Ionized X-ray winds in the radio galaxy Mrk 6

Active galactic nucleus (AGN) outflows including jets and ionized winds have been key phenomena such as jet collimation and AGN feedback to the host galaxy in astrophysics. Radio galaxies, a type of AGN with misaligned jets, have provided valuable insights into the properties and relationships of these outflows. However, several aspects regarding AGN outflows remain unclarified, such as the relationship between jets and ultrafast outflows (UFOs) and the differences between the properties of radio-loud AGN disk winds and radio-quiet AGN ionized winds. To clarify these aspects, radio galaxies containing UFOs and warm absorbers (WAs) must be studied. Currently, both UFOs and WAs have been reported in only two radio galaxies, 3C 390.3 and 4C $+$74.26. To enhance our understanding of the connection between the jets and ionized winds, we conducted a study on Mrk 6, a potential candidate for the third case of a radio galaxy displaying these characteristics. Using X-ray spectra obtained from {\it XMM-Newton}, we performed photoionization modeling using the SPEX code. The best-fit model analysis results revealed the presence of a UFO component with a relatively low ionization parameter (Fe {\sc xix}/{\sc xviii} lines blueshifted by $-34700^{+400}_{-200}~{\rm km~s^{-1}}$) and a WA component with an outflow velocity of $-7600 \pm 200~{\rm km~s }^{-1}$. To further confirm the nature of the UFO and WA in Mrk 6, we simulated the X-ray imagining and spectroscopy mission spectra.

AGNs can be classified by radio intensity (or radio loudness).The radio intensity  is defined based on Kellermann et al. (1989) as  =  6 cm / opt using the radio flux ( 6 cm ) and the optical flux log  opt = (48.36− )/2.5 (Schmidt & Green 1983).For radio-loud AGN,  ≥ 10, and for radio-quiet AGN,  < 10.Although the spectral energy distribution (SED) of radio-loud and radio-quiet AGNs are similar across the infrared to X-ray spectrum, their radio and gamma radiations differ owing to the strong relativistic jets emitted by radioloud AGNs (e.g.Urry & Padovani 1995;Kayanoki & Fukazawa 2022).In principle, radio galaxies are radio-loud AGNs with jets misaligned to the line of sight, which facilitate the observation of the jets as well as the central core in such galaxies.Therefore, radio galaxies are deemed as the most suitable objects for investigating the environment surrounding SMBHs and characterizing the relationship between the jet and accretion disk and that between the jet and disk winds.Recently, several scholars have conducted statistical studies on radio galaxies, e.g., X-ray absorption in radio galaxies in Kayanoki & Fukazawa (2022), UFOs in radio galaxies in Tombesi et al. (2010a), Tombesi et al. (2012b), Gofford et al. (2013), Tombesi et al. (2013) and Tombesi et al. (2014), and WAs in radio galaxies in Reeves et al. (2009), Torresi et al. (2010), Torresi et al. (2012), Tombesi et al. (2016), andMehdipour &Costantini (2019).Notably, WAs are seldom observed in radio galaxies and might be related to the power of the jets (Mehdipour & Costantini 2019); however, a direct physical relationship has not yet been established.
Jets are presumed to be magnetically powered; however, the collimation of jets in the vicinity of SMBHs is not understood.A quasianalytical study suggested that disk winds may promote the collimation of jets near SMBHs (Fukumura et al. 2014;Globus & Levinson 2016;Blandford et al. 2019).Moreover, as reported by Mehdipour & Costantini (2019), WAs are related to the jet intensity (radioloudness).However, several questions regarding the AGN outflows remain unsolved, such as the possible variations in the properties of disk winds in radio-loud AGNs and radio-quiet AGNs, including the relationship between jets and UFOs.To resolve these questions, the AGN with observed jets, UFOs, and WAs, i.e., radio galaxies containing both UFOs and WAs, provide vital insights.However, only two radio galaxies have been reported to contain both UFOs and WAs, 3C 390.3 and 4C +74.26 (Tombesi et al. (2014) for UFOs, Mehdipour & Costantini (2019) for WAs).Therefore, understanding the physical state of the disk winds in radio galaxies is crucial for resolving these problems.Mrk 6 is the Seyfert 1.5-type AGN in a lenticular galaxy (Kharb et al. 2014).-band magnitude and 6 cm radio flux ( 6 cm ) of Mrk 6 is  = 15.16 and  6 cm = 0.115 Jy (Véron-Cetty & Véron 2010), Mrk 6 can be classified as a radio-loud galaxy ( = 30).Specifically, it has a black hole mass of (1.8±0.2) ×10 8  ⊙ (Doroshenko et al. 2012) with  = 0.019 (de Vaucouleurs et al. 1991).In Mehdipour & Costantini (2019), the X-ray data analysis of the European Photon Imaging Camera (EPIC, Strüder et al. 2001;Turner et al. 2001) and Reflection Grating Spectrometer (RGS, den Herder et al. 2001) detectors on the XMM-Newton satellite for 16 radio galaxies reported WAs, among which Mrk 6 exhibited the highest density of WAs.Furthermore, Schurch et al. (2006) reported that Mrk 6 contains disk winds or an ionized skin or an atmosphere of the molecular torus.However, as the UFO of the Mrk 6 has not yet been discussed, this study analyzed the Mrk6 disk winds in more detail.The cosmological parameters ( 0 , Ω  , Ω Λ ) used in this study are  0 = 70 km s −1 Mpc −1 , Ω  = 0.3, Ω Λ = 0.7.

