Quasar absorption outflows are seen in a large fraction of quasar spectra (40%), often detected via blueshifted absorption troughs in the rest frame of quasars (Hewett & Foltz 2003;Dai et al. 2008;Knigge et al. 2008). These outflows are often mentioned as likely candidates for producing active galactic nucleus (AGN) feedback (e.g., Silk & Rees 1998;Scannapieco & Oh 2004;Ciotti et al. 2009;Yuan et al. 2018;Vayner et al. 2021). According to theoretical models, outflow systems require a kinetic luminosity ( E k ) of at least ∼0.5% (Hopkins & Elvis 2010) or ∼5% (Scannapieco & Oh 2004) of the quasar's Eddington luminosity (L Edd ) to contribute to AGN feedback. Outflow systems that fit these criteria have been found (e.g., Moe et al. 2009;Arav et al. 2013;Chamberlain et al. 2015;Xu et al. 2019Xu et al. , 2020a;;Miller et al. 2020a;Xu et al. 2020b;Miller et al. 2020b;Arav et al. 2020).
The kinetic luminosity of a quasar's outflow system is dependent on its distance from its central source (R), which we can find by measuring both the electron number density (n e ) and ionization parameter (U H ) (Borguet et al. 2012a). Our group and others have used this method to find the distances of outflow systems in the past (de Kool et al. 2001;Hamann et al. 2001;de Kool et al. 2002;Gabel et al. 2005;Borguet et al. 2012a;Xu et al. 2018;Miller et al. 2020a;Arav et al. 2020). Using the ratios between excited and resonance state column densities of ionized species (N ion ) can lead us to a value of n e (Arav et al. 2018). This paper presents one such determination of the R and E k values of three outflow components found in the Very Large Telescope (VLT)/UVES spectrum of SDSS J024221.87+004912.6 (hereafter J0242+0049).
The analysis of J0242+0049 shown in this paper is based on data from the VLT/UVES Spectral Quasar Absorption Database (SQUAD) published by Murphy et al. (2019), which contains the spectra of 475 quasars. Analysis of more SQUAD objects will be conducted in the future.
The UVES data of J0242+0049 are from program 075.B-0190 (A), which Hall et al. (2007) used to identify a high-velocity C IV broad absorption line (BAL) at z ≈ 1.88 (v ≈ -18,000 km s -1 ), as well as two mini-BAL outflows and one narrow absorption line (NAL) outflow at lower velocities; we have identified all four independently. Comparing the UVES spectrum to SDSS spectra from previous epochs, Hall et al. (2007) have identified a shift in the velocity of the high-velocity BAL, which could potentially be explained by acceleration. They have also found potential line locking in the Si IV absorption doublets of the two lower-velocity mini-BAL systems. In addition to the analysis of the UVES data, we conduct a follow-up to their observation of the velocity shift using SDSS observation data from more recent epochs.
This paper is structured as follows. Section 2 discusses the observation of J0242+0049, as well as the data acquisition process. In Section 3, we present the ionic column density measurements and the process of finding n e and U H . Section 4 shows the results of the analysis, including the energetics parameters of the outflow systems. We also show observations of the high-velocity BAL from recent SDSS epochs. Section 5 provides a discussion of the results, and Section 6 summarizes and concludes the paper. For this analysis, we adopt a cosmology of h = 0.696, Ω m = 0.286, and Ω Λ = 0.714 (Bennett et al. 2014) and use the Python astronomy package Astropy (Astropy Collaboration et al. 2013, 2018) for cosmological calculations.
