A Catalog of High-Velocity CIV Mini-BALs in the VLT-UVES and Keck-HIRES Archives
DDraft version February 3, 2021
Typeset using L A TEX twocolumn style in AASTeX63
A Catalog of High-Velocity C iv Mini-BALs in the VLT-UVES and Keck-HIRES Archives
Chen Chen,
1, 2
Fred Hamann, Bo Ma, and Michael Murphy School of Physics & AstronomySun Yat-Sen UniversityZhuhai 519000, China Department of Physics & AstronomyUniversity of CaliforniaRiverside, CA 92521, USA Centre for Astrophysics and SupercomputingSwinburne University of TechnologyHawthorn, Victoria 3122, Australia (Received; Revised; Accepted)
Submitted to ApJABSTRACTWe present a catalog of high-velocity C iv λ iv mini-BALs based on smooth rounded BAL-like profiles with velocity blueshifts < − − andwidths in the range 70 (cid:46) FWHM(1548) (cid:46) − (for λ ∼
9% after correcting for incomplete velocity coverage. However, the numbersof systems rise sharply at lower velocities and narrower FWHMs, suggesting that many outflow linesare missed by our study. All of the systems are highly ionized based on the strong presence of N v andO vi and/or the absence of Si ii and C ii when within the wavelength coverage. Two of the mini-BALsystems in our catalog, plus three others at smaller velocity shifts, have P v λ iv absorption and total hydrogen column densities (cid:38) cm − . Most ofthe mini-BALs are confirmed to have optical depths (cid:38) iv and 0.03 in Si iv , corresponding to outflowabsorbing structures < .
002 pc across. When multiple lines are measured, the lines of less abundantions tend to have narrower profiles and smaller covering fractions indicative of inhomogeneous absorberswhere higher column densities occur in smaller clumps. This picture might extend to BAL outflowsif the broader and generally deeper BALs form in either the largest clumps or collections of manymini-BAL-like clumps that blend together in observed quasar spectra. INTRODUCTIONOutflows from quasar accretion disks might combinewith starburst-driven winds to provide important feed-back that regulates star formation and mass assemblyin the host galaxies (Silk & Rees 1998; Kauffmann &Haehnelt 2000; King 2003; Scannapieco & Oh 2004; DiMatteo et al. 2005; Hopkins et al. 2008; Ostriker et al.2010; Debuhr et al. 2012; Rupke & Veilleux 2013; Rupke
Corresponding author: Bo [email protected] et al. 2017; Cicone et al. 2014; Weinberger et al. 2017).The kinetic power needed from quasar outflows to pro-duce important feedback effects on their own are be-lieved to be just (cid:38) (cid:38) iv λ α forest. BALs are often defined by C iv ve- a r X i v : . [ a s t r o - ph . GA ] F e b Chen et al. locity widths (cid:38) − at flow speeds from a fewthousand up to a several tens of thousands of km s − (Anderson et al. 1987; Weymann et al. 1991; Reichardet al. 2003; Trump et al. 2006). However, quasar out-flows produce a wide variety of absorption lines of dif-ferent velocity widths. These include narrow absorptionlines (NALs) with widths that are nominally less thana few hundred km s − , and so-called “mini-BALs” thathave widths intermediate between BALs and NALs (e.g.,Hamann et al. 1997b; Hamann & Sabra 2004; Hamannet al. 2011; Misawa et al. 2007a, 2014). Quasar spec-troscopic surveys have shown that mini-BALs and out-flow NALs (as opposed to NALs that form in interven-ing gas unrelated to the quasars) are substantially morecommon than BALs in quasar spectra. In particular,roughly 50% of bright quasars in the Sloan Digital SkySurvey (SDSS, Schneider et al. 2010) exhibit a C iv NALor mini-BAL outflow absorption line while only ∼ ∼
20% contain a classic C iv BAL (Trump et al. 2006;Rodr´ıguez Hidalgo 2008; Nestor et al. 2008; Wild et al.2008; Misawa et al. 2007a; Ganguly & Brotherton 2008;Knigge et al. 2008; Hamann et al. 2012; Pˆaris et al. 2017;Guo & Martini 2019, and refs. therein).Mini-BALs are the least studied of these outflow fea-tures. They appear across same range of velocity shiftsas BALs, nominally from near 0 to ∼ − form, with prominent absorption in higher ions such asC iv , N v λ vi λ iv ‘absorption index’ (AI,Hall et al. 2002; Trump et al. 2006; Pˆaris et al. 2017)produce samples severely contaminated by unrelated in-tervening absorption lines (e.g., damped Ly α systemsor blends of multiple NALs that have nothing to dowith quasar outflows) unless high minimum values on AIare set to avoid this contamination (e.g., Rodr´ıguez Hi-dalgo 2008; Ganguly & Brotherton 2008; Hamann et al.2019a,b). However, large AI values exclude interestingnarrow mini-BALs near the ambiguous boundary withNALs, e.g., with velocity widths less than a few hundredkm s − . Studies of small quasars samples using higher-resolution echelle spectra confirm that quasar outflowsproduce a continuous range of line widths from mini-BALs to NALs, while the survey statistics mentionedabove suggest that narrower outflow lines are more com-mon than broad ones (e.g., Hamann et al. 1997b,a; Aravet al. 2001; Gabel et al. 2005; Simon & Hamann 2010a;Misawa et al. 2010; Hamann et al. 2011).Another obstacle for studies of broad outflow absorp-tion lines (all BALs and some mini-BALs) is that dou-blet lines like C iv λ − ) can blend together so their strength ratios cannotbe measured for optical depth constraints. This is prob-lematic because partial covering of the background lightsource appears to be common in quasar outflows, lead-ing to observed line strengths that depend on the line-of-sight