OBSERVATION AND DATA REDUCTION
Mrk 6 was observed with XMM-Newton in March 2001, April 2003, and October 2005 (Table 1).In this study, we used the RGS data for the soft X-ray spectra and the positive-negative junction (pn) chargedcoupled device (CCD) camera (Strüder et al. 2001) of the EPIC for the hard X-ray spectra, because EPIC/MOS data does not significantly increase the statistics.The XMM-Newton data were reprocessed and analyzed using Science Analysis System version 19.1.0and the calibration data from current calibration files (CCF) were updated on November 18, 2021.After eliminating the data at instances unsuitable for analysis because of background flares, the exposure durations of the EPIC detector were shorter than those of the RGS detector data (Table 1).Moreover, the observations (IDs 0061540101 and 0305600501) were unsuitable for disk wind analysis, because their extremely short exposure durations produce excessively large errors in the spectral data points.Herein, the analysis was performed For the EPIC/pn data reduction, the epchain pipeline task was used.In EPIC/pn (> 0.50 counts s −1 ), periods with high flaring were filtered out by applying the #XMMEA_EP filters.The EPIC/pn spectra were extracted from a circular region of the center coordinate (06:52:11.6,+74:25:35.1)with a radius of 35.750 arcsec.Moreover, the background was extracted from a neighboring source-free rectangle region of center coordinates (06:52:05.0,+74:22:05.9),length and width of 96 arcsec and 262 arcsec, respectively, and a tilt of 342.29 • in the same CCD.Overall, the pile-up was evaluated to be negligible.Thereafter, the response matrices were generated for the EPIC/pn spectra of each observation using the rmfgen and arfgen tasks.For the RGS data, the rgsproc pipeline task was used for data reduction, and the source and background spectra were extracted to generate the response matrices.Moreover, the time intervals with background count rates > 3 were filtered out with the Chandra Interactive Analysis of Observations (CIAO, Fruscione et al. 2006) version 4.14 lc_sigma_clip task.To rectify the cross-calibration between the EPIC/pn and RGS spectra, the RGS1 and RGS2 spectra were rescaled by a factor of 0.91 and 0.89, respectively, to correspond to the flux level in the 1.2 − 1.3 keV energy range.