The quasar J0242+0049 (J2000: R.A. = 02:42:22, decl. = +00:49:12.6; z = 2.06) (Pâris et al. 2018) was observed in 2005 September 5, with the VLT/UVES as part of program 075.B-0190(A), with resolution R ; 40,000 and wavelength coverage from 3291 to 9300 Å (Hall et al. 2007). The systemic redshift z = 2.06 given by Murphy et al. (2019) is consistent with the value we find based on the Mg II emission line in the SDSS spectrum of the MJD = 57758 epoch. The spectral data was reduced and normalized by its continuum and emission by Murphy et al. (2019) as part of their SQUAD database. Broad and narrow absorption lines have been found in the spectrum of J0242+0049 by Hall et al. (2007), which we identify here as NAL S1 at -1200 km s -1 (Lyα FWHM = 240 km s -1 ), mini-BAL S2 at -1800 km s -1 (N V FWHM = 900 km s -1 ), mini-BAL S3 at -3500 km s -1 (N V FWHM = 720 km s -1 ), and the aforementioned BAL S4 at -18,000 km s -1 , as shown in the full spectrum in Figure 1. Following Weymann et al. (1991), a BAL is a continuous absorption feature below 0.9 normalized intensity over 2000 km s -1 , a mini-BAL is the same but between 500 and 2000 km s -1 (Hamann & Sabra 2004), and an NAL is an absorption feature with width below 500 km s -1 . We measure the width of S4 at 0.9 normalized intensity to be 2200 km s -1 , which is above the threshold of a BAL, with a balnicity index, as defined by Weymann et al. (1991), of 660 km s -1 . Chen et al. (2021) have identified four C IV absorption systems, two of which coincide with systems S3 and S4. We label the other two as systems A and B and show them in Figure 1. The focus of this paper is on the four systems S1, S2, S3, and S4. We do not discuss systems A and B because they only show absorption in C IV, which does not lend itself to further analysis. The outflows show absorption from low-ionization species such as Si II, C II, and Fe II, as well as lines of Ly α, C IV, N V, P V, Mg II, Al II, and Al III. For the purpose of measuring the ionic column densities, we convert the normalized spectrum data from wavelength to velocity space via the systemic redshift of the quasar, as shown in Figure 2. Note that S2 appears to be composed of at least seven subcomponents, as seen in plot (l) of Figure 2. The components are blended in the absorption troughs of C IV and Si IV, and due to the shallowness of the C II * troughs, it is impossible to decompose it into the different subcomponents. For this reason, they are treated as a singular absorption system for the sake of analysis in this paper.
For the velocity-shift analysis, SDSS spectra from MJD = 52177, 52199, 55455, and 57758 were retrieved and corrected for galactic extinction with E(B -V ) = 0.0269 (Schlafly & Finkbeiner 2011). The spectra from both the BOSS and SDSS spectrographs have spectral resolutions of R ≈ 2000 Normalized spectrum plotted in velocity space for each ion in the absorption systems. The green, red, and blue vertical lines represent the velocity of systems S1, S2, and S3, respectively. The dotted vertical lines show the integration ranges used for the calculation of the ionic column densities. The horizontal dashed line represents the continuum level. Intervening absorption systems that contaminate the blue spectra are marked with cyan vertical lines, while intervening systems contaminating the red spectra are marked with brown vertical lines. Note in plot (f) that the S1 integration range for Si II 1265 Å is contaminated with the Mg II 2796 Å absorption of the z = 0.3783 intervening system. Plot (l) shows the structure of the S2 absorption trough of Si II and C II, on a narrower velocity scale. (Schneider et al. 2010;Smee et al. 2013;Pâris et al. 2018). More details on the SDSS spectra can be found in Table 1.
To find the physical characteristics of the outflow systems, we first find the column densities of the observed ions (N ion ). The simplest method for measuring column densities is by assuming the apparent optical depth (AOD) of a uniformly covered homogeneous source, as demonstrated by Savage & Sembach (1991). When calculating column density under this assumption, we first assume the relation between intensity and optical depth as follows (see Equation (1) of Savage & Sembach 1991): ( ) ( ) ( ) ( ) l l = t l -I I e , 1 0 where I(λ) is the intensity, I 0 (λ) is the intensity without absorption, and τ(λ) is the optical depth as a function of wavelength. When writing optical depth as a function of outflow velocity, it has a relation with column density N(v) of (see Equation (8) of Savage & Sembach 1991):
where m e is the mass of an electron, e is the elementary charge, and f and λ are the oscillator strength and wavelength of the transition line, respectively. Finding N(v) and integrating it over the velocity range of the absorption trough yields the column density based on the AOD assumption. The AOD method is used to find lower limits of N ion for singlets or contaminated doublets, or upper limits when there are no discernible absorption troughs.
