagating inward to the inner accretion disk (Lyubarskii 1997; Kotov et al. 2001; Arévalo & Uttley 2006; Uttley & Malzac 2024). These time lags provide critical insights into the geometry and dynamics of the regions very close to the black hole, helping us to map out the configuration of the inner accretion environment.

Besides accreting a large amount of gas, AGNs can release part of the accumulated energy via the ejection of outflows (for review see Fabian 2012; Laha et al. 2021), some of which may have sufficient kinetic power (Lkin > 0.5 − 5%LEdd, LEdd is the Eddington luminosity) to impact the evolution of the host galaxy and growth of SMBHs (e.g., Di Matteo et al. 2005; Hopkins & Elvis 2010). Ultrafast outflows (UFOs) are an extreme subclass of AGN winds with mildly relativistic speeds (> 10 000 km/s), which are commonly discovered by strongly blueshifted absorption lines in X-ray spectra (Chartas et al. 2002, 2003, 2021; Pounds et al. 2003; Tombesi et al. 2010; Patrick et al. 2012; Gofford et al. 2013; Matzeu et al. 2023; Yamada et al. 2024; Xu et al. 2024). The ultrafast velocity suggests that UFOs are launched from the accretion disk within a few hundred gravitational radii close to the central SMBH (e.g., Tombesi et al. 2013), the same region probed by the X-ray spectral-timing analysis. Therefore, UFOs are expected to leave imprints on the X-ray variability of AGNs.

The variability study of UFOs has shown that UFOs produce characteristic positive “spikes” of enhanced variability in variance spectra, which exhibit the energy dependence of variability (Parker et al. 2017a, 2018, 2021; Igo et al. 2020; Härer et al. 2021). These features result from the UFO response to the continuum, where absorption lines from the UFO disappear as the X-ray flux rises (Parker et al. 2017b; Pinto et al. 2018), leading to a higher variability amplitude in energy bands affected by absorption than others dominated by the continuum.

The contribution of UFOs to the X-ray reverberation lags in the high-frequency domain was studied by Mizumoto et al. (2018, 2019), showing that the high-frequency Fe K lag may arise from the scattering in a highly ionized disk wind instead of the reflection onto the accretion disk. This alternative explanation could also reproduce the observed Fe K lags in 1H 0707-495 and Ark 564. However, the impact of UFOs on low-frequency hard lags in AGNs has not yet been investigated thoroughly. A transient UFO event was once observed by a large XMM-Newton campaign in PG 1211+143 (Pounds et al. 2016a,b), wherein Lobban et al. (2018) noticed that the low-frequency lag weakened in the presence of the UFO, which was explained by a change or disruption in the inner accretion flow before the reemergence or an additional soft lag introduced by other physical processes.

We would like to explore whether the coexistence of UFOs and weak low-frequency hard lags is common, and if so whether they are coincidental or causally related. To answer these questions, we have revisited the archival XMM-Newton dataset and selected two variable type 1 AGNs with transient UFOs; that is, PG 1448+273 and IRAS 13224-3809 (hereafter PG 1448 and IRAS 13224), to perform the X-ray lag analysis. The data of PG 1211+143 (hereafter PG 1211) have also been reanalyzed to reproduce the results in Lobban et al. (2018). The choice of XMM-Newton is because it has a higher orbit, and thus a longer orbital period (∼48 h), than other X-ray satellites, enabling us to probe the long-timescale variability of AGN. The targets are required to be type 1 AGNs to avoid the influence of large-scale obscuration (i.e., torus, Krolik & Begelman 1988; Netzer 2013) on X-ray variability and have at least two XMM-Newton observations with net exposures of > 30 ks with and without UFOs separately for comparison.

PG 1448 is a luminous (Lbol ∼ 1045.24 erg/s, Rakshit et al. 2020), nearby (z = 0.0645), narrow-line Seyfert 1 (NLS1) galaxy, hosting a SMBH of MBH ∼ 107 M⊙ (Vestergaard & Peterson 2006), indicating its accretion rate is close to the Eddington limit. PG 1448 was observed by XMM-Newton in 2017 (∼75 ks net EPIC-pn exposure) at a low-flux state (F2 − 10 keV = 1.3  ×  10−12 erg/s/cm2) with the presence of a UFO (∼0.09c), measured by the simultaneous detection of blueshifted soft X-ray and Fe K absorption lines (Kosec et al. 2020; Laurenti et al. 2021). Recently, PG 1448 was observed by NuSTAR for 130 ks net exposure (250 ks duration) joint with XMM-Newton for ∼60 ks net exposure. The observation was performed at a bright state (F2 − 10 keV = 4.8  ×  10−12 erg/s/cm2), revealing strong X-ray variability with a generally featureless X-ray spectrum. Although the last 60 ks of the NuSTAR observation (without a simultaneous XMM-Newton exposure) exhibited a flux decline and the emergence of a UFO, no UFO was detected during the XMM-Newton observation (Reeves et al. 2023).

IRAS 13224 is a nearby (z = 0.066, Allen et al. 1991), variable (e.g., Ponti et al. 2012) NLS1 galaxy, hosting an SMBH of MBH ∼ 2  ×  106 M⊙ (Zhou & Wang 2005; Alston et al. 2020). This target was extensively observed by XMM-Newton for a total of ∼2 Ms, including a ∼1.5 Ms campaign in 2016. The UFOs (∼0.236c) found in both the soft and hard X-ray bands of IRAS 13224 show a strong dependence on the source flux, where the strength of the absorption decreases as the flux increases, explained by the over-ionization of plasma (Parker et al. 2017b; Pinto et al. 2018). High-frequency soft lag and low-frequency hard lag were significantly detected (e.g., Ponti et al. 2010; Kara et al. 2013c; Hancock et al. 2022), where the reverberation lag is positively correlated with luminosity, indicating a variable corona height (Alston et al. 2020).

PG 1211 is a nearby (z = 0.0809, Marziani et al. 1996) and variable (e.g., Ponti et al. 2012) NLS1 galaxy or quasar, hosting a black hole of MBH ∼ 1.5  ×  108 M⊙ (Peterson et al. 2004). It is bright in both the optical and X-ray bands with an X-ray luminosity on the order of ∼1044 erg/s (Lbol ∼ 5  ×  1045 erg/s, Pounds et al. 2003). The source is an archetypal example of an AGN exhibiting a UFO in both UV (Kriss et al. 2018) and X-ray (e.g., Pounds et al. 2003, 2016a; Pounds & Reeves 2009; Danehkar et al. 2018; Reeves et al. 2018; Pounds & Nayakshin 2023) bands at velocities between ∼0.066 and 0.2c. The large XMM-Newton campaign in 2014 (∼630, ks) observed a complete UFO obscuration event from its emergence to its disappearance. The high-frequency reverberation lag and variable low-frequency hard lag were discovered from the same observations (Lobban et al. 2016, 2018).