SPECTRAL FITTING AND RESULT
In this study, the spectral analysis was performed using the SPEX code v3.06.01 (Kaastra et al. 2020).For spectral analysis, EPIC/pn spectra were used in the energy range of 1.77−10.0keV, and the obin command of the SPEX code was applied to perform optimal binning according to Kaastra & Bleeker (2016).The two RGS spectra were utilized for the 7.0 − 27.0 Å wavelength range and binned by a factor of 2 to ensure that the bin size was approximately 3rd of the RGS energy resolution.In particular, the spectral fitting was performed on the RGS and EPIC/pn spectra using the C-statistic (Kaastra 2017).In all the fittings, performed hereinafter, all models were multiplied by the neutral absorption model (hot model) to account for the absorption by our Galaxy.The column density of our Galaxy was set to  H = 7.63 × 10 20 cm −2 (HI4PI Collaboration 2016).For the reflection and absorption models, we considered the protosolar abundances of Lodders et al. (2009).The error for each parameter refers to the 68.3% confidence level.

Search for emission and absorption lines
First, to inspect the emission and absorption lines in the spectra, we analyzed the EPIC/pn spectra in the 5 − 10 keV energy band using the power-law (pow) model, the power-law (pow) + emission line model, and the power-law (pow) + emission line (gaus) + absorption line (gaus) model (Table 2).The EPIC/pn spectra in the 5−10 keV energy band and the power-law (pow) + emission line (gaus) + absorption Fit the EPIC/pn hard X-ray data (5-10 keV).ΔC-stat: Difference in C-statistics between the simple power-law model.gauss model used for obtaining emission and absorption lines.(1) Photon index of the X-ray power-law component.( 2) Normalization of the X-ray power-law component in 10 51 photons s −1 keV −1 at 1 eV.
5.5 6.0 6.5 7.0 7.5 8.0 Observed Energy (keV) line model (gaus) are illustrated in Figure 1.The Fe emission line at 6.4 keV and absorption line at ∼ 7 keV were observed in Figure 1.
In addition to Table 2, the model with the absorption lines displayed a more accurate C-statistic (Δ = −5.9, this would be a marginal detection) in Figure 1, thereby suggesting the existence of UFOs.The emission line at 6.4 keV was considered to be fluorescence X-rays owing to the neutral or low ionized iron in AGN, whereas the ∼ 7 keV absorption lines were considered to be caused by a highly ionized gas.

Continuum
Furthermore, we investigated the features (photon index and flux) of the continuum component of the X-ray spectra.As the high-energy band is considered to be weakly affected by absorption, the EPIC/pn data were fitted with a model in which the power law was multiplied by the absorption (hot) by the neutral gas of the host galaxy.In addition, as the 6.4 keV Fe emission lines possibly existed based on the results of the previous subsection, the power-law (pow) + reflection component (refl) multiplied by the absorption (hot) of the neutral gas of host galaxy was fitted to the EPIC/pn data.Moreover, the RGS and EPIC/pn data were fitted with a model that multiplied the modified blackbody component (mbb) + power-law (pow) + reflection component (refl) by the absorption (hot) of the neutral gas in the host galaxy.
Compared to the power-law model with the power-law (pow) + reflection component (refl) model, the C-statistic for the total spectrum (EPIC/pn + RGS spectrum) is more suitable for the power-law model without the reflection component (Δ C = 100.5).Nonetheless, the C-statistic for the EPIC/pn data spectrum is more appropriate for the model with the reflection component.A comparison of the C-statistics of the EPIC/pn data spectra indicated the considerably superior performance of the model incorporating the reflective component.This is because the reflection model can adequately explain the 6.4 keV Fe emission lines.However, the power-law (pow) + reflection component (refl) model does not adequately represent the RGS (soft X-ray spectrum).Accordingly, the modified blackbody component (mbb) was added to the continuum.As listed in Table 3, the C-statistic for the total spectrum (EPIC/pn + RGS spectrum) was 2286.8, which is a relatively more accurate result.Based on the aforementioned findings, we adopted the modified blackbody model (mbb) + power-law (pow) + reflection (refl) as the continuum component.In the following subsection, we investigated the suitability of the neutral gas hot model (de Plaa et al. 2004;Steenbrugge et al. 2005) and the ionized gas pion model (Miller et al. 2015;Mehdipour et al. 2016) as the absorbed component.