When there are multiple lines of the same ion and energy state, we can use the partial covering (PC) method, which assumes a homogeneous source partially covered by the outflow (Barlow et al. 1997;Arav et al. 1999aArav et al. , 1999b) ) and solves for a velocity-dependent covering factor (de Kool et al. 2002;Arav et al. 2005), to improve our measurements by taking phenomena such as nonblack saturation into account (Edmonds et al. 2011;Borguet et al. 2012). When calculating the PC-based column density of an ion with a doublet of transition lines, we find the covering fraction C(v) via the following relations (see Equations (2) and (3) of Arav et al. 2005):
where I R (v) and I B (v) are the normalized intensities of the red and blue absorption features, respectively, and τ(v) is the optical depth of the red component.
We choose integration ranges that cover visible absorption in the data, as can be seen in Figure 2, while minimizing the effects of blending and contamination. For instance, for Si IV, we use the blue line for S3 and the red line for S2. Si II * of S1 shows contamination due to an intervening absorption feature, so we use the measured column density as an upper limit for the sake of our analysis. The C IV of S2 is heavily blended between the red and blue features, so we choose a velocity range in which the blue and red spectra do not overlap with each other in order to find a lower limit of the column density.
Attempting a Gaussian fit of the C IV absorption of S2 yields a poor fit due to the saturation of the trough. Calculating the column density based on the fit results in a lower limit of 2400 × 10 12 cm -2 , compared to the measured lower limit of 3900 × 10 12 cm -2 . This difference does not affect the solution of the hydrogen column density and photoionization parameter as described in Section 3.2.
The measured column density values can be found in Table 2. Note that most adopted values in Table 2 are upper or lower limits. The errors in the column densities are propagated from the errors in the normalized flux from the data, binned along with the data into segments of Δv = 10 km s -1 for numerical integration. A 20% error is added in quadrature for the column density values adopted for photoionization analysis (see the last column of Table 2) to take into account the uncertainty in the modeled continuum level (Xu et al. 2018).
We use a grid of photoionization models created using the spectral synthesis code Cloudy (version c17.00; Ferland et al. 2017), in order to find the hydrogen column density (N H ) and ionization parameter (U H ) that best fit the measured ionic column densities, following the method of previous works (e.g., Miller et al. 2018;Xu et al. 2019;Miller et al. 2020a).
We use Cloudy to create a grid of simulated models that correspond to different N H and U H values, assuming solar metallicity, and the spectral energy distribution (SED) of quasar HE 0238-1904 (hereafter HE0238;Arav et al. 2013).
The N H and U H parameters determine the ionic column densities of each model, which we compare with the measured column densities shown in Table 2. For S2, including the lower bound of the Fe II column density in the analysis introduced an N H and U H solution that was contradictory to the constraints from the other ions. We suspect that this is because the Fe abundance of the system does not match solar metallicity (Z e ), requiring a metallicity of ∼10 Z e . This is in approximate agreement with the highest outflow metallicity found by Gabel et al. (2006; Z ≈ 5 Z e ). For this reason, we model our solution using the other ions but excluding Fe II. The N log H and U log H values from this analysis are shown in Table 3, as well as in Figure 3.