The paper is organized as follows. Observations and data reduction used in this analysis are described in Sect. 2. Details of our Fourier analysis and results are shown in Sect. 3. Finally, we discuss the results and provide our conclusions in Sects. 4 and 5, respectively.

2. Data reduction

Basic information about the analyzed XMM-Newton observations of three AGNs is shown in Table 1. The observations are classified as those with or without UFOs according to the literature (Parker et al. 2017b; Pinto et al. 2018; Lobban et al. 2018; Reeves et al. 2023). To increase the signal-to-noise ratio (S/N), we stacked observations within the same class for our analysis.

Table 1.

XMM-Newton observations of three AGNs analyzed in this work.

For this analysis, we used the data from the XMM-Newton EPIC-pn camera because of its high effective area and fast readout (Jansen et al. 2001; Strüder et al. 2001). Data were reduced following standard threads with the XMM-Newton Science Analysis System (SAS v20.0.0) and calibration files available by February 2023. We reduced the EPIC-pn data using EPPROC and filtered the time intervals affected by the background solar flares, which show the count rates larger than 1 counts/sec in the 10–12 keV. This filter is higher than the regular threshold (0.5 count/sec), reducing interruptions in the extracted light curves caused by background flares, so that we can obtain a long continuous light curve to probe the low-frequency domain. We estimated the effect of the introduced background flares to be negligible by comparing results from this filter and those from the stricter filter with interpolated and Poisson-noise-added gaps. The net EPIC-pn exposure time and count rate are listed in Table 1. Meanwhile, for flares of < 500 s in duration, the source light curve was cut out and interpolated between the flare gap by adding Poisson noise using the mean of neighboring points. The interpolation fraction is typically < 1%. For gaps longer than 500 s, the data were treated as separate segments. The source and background regions were selected by a circular region with a radius of 30 arcsec centered on the source and offset but near it, respectively. The observation affected by the pile-up effect is marked by an asterisk in Table 1, and the corresponding source region has been replaced by an annulus with an inner radius of 5 arcsec and an outer radius of 30 arcsec. The background-subtracted light curves of AGNs were extracted by the EPICLCCORR package and shown in the top panels of Fig. A.1.

3. Results

Observations within the same class (i.e., with or without UFOs) were stacked, resulting in two stacked observations for each target, although the classification of IRAS 13224 observations is not explicit in the literature. However, given the flux-dependent UFO detection in IRAS 13224, UFOs are expected to be significantly detected in faint states and to become negligible in bright states. Therefore, the three brightest observations are classified as those without UFOs and the three faintest exposures are regarded as observations with UFOs. The presence or absence of UFOs in stacked spectra labeled by “low/high-flux w/wo UFO” are illustrated by the data-to-model ratios to the best-fit broadband continuum model (shown in Fig. 1), also justifying our expectation for UFOs in IRAS 13224. The continuum model consists of an absorbed power law, a soft excess, and a reflection component. The unfolded stacked EPIC-pn spectra with respect to a power law with an index of Γ = 2 of each AGN are shown in the first column of Fig. 2. In PG 1448 and IRAS 13224, the difference between stacked spectra with and without UFOs comes from both the continuum shape and the UFO absorption features (> 7 keV), while that difference in PG 1211 only occurs in the soft X-ray band, mainly dominated by the transient UFO.

thumbnail Fig. 1.

Data-to-model ratios of stacked EPIC (PG 1448 and IRAS 13224) and RGS (PG 1211) spectra to the best-fit broadband continuum model (see details in Sect. 3). The “high-flux w/o UFO” and “low-flux w/ UFO” results are marked by royal blue and black points, respectively. The energy band distinctly affected by the UFO is marked by the gray region.

3.1. Lag-frequency spectra

We used the package pylag1 (Wilkins 2019) to compute the X-ray time-lag spectrum (e.g., Uttley et al. 2014). Briefly, we performed the Fourier transform of the light curves in two different energy bands and multiplied the transform from the soft band and the complex conjugate of that from the hard band, called the cross-spectrum. In this paper, we take the conventional 0.3–1.0 keV as the soft band and 1–4 keV as the hard band, except for the 0.7–1.5 keV and 2–10 keV bands for PG 1211 for consistency with the literature (Lobban et al. 2018). The segment lengths are typically 50 ks for PG 1448 and 70 ks for IRAS 13224 and PG 1211, respectively, so that we can probe the lowest frequencies possible. The minimum frequency is defined by the inverse of the segment length, and the maximum frequency is set by the frequency at which the power spectral density (PSD, see details in Appendix A) becomes dominated by Poisson noise. The argument of the cross-spectrum is the phase lag between the two light curves. The time lag is τ(f) = ϕ(f)/2πf, where ϕ(f) is the frequency-dependent phase lag. The uncertainty of the time lag depends on the coherence of two light curves (see details in Uttley et al. 2014). We binned the lag-frequency spectrum into equal logarithmic frequency bins and the number of frequency bins depends on the quality of the data (i.e., more bins for higher-quality data).

The lag-frequency spectra of the stacked observations in each AGN are shown in the second column of Fig. 2. The black and royal blue arrows separately refer to frequencies, where the PSD of the highest energy bin (in lag-energy spectra) becomes dominated by Poisson noise, for the “low/high-flux w/wo UFO” observations. All lag-frequency spectra in Fig. 2 successfully reveal the high-frequency soft reverberation lag (e.g., Fabian et al. 2009; Kara et al. 2013c, 2016; De Marco et al. 2013) and low-frequency hard continuum lag. The soft reverberation lag in IRAS 13224 is not obvious due to its tiny lag amplitude (∼100 s), which is aligned with the literature (e.g., Kara et al. 2013c; Alston et al. 2020). The lag-frequency spectra of “low-/high-flux w/wo UFO” observations are more or less consistent within the uncertainties. However, it generally indicates that the “high-flux w/o UFO” observations have a larger low-frequency hard lag than “low-flux w/ UFO” observations.

thumbnail Fig. 2.

X-ray time lags for PG 1448 (top), IRAS 13224 (middle), and PG 1211 (bottom). The “low-/high-flux w/wo UFO” observations are marked in black and royal blue, respectively. The first column shows the unfolded stacked flux energy spectrum with respect to a power law of Γ = 2 of each AGN. The second column shows the lag-frequency spectrum between soft and hard energy bands. The arrows separately refer to the frequencies, at which the PSD of the highest-energy bin (in the lag-energy spectra) becomes dominated by Poisson noise, for the corresponding stacked observations. The third and fourth columns present the low- (third) and high-frequency (fourth) lag-energy spectra separately at particular frequencies, marked in the plots. The shaded blue regions in the low-frequency lag-energy spectra show the 1σ spread in lags from MC simulations (see details in Sect. 3.2).