Absorption component
We investigated whether the neutral gas (hot) model or the ionized gas (pion) model (disk wind model) is appropriate as the absorber.As discussed in Section 3.2, the M1 model multiplying the neutral gas (hot) model to the continuum did not explain the ∼ 7 keV absorption line.Therefore, we adopted the photoionization model pion.The photoionization equilibrium of the pion model is calculated self-consistently using the available plasma routines of SPEX 1 .Therefore, the pion model could calculate the SED and SED filtering (the second PION sees a filtered SED) with the photoionization equilibrium simultaneously.However, the photoionization equilibrium of the other photoionization model (e.g. the xabs model) is pre-calculated for a grid of -values by an external code, for instance, Cloudy or XSTAR.For the pion model, the hydrogen density were set to 1 cm −3 and the broadening velocity were set to 100 km s −1 .These two values are the default values of the pion model.
First, we added a single pion component to the M1 model (M2 model).Compared to the M1 model, the M2 model improved the C-statistic.However, the ionizing absorption component of the M2 model exhibited soft X-ray absorption and did not represent the ∼ 7 keV absorption line.Subsequently, the neutral gas (hot) component in the M2 model was modified to an ionized gas (pion) component (M3 model).Initially, the M3 model was assumed to contain one neutral gas (hot) and two ionized gases.As the column density for the hot model was obtained as zero from the fitting results, the M3 model potentially did not require a neutral gas (hot).Compared to the M2 model, the M3 model improved the C-statistic by 61.7.The M3 model exhibited the soft X-ray absorption (Figure 4) and ∼ 7 keV absorption lines (however, the ∼ 7 keV absorption lines were weak (Figure 5).Although the >7 keV absorption lines of UFOs are typically considered to be Fe xxv/xxvi absorption lines (e.g.Tombesi et al. 2014).However, the M3 model has a low ionization parameter and it ascribed the ∼ 7 keV absorption lines to Fe xix/xviii.Therefore, the M3 model has weak absorption lines.Thus, the parameters of the M3 model were modified and fitted again.The obtained result may be a local minimum (M4 model).Although the M4 model more accurately represented the ∼ 7 keV absorption lines (Figure 5), its C-statistic of 36.6 was inferior to that obtained with the M3 model.However, the ∼ 7 keV absorption lines can be possibly attributed to the Fe xxv/xxvi absorption lines, the spectra obtained herein suggested that the ∼ 7 keV absorption lines are Fe xix/xviii.As such, the data from the X-ray imagining and spectroscopy mission (XRISM) can be utilized to verify whether the ∼ 7 keV absorption lines in Mrk 6 are caused by Fe xxv/xxvi or Fe xix/xviii.The fitting results for the M1, M2, M3, and M4 models are summarized in Table 3.
At this point, we tried thawing the covering factor and Fe abundance of the UFO component for both the M3 model and the M4 model.First, we thawed the covering factor (C f ) of the UFO component for the M3 model and the M4 model.For the M3 model, the covering factor changed to C f = 0.7 and the C-stat improved significantly (Δ = −4.3).On the other hand, for the M4 model, the covering factor did not change significantly.Therefore, we thawed the covering factor model for the M3 model.Next, we thawed the Fe abundance of the UFO component for the M3 model and the M4 model.For the M3 model, the Fe abundance did not change significantly.On the other hand, for the M4 model, the Fe abundance changed to 3.3 solar and the C-stat improved significantly (Δ = −4.7).As a result, the M3 model has the lowest C-stat value (Table 3) among the above 4 cases.Therefore, the M3 model with a covering factor of 0.7 was deemed to be the best model.
To investigate whether the M3 model and M4 model are not a local minimum, we varied the value of outflow velocity in the range 0 − 0.35  while performing the fitting and checked the C-stat.In this test, both the M3 model and the M4 model look not local minimum.(We could not determine whether the M4 model is just a local minimum or not.)In addition, we made contour maps for the M3 model of the UFO component (Fig 2).We did the likelihood ratio test to estimate the statistical confidence level of the M3 model with PN spectra.The likelihood ratio is (PN C-stat of the M2 model) − (PN C-stat of the M3 model) = 183.0− 166.8 = 16.2.PN spectra have 162 bins, and then we decided that we could use the likelihood ratio test for PN spectra and approximate the test following the  2 distribution.Since the difference in degrees of freedom between the M2 model and the M3 model is 3, the -value of the M3 model should be a  < 0.001.From the above, the confidence level of the M3 model is > 99.9%.
In addition, to investigate the number of layers of the ionized gas, the M3 model was multiplied by an additional component of the ionized gas (pion) model (M5 model).Similar to the M3 model, the column density of the neutral absorption component was zero in this model.Thus, the neutral absorption (hot) component was not required in the M5 model.Compared to the M3 model, the M5 model improved the C-statistic by only 7.9 with the 3 degrees of freedom.Thus, the third ionized gas (pion) was regarded as insignificant.(Even if the ionized gas (pion) is included within the M4 model was not significant.) According to the results obtained by the best model (M3 model), an ionized gas exhibited high ionization energy of log  (erg cm s −1 ) = 2.1 +0.5 −0.1 , high column density of  H = 3.5 +1.6 −0.9 × 10 22 cm −2 , and an extremely fast outflow velocity of  out = −34700 +400 −200 km s −1 .Thus, this ionized gas component was identified to be a UFO.The other ionized gas has a lower ionization parameter of log  (erg cm s −1 ) = 0.51 +0.19  −0.15 , lower column density of  H = 8.1 +0.8 −0.6 × 10 21 cm −2 and a lower outflow velocity of  out = −7600 ± 200 km s −1 .Thus, this ionized gas component was deemed as a WA.
We evaluated the absorption lines of UFOs and WAs.The UFO absorption lines above 6.0 keV with  > 2 eV and soft X-ray WA absorption lines with  > 2 eVare illustrated in Figure 7 and Figure 8, respectively.In addition, the information on UFO absorption is likely better identified with presented in Table 4.