Table 1 SDSS Spectra Information Epoch in MJD Spectrograph Plate Fiber Observed Date Wavelength Coverage (Å) 52177 SDSS 707 332 2001 Sep 25 3824-9215 52199 SDSS 706 617 2001 Oct 17 3820-9202 55455 BOSS 4240 754 2010 Sep 16 3590-10382 57758 BOSS 9381 79 2017 Jan 5 3573-10334
The electron number density and, by extension, the distance of the outflow systems from the central source, can be found by determining the abundance ratios, measured via column densities, between excited and resonance states of low-ionization species (Moe et al. 2009). We use the CHIANTI 9.0.1 Database (Dere et al. 1997(Dere et al. , 2019) ) to model the relationship between the ratio of excited and resonance state ion abundances, and the electron number density, based on collisional excitation. We overlay this relation with the ratios based on the measured column densities, as shown in Figure 4. For this object, we use the ratios N(Si II * )/N (Si II), N(C II * )/N(C II), and N(Fe II * )/N(Fe II), where N(ion) is the column density of a particular ion.
For S3, we have an upper limit given by the C II ratio and a measurement from the Si II ratio, which agree with one another. Our measurements of Fe II are dominated by noise and, as such, are not included in the n e measurement. Taking the ratio of N(Si II * )/N (Si
II), we find that [ ] = -+ n log 3.3 cm e 0.4 0.8 3 . S2 provides us a measurement from C II and upper limits from Si II and Fe II. From the N(C II * )/N(C II) ratio, we find [ ] = -+ n log 0.25 cm e 0.2 0.2 3 . S1 only gives us a lower limit from C II, as Si II * is contaminated by an intervening line and cannot give us a reliable ratio between N(Si II * ) and N(Si II). Thus, we get a lower limit for the electron number density, [ ] > - n log 2.0 cm e 0.45 3 .
In order to find the distance of the outflow systems, we use the definition for the ionization parameter
where Q H is the rate of ionizing photons, R is the distance of the outflow from the central source, and n H is the hydrogen number density, which is estimated as n e ≈ 1.2n H for highly ionized plasma (Osterbrock & Ferland 2006). Because we have a solution for U H from our photoionization analysis, as well as the n e for each outflow from the excited to resonance state ratios, we can find R after determining the value of Q H . We determined Q ) is at least an order of magnitude larger than that of S3 (-3500 km s -1 , = -+ R 1.2 0.9 0.8 kpc) or S1 (-1200 km s -1 , R < 5.4 +7.3 kpc).
Table 2 J0242+0049 Outflow Ionic Column Densities Troughs AOD PC Adopted S1, v = -1200 km s -1 H I -+ 177.0 1.8 1.9 >180 -40 N V -+ 467 5 5 >470 -90 P V -+ 56 4 4 >50 -10 C II total -+ 26 2 2 >26 -5 C II 1335 -+ 10.8 1.2 1.5 C II * 1336 -+ 14.7 1.3 1.5 C IV -+ 350 4 5 >350 -70 Si II total -+ 5.5 0.4 0.3 <5.5 +1.2 Si II 1260 -+ 1.3 0.2 0.3 <1.3 +0.4 Si II * 1265 -+ 2.8 0.2 0.2 <2.8 +0.6 Si IV -+ 92.6 1.0 1.0 >90 -20 Mg II -+ 2.7 0.3 0.3 >2.7 -0.6 Al II -+ 0.3 0.08 0.09 <0.3 +0.1 Al III -+ 3.3 0.3 0.3 -+ 4.9 0.5 0.8 -+ 4.9 1.1 1.3 Fe II -+ 1.9 0.3 0.4 <1.9 +0.5 S2, v = -1800 km s -1 H I -+ 1680 10 180 >1680 -340 N V -+ 4620 30 30 >4620 -920 P V -+ 450 10 10 >450 -90 C II total -+ 740 10 90 >740 -150 C II 1335 -+ 690 10 90 C II * 1336 -+ 50 2 2 C IV -+ 3910 20 330 >3910 -780 Si II total >80 -20 Si II 1260 -+ 77.6 1.1 