3.2. Lag-energy spectra

The lag-energy spectra were computed in a similar way to the lag-frequency spectrum. The only difference is that, at this moment, the lag is calculated in a particular frequency range between a narrow band of interest and a large reference band. To obtain the best S/N, we chose the reference band to be the entire 0.3–10 keV, and removed the band of interest so that the noise was not correlated. The frequency ranges for the lag-energy spectra were chosen based on the lag-frequency spectra and marked in the plots. As a conservative approach, the upper bound of the high-frequency lag-energy spectrum was set by the frequency at which the PSD of the hardest energy band becomes dominated by Poisson noise (indicated by arrows in Fig. 2).

For observations unaffected by obscuration, the lag-energy spectra present an increasing lag with energy at low frequencies and a soft excess lag plus an iron K reverberation lag at high frequencies, which is consistent with previous results for IRAS 13224 and PG 1211 (Kara et al. 2013c; Lobban et al. 2018; Hancock et al. 2022). On the other hand, this is also the first time these typical AGN X-ray time lags have been detected in PG 1448. Regarding observations obscured by UFOs, we reproduced the same results of PG 1211 discovered in Lobban et al. (2018) that the low-frequency (1.4 − 3.5  ×  10−5 Hz) hard lag changes to flatter. Furthermore, we find that this phenomenon also appears in the other two AGNs, which is compatible with the results of lag-frequency spectra. In the high-frequency domain, the soft lags still exist, while the iron K reverberation lag seems to disappear or become weak. However, due to the smaller number of counts in the low-flux state, especially in the Fe K band, we cannot conclude that the origin of the disappearance or weakness of the Fe K reverberation lag is physical or just limited by data quality.

To quantify the significance of the deviation between the low-frequency lag-energy spectra of two states, we adopted the χ2 statistics by assuming one spectrum as the expected values and the other as the observed values. The calculated null hypothesis probability, P null Cal $ P_{\mathrm{null}}^{\mathrm{Cal}} $, and corresponding significance, σCal, are shown in Table 2, where all cases exhibit a deviation significance larger than 2.7σ. The deviation is also confirmed by Monte Carlo (MC) simulations. Following the method described in Timmer & König (1995), we used the GaussianResponse task in pylag package to simulate light curve pairs based on the observed PSD in each energy bin with a time lag described by a Gaussian response function. The Gaussian response function centers on the observed lag with the width of the lag uncertainty. The shaded blue regions in Fig. 2 represent the 1σ uncertainties in lag from 1000 MC simulated light curve pairs in each energy band, which are comparable with our calculated uncertainties (Uttley et al. 2014). The null hypothesis probability, P null MC $ P_{\mathrm{null}}^{\mathrm{MC}} $, and significance, σMC, of deviation were then calculated based on uncertainties obtained from MC simulations (see Table 2), where the significance is at least larger than 2.8σ. Our results show that the lag-energy spectra of the “low-flux w/ UFO” observations differ remarkably from those of the “high-flux w/o UFO” observations.

Table 2.

Significance of the deviation between low-frequency lag-energy spectra of two stacked observations in AGNs with a transient UFO.

4. Discussion

Through the spectral-timing analysis of archival XMM-Newton observations on three AGNs with transient UFOs, we find that suppression of low-frequency hard lags during low-flux UFO-affected states is not unique in PG 1211, since the other two AGNs manifest it with a transient UFO obscuration event. However, whether the suppressed continuum lags are related to UFOs is still unknown because UFOs and low-flux states commonly appear simultaneously. Perhaps the lower lag amplitude results purely from the low flux, related to the continuum variation, and X-ray time lags are flux-dependent, as was suggested by Alston et al. (2013).

4.1. Whether UFOs are causally related to suppressed low-frequency hard lags

We therefore searched for another three bright targets with large X-ray variability but without UFOs in archival XMM-Newton observations: Ark 564, NGC 7469, and Mrk 335. Their X-ray time properties have been well explored, and typical low-frequency hard lags and high-frequency iron K reverberation lags have been found in all of them (e.g., Kara et al. 2013a, 2016; Mastroserio et al. 2020; Lewin et al. 2022). It should also be noted that despite the absence of UFOs in these targets, slower and less ionized absorbers were discovered in all of them (Giustini et al. 2015; Longinotti et al. 2013; Khanna et al. 2016; Mehdipour et al. 2018; Grafton-Waters et al. 2020; Liu et al. 2021).

Following the same data reduction and analysis routines, we selected archival XMM-Newton observations of these three targets with distinct fluxes and stacked them into low-flux and high-flux subclasses. The basic information on selected observations is listed in Table B.1. The light curves in the EPIC-pn band and the corresponding PSD are shown in Fig. B.1, exhibiting frequencies above which the Poisson noise becomes dominant. The segment lengths are 50 ks for Ark 564, and 80 ks for NGC 7469 and Mrk 335, respectively. According to the literature, the origins of the flux variation among these targets are different: the variation in Ark 564 originates mainly from the variable normalization of the broadband continuum (Giustini et al. 2015); the variation in NGC 7469 results from the variable normalization of the soft excess component (Peretz et al. 2018; Middei et al. 2018); and the variation in Mrk 335 is contributed to by both the variable intrinsic continuum and ionized absorbers (Gallo et al. 2015; Komossa et al. 2020).

The stacked flux-energy, lag-frequency, and lag-energy spectra are shown in Fig. B.2. The details of the lag-frequency and high-frequency lag-energy spectra are described in Appendix B and, briefly speaking, the flux variability has a negligible influence on them. For the low-frequency lag-energy spectra, the typical hard lags that increase with energy are observed and are consistent with each other within spectra of two flux states for every AGN. The null hypothesis probability and the significance of deviation are listed in Table 3, where no case exhibits a deviation significance greater than 1.2σ. This behavior is quite different from previous results of AGNs with UFOs, suggesting that flux changes, originating from the continuum variations, do not affect the low-frequency hard lags, and the suppression of low-frequency hard lags might be linked to UFOs. We note a tentative deviation in the low-frequency lag-energy spectrum of Mrk 335, where the deviation occurs in the highest energy bin, leading to a deviation significance of ∼1σ. However, it should be noted that the absorbers in Mrk 335 have large velocities of ∼5000 to 7000 km/s (Longinotti et al. 2019), making them distinct from typical warm absorbers (WAs, v < 1000 km/s, Laha et al. 2021) but closer to UFOs (v > 10 000 km/s). They are therefore called obscuring outflows. If the hypothesis that the presence of UFOs can suppress the low-frequency hard lags is true, this weak indication may be related to the similarity between obscuring outflows in Mrk 335 and typical UFOs.