UFO
As reported in prior research, 3C 390.3 and 4C +74.26 are the only two galaxies where both UFOs and WAs have been observed, with Mrk 6 being the third case considered in this study (Table 5).Compared to the Mrk 6 UFOs with 3C 390.3 and 4C +74.26, the ionization was lower, and the column density and outflow velocity are comparable (Table 5).Although, both 3C 390.3 and 4C +74.26 have lower limits for the column density (Tombesi et al. 2014).
The ionization parameter is defined as  =  ion / H  2 (Tarter et al. 1969) using the hydrogen density  H , disk wind distance , and luminosity of the light source  ion .Thus, the distance of the disk wind was calculated as Based on the M3 model result, the UFO ionization parameter was log 10  = 2.1 +0.5 −0.1 erg cm s −1 and the non-filtered (before getting absorbed by the UFO pion component) ionized luminosity was calculated to  ion / = 3.2 × 10 41 cm −1 .Therefore, the distance of the UFO from the SMBH could be estimated as This is equivalent to 95  H 10 8 cm −3 −0.5  S , where the Schwarzschild radius  S = 2  BH / 2 = 5.3 × 10 13 cm.Assuming that the wind has a higher velocity than the escape velocity.Therefore, the lower limit of the distance of the disk wind was evaluated as using the gravitational constant , SMBH mass  BH , and outflow velocity  out .Based on the analysis results, the outflow velocity of the UFO was  out = 34700 km s −1 , and therefore, the lower limit of the UFO distance was calculated as  min = 4.0 × 10 15 cm (= 75  S ).
Assuming that the disk wind width Δ did not exceed the distance  to the SMBH, the upper limit of the disk wind distance was computed as    For the EPIC/pn spectra, we used the obin command for optimal binning.RGS spectra were combined by rgscombine command and the binning factor was set to 5.
considering the covering factor  v of the disk wind (Wang et al. 2022).M3 model assumed a covering factor of  v = 1, such that the upper limit of the UFO distance was  max = 7.2 × 10 16 cm (= 1.3×10 3  S ), considering a column density of  H = 3.5×10 22 cm −2 .
Given the observational and model uncertainties, here you can say that the range is consistent with what expected from disk winds.Furthermore, we evaluated the range of hydrogen density from that of UFOs.Based on the definition of ionization, the density of   the disk wind can be evaluated as (5) The range of the UFOs was 4.0 × 10 15 − 7.2 × 10 16 cm, and thus, the range of the hydrogen density was calculated as 4.8 × 10 5 − 1.6 × 10 8 cm −3 .