covering fractions in addition to the line opticaldepths and column densities (Sections 3 and 5 below,Hamann et al. 1997b; Arav et al. 2005). In particular,highly saturated lines produced in high-column densityoutflows can have shallow observed troughs (not reach-ing zero intensity) in quasar spectra, leading to largeuncertainties (generally underestimates) of key outflowproperties like the total column densities, mass lossrates, and kinetic energy yields (see also Moravec et al.2017; Hamann et al. 2019a,b, for recent discussions).Severe line saturation and large total column densitiescan sometimes be identified via measurements of lines Catalog of C iv Mini-BALs v λ v and C iv ions coexist spatiallyin quasar outflows, but the P v line optical depths areroughly ∼ iv due to the lowerabundance (assuming roughly solar abundance ratiosand ionization by a standard quasar spectrum). Mea-surements of P v absorption lines therefore indicate thatstrong transitions like C iv , Si iv , O vi , and others arehighly saturated and that the total column densities inthe outflows are conservatively log N H (cm − ) (cid:38) (Hamann 1998; Leighly et al. 2011; Borguet et al. 2013;Chamberlain et al. 2015). The recent study by (Hamannet al. 2019a) using median composite spectra of quasarsin the BOSS survey shows that P v absorption is com-mon in all types of weak and strong BAL outflows andthat it is weaker but still present in typical mini-BALsystems. There are also numerous measurements of P v absorption in individual BAL quasars (e.g., Capellupoet al. 2017). However, P v is much more difficult todetect in weaker and narrower mini-BAL systems dueto blending problems in the Ly α forest. There is, toour knowledge, only one reported case of P v absorp-tion in a mini-BAL outflow (in the low-redshift quasarPG1411+442, Hamann et al. 2019b).These difficulties provide further motivation to findand study narrow outflow lines where the common dou-blets like C iv are resolved. In this paper, we presenta catalog of C iv mini-BALs and other narrow outflowlines in archival high-resolution and high signal-to-noiseratio spectra of 638 quasars measured with the UV-Visual Echelle Spectrograph (UVES) on the Very LargeTelescope (VLT) or the High Resolution Echelle Spec-trometer (HIRES) on the W. M. Keck Observatory tele-scope. This unique dataset allows us to detect andmeasure C iv mini-BALs that are considerably weakerand narrower than any previous outflow line survey. Wedeliberately cross the ambiguous boundary from mini-BALs to NALs to include in our catalog all C iv linesnarrower than BALs that can be readily attributed toquasar outflows based on their smooth rounded BAL-like profiles. Many of the lines included would be consid-ered NALs or “outflow NALs” in previous work. How-ever, for convenience and because of their related originsin quasar outflows, we will refer hereafter to all of theC iv outflow lines in our catalog as mini-BALs. We alsoreport on other lines detected in the same outflow sys-tems, notably Si iv , N v , O vi , and P v , which provideimportant constraints of the outflow ionizations, cover-ing fractions, and column densities. Section 2 below describes the UVES and HIRESdatasets and our mini-BAL selection criteria. Section3 describes fits to the C iv mini-BAL profiles that yieldmeasurements of the line full widths at half maximum(FWHMs), rest equivalent widths (REWs), and cover-ing fractions if available. Section 4 presents the basic re-sults and statistical analysis. Section 5 provides a briefsummary and discusses the broader implications of ourstudy for nature of quasar outflows and their possiblerole in feedback to galaxy evolution. Throughout thispaper, we adopt a cosmology with H = 71 km s − Mpc − , Ω M = 0 .
27 and Ω Λ = 0 . DATASETS & MINI-BAL SELECTION2.1.
Quasar Spectra
The parent sample of quasars for our study comes fromtwo datasets of archival high-resolution spectra obtainedwith VLT-UVES (Murphy et al. 2019) and Keck-HIRES(O’Meara et al. 2015). The VLT-UVES spectra fromMurphy et al. (2019) are all 468 quasars (468 spectra)first observed with UVES before mid-2008. All expo-sures of those 468 quasars recorded before late 2016 werecombined to form the final UVES spectra. The Keck-HIRES spectra from O’Meara et al. (2015) are the firstdata release of the KODIAQ survey containing all 170quasars (247 spectra, some quasars have multiple spec-tra obtained at different times) observed with HIRESbetween 2004 and 2012. The spectral resolutions arein the range 22 , (cid:46) R (cid:46) ,
000 for VLT-UVES and36 , (cid:46) R (cid:46) ,
000 for Keck-HIRES.We exclude spectra that have low signal-to-noise ra-tios, roughly SNR (cid:46) ∼ ∼ iv outflowlines. We also exclude spectra containing broad BALsor FeLoBALs that substantially cover these same im-portant wavelengths and thus limit our search for mini-BALs. A fraction of the KODIAQ spectra, roughly 5 to10 percent, needed to be excluded because they containmany spectral pixels without data due to problems indata reductions.Our final samples with suitable spectra include 450quasars (450 spectra) from UVES and 150 quasars (189spectra) from HIRES. Figure 1 shows the numbers ofquasars with spectral coverage as a function of velocityshift relative to C iv λ ∼
330 in Figure 1because some of the spectra do not cover the full velocityrange shown in this plot.
Chen et al. N u m b e r o f Q u a s a r s Velocity Shift (km/s)
Figure 1.