1.3 >80 -20 Si II * 1265 -+ 2.9 0.3 0.3 <3 +0.7 Si IV -+ 1410 10 100 >1410 -280 Mg II -+ 84 0.9 1.0 -+ 90.7 1.0 1.0 -+ 90 20 20 Al II -+ 10.3 0.1 0.2 >10 -2 Al III -+ 48.4 0.9 0.9 -+ 55.6 0.8 0.9 -+ 55 10 10 Fe II total >12 -2.5 Fe II 2600 -+ 12.2 0.5 0.5 >12 -2.5 Fe II * 2612 -+ 1.0 0.5 0.5 <1.0 +0.5 S3, v = -3500 km s -1 H I -+ 417.7 3.4 3.7 >420 -80 N V -+ 3780 20 20 >3780 -760 P V -+ 390 10 10 >390 -80 C II total -+ 94.2 3.3 3.5 >90 -20 C II 1335 -+ 32 2.2 2.4 C II * 1336 -+ 62.1 2.5 2.5 C IV -+ 2180 10 10 >2180 -440 Si II total -+ 5.1 0.3 0.4 >5.1 -1.1 Si II 1260 -+ 2.5 0.3 0.3 Si II * 1265 -+ 2.6 0.2 0.2 Si IV -+ 285.8 1.4 1.5 >290 -60 Mg II -+ 17.9 0.6 0.6 >18 -4 Al II -+ 2 0.2 0.2 <2 +0.4 Al III -+ 19.7 0.4 0.4 >20 -4 Fe II total -+ 6.7 0.9 0.7 <6.7 +1.5 Fe II 2600 -+ 2.2 0.5 0.4 Fe II * 2612 -+ 4.5 0.7 0.6
Notes. Units are in 10 12 cm -2 . Values have been calculated by numerical integration over bins with width Δv = 10 km s -1 . Note that most of the adopted values are upper or lower limits.
Once we have the distance of the outflow, we can find the mass flow rate (Borguet et al. 2012b),
H p and the kinetic luminosity,
assuming a partially filled shell, where Ω is the global covering factor (fraction of the total solid angle of the quasar that the outflow covers), μ = 1.4 is the mean atomic mass per proton, m p is the proton mass, and v is outflow velocity. For the global covering factor, we assume Ω = 0.2, the portion of quasars from which C IV BALs are found (Hewett & Foltz 2003). As explained by Dunn et al. (2010), this is a reasonable assumption despite the relative rarity of quasars showing singly ionized absorption troughs such as Si II, due to the likelihood that such quasars are regular BAL quasars seen from specific lines of sight. The resulting kinetic luminosity calculations yield [ ] = --+ -+ E log erg s 45.42 , 45.82 K 1 0.64 1.33 0.32 0.37 for Table 3 Physical Properties of the J0242+0049 Outflow Systems Outflow System S1 = -1200 km s -1 S2 = -1800 km s -1 S3=-3500 km s -1 log(N H ) -+ 21.41 0.70 0.38 -+ 21.27 0.58 0.64 -+ 21.78 0.24 1.30 [cm -2 ] log(U H ) --+ 0.86 0.59 0.33 --+ 1.30 0.48 0.49 --+ 0.83 0.18 0.95 [dex] log(n e ) >2.00 -0.45 -+ 0.25 0.20 0.20 -+ 3.30 0.40 0.75 [cm -3 ] Distance <5.4 +7.3 -+ 67 31 55 -+ 1.2 0.9 0.8 [kpc] M <480 +300 -+ 6500 3400 8900 -+ 700 30 2900 [M e yr -1 ] Mv <3.6 +2.3 -+ 74 39 100 -+ 16 0.7 60 [10 36 erg cm -1 ] ( ) E log K <44.33 +0.21 -+ 45.82 0.32 0.37 -+ 45.43 0.02 0.7 [erg s -1 ] E L K e d d <0.18 +0.16 -+ 5.5 3.1 8.8 -+ 2.3 0.8 9.9
[%]
Note. A temperature of 10,000 K is assumed. S3 and S2, respectively, as well as an upper limit of < + E log 44.33 K 0.53 for S1. In addition, we calculate the momentum flux ( Mv) of each outflow system (see Table 3) and compare it to the single-scattering limit of the quasar ( )
= -+ -6.44 10 erg cm L c 0.61 0.61 36 1 bol
. The single-scattering limit assumes the scenario in which absorption of photon momentum drives acceleration (Abbott 1982;Arav & Li 1994). The momentum flux of S1 is smaller than the single-scattering limit, while those of S2 and S3 are above the limit. As S2 has a momentum flux an order of magnitude higher than the singlescattering limit, this implies the possibility of a multiplescattering scenario (Lucy & Abbott 1993).