WA
WAs are often reported to contain multiple components, e.g., NGC 3516 has 3 WAs (Mehdipour et al. 2010), NGC 7469 has 4 WAs (Mehdipour et al. 2018), NGC 5548 has 6 WAs (Mao et al. 2017), NGC 3783 has 9 WAs (Mao et al. 2019) and NGC 3227 has 4 WAs (Wang et al. 2022).Both 3C 390.3 and 4C +74.26 have two WAs components each (Mehdipour & Costantini 2019).However, for Mrk 6, the second WA component is not significant.Thus, Mehdipour & Costantini (2019) used only one WA component along with a neutral absorption component.Compared to the Mrk 6 WAs with 3C 390.3 and 4C +74.26, the ionization was low and the outflow velocity was high, however, the column densities did not vary significantly (Table 6).
Based on the results of the analysis with the M3 model, the ionization parameter of WA was log 10  (erg cm s −1 ) = 0.51 +0.19  −0.15 and  which is equivalent to 1.8  H 10 5 cm −3 −0.5 pc.The fitting results exhibited that the outflow velocity of the WA was  out = 7600 km s −1 , and the lower limit of the WA distance was  min = 8.3 × 10 16 cm (= 2.7 × 10 −2 pc) using Equation (3).As the column density of the WA was  H = 8.1 +0.8 −0.6 × 10 21 cm −3 according to the analysis and the coverage factor was assumed to be  v = 1, the upper limit of the WA distance was calculated to be  max = 3.9 × 10 20 cm (= 1.3 × 10 2 pc) using Equation (4).Therefore, the WAs of Mrk 6 originated at 1.8  H 10 5 cm −3 −0.5 pc, and the range of the WAs was 2.7 × 10 −2 − 1.3 × 10 2 pc.WAs are considered to originate in the 0.1 − 1000 pc (Laha et al. 2021), and the present findings are consistent with this result.
Similar to the case with UFOs, the hydrogen density of the WA was calculated as well.The WA distance range was calculated to be 8.3 × 10 16 − 3.9 × 10 20 cm, such that the hydrogen density of WA was calculated as 2.1 × 10 1 − 4.6 × 10 8 cm −3 using Equation ( 5).The hydrogen number density of WAs was estimated to be ∼ 10 4 − 10 11 cm −3 (Laha et al. 2021, and references therein), and the present result was wider than the range.

XRISM simulation
As the EPIC/pn instrument of XMM-Newton exhibited a low energy resolution, the absorption lines could not be clearly observed (Figure 3) and the ionized wind parameters (ionization parameter, hydrogen column density, and outflow velocity) yielded large errors (Table 3).However, XRISM/Resolve (Tashiro et al. 2020; XRISM Science Team 2022) provides an ultrahigh energy resolution, which facilitates the prominent observation of the absorption line and reduces the uncertainties of the parameters and XRISM to verify whether Mrk 6 contained only one WA component.Thereafter, we simulated the observation in XRISM/Resolve based on the M3 model, which was determined to be the best-fit model.The XRISM/Resolve response 6.6 6.7 6.8 6.9 7.0 7.1 7. file was used with an energy resolution of 7 eV version 2019a 2 .For the simulations and fitting, the SPEX code was used, and the exposure time was set to 150 ks.In the analysis, the energy range was set to 0.34 − 10.0 keV.For comparison, we simulated the M4 model to obtain results with Fe xxv/xxvi absorption lines at ∼ 7 keV.
Based on the findings listed in Table 7, the errors of the disk wind parameters in both the M3 and M4 models were smaller than those reported in Table 3.In particular, the error in the column density  H , which is a vital parameter for the significant detection of the disk wind, was reduced.The ∼ 7 keV absorption lines can be more prominently observed (Figure 9).In addition, the soft X-ray spectra error bar was more accurate (Figure 10).Therefore, XRISM/Resolve can more accurately detect the disk wind.As observed in Figure 9