Number of quasars measured at different ve-locity shifts in the VLT-UVES, Keck-HIRES and combineddatasets. The velocities are relative to C iv λ The C iv mini-BAL catalog Our goal is to create a complete catalog of moderatelybroad C iv outflow lines (narrower than BALs) acrossthe velocity range plotted in Figure 1. Again, for conve-nience, we refer to all of these lines as mini-BALs. Werequire, specifically, that the lines have widths in therange 70 (cid:46) FWHM(1548) (cid:46) − (for just the λ iv doublet) at velocity shiftsv (cid:46) − − in the quasar rest frame (where neg-ative values indicate a blueshift). The upper limit onFWHM(1548) excludes BALs based on the common def-inition of BALs in the “balnicity” index (Weymann et al.1991). This maximum FWHM threshold is a soft limit,and we specifically search for broader outflow lines atlarge velocities (cid:46) − ,
000 km s − because those sys-tems are not counted in previous large BAL surveys(e.g., Trump et al. 2006; Pˆaris et al. 2017, and refs.therein). The minimum FWHM threshold is neededto avoid large numbers of narrow absorption lines thatform in cosmologically intervening gas or galaxies un-related to the quasars. We chose the specific thresholdFWHM (cid:38)
70 km s − after careful inspection of the spec-tral dataset; it is large enough to easily identify broadsmooth line troughs characteristic of outflows while alsobeing large enough to reject narrower lines that are morelikely to be intervening and unrelated to the quasars. Weare also guided by previous C iv NAL studies that at-tempt to identify outflow systems based on partial cover-ing signatures (Misawa et al. 2007a; Simon 2011). Thosestudies show that the lines without partial covering, andtherefore unclassified but more likely to be intervening,have FWHM that are typically (cid:46)
25 km s − . We imposea minimum velocity shift to avoid narrow “associated” absorption lines (AALs), near the quasar redshifts, thatcan appear in complex blends or have unusual absorp-tion troughs that, in spite of broad widths, have squareprofiles that do not appear BAL-like and are not likelyto form outflows like BALs and mini-BALs (see Hamannet al. 2001, for an example of the type broad low-velocitysystem we avoid with this velocity cut). We chose thespecific cutoff at ∼ − − to avoid specific caseslike this evident in our dataset.We construct the mini-BAL catalog by first visuallyinspecting every spectrum in our sample to select mini-BAL candidates based on the criteria above. We then fitthe line profiles and search for other lines at the same ve-locity shift. C iv λ − . For broad systems with unresolved doublets, weidentify C iv absorption by its velocity shift in the quasarframe and the presence or absence of other mini-BALsat the same shift (see Rodr´ıguez Hidalgo 2008; Hamannet al. 2013, for more discussion). We find that the con-tinuum fitting performed by O’Meara et al. (2015) andMurphy et al. (2019) to produce normalized spectra canremove real mini-BALs with broad profiles and intro-duce anomalous broad absorption dips, e.g., in the wingsor peaks of broad emission lines where there spectralslopes change dramatically. We avoid these problem forthe VLT-UVES sample by visually inspecting both thenormalized and unnormalized spectrum of every quasar.For the Keck-HIRES sample, only normalized spectraare available in the KODIAQ data release, but we checkall of the broad mini-BAL candidates by inspecting theunnormalized spectra available as quick-look reductionsin the online Keck Observatory Archive. This allowsus to reject spurious mini-BALs identified as candidatesin the normalized spectra. However, we do not inspectthe unnormalized Keck-HIRES spectra to select mini-BAL candidates and, therefore, some broad mini-BALsmight be missed in that sample if they are removed bythe normalization fits. For quasars with BALs in theirspectra, we include mini-BALs in our catalog only if theyhave clearly distinct features, e.g., residing fully outsidethe BAL troughs or they are clearly distinct features inthe weak wing of a BAL trough. These are subjectivechoices affecting a small number mini-BALs, which wediscuss further below.Next we inspect every candidate mini-BAL system forindications that it might be an unrelated intervening ab-sorption line system. The most obvious contaminants wefind among our mini-BAL candidates are C iv lines thatbelong to damped Ly α (DLA) or Lyman limit systems(LLSs), which form in intervening galaxies or extendedhalos and can have C iv lines broad enough to overlap Catalog of C iv Mini-BALs iv doublets in DLA systems that haveFWHM(1548) >
250 km s − . We automatically rejectall C iv systems if they have broad and deep Ly α linesindicative of DLAs or LLSs within our spectral coverage.If the Lyman lines are not covered, we search for low-ionization metal lines that are common in DLAs andLLSs, such as Si ii λ i λ ii λ ii λ iv line profilesand reject systems that resemble the C iv lines found inDLA systems, e.g., with deep and narrow profiles withsteep vertical sides, or highly asymmetric and clearlycomposed of blended narrow components. Overall, wereject systems with any reasonable possibility of beingintervening so that our final mini-BAL catalog is free ofcontamination. This conservative approach means thatsome real outflow mini-BALs (with unusual profiles orother characteristics) could be excluded from our survey.However, the number of mini-BAL candidates rejectedbecause they might belong to a DLA system, e.g., notobviously in DLAs by the criteria above, is small com-pared to the total number of mini-BALs in our finalcatalog. Thus, we estimate that our catalog is > (cid:38) .
03 ˚A for lines with FWHM(1548) ∼ − (or roughly double that REW for lines that aretwice as broad, etc.). LINE PROFILE FITSThe first step in our fitting procedure is to carefullyexamine the continuum placement provided with theUVES and HIRES archive spectra in the vicinity ofevery line we will fit. In some cases (e.g., for broadmini-BALs), the continuum extrapolated across the ab-sorption lines appears poorly constrained or too sharplyvarying compared to the unnormalized spectrum. Inthose cases, we redo the continuum fit using a simplepower law constrained locally near the line. We fit ev-ery candidate C iv mini-BAL using a Gaussian opticaldepth profile given by, τ v = τ e − ( v − v ) /b , (1)where τ v is the optical depth at velocity v , τ is the linecenter optical depth, v is the line center velocity, and b is the Doppler parameter. The doublet components inC iv λ b value. We account for partial covering -23200 -22600 -2200000.51 J103514+544040 FWHM=263 km/s-32500 -31900 -3130000.51 J055246-363727 FWHM=337 km/s-42900 -42300 -4170000.51 J233156-090802 FWHM=346 km/s-11600 -11000 -1040000.51 J095309+523029 FWHM=378 km/s-13800 -13200 -1260000.51 J175603+574848 FWHM=516 km/s Velocity Shifts (km/s) N o r m a li z e d F l u x Figure 2.