Table 3.

Significance of the deviation between low-frequency lag-energy spectra of two stacked observations in AGNs without UFOs (Similar to Fig. 3).

Similar fast obscuring outflows were also observed in NGC 3783 and Mrk 817 and expected from the outer disk or broad-line region. In De Marco et al. (2020), NGC 3783 shows negative low-frequency lags in obscured observations and positive low-frequency lags in unobscured exposures, which is consistent with what we observed in targets with variable UFOs. In Lewin et al. (2024), Mrk 817 exhibits low-frequency hard lags up to 800s when the source is partially obscured but relatively bright compared with other heavily obscured long XMM-Newton observations. As an SMBH with a mass of 3.85  ×  107 M⊙, the amplitude of its low-frequency lag is smaller than those at similar masses (e.g., PG 1448) and comparable with the results for IRAS 13224 (MBH ∼ 2  ×  106 M⊙), if we assume that low-frequency lags scale with SMBH masses. This indicates that obscuring outflows might have a moderate influence (between UFOs and WAs) on suppressing low-frequency hard lags, pointing toward the suppression strength being related to outflow properties (awaiting for further works).

As a result, by comparing X-ray time lags at different flux states of AGNs without UFOs, we find no evidence that the suppressed low-frequency hard lags originate from the faint continuum flux. The results of Mrk 335, combined with archival discoveries of Mrk 817 and NGC 3783, tentatively reveal that obscuring outflows with velocities close to those of UFOs might have a moderate influence on low-frequency hard lags. In fact, our discoveries are compatible with the results of NGC 4051 discovered by Alston et al. (2013), because some ionized winds in NGC 4051 are also close to UFOs with high velocities (v > 3000 km/s, Pounds & King 2013) and the major differences between the high- and low-flux X-ray spectra come from the varying ionized winds (Pounds et al. 2004).

4.2. Explanations for varying low-frequency hard lags

Based on our results and discussions, the suppressed hard continuum lags are likely related to the emergence of fast-moving outflows, not necessarily related to the source flux, while the underlying mechanism remains unclear.

4.2.1. Scenario I: Additional lags associated with outflows

One possible interpretation is that different physical processes dominate on different timescales. Silva et al. (2016) once investigated the NGC 4051 data using a time-dependent photoionization code to estimate the influence of outflowing ionized absorbers on X-ray time lags. They found that WAs in NGC 4051 can produce low-frequency soft X-ray lags, where the delay arises from the long radiative recombination time of WAs when the plasma varies in response to the ionizing continuum. The phenomenon of low-frequency soft X-ray lags was also observed in NGC 1365 and was explained by different light-crossing timescales of hard and soft X-ray photons in WAs (Kara et al. 2015). As a result, in this scenario, low-frequency X-ray lags comprise both signals of hard X-ray continuum lags related to the accretion flow and soft X-ray lags associated with the ionized absorbers. Similar effects might also arise in our results if soft X-ray UFOs (detected in all three targets) can enhance the time lags in the soft X-ray bands, leading to a seemingly flat lag-energy spectrum.

In general, UFOs are closer to SMBHs than WAs (e.g., Laha et al. 2021), and are expected to have a higher number density. Therefore, X-ray time lags introduced by the recombination of UFOs should have shorter timescales than those of WAs, since the recombination timescale is inversely proportional to the plasma number density (Krolik & Kriss 1995). The recombination timescales of UFOs were calculated using the rec_time tool in the SPEX package (Kaastra et al. 1996) with assumed number densities of ne = 1010 cm−3 for PG 1448 (Reeves et al. 2023), ne = 1011 cm−3 for IRAS 13224, and ne > 108 cm−3 for PG 1211 (Reeves et al. 2018). All densities were estimated with the definition of the ionization parameter, ξ = Lion/neR2, where the location of the outflow, R, was constrained by the transient timescale of the UFO, Δt, and the variation in the column density, ΔNH, reported in the literature, R5/2 = (GM)1/2(LionΔt/ΔNHξ). Recombination timescales are listed in Table 4. Our estimations show that delays from recombination are negligible and cannot contribute to the flat lag-energy spectra in the low-frequency domain.

Table 4.

Estimates of the recombination time, absorber size, and light-crossing time of UFOs in AGNs.

The other possible contributor is the delay resulting from different light-crossing timescales for soft and hard X-rays in UFOs. We estimated the light-crossing time by assuming an extreme situation: that UFOs are transparent to hard X-rays but optically thick to soft X-rays. The light-crossing time was then related to the absorber size, ΔR, which could be derived through ΔR = (GM/R)1/2Δt by assuming that the transverse velocity is equal to the Keplerian velocity. The resulting absorber size and light-crossing time are listed in Table 4. The light-crossing times are on an order of magnitude that is comparable with observed low-frequency lags, indicating that the additional lags associated with different light-crossing timescales of photons are promising for explaining the observed phenomenon. However, we caution that those values are generally several times lower than observed lags under a conservative condition, which weakens the likelihood of this mechanism.

4.2.2. Scenario II: Inhibited disk-corona energy transfer

Another intriguing possible scenario is that the energy transfer between the disk and corona is inhibited by outflows, preventing the propagation of accretion fluctuations if low-frequency hard lags originate from the inward-propagating mass accretion fluctuation generated in large radii (e.g., Uttley & Malzac 2024). Since UFOs, with their extreme velocities, are expected to originate from the inner accretion disk, their launching can carry away a portion of the inflowing materials from the inner disk. In this scenario, the reduction in lags results from the removal of the UV seed photons carrying fluctuations and/or the obscuration of seed photons by UFOs, leading to fewer fluctuation signals propagating into the hot corona. This explanation is supported by the differences between UFO-resolved coherence spectra at low frequencies, where the band choices are the same as those in Figs. 2 and B.2, shown in Fig. 3. In AGNs with transient UFOs (i.e., upper panels in Fig. 3), the entire coherence spectra of UFO-obscured observations are tentatively weaker than those without UFOs, although the results of PG 1448 are compatible within uncertainties due to limited exposures, while in AGNs without UFOs (lower panels), the coherence spectra at different fluxes remain consistent. These results suggest that UFOs can prevent the energy transportation between the disk and the hot corona, reducing time lags.

thumbnail Fig. 3.