SUMMARY
This study analyzed the disk wind of the radio galaxy Mrk 6 using XMM-Newton archival data with the SPEX code.Based on the results of the M3 model, the existence of both UFOs and WAs was suggested in Mrk 6.In addition to the radio galaxies 3C 390.3, 4C +74.26,Mrk 6 is the third radio galaxy in which both UFOs and WAs have been observed.Compared to the Mrk 6 disk wind with 3C 390.3 and 4C +74.26, the UFOs exhibited low ionization, however, the UFO and WA parameters are broadly consistent with the ones reported for the other two radio galaxies.The distance from SMBH of UFOs and WAs are 95

Figure 1 .
Figure1.EPIC/pn spectra in 5 − 10 keV energy band and the power-law + emission line + absorption line model; black cross point and red line represent EPIC/pn spectra and power-law + gaussian (emission line) + gaussian (absorption line), respectively.For the EPIC/pn spectra we used the obin command for optimal binning.

Fit
the total spectra (0.34-10 keV) with a continuum (modified blackbody model + power-law + reflection) and the absorption component.The cut-off energy of the power law and reflection was set to 120 keV.ΔC-stat: Compare the C-stat with M1 model.M1: With hot model.M2: With one pion model and one hot model.M3, M4: With two pion models.(1) Photon index of the X-ray power-law component.(2) Normalization of the X-ray power-law component in 10 51 photons s −1 keV −1 at 1 keV.(3) Scale parameter of the reflection component.(4) Column density  H of neutral ISM gas component in the host galaxy of the AGN in 10 22 cm −2 .(5) Normalization of the modified blackbody in 10 34 cm 0.5 .(6) Temperature of the modified blackbody in eV.(7) Logarithm of the ionization parameter  of the ionized wind components in erg cm s −1 .(8) Column density  H of the ionized wind components in 10 22 cm −2 .(9) Velocity of the outflowing ionized wind components in km s −1 .(10) Covering factor of the pion component.(11) Fe abundance of the pion component.

Figure 3 .
Figure3.All spectra (RGS + EPIC/pn) and model curve, wherein the black cross point, orange point, and red line represent EPIC/pn spectra, combined RGS spectra, and the M3 model, respectively.For the EPIC/pn spectra, we used the obin command for optimal binning.RGS spectra were combined by rgscombine command and the binning factor was set to 5.

Figure 4 .
Figure 4. Soft X-ray spectra (RGS) and model curve, wherein the orange point, blue dotted line, magenta dash-dotted line, red line, and green line represent the combined RGS spectra, M1 model, M2 model, M3 model, and M4 model, respectively.RGS spectra were combined by rgscombine command and the binning factor was set to 5.

Figure 5 .Figure 6 .Figure 7 .Figure 8 .
Figure 5. Energy band spectra (EPIC/pn)of 5 − 10 keV and model curve, wherein the black cross point, blue-dotted line, magenta dash-dotted line, red line, and green line represent the EPIC/pn spectra, M1 model, M2 model, M3 model, and M4 model, respectively.For the EPIC/pn spectra, the obin command was used for optimal binning.
et al. (2014).The xabs model was used for fitting.(1) Logarithm of the ionization parameter  of the ionized wind components in erg cm s −1 .(2) Column density  H of the ionized wind components in ×10 22 cm −2 .(3) Velocity of the outflowing ionized wind components km s −1 .
& Costantini (2019).The pion model was used for fitting.(1) Logarithm of the ionization parameter  of the ionized wind components in erg cm s −1 .(2) Column density  H of the ionized wind components in ×10 22 cm −2 .(3) Velocity of the outflowing ionized wind components km s −1 .thefiltered (after getting absorbed by the UFO pion component and before getting absorbed by the WA pion component) ionized luminosity was calculated to  ion / = 1.0 × 10 43 cm −1 .Therefore, with Equation (1), the distance of WA from SMBH was