Normalized spectra showing examples of C iv ab-sorption lines in DLAs or LLSs, plotted on a velocity scalerelative to the quasar redshift (O’Meara et al. 2015; Murphyet al. 2019). The blue dotted lines mark the approximatecentroids of the two C iv doublet components. The linesplotted all have FWHM >
250 km s − , illustrating the over-lap with mini-BALs in FWHM parameter space, but theirmembership to DLA or Lyman Limit systems is confirmedby measurements of Ly α and other lines at the same redshift. Chen et al. in our line fits, but we assume for simplicity that thebackground light source has a spatially uniform bright-ness and the absorbing medium is homogeneous, withthe same optical depth along every sightline, so the ob-served intensity at velocity v is given by I v I = 1 − C v + C v e − τ v , (2)where I is the continuum intensity, I v is the measuredintensity at velocity v , and C v is the covering fraction ofthe absorbing medium across the emission source suchthat 0 < C v ≤ C v = C . This is justified byour fits below showing that, in general, the effects ofany velocity dependence in C v are negligible comparedto the overall value of the covering fraction captured bythe constant C .An important caveat to Equation (2) is that quasaroutflows are often spatially inhomogeneous, with a rangeof column densities and line optical depths across theprojected area of the emission source (see Section 5,also Barlow et al. 1997; de Kool et al. 2002; Hamannet al. 2001; Hamann & Sabra 2004; Arav et al. 2002,2005; Hamann et al. 2019a,b, and refs. therein). Thissituation leads to optical depth-dependent covering frac-tions, as derived from Equation (2). For example, if theabsorber spans a range of optical depths from thick tothin at different spatial locations, then 1) stronger tran-sitions will yield larger covering fractions because theyare optically thick over larger areas, and 2) the relativestrengths of different absorption lines (as in the C iv doublet) will depend on both the magnitude and spe-cific spatial distribution of the line optical depths acrossthe emission source. The actual spatial distributionscannot be determined from a simple doublet analysis.However, previous studies have shown that, even withinhomogeneous absorbers, Equation (2) yields useful es-timates of the τ (cid:38) iv mini-BALsincluded in our final catalog (described in Section 4 be-low). We measure v , τ and C from Equations (1)and (2) by fitting the C iv doublet lines simultaneously,with their τ values fixed to the ∼ iv doublet lines in the mini-BALare broad enough to be fully blended together, 2) themini-BAL is badly blended with another mini-BAL or unrelated lines, and 3) the mini-BALs have complex pro-files not well-characterized by a Gaussian optical depthfunction. We deal with each of these complicated casesas follows: Case 1)
We fit broad blended mini-BALs doublets as-suming the lines have a ∼ τ and C (equalto the line depth below the continuum). If the broadblended mini-BALs, instead, have roughly flat-bottomtroughs, good fits require τ (cid:29) C ≈ − I v /I . Case 2)
We fit all C iv mini-BALs simultaneously ifthere are two or more such systems blended together.Examples of this are the low-velocity mini-BALs inJ131215+423900 and J211654-433234 (Figure 3). Weidentify and fit distinct mini-BALs in these blends onlyif they are clearly distinct by visual inspection, e.g., withwell-measured absorption minima separated by (cid:38) − . In some cases, we fit one C iv doublet to whatis probably a more complex blend of mini-BALs. Thisprocedure is somewhat subjective, but we are deliber-ately conservative to avoid over-counting the numbermini-BALs in our dataset. For C iv mini-BALs blendedwith unrelated absorption lines (at some other redshift),we simply mask out the spectral regions containing theunrelated lines before fitting. This is generally straight-forward. However, there are rare cases where the fits arepoorly constrained because the masked regions are largedue to BALs near the mini-BALs (e.g., for the mini-BALs at − − in J235702-004824 and roughly − − in J221531-174408 in Figure 3) or be-cause part of the spectrum is missing due to gaps inwavelength coverage between the echelle orders (e.g., forthe mini-BAL at − − in J102325+514251 inFigure 3). Case 3)
We fit mini-BALs with complex/non-Gaussian profiles using the minimum number of Gaus-sian components to provide a good fit. Again, our goal isto avoid over-counting the mini-BALs in our study. Asnoted for Case 2 above, distinct components must havewell-measured absorption minima separated by (cid:38) − . The fits to these complex profiles are not ideal, Catalog of C iv Mini-BALs − ,
966 km s − in J024221+004912, the one at − − in J104642+053107, the one at − − in J211654-433234, to highly-structured blend ofmini-BALs at − − of J104032-272749 (Fig-ure 3).All of the broad and blended mini-BALs are flagged inthe catalog notes in Tables 1 and 2. Additional notes onspecific cases with complex blends, under-counted out-flow lines, or problematic/uncertain continuum place-ments are provided in Appendix A. RESULTS & ANALYSISTables 1 and 2 provide our final catalog of C iv mini-BALs in the VLT-UVES and Keck-HIRES datasets, re-spectively. The measured quantities in these tables arethe velocity shift, v , covering fraction C , rest equiva-lent widths, REW, given separately for the C iv λ iv λ σ errors output by thefitting software; these are due mainly to pixel-to-pixelnoise fluctuations in the spectra. The tables also includeNotes to indicate specific cases with potentially largeruncertainties due blends, uncertain continuum fits, orother problems. These notes are explained in the tablecaption. One of those designations is “cmplx” to indi-cate that the recorded mini-BALs belong to a complexof related outflow lines. We define an absorption-linecomplex as three or more C iv mini-BALs that appearto be related based on similar profiles and roughly sim-ilar velocity shifts ( (cid:46) − apart). These com-plexes sometimes include outflow lines not in our mini-BAL catalog because they are too narrow or at velocitiesv > − − (see Section 4.3 for specific examplesand discussion). Chen et al. -15400 -14900 -144000.80.91.01.1