Poisson noise-corrected coherence-energy spectra of UFO-(upper) or flux-resolved (lower) observations for targets analyzed in this work at low frequencies. In AGNs with transient UFOs (upper), the “low-flux w/ UFO” (black) coherence spectra are in general lower than the “high-flux w/o UFO” coherence spectra (royal blue), while the coherence spectra remain unchanged at different flux levels for AGNs without UFOs (lower).

In this scenario, X-ray lags in the high-frequency domain are also expected to be affected by UFOs because coronal X-ray emission is correlated with mass accretion (e.g., Lusso et al. 2010; Svoboda et al. 2017). The reduction in accretion mass by outflows would decrease the coronal X-ray emission, leading to weaker signals of X-ray reprocessing in the accretion disk. Our results seem to agree with this expectation, that in AGNs with UFOs the Fe K reverberation lags of absorbed observations are less significant than those in unobscured states (see Fig. 2), while they remain almost unchanged in AGNs without UFOs (see Fig. B.2). However, it is still unclear whether their disappearance is physically related to UFOs or depends on chosen frequency bands or specific observations, and the soft lags remain constant. In addition, it is still possible that the Fe K reverberation lags occur in absorbed states if the coronal variability, in some cases, has not started to be affected by the removal of propagating fluctuations generated in larger radii.

However, supporting this hypothesis requires a dedicated theoretical simulation, incorporating accretion flows with mass fluctuation propagation and outflows launched from the inner accretion disk. This work has not been accomplished yet and should be done in the future for comparison with observational results.

5. Conclusions

Working on archival XMM-Newton data, we have performed a UFO-resolved spectral-timing analysis of three AGNs with transient UFOs to investigate the potential influence of UFOs on X-ray time lags. As a result, we have shown that low-frequency hard lags become weak, accompanied by UFOs, in all three targets. Our investigations of another three variable AGNs without UFOs indicate that suppressed low-frequency hard lags might be unrelated to flux variability. Results of Mrk 335 with fast-moving obscuring outflows, combined with archival results in Mrk 817 and NGC 3783, even indicate an intermediate state of effects between UFOs and slow WAs. We propose an additional time delay associated with UFOs or the inhibited energy and variability transfer between the disk and corona for explanations. Our work indicates that any analysis of X-ray time lags in AGNs must account for the potential effects of ionized absorbers if they are present. The same investigation like this work should be extended to more sources, particularly with various black hole masses and accretion rates, to justify proposed explanations.

Acknowledgments

The authors thank Phil Uttley, Matteo Lucchini, Peter Kosec, and the anonymous referee for constructive suggestions. This research has made use of data obtained with the XMM-Newton, an ESA science mission funded by ESA Member States and the USA (NASA). YX, CP, and SB acknowledge support for PRIN MUR 2022 SEAWIND 2022Y2T94C (funded by NextGenerationEU) and INAF LG 2023 BLOSSOM. EK acknowledges the XRISM Participating Scientist Program for support under NASA grant 80NSSC20K0733. FT acknowledges funding from the European Union - Next Generation EU, PRIN/MUR 2022 (2022K9N5B4).