Following up on the results reported by Hall et al. (2007), we examine the velocity shift of the C IV BAL of S4. Using two Gaussian profiles, one broad and shallow, and the other narrow and deep, we modeled the absorption in each of the five epochs, as shown in Figure 5. We can see that the centroid velocity of the narrow Gaussian monotonically grows, while the equivalent width becomes smaller from epoch to epoch. Detailed information on the centroid velocities and equivalent widths per epoch can be seen in Table 4.
Assuming acceleration along the line of sight, based on the centroid velocities of the narrow Gaussian, the average acceleration between the observations in 2001 September and 2017 January would be a = -0.25 ± 0.13 cm s -2 in the quasar's rest frame, which agrees within error with the acceleration a = -0.154 ± 0.025 cm s -2 between 2001 September 1 and 2005 September found by Hall et al. (2007). Due to the shrinking of the trough, we must take into consideration effects other than the lineof-sight acceleration, such as changes in photoionization, as discussed by Xu et al. (2020c).
As previously mentioned in the introduction, the kinetic luminosity ( E k ) of the outflow systems must be at least ∼0.5% (Hopkins & Elvis 2010) or ∼5% (Scannapieco & Oh 2004) of the source quasar's Eddington luminosity (L Edd ) to contribute to AGN feedback. In order to find this ratio, we must first find the Eddington luminosity. We compute the mass of the black hole using the Mg II-based mass equation in Bahk et al. (2019), with the FWHM of the Mg II emission feature in the SDSS spectrum. To account for the Fe II emission throughout the spectrum, we use the Fe II template by Tsuzuki et al. (2006) and run a best-fit algorithm to match the features in the spectrum, as done by Woo et al. (2018). This yields a black hole mass of
= -+ M M 9.7 10 BH 3.4 4.9 8 , corresponding to an Eddington luminosity of = -+ -L 1.2 10 erg s Edd 0.4 0.6 47
1 . We expect the Fe II emission's effect on the absorption to be small, n log e . The curves marked Si II, C II, and Fe II are the theoretical ratios modeled with CHIANTI, assuming a temperature of 10,000 K. The crosses on the curves show the ranges of the C II, Si II, and Fe II column density ratios, based on the measured AOD column densities. The green, red, and blue correspond to systems S1, S2, and S3, respectively. Arrows indicate either upper or lower limits in n log e depending on the direction of the arrow. The upper limit of the N(Si II * )/N(Si II) ratio for S3 is marked with a tick, as it overlaps with the error bars of the C II ratio of the same system.
as the fitted emission template from Tsuzuki et al. (2006) is <20% of the continuum level of the SDSS spectrum of MJD = 57758, leaving us with column densities that agree with our measured values within error.
Taking the ratio between the kinetic luminosity of each outflow system and the Eddington luminosity of the quasar, we find that S2 and S3 are well above the 0.5% threshold from Hopkins & Elvis (2010) and S2 is above the 5% threshold by Scannapieco & Oh (2004), while S1ʼs kinetic luminosity is below 0.18% of the Eddington luminosity, as seen in Table 3. We can thus conclude that S2 and S3 are energetic enough to contribute to AGN feedback.