J000448-415728 C =0.22FWHM=316 km/s -22600 -19200 -157000.900.951.001.051.10 J000448-415728 C -5500 -5000 -45000.80.91.01.1 J005758-264314 C =0.13FWHM=170 km/s -19800 -16500 -132000.500.751.001.251.50 J005758-264314 C -5200 -4400 -35000.00.51.01.5 J013405+005109 C =0.64FWHM=417 km/s -6900 -6400 -59000.81.01.2 J013405+005109 C =0.17FWHM=261 km/s -5300 -4800 -43000.51.01.5 J015327-431137 C =0.49FWHM=226 km/s -6600 -6100 -56000.951.001.05 J015327-431137 C =1FWHM=113 km/s -8700 -7800 -70000.60.81.01.2 J015327-431137 C =0.29, 0.21FWHM=169, 264 km/s -5200 -4000 -29000.00.51.01.5 J024221+004912 C =0.91FWHM=564 km/s -12200 -11700 -112000.51.01.5 J024221+004912 C =0.5FWHM=175 km/s -14300 -13800 -133000.91.01.1 J024221+004912 C =0.1FWHM=91 km/s -21200 -17600 -140000.00.51.01.5 J024221+004912 C -14300 -13800 -133000.91.01.1 J042353-261801 C =0.1FWHM=121 km/s -20500 -19300 -182000.500.751.001.25 J042353-261801 C =0.3, 0.14FWHM=448, 130 km/s -27800 -24500 -212000.60.81.01.21.4 J042353-261801 C -9300 -8800 -83000.60.81.01.21.4 J064326-504112 C =0.28FWHM=206 km/s -23300 -18000 -126000.60.81.01.2 J091127+055054 C -25900 -23500 -211000.60.81.01.21.4 J103909-231326 C -15200 -13900 -126000.00.51.01.5 J103921-271916 C =0.71, 0.86FWHM=153, 498 km/s -44000 -42400 -409000.51.01.5 SiIV
J103921-271916 C -17100 -15700 -142000.51.01.5 J104032-272749 C =0.21, 0.37FWHM=235, 497 km/s -24700 -20700 -168000.51.01.5 J104032-272749 C -12200 -11100 -100000.51.01.5 J104033-272308 C =0.48FWHM=572 km/s Velocity Shifts (km/s) N o r m a li z e d F l u x Figure 3.
Normalized C iv mini-BAL profiles plotted on a velocity scale relative to the quasar emission-line redshift. Thespectra are shown in black. Our fits to the lines are shown in red. The dash lines with different colors indicate separate mini-BAL systems, and we fit the separate mini-BAL systems in each panel simultaneously. The green boxes highlight the systemswith definite C < .
1, where we exclude the ones with only lower limits and the ones with uncertain measurements, and theorange boxes highlight the systems with extremely high speeds, v < − ,
000 km/s. Observing dates are given (in blue text) ifquasars observed more than once. The velocities pertain to the short-wavelength lines in the doublets.
Catalog of C iv Mini-BALs -10300 -9800 -93000.00.51.01.5 J104642+053107 C =0.67FWHM=111 km/s -11100 -9800 -85000.00.51.01.5 J104733+052454 C -20500 -17300 -141000.51.01.5 J104733+052454 C -12200 -10000 -78000.00.51.01.5 J112010-134625 C =0.73, 0.47FWHM=722, 359 km/s -13700 -13200 -127000.500.751.001.251.50 J112010-134625 C =0.3FWHM=301 km/s -27400 -26900 -264000.900.951.001.05 J112010-134625 C =0.09FWHM=395 km/s -12200 -11700 -112000.51.01.5 J113010+041128 C =1FWHM=211 km/s -5100 -4000 -29000.60.81.01.21.4 J115122+020426 C -7500 -5800 -42000.00.51.01.5 J115122+020426 C -11600 -8700 -59000.00.51.01.5 J115122+020426 C =0.14, 0.65FWHM=265, 1411 km/s -9700 -7700 -56000.00.51.01.5 J115944+011206 C -45100 -43900 -428000.91.01.1 J115944+011206 C -16000 -14300 -126000.51.01.5 J122848-010414 C -14100 -11800 -94000.81.01.2 J134258-135559 C -12500 -9900 -73000.81.01.2 J134427-103541 C -4300 -3800 -33000.900.951.001.051.10 J144653+011356 C =0.08FWHM=208 km/s -5200 -4400 -36000.51.01.5 within 4000 km/sJ151352+085555 C =0.45FWHM=406 km/s -6800 -6300 -58000.91.01.1 J151352+085555 C =0.1FWHM=145 km/s -5700 -5000 -42000.500.751.001.251.50 J211654-433234 C =0.51, 0.26FWHM=197, 190 km/s -8200 -7100 -60000.00.51.01.5 J211654-433234 C -13300 -12800 -123000.00.51.01.5 J211654-433234 C =0.56FWHM=280 km/s -9300 -8300 -73000.81.01.2 J212329-005052 C =0.24, 0.18FWHM=103, 70 km/s -14000 -12400 -108000.81.01.2 J212329-005052 C =0.16, 0.22, 1FWHM=90, 199, 174 km/s Velocity Shifts (km/s) N o r m a li z e d F l u x Figure 3. continued. Chen et al. -17400 -16500 -156000.81.01.2
J214159-441325 C =0.21, 0.20FWHM=109, 341 km/s -21300 -20500 -196000.60.81.01.21.4 J214159-441325 C =0.25FWHM=433 km/s -13500 -13000 -125000.81.01.2 J214225-442018 C -22700 -22200 -217000.500.751.001.251.50 J214225-442018 C =0.42FWHM=352 km/s -8800 -6900 -50000.00.51.01.5 J221531-174408 C -11900 -10800 -97000.60.81.01.2 J221531-174408 C =0.21FWHM=570 km/s -15300 -14000 -127000.91.01.1 J221531-174408 C -4600 -4100 -36000.951.001.05 J222006-280323 C =0.06FWHM=196 km/s -10700 -8400 -61000.51.01.5 J232046-294406 C =0.43, 0.40FWHM=549, 596 km/s -24400 -23000 -216000.81.01.2 J234628+124858 C -5000 -4500 -40000.51.01.5 J235534-395355 C =0.49FWHM=80 km/s -4800 -4300 -38000.00.51.01.5 J235702-004824 C -11800 -9700 -76000.51.01.5 J235702-004824 C -6000 -5500 -50000.80.91.01.1 J100841+362319 C =0.14FWHM=368 km/s -10900 -10400 -99000.91.01.1 J101723-204658 C =0.12FWHM=102 km/s -4800 -4300 -38000.51.01.5 J102325+514251 C -6200 -5700 -52000.80.91.01.1 J102325+514251 C =0.12FWHM=292 km/s -4700 -4200 -37000.51.01.5 J115940-003203 C =0.44FWHM=317 km/s -7000 -6500 -60000.91.01.1 J115940-003203 C =0.08FWHM=210 km/s -8300 -7800 -73000.500.751.001.251.50 J115940-003203 C =0.33FWHM=314 km/s -10100 -9600 -91000.500.751.001.251.50 J115940-003203 C =0.34FWHM=278 km/s -7700 -5900 -41000.00.51.01.5 J131215+423900 C -21500 -18100 -147000.51.01.5 J131215+423900 C -5300 -3900 -25000.00.51.01.5 J141719+413237 C =0.93, 0.89FWHM=196, 512 km/s Velocity Shifts (km/s) N o r m a li z e d F l u x Figure 3. continued.