References

  1. Allen, D. A., Norris, R. P., Meadows, V. S., & Roche, P. F. 1991, MNRAS, 248, 528 [NASA ADS] [CrossRef] [Google Scholar]
  2. Alston, W. N., Vaughan, S., & Uttley, P. 2013, MNRAS, 435, 1511 [NASA ADS] [CrossRef] [Google Scholar]
  3. Alston, W. N., Fabian, A. C., Kara, E., et al. 2020, Nat. Astron., 4, 597 [NASA ADS] [CrossRef] [Google Scholar]
  4. Arévalo, P., & Uttley, P. 2006, MNRAS, 367, 801 [Google Scholar]
  5. Ballantyne, D. R., Sudhakar, V., Fairfax, D., et al. 2024, MNRAS, 530, 1603 [NASA ADS] [CrossRef] [Google Scholar]
  6. Bartlett, M. S. 1948, Nature, 161, 686 [NASA ADS] [CrossRef] [Google Scholar]
  7. Belloni, T., & Hasinger, G. 1990, A&A, 227, L33 [NASA ADS] [Google Scholar]
  8. Chartas, G., Brandt, W. N., Gallagher, S. C., & Garmire, G. P. 2002, ApJ, 579, 169 [Google Scholar]
  9. Chartas, G., Brandt, W. N., & Gallagher, S. C. 2003, ApJ, 595, 85 [Google Scholar]
  10. Chartas, G., Cappi, M., Vignali, C., et al. 2021, ApJ, 920, 24 [NASA ADS] [CrossRef] [Google Scholar]
  11. Danehkar, A., Nowak, M. A., Lee, J. C., et al. 2018, ApJ, 853, 165 [NASA ADS] [CrossRef] [Google Scholar]
  12. De Marco, B., Ponti, G., Cappi, M., et al. 2013, MNRAS, 431, 2441 [Google Scholar]
  13. De Marco, B., Adhikari, T. P., Ponti, G., et al. 2020, A&A, 634, A65 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  14. Di Matteo, T., Springel, V., & Hernquist, L. 2005, Nature, 433, 604 [NASA ADS] [CrossRef] [Google Scholar]
  15. Done, C., Davis, S. W., Jin, C., Blaes, O., & Ward, M. 2012, MNRAS, 420, 1848 [Google Scholar]
  16. Fabian, A. C. 2012, ARA&A, 50, 455 [Google Scholar]
  17. Fabian, A. C., Rees, M. J., Stella, L., & White, N. E. 1989, MNRAS, 238, 729 [Google Scholar]
  18. Fabian, A. C., Zoghbi, A., Ross, R. R., et al. 2009, Nature, 459, 540 [Google Scholar]
  19. Gallo, L. C., Wilkins, D. R., Bonson, K., et al. 2015, MNRAS, 446, 633 [NASA ADS] [CrossRef] [Google Scholar]
  20. García, J., & Kallman, T. R. 2010, ApJ, 718, 695 [CrossRef] [Google Scholar]
  21. García, J. A., Kara, E., Walton, D., et al. 2019, ApJ, 871, 88 [CrossRef] [Google Scholar]
  22. George, I. M., & Fabian, A. C. 1991, MNRAS, 249, 352 [Google Scholar]
  23. Giustini, M., Turner, T. J., Reeves, J. N., et al. 2015, A&A, 577, A8 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  24. Gofford, J., Reeves, J. N., Tombesi, F., et al. 2013, MNRAS, 430, 60 [Google Scholar]
  25. Grafton-Waters, S., Branduardi-Raymont, G., Mehdipour, M., et al. 2020, A&A, 633, A62 [EDP Sciences] [Google Scholar]
  26. Haardt, F., & Maraschi, L. 1993, ApJ, 413, 507 [Google Scholar]
  27. Hancock, S., Young, A. J., & Chainakun, P. 2022, MNRAS, 514, 5403 [NASA ADS] [CrossRef] [Google Scholar]
  28. Härer, L., Parker, M. L., Joyce, A., et al. 2021, MNRAS, 500, 4506 [Google Scholar]
  29. Hopkins, P. F., & Elvis, M. 2010, MNRAS, 401, 7 [NASA ADS] [CrossRef] [Google Scholar]
  30. Igo, Z., Parker, M. L., Matzeu, G. A., et al. 2020, MNRAS, 493, 1088 [NASA ADS] [CrossRef] [Google Scholar]
  31. Jansen, F., Lumb, D., Altieri, B., et al. 2001, A&A, 365, L1 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  32. Kaastra, J. S., Mewe, R., & Nieuwenhuijzen, H. 1996, in UV and X-ray Spectroscopy of Astrophysical and Laboratory Plasmas, 411 [Google Scholar]
  33. Kara, E., Fabian, A. C., Cackett, E. M., et al. 2013a, MNRAS, 434, 1129 [NASA ADS] [CrossRef] [Google Scholar]
  34. Kara, E., Fabian, A. C., Cackett, E. M., et al. 2013b, MNRAS, 428, 2795 [NASA ADS] [CrossRef] [Google Scholar]
  35. Kara, E., Fabian, A. C., Cackett, E. M., Miniutti, G., & Uttley, P. 2013c, MNRAS, 430, 1408 [NASA ADS] [CrossRef] [Google Scholar]
  36. Kara, E., Zoghbi, A., Marinucci, A., et al. 2015, MNRAS, 446, 737 [NASA ADS] [CrossRef] [Google Scholar]
  37. Kara, E., Alston, W. N., Fabian, A. C., et al. 2016, MNRAS, 462, 511 [NASA ADS] [CrossRef] [Google Scholar]
  38. Khanna, S., Kaastra, J. S., & Mehdipour, M. 2016, A&A, 586, A2 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  39. Komossa, S., Grupe, D., Gallo, L. C., et al. 2020, A&A, 643, L7 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  40. Kosec, P., Zoghbi, A., Walton, D. J., et al. 2020, MNRAS, 495, 4769 [CrossRef] [Google Scholar]
  41. Kotov, O., Churazov, E., & Gilfanov, M. 2001, MNRAS, 327, 799 [Google Scholar]
  42. Kriss, G. A., Lee, J. C., Danehkar, A., et al. 2018, ApJ, 853, 166 [NASA ADS] [CrossRef] [Google Scholar]
  43. Krolik, J. H., & Begelman, M. C. 1988, ApJ, 329, 702 [Google Scholar]
  44. Krolik, J. H., & Kriss, G. A. 1995, ApJ, 447, 512 [NASA ADS] [CrossRef] [Google Scholar]
  45. Laha, S., Reynolds, C. S., Reeves, J., et al. 2021, Nat. Astron., 5, 13 [NASA ADS] [CrossRef] [Google Scholar]
  46. Laurenti, M., Luminari, A., Tombesi, F., et al. 2021, A&A, 645, A118 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  47. Lewin, C., Kara, E., Wilkins, D., et al. 2022, ApJ, 939, 109 [Google Scholar]
  48. Lewin, C., Kara, E., Barth, A. J., et al. 2024, ApJ, 974, 271 [NASA ADS] [CrossRef] [Google Scholar]
  49. Liu, H., Parker, M. L., Jiang, J., et al. 2021, MNRAS, 506, 5190 [NASA ADS] [CrossRef] [Google Scholar]
  50. Lobban, A. P., Vaughan, S., Pounds, K., & Reeves, J. N. 2016, MNRAS, 457, 38 [NASA ADS] [CrossRef] [Google Scholar]
  51. Lobban, A. P., Vaughan, S., Pounds, K., & Reeves, J. N. 2018, MNRAS, 476, 225 [NASA ADS] [CrossRef] [Google Scholar]
  52. Longinotti, A. L., Krongold, Y., Kriss, G. A., et al. 2013, ApJ, 766, 104 [NASA ADS] [CrossRef] [Google Scholar]
  53. Longinotti, A. L., Kriss, G., Krongold, Y., et al. 2019, ApJ, 875, 150 [Google Scholar]
  54. Lusso, E., Comastri, A., Vignali, C., et al. 2010, A&A, 512, A34 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  55. Lyubarskii, Y. E. 1997, MNRAS, 292, 679 [Google Scholar]
  56. Marziani, P., Sulentic, J. W., Dultzin-Hacyan, D., Calvani, M., & Moles, M. 1996, ApJS, 104, 37 [Google Scholar]