Unlike in objects analyzed in other papers (e.g., Xu et al. 2020a;Miller et al. 2020a), we do not have lines from the veryhigh-ionization phase. Thus, while there may be a very-highionization phase, we cannot tell from the information we have.
Following the examination of the S4 C IV BAL at different epochs, we looked to systems S1, S2, and S3 for time-variability. As shown in Figure 6, the Si IV trough depth becomes increasingly shallower over time, which may be explained by the same ionization effects that affect the S4 C IV BAL shown in Figure 5, discussed by Xu et al. (2020c). As the ionization parameter U H changes, ions of particular ionization states become more or less abundant over time. Because the C IV of S4, along with the Si IV of S1, S2, and S3, decreases monotonically, this supports the assertion that the changes in the troughs are due to changes in the ionization parameter. Further observation and analysis will be required to confirm these effects.
An alternative to using the SED of HE0238 would be to use the theoretical SED as defined by Mathews & Ferland (1987), which is based on the He II line. The HE0238 SED is based on observation of a high-quality spectrum that stretches into the far-UV range, better representing a quasar spectrum (Arav et al. 2013). Just like in other objects (e.g., Xu et al. 2018;Miller et al. 2020a), higher metallicity drops the values of the energetics parameters, for instance, raising the metallicity to four times solar metallicity, using abundance ratios from Ballero et al. (2008), changes the photoionization solution of S2 to = -- + U log 1.5 H 0.3 0.3 , and [ ] = -+ -N log 20.5 cm H 0.4 0.4 2 , and lowering the mass flow rate and kinetic luminosity to ˙ = -+ -M M 1300 yr 400 600 1 and [ ] = -+ -E log 45.13 erg s K 0.18 0.17 1 , respectively. Using the SED by Mathews & Ferland (1987) with solar metallicity changes the solution to = -= -+ -+ U N log 1.5 , log 21.2 H H 0.4 0.4 0.5 0.5 , which is in agreement with the values in Table 3 within error.
It is possible that the SED seen by one outflow system can be attenuated by another, resulting in a smaller H and, by extension, a smaller distance R. In particular, as S2 is farther out than the other mini-BAL system S3, it is likely that the SED seen by S2 is obscured by S3 (e.g., Bautista et al. 2010;Sun et al. 2017;Miller et al. 2018Miller et al. , 2020c)). We used the method described by Miller et al. (2018) to test the effects of attenuation by S3. We used Cloudy to model the attenuated SED by S3 by inputting the relevant N H and U H values of S3 shown in Table 3. We then use that attenuated SED to find the resulting Q H and R of S2
. The reduced values for the parameters are = -+ -Q 4.9 10 s H 0.5 0.5 56 1 and = -+ R 43 20 35
kpc, which is a ∼30% decrease in the distance of S2. We choose S3 as the attenuation source, as its stronger features compared to S1 suggest that the attenuation effect from S3 would be larger than that of S1. We are unable to calculate the attenuation by S4, as we cannot obtain N H or U H from its singular C IV absorption trough.
There have been several prior studies of quasar outflow acceleration, including that of the acceleration of the outflow of quasar SDSS J1042+1646 conducted by Xu et al. (2020c), based on the acceleration seen in Ne VIII λλ770, 780. The bolometric luminosity of SDSS J1042+1646 is estimated to be ∼1.5 × 10 47 erg s -1 , which is comparable to that of J0242+0049 (1.9 × 10 47 erg s -1 ). The average acceleration of S4 that we have found (a ≈ -0.25 cm s -2 ) is roughly an order of magnitude smaller than that by Xu et al. (2020c; a = -1.52 cm s -2 ), which suggests that if S4 is truly accelerating, the acceleration of quasar outflows can cover a wide range.