Catalog of C iv Mini-BALs -11400 -10900 -104000.00.51.01.5 J160455+381214 C =0.36FWHM=269 km/sApr 2005 -11400 -10900 -104000.00.51.01.5 J160455+381214 C =0.3FWHM=276 km/sMay 2009 -5000 -4500 -40000.500.751.001.25 J160843+071508 C =0.36FWHM=161 km/sApr 2007 -5000 -4500 -40000.51.01.5 J160843+071508 C =0.4FWHM=173 km/sJul 2008 -23600 -23100 -226000.751.001.25 J162453+375806 C =0.25FWHM=235 km/s -5500 -5000 -45000.51.01.5 J175603+574848 C =0.65FWHM=141 km/sOct 2004 -5500 -5000 -45000.00.51.01.5 J175603+574848 C =0.64FWHM=153 km/sAug 2006 -5500 -5000 -45000.51.01.5 J175603+574848 C =0.61FWHM=160 km/sJul 2008 -9300 -8300 -73000.51.01.5 J212329-005052 C =0.31, 0.33FWHM=104, 74 km/sAug 2006 -14000 -12400 -108000.51.0 J212329-005052 C -8400 -7600 -68000.500.751.001.25 J231324+003444 C =0.27, 0.34FWHM=233, 170 km/s -9700 -9200 -87000.91.01.1 J231324+003444 C -13300 -10700 -82000.00.51.01.5 J231324+003444 C -24000 -22900 -217000.51.01.5 J231324+003444 C =0.56FWHM=588 km/s -32500 -30600 -288000.81.01.2 J231324+003444 C =0.13FWHM=936 km/s Velocity Shifts (km/s) N o r m a li z e d F l u x Figure 3. continued. Chen et al.
Table 1 . C iv mini-BALs in VLT-UVES. Columns show the quasar names, emission-line redshifts, z em , velocity shifts, v , of theabsorption-line centroid from z em , rest equivalent widths, REWs, for λ λ iv doublet, FWHMs for the separateC iv lines, covering fractions, C , inferred from the line fits, and Notes on individual features: ‘broad’ = broad profiles where theC iv doublet components are blended together and indistinguishable as separate troughs; ‘bl1’ = asymmetric non-gaussian profilethat we fit approximately with a single gaussian component; ‘bl2’ = blended with distinct unrelated lines (if the blended system isalso a measured mini-BAL, we identify the blended system by its velocity shift in the notation bl2-xxxx); ‘bl-BAL’ = in the wingof a much broader C iv BAL; ‘cmplx’ = appears to belong to a complex of related outflow lines, see Section 4.0; ‘BALQSO’ = thequasar has a C iv BAL; ‘unc’ = the fit data listed have uncertainties larger than indicated due to blends or uncertain continuumplacement; ‘mBAL?’ = the fitted feature appears to be a mini-BAL but it has some characteristics resembling C iv lines in DLAsystems; ‘var’ = variability is identified between two or more VLT-UVES and/or Keck-HIRES observations included in our studyvia visual inspection (these variable systems have multiple table entries distinguished in the Notes by the observation dates). QSO z em v REW( λ λ λ C Notes(km s − ) (˚A) (˚A) (km s − )J000448-415728 2.760 -15136 0 . ± .
03 0 . ± .
02 316 . ± . . ± .
04 bl1-19565 0 . ± .
01 0 . ± .
02 1712 . ± . (cid:38) .
06 broadJ005758-264314 3.655 -5273 0 . ± .
01 0 . ± .
01 169 . ± . . ± .
01 bl2-16876 2 . ± .
07 1 . ± .
06 1636 . ± . (cid:38) .
34 broad, bl2, uncJ013405+005109 1.520 -4501 1 . ± .
04 0 . ± .
05 416 . ± . . ± .
02 bl2-6666 0 . ± .
03 0 . ± .
03 260 . ± . . ± .
02 uncJ015327-431137 2.740 -5080 0 . ± .
01 0 . ± .
04 226 . ± . . ± .
01 cmplx-6402 0 . ± .
01 0 . ± .
00 113 . ± . . . ± .
00 0 . ± .
00 264 . ± . . ± .
00 bl2-8387, cmplx-8387 0 . ± .
02 0 . ± .
02 169 . ± . . ± .
06 bl2-7789, cmplxJ024221+004912 2.068 -4189 2 . ± .
02 2 . ± .
07 564 . ± . . ± . . ± .
01 0 . ± .
09 174 . ± . . ± . . ± .
01 0 . ± .
02 90 . ± . . ± . . ± .
48 2 . ± .
32 1802 . ± . (cid:38) .
56 broad, bl2, uncJ042353-261801 2.277 -14013 0 . ± .
01 0 . ± .