  57. Mastroserio, G., Ingram, A., & van der Klis, M. 2020, MNRAS, 498, 4971 [NASA ADS] [CrossRef] [Google Scholar]
  58. Matzeu, G. A., Brusa, M., Lanzuisi, G., et al. 2023, A&A, 670, A182 [Google Scholar]
  59. Mehdipour, M., Kaastra, J. S., Costantini, E., et al. 2018, A&A, 615, A72 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  60. Middei, R., Bianchi, S., Cappi, M., et al. 2018, A&A, 615, A163 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  61. Mizumoto, M., Done, C., Hagino, K., et al. 2018, MNRAS, 478, 971 [NASA ADS] [CrossRef] [Google Scholar]
  62. Mizumoto, M., Ebisawa, K., Tsujimoto, M., et al. 2019, MNRAS, 482, 5316 [NASA ADS] [CrossRef] [Google Scholar]
  63. Netzer, H. 2013, The Physics and Evolution of Active Galactic Nuclei (Cambridge, UK: Cambridge University Press) [Google Scholar]
  64. Nowak, M. A., & Vaughan, B. A. 1996, MNRAS, 280, 227 [NASA ADS] [Google Scholar]
  65. Papadakis, I. E., Nandra, K., & Kazanas, D. 2001, ApJ, 554, L133 [NASA ADS] [CrossRef] [Google Scholar]
  66. Parker, M. L., Alston, W. N., Buisson, D. J. K., et al. 2017a, MNRAS, 469, 1553 [NASA ADS] [CrossRef] [Google Scholar]
  67. Parker, M. L., Pinto, C., Fabian, A. C., et al. 2017b, Nature, 543, 83 [Google Scholar]
  68. Parker, M. L., Reeves, J. N., Matzeu, G. A., Buisson, D. J. K., & Fabian, A. C. 2018, MNRAS, 474, 108 [Google Scholar]
  69. Parker, M. L., Alston, W. N., Härer, L., et al. 2021, MNRAS, 508, 1798 [NASA ADS] [CrossRef] [Google Scholar]
  70. Patrick, A. R., Reeves, J. N., Porquet, D., et al. 2012, MNRAS, 426, 2522 [NASA ADS] [CrossRef] [Google Scholar]
  71. Peretz, U., Behar, E., Kriss, G. A., et al. 2018, A&A, 609, A35 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  72. Peterson, B. M., Ferrarese, L., Gilbert, K. M., et al. 2004, ApJ, 613, 682 [Google Scholar]
  73. Petrucci, P. O., Ursini, F., De Rosa, A., et al. 2018, A&A, 611, A59 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  74. Petrucci, P. O., Gronkiewicz, D., Rozanska, A., et al. 2020, A&A, 634, A85 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  75. Pinto, C., Alston, W., Parker, M. L., et al. 2018, MNRAS, 476, 1021 [NASA ADS] [CrossRef] [Google Scholar]
  76. Ponti, G., Gallo, L. C., Fabian, A. C., et al. 2010, MNRAS, 406, 2591 [CrossRef] [Google Scholar]
  77. Ponti, G., Papadakis, I., Bianchi, S., et al. 2012, A&A, 542, A83 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  78. Porquet, D., Reeves, J. N., Matt, G., et al. 2018, A&A, 609, A42 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  79. Porquet, D., Reeves, J. N., Grosso, N., Braito, V., & Lobban, A. 2021, A&A, 654, A89 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  80. Pounds, K. A., & King, A. R. 2013, MNRAS, 433, 1369 [NASA ADS] [CrossRef] [Google Scholar]
  81. Pounds, K., & Nayakshin, S. 2023, MNRAS, submitted [arXiv:2310.18105] [Google Scholar]
  82. Pounds, K. A., & Reeves, J. N. 2009, MNRAS, 397, 249 [NASA ADS] [CrossRef] [Google Scholar]
  83. Pounds, K. A., King, A. R., Page, K. L., & O’Brien, P. T. 2003, MNRAS, 346, 1025 [Google Scholar]
  84. Pounds, K. A., Reeves, J. N., King, A. R., & Page, K. L. 2004, MNRAS, 350, 10 [NASA ADS] [CrossRef] [Google Scholar]
  85. Pounds, K. A., Lobban, A., Reeves, J. N., Vaughan, S., & Costa, M. 2016a, MNRAS, 459, 4389 [Google Scholar]
  86. Pounds, K., Lobban, A., Reeves, J., & Vaughan, S. 2016b, MNRAS, 457, 2951 [NASA ADS] [CrossRef] [Google Scholar]
  87. Rakshit, S., Stalin, C. S., & Kotilainen, J. 2020, ApJS, 249, 17 [Google Scholar]
  88. Reeves, J. N., Lobban, A., & Pounds, K. A. 2018, ApJ, 854, 28 [Google Scholar]
  89. Reeves, J. N., Braito, V., Porquet, D., et al. 2023, ApJ, 952, 52 [NASA ADS] [CrossRef] [Google Scholar]
  90. Ross, R. R., & Fabian, A. C. 2005, MNRAS, 358, 211 [CrossRef] [Google Scholar]
  91. Shakura, N. I., & Sunyaev, R. A. 1973, A&A, 24, 337 [NASA ADS] [Google Scholar]
  92. Silva, C. V., Uttley, P., & Costantini, E. 2016, A&A, 596, A79 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  93. Strüder, L., Briel, U., Dennerl, K., et al. 2001, A&A, 365, L18 [Google Scholar]
  94. Sunyaev, R. A., & Titarchuk, L. G. 1980, A&A, 86, 121 [NASA ADS] [Google Scholar]
  95. Svoboda, J., Guainazzi, M., & Merloni, A. 2017, A&A, 603, A127 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  96. Timmer, J., & König, M. 1995, A&A, 300, 707 [Google Scholar]
  97. Tombesi, F., Cappi, M., Reeves, J. N., et al. 2010, A&A, 521, A57 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  98. Tombesi, F., Cappi, M., Reeves, J. N., et al. 2013, MNRAS, 430, 1102 [Google Scholar]
  99. Uttley, P., & Malzac, J. 2024, MNRAS, submitted [arXiv:2312.08302] [Google Scholar]
  100. Uttley, P., Cackett, E. M., Fabian, A. C., Kara, E., & Wilkins, D. R. 2014, A&ARv, 22, 72 [NASA ADS] [CrossRef] [Google Scholar]
  101. van der Klis, M. 1989, in NATO Advanced Study Institute (ASI) Series C, eds. H. Ögelman, & E. P. J. van den Heuvel, Timing Neutron Stars, 262, 27 [NASA ADS] [Google Scholar]
  102. Vestergaard, M., & Peterson, B. M. 2006, ApJ, 641, 689 [Google Scholar]
  103. Wilkins, D. R. 2019, MNRAS, 489, 1957 [NASA ADS] [CrossRef] [Google Scholar]
  104. Xiang, X., Ballantyne, D. R., Bianchi, S., et al. 2022, MNRAS, 515, 353 [NASA ADS] [CrossRef] [Google Scholar]
  105. Xu, Y., García, J. A., Walton, D. J., et al. 2021, ApJ, 913, 13 [NASA ADS] [CrossRef] [Google Scholar]
  106. Xu, Y., Pinto, C., Rogantini, D., et al. 2024, A&A, 687, A179 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  107. Yamada, S., Kawamuro, T., Mizumoto, M., et al. 2024, ApJS, 274, 8 [NASA ADS] [CrossRef] [Google Scholar]
  108. Zhou, X.-L., & Wang, J.-M. 2005, ApJ, 618, L83 [NASA ADS] [CrossRef] [Google Scholar]
  109. Zoghbi, A., Fabian, A. C., Uttley, P., et al. 2010, MNRAS, 401, 2419 [Google Scholar]