To give context to the study of outflow S2, we review a few outflows with a similarly large R and/or E K . Analysis of a molecular outflow of quasar SDSS J1148+5251 at a distance R ∼ 15 kpc conducted by Maiolino et al. (2012) ´-E 1.9 10 erg s K 45 1
. Liu et al. (2013) analyzed the ionized gas around 11 radioquiet quasars via the [O III] λ5007Å emission. These outflows were found between ∼10-20kpc from the central source, had velocities of up to -1000 km s -1 , and had an estimated n e ∼ 1.2 cm -3 . The outflows had an estimated range of E K from 4 × 10 44 to 3 × 10 45 erg s -1 , and M from 2 × 10 3 to 2 × 10 4 M e yr -1 . These numbers are within a factor of a few of the values we find for S1, S2, and S3 (see Table 3), which suggests we may find similar outflows in absorption.
In their analysis of SDSS J1051+1247, Miller et al. (2020a) found an outflow system with = ´- E 3 10 erg s
K 45 1 . The mass flow rate ( ˙ = -M M 6500 yr 1 ) and kinetic luminosity ( =
´-E 6.6 10 erg s K 45 1
) of S2 align with these values and those of the objects mentioned above, within a margin of error. Xu et al. (2020a) claim the most energetic quasar outflow measurement to date from quasar SDSS J1042+1646 ( = ´- E 5 10 erg s
), and this claim remains uncontested. While the distance of S2 from the quasar is unprecedentedly large, there exists a theoretical model that may be supported by this observation. Faucher-Giguère et al. (2012) provide an argument that FeLoBALs, absorption systems with signs of Fe II, may be formed in situ at distances of several kiloparsecs. They clarify that while their model focuses on the formation of FeLoBALs at large distances, other classes of outflows may form as described by it.
This paper has presented the analysis of three absorption systems of quasar SDSS J0242+0049, dubbed S1, S2, and S3, from VLT/UVES observational data, as well as the velocity shift of the S4 C IV BAL across five different epochs. From the absorption troughs we identified, we measured the column densities of 11 ions in each system as shown in Table 2. Through photoionization analysis using the measured column densities, we found the best-fit solutions to U H and N H for each system. ) have been calculated by integrating over the Gaussians in velocity space. The parameters for the wider Gaussians are more affected by the continuum models for each epoch. Note that the uncertainty in the centroid velocity of the MJD = 53619 epoch is significantly smaller than those of the other epochs, due to the higher signal-to-noise ratio and resolution of the data.
The abundance ratios between the excited and resonance states of ions Si II and C II were used to find the electron number density n e of the three systems S1, S2, and S3, as shown in Figure 4. Equations (5), (6), and (7) were used to find the distance from the central source, the mass flow rate, and the kinetic luminosity of each system respectively. The ratios between the kinetic luminosities and the quasar's Eddington luminosity were found in order to evaluate their AGN feedback contribution, the results of which can be seen in Table 3. From this analysis, we have found that S2 and S3 have sufficient kinetic luminosity for AGN feedback contribution. Most notable in this result is the distance of S2 R = 67 kpc, farther than the absorption system of 3C 191 found at R = 28 kpc by Hamann et al. (2001), making this the farthest reported distance of a mini-BAL absorption outflow from its central source.
Following the analysis of the three systems, we examined the change in velocity and equivalent width of the S4 C IV BAL, as shown in Figure 5, based on the UVES spectrum, as well as different SDSS observations. As seen in Table 4, there has been a monotonic increase in the line-of sight velocity, as well as a decrease in equivalent width, with the trough being a factor of 6 weaker at the epoch of 2017 January compared to that of 2001 September.
Through further observation and analysis, we expect to shed more light on the time-variability of the S4 C IV BAL, as well as that of systems S1, S2, and S3.
The Astrophysical Journal, 927:176 (10pp), 2022 March 10 Byun,Arav, & Hall