02 121 . ± . . ± . . ± .
02 0 . ± .
02 129 . ± . . ± . . ± .
03 0 . ± .
03 448 . ± . . ± . . ± .
23 1 . ± .
30 1659 . ± . (cid:38) .
26 broad, uncJ064326-504112 3.090 -9017 0 . ± .
01 0 . ± .
03 206 . ± . . ± . . ± .
03 2 . ± .
07 2692 . ± . (cid:38) .
25 broadJ103909-231326 3.130 -23808 1 . ± .
05 1 . ± .
11 1213 . ± . (cid:38) .
27 broadJ103921-271916 2.230 -13684 2 . ± .
02 1 . ± .
19 497 . ± . . ± .
00 cmplx, bl2-14816 0 . ± .
01 0 . ± .
18 153 . ± . . ± .
01 cmplx, bl2-42626 0 . − .
42 564 . − . (cid:38) .
49 broad, bl2 with Si iv , uncJ104032-272749 2.320 -15623 0 . ± .
02 0 . ± .
03 496 . ± . . ± .
01 bl2-16568, cmplx-16568 0 . ± .
01 0 . ± .
01 235 . ± . . ± .
01 bl2-15623, cmplx-19506 0 . ± .
02 0 . ± .
08 396 . ± . . ± .
01 cmplx-21767 4 . ± .
06 3 . ± .
15 1966 . ± . (cid:38) .
40 broad, cmplx, bl1, uncJ104033-272308 1.937 -11259 1 . ± .
04 0 . ± .
08 571 . ± . . ± .
01 bl-BAL, BALQSOJ104642+053107 2.698 -10018 0 . ± .
01 0 . ± .
10 110 . ± . . ± . . ± .
03 0 . ± .
02 158 . ± . (cid:38) . . ± .
02 0 . ± .
03 307 . ± . . ± .
01 bl2-10552, cmplx-10552 0 . ± .
04 — 183 . ± . (cid:38) . . ± .
04 1 . ± .
60 1593 . ± . (cid:38) . . ± .
02 0 . ± .
05 359 . ± . . ± .
01 bl2-10806, BALQSO-10806 2 . ± .
02 2 . ± .
13 721 . ± . . ± .
00 bl2-10002, BALQSO-13416 0 . ± .
01 0 . ± .
07 300 . ± . . ± .
00 unc, bl-BAL, BALQSO-27118 0 . ± .
02 0 . ± .
02 395 . ± . . ± .
02 unc, bl-BAL, BALQSO
Table 1 continued
Catalog of C iv Mini-BALs Table 1 (continued)
QSO z em v REW( λ λ λ C Notes(km s − ) (˚A) (˚A) (km s − )J113010+041128 3.930 -11975 0 . ± .
14 0 . ± .
09 211 . ± . . . ± .
02 0 . ± .
02 360 . ± . . ± .
01 bl2-4529, bl2, cmplx-4529 0 . ± .
17 0 . ± .
10 194 . ± . (cid:38) .
30 bl2-4107, cmplx-6046 0 . ± .
07 0 . ± .
06 292 . ± . (cid:38) .
38 bl1, bl2-6578, cmplx-6578 1 . ± .
19 1 . ± .
14 520 . ± . (cid:38) .
71 bl1, bl2-6046, cmplx-9024 4 . ± .
10 2 . ± .
10 1411 . ± . (cid:38) .
65 broad, cmplx, unc-10703 0 . ± .
02 0 . ± .
03 265 . ± . . ± .
01 cmplx, uncJ115944+011206 2.000 -7917 2 . ± .
23 1 . ± .
15 1033 . ± . (cid:38) .
65 broad-44093 0 . ± .
02 0 . ± .
07 579 . ± . (cid:38) .
09 broad, uncJ122848-010414 2.655 -14474 1 . ± .
18 0 . ± .
12 849 . ± . (cid:38) .
40 broadJ134258-135559 3.190 -12070 0 . ± .
07 0 . ± .
05 1168 . ± . (cid:38) .
15 broadJ134427-103541 2.134 -10226 0 . ± .
07 0 . ± .
06 1293 . ± . (cid:38) .
12 broadJ144653+011356 2.206 -4004 0 . ± .
01 0 . ± .
01 208 . ± . . ± .
01 cmplx, uncJ151352+085555 2.904 -4922 0 . ± .
01 0 . ± .
04 405 . ± . . ± .
01 bl2, cmplx-6564 0 . ± .
01 0 . ± .
01 144 . ± . . ± .
01 cmplxJ211654-433234 2.053 -5180 0 . ± .
03 0 . ± .
03 189 . ± . . ± .
04 bl2-5409, cmplx-5409 0 . ± .
07 0 . ± .
05 197 . ± . . ± .
10 bl2-5180, cmplx-7274 2 . ± .
05 1 . ± .
10 554 . ± . (cid:38) .
75 broad, bl1, cmplx-13061 0 . ± .
03 0 . ± .
10 280 . ± . . ± . . ± .
01 0 . ± .
02 69 . ± . (cid:38) .
18 bl2, cmplx, var (Aug 2008)-9148 0 . ± .
02 0 . ± .
02 102 . ± . . ± .
05 cmplx, var (Aug 2008)-11521 0 . ± .
02 0 . ± .
02 174 . ±
52 1 . . ± .
01 0 . ± .
03 199 . ± . . ± .
02 cmplx, bl2, var (Aug 2008)-13568 0 . ± .
01 0 . ± .
02 89 . ± . . ± .
02 cmplx, bl2, var (Aug 2008)J214159-441325 3.170 -16414 0 . ± .
01 0 . ± .
05 340 . ± . . ± .
01 bl2, cmplx-17319 0 . ± .
01 0 . ± .
04 109 . ± . . ± .
02 cmplx-20573 0 . ± .
01 0 . ± .
08 432 . ± . . ± .
01 cmplxJ214225-442018 3.230 -13239 0 . ± .
10 0 . ± .
06 135 . ± . (cid:38) .