Appendix A: Supplementary results of AGN with a transient UFO

The background-subtracted light curves are extracted by the package EPICLCCORR and are shown in the top panels of Fig. A.1. Observations are classified as the “low-flux w/ UFO” and “high-flux w/o UFO” states, marked in black and royal blue respectively.

thumbnail Fig. A.1.

EPIC-pn (0.3–10 keV) light curves (top) and PSD (bottom) of PG 1448 (left), IRAS 13224 (middle), and PG 1211 (right). The “low-/high-flux w/wo UFO” observations are marked by the black and royal blue, respectively.

For each AGN with UFOs, the stacked PSD spectra of each classification (in 0.3–10 keV energy band) were computed by the pylag package Bartlett 1948; van der Klis 1989; Uttley et al. 2014) and were normalized to units of fractional variance per hertz (Belloni & Hasinger 1990). The minimum frequency is defined by the inverse of the length of the observation (though we did not include the lowest-frequency bin as this is often biased by red noise leakage), and the maximum frequency is set by the frequency at which the PSD becomes dominated by Poisson noise. We binned the PSD in equal logarithmic frequency bins, and the number of frequency bins depends on the quality of the data (i.e., more bins for higher-quality data). The computed PSD functions of each AGN are shown in the bottom panels of Fig. A.1. Here, the Y-axis is in units of frequency*periodogram to illustrate the region dominated by Poisson noise, where the quantity of the Y-axis increases with frequency.

Appendix B: Supplementary results of targets without a transient UFO

The basic information on the selected AGN observations without transient UFOs is listed in Table 3, including the observation classification, observation IDs, net EPIC-pn exposure time, and the corresponding count rates. Observations in each AGN are classified as ‘low-flux’ and ‘high-flux’ states for the investigation discussed in Sect. 4.1 that whether the varying low-frequency hard lags are related to the source luminosity.

Table B.1.

XMM-Newton observations of the three AGN without a transient UFO mentioned in Sect. 4.

The background-subtracted light curves of AGN without transient UFOs and corresponding periodogram are shown in Fig. B.1, similar to Fig. A.1. The frequencies at which the Poisson noise becomes dominant can be obtained from our PSD spectra.

thumbnail Fig. B.1.

EPIC-pn light curves (top) and PSD (bottom) of Ark 564 (left), NGC 7469 (middle), and Mrk 335 (right). The “low-/high-flux” observations are marked in black and royal blue, respectively.

For each AGN, the stacked flux energy spectra of observations at different flux states are shown in the first column of Fig. B.2. The spectral variations are different among these AGN. The spectral variation of Ark 564 appears in the entire EPIC band without significant changes in the spectral slope, while that of NGC 7469 only occurs in the soft X-ray band, revealing that the soft excess mainly contributes to its spectral and flux variations. In Mrk 335, changes take place in both the spectral slope and spectral normalization, indicating that the variability originates from both broadband continuum variations and ionized absorbers (Gallo et al. 2015; Komossa et al. 2020).

thumbnail Fig. B.2.

X-ray time lags for Ark 564 (top), NGC 7469 (middle), and Mrk 335 (bottom).

The stacked flux-energy, lag-frequency, and lag-energy spectra are shown in Fig. B.2, while the low-frequency lag-energy spectra are discussed in detail in Sect. 3.2. The lag-frequency spectra of stacked observations at different flux states are generally consistent within uncertainties, with the exception that lags of NGC 7469 flip in the 4.4–10 × 10−5 Hz band, illustrated in its high-frequency lag-energy spectrum. In the high-frequency lag-energy spectra, the Fe K reverberation lag occurs in most cases regardless of the brightness, except for the high-flux state of Ark 564. We emphasize that the lags in Fig. B.2 are the stacked results, and the disappeared Fe K lags do not mean Ark 564 is special since some Fe K reverberation results are from specific observations, and not from all available observations (Kara et al. 2016). For example, when we calculate the lag-energy spectra within the same frequency band for individual high-flux observations, the Fe K lag only occurs in the observation performed on 01/07/2011, and it thus disappears in the averaged lags. The lags in the soft X-ray band are consistent with the lag-frequency spectra, where soft X-ray lags appear in nearly all spectra except for the high-flux spectrum of NGC 7469, the origin of which is beyond the scope of this work.

All Tables

Table 1.

XMM-Newton observations of three AGNs analyzed in this work.

In the text
Table 2.

Significance of the deviation between low-frequency lag-energy spectra of two stacked observations in AGNs with a transient UFO.

In the text
Table 3.

Significance of the deviation between low-frequency lag-energy spectra of two stacked observations in AGNs without UFOs (Similar to Fig. 3).

In the text
Table 4.

Estimates of the recombination time, absorber size, and light-crossing time of UFOs in AGNs.

In the text
Table B.1.

XMM-Newton observations of the three AGN without a transient UFO mentioned in Sect. 4.

In the text

All Figures

thumbnail Fig. 1.

Data-to-model ratios of stacked EPIC (PG 1448 and IRAS 13224) and RGS (PG 1211) spectra to the best-fit broadband continuum model (see details in Sect. 3). The “high-flux w/o UFO” and “low-flux w/ UFO” results are marked by royal blue and black points, respectively. The energy band distinctly affected by the UFO is marked by the gray region.
In the text
thumbnail Fig. 2.

X-ray time lags for PG 1448 (top), IRAS 13224 (middle), and PG 1211 (bottom). The “low-/high-flux w/wo UFO” observations are marked in black and royal blue, respectively. The first column shows the unfolded stacked flux energy spectrum with respect to a power law of Γ = 2 of each AGN. The second column shows the lag-frequency spectrum between soft and hard energy bands. The arrows separately refer to the frequencies, at which the PSD of the highest-energy bin (in the lag-energy spectra) becomes dominated by Poisson noise, for the corresponding stacked observations. The third and fourth columns present the low- (third) and high-frequency (fourth) lag-energy spectra separately at particular frequencies, marked in the plots. The shaded blue regions in the low-frequency lag-energy spectra show the 1σ spread in lags from MC simulations (see details in Sect. 3.2).
In the text
thumbnail Fig. 3.

Poisson noise-corrected coherence-energy spectra of UFO-(upper) or flux-resolved (lower) observations for targets analyzed in this work at low frequencies. In AGNs with transient UFOs (upper), the “low-flux w/ UFO” (black) coherence spectra are in general lower than the “high-flux w/o UFO” coherence spectra (royal blue), while the coherence spectra remain unchanged at different flux levels for AGNs without UFOs (lower).
In the text
thumbnail Fig. A.1.

EPIC-pn (0.3–10 keV) light curves (top) and PSD (bottom) of PG 1448 (left), IRAS 13224 (middle), and PG 1211 (right). The “low-/high-flux w/wo UFO” observations are marked by the black and royal blue, respectively.
In the text
thumbnail Fig. B.1.

EPIC-pn light curves (top) and PSD (bottom) of Ark 564 (left), NGC 7469 (middle), and Mrk 335 (right). The “low-/high-flux” observations are marked in black and royal blue, respectively.
In the text
thumbnail Fig. B.2.

X-ray time lags for Ark 564 (top), NGC 7469 (middle), and Mrk 335 (bottom).
In the text

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

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

Initial download of the metrics may take a while.