Figure 9 .
Figure 9.The 6.6 − 7.3 keV energy band simulated spectra and model curve.The magenta point, black cross point, blue line, and red dashed line represent the XRISM simulated spectra based on the M3 model, and the XRISM simulated spectra based on the M4 model, the M3 model, and the M4 model, respectively.For plotting, the binning factor was set to 10 for the simulated XRISM spectra.
, the Fe xix/xviii absorption lines are weak, and the Fe xxv/xxvi absorption lines are strong.Furthermore, the Fe xxv/xxvi and Fe xix/xviii absorption lines appeared prominently at various energies.Thus, we can employ the XRISM/Resolve to decide whether the ∼ 7 keV absorption lines in Mrk 6 can be attributed to Fe xxv/xxvi or Fe xix/xviii.

Fit
the total spectra (0.34-10 keV) with a continuum (modified blackbody model + power-law + reflection) and the absorption component.The cut-off energy of the power law and reflection was set to 120 keV.(1) Photon index of the X-ray power-law component.(2) Normalization of the X-ray power-law component in 10 51 photons s −1 keV −1 at 1 keV.(3) Scale parameter of the reflection component.(4) Normalization of the modified blackbody in 10 34 cm 0.5 .(5) Temperature of the modified blackbody in eV.(6) Logarithm of the ionization parameter  of the ionized wind components in erg cm s −1 .(7) Column density  H of the ionized wind components in 10 22 cm −2 .(8) Velocity of the outflowing ionized wind components) in km s −1 .(9) Covering factor of the pion component.(10) Fe abundance of the pion component.

Figure 10 .
Figure 10.The 0.68 − 0.94 keV energy band simulated spectra and model curve.The magenta point, black cross point, blue line, and red dashed line represent the XRISM simulated spectra based on the M3 model, and the XRISM simulated spectra based on the M4 model, the M3 model, and the M4 model, respectively.For plotting the simulated XRISM spectra, we used the obin command for optimal binning consistent with the typical ones.The M3 model attributes the 7 keV absorption lines to Fe xix/xviii.For comparison, the M3 model was re-fitted with varying parameters to obtain a model that attributed the 7 keV absorption lines to Fe xxv/xxvi (M4 model) and it may represent a local minimum of the C-stat.Although the ∼ 7 keV absorption lines can be attributed to Fe xxv/xxvi absorption lines, the spectra derived herein suggested that the ∼ 7 keV absorption lines are actually Fe xix/xviii absorption lines.Finally, we performed simulations of observations with the XRISM satellite and determined that the errors in the pion model parameters were extremely small using XRISM.Therefore, XRISM can be used to detect significant disk winds.In addition, based on XRISM simulations we investigated whether the ∼ 7 keV absorption lines in Mrk 6 were caused by Fe xxv/xxvi or Fe xix/xviii.

Table 2 .
Emission line and absorption line search

Table 4 .
The UFO absorption lines from the M3 model and the M4 model ( > 2 eV).
These lines are based on the M3 model ( out = −34700 km/s) and M4 model ( out = −12000 km/s) for each.The red dashed line and blue dashed-dotted line in Fig. 7 are the absorption lines shown here.)+Figure 2. Contour maps based on the M3 model (upper left panel: the ionization parameter vs. column density of UFO component.upperright panel: ionization parameter vs. outflow velocity of UFO component.bottomleft panel: the ionization parameter vs. column density of WA component.bottomright panel: ionization parameter vs. outflow velocity of WA component.).Black cross point, black line, red line, green line, blue line, and cyan line show the best-fit parameter, 68.3% confidence level, 90% confidence level, 95.4% confidence level, 99% confidence level, and 99.99% confidence level, respectively.

Table 6 .
Comparison of parameters of WAs in the 3C 390.3 and 4C +74.26.