20 mBAL?, bl1, unc-22453 0 . ± .
04 0 . ± .
09 351 . ± . . ± . . ± .
02 1 . ± .
02 440 . ± . (cid:38) .
63 bl2-7711, bl-BAL, cmplx, BALQSO-7711 1 . ± .
21 0 . ± .
15 517 . ± . (cid:38) .
75 bl2-7155, bl-BAL, cmplx, BALQSO-10971 0 . ± .
01 0 . ± .
03 570 . ± . . ± .
00 BALQSO-14179 0 . ± .
01 0 . ± .
02 665 . ± . (cid:38) .
09 broad, bl2, unc, BALQSOJ222006-280323 2.406 -4355 0 . ± .
01 0 . ± .
01 196 . ± . . ± .
01 uncJ232046-294406 2.401 -8148 1 . ± .
05 0 . ± .
06 595 . ± . . ± . . ± .
04 0 . ± .
05 549 . ± . . ± .
01 bl1J234628+124858 2.515 -23164 0 . ± .
05 0 . ± .
06 691 . ± . (cid:38) .
20 broadJ235534-395355 1.580 -4756 0 . ± .
01 0 . ± .
03 79 . ± . . ± . . ± .
07 — 312 . ± . (cid:38) .
81 bl2, bl-BAL, BALQSO-9934 1 . ± .
18 1 . ± .
13 1053 . ± . (cid:38) .
35 broad, BALQSO
Table 2 . C iv mini-BALs in Keck-HIRES. See Table 1 for descriptions of the table contents. QSO z em v REW( λ λ λ C Notes(km s − ) (˚A) (˚A) (km s − )J100841+362319 3.126 -5728 0 . ± .
03 0 . ± .
03 321 . ± . . ± .
02 bl1, cmplxJ101723-204658 2.545 -10666 0 . ± .
00 0 . ± .
00 101 . ± . . ± .
05 uncJ102325+514251 3.447 -4544 0 . ± .
05 — 288 . ± . (cid:38) .
45 bl1, cmplx-5931 0 . ± .
01 0 . ± .
02 292 . ± . . ± .
01 bl1, cmplxJ115940-003203 2.035 -4469 0 . ± .
08 0 . ± .
08 316 . ± . . ± .
06 bl2, cmplx
Table 2 continued Chen et al.
Table 2 (continued)
QSO z em v REW( λ λ λ C Notes(km s − ) (˚A) (˚A) (km s − )-6730 0 . ± .
01 0 . ± .
01 210 . ± . . ± .
01 bl1, cmplx-8045 0 . ± .
06 0 . ± .
07 314 . ± . . ± .
04 cmplx-9835 0 . ± .
05 0 . ± .
06 278 . ± . . ± .
03 bl2, cmplxJ131215+423900 2.567 -5459 0 . ± .
01 0 . ± .
05 170 . ± . . ± .
01 cmplx, bl2-6433 0 . ± .
02 0 . ± .
08 211 . ± . . ± .
01 cmplx, bl2-6570, bl2-6872-6570 0 . ± .
01 0 . ± .
06 271 . ± . . ± .
01 cmplx, bl2-6433-6872 0 . ± .
34 — 267 . ± . (cid:38) .
48 cmplx, bl2-6433, bl2-7273-7273 0 . ± .
14 — 279 . ± . (cid:38) .
10 cmplx, bl2-6872-18030 0 . ± .
56 0 . ± .
35 673 . ± . (cid:38) .
21 broad-19778 2 . ± .
06 2 . ± .
13 1337 . ± . (cid:38) .
43 broad, bl1, uncJ141719+413237 2.024 -4136 2 . ± .
03 1 . ± .
04 512 . ± . . ± .
01 bl2-4483, cmplx, BALQSO-4483 0 . ± .
05 0 . ± .
06 196 . ± . . ± .
02 bl2-4136, cmplx, BALQSOJ160455+381214 2.551 -11104 0 . ± .
01 0 . ± .
06 269 . ± . . ± .
01 var (Apr 2005), cmplx-11098 0 . ± .
01 0 . ± .
05 275 . ± . . ± .
01 var (May 2009), cmplxJ160843+071508 2.877 -4710 0 . ± .
02 0 . ± .
02 161 . ± . . ± .
04 var (Apr 2007), cmplx-4705 0 . ± .
03 0 . ± .
03 172 . ± . . ± .
04 var (Jul 2008), cmplxJ162453+375806 3.380 -23352 0 . ± .
01 0 . ± .
09 234 . ± . . ± .
01 bl2J175603+574848 2.110 -5274 0 . ± .
05 0 . ± .
04 141 . ± . . ± .
09 var (Oct 2004), bl2, cmplx-5276 0 . ± .
03 0 . ± .
04 153 . ± . . ± .
04 var (Aug 2006), bl2, cmplx-5274 0 . ± .
02 0 . ± .
04 160 . ± . . ± .
03 var (Jul 2008), bl2, cmplxJ212329-005052 2.262 -8138 0 . ± .
01 0 . ± .
01 74 . ± . (cid:38) .
33 bl2, cmplx, var (Aug 2006)-9117 0 . ± .
01 0 . ± .
01 103 . ± . . ± .
01 cmplx, var (Aug 2006)-11501 0 . ± .
05 0 . ± .
04 178 . ± . . ± .
11 cmplx, bl1, unc, var (Aug 2006)-12337 0 . ± .
01 0 . ± .
06 197 . ± . . ± .
01 cmplx, bl2, var (Aug 2006)-13546 0 . ± .
05 0 . ± .
02 82 . ± . (cid:38) .
10 cmplx, bl2, unc, var (Aug 2006)J231324+003444 2.083 -7490 0 . ± .
07 0 . ± .
05 169 . ± . . ± .
14 bl2, cmplx-8245 0 . ± .
02 0 . ± .
02 233 . ± . . ± .
01 cmplx-9438 0 . ± .
06 0 . ± .
05 397 . ± . (cid:38) .
10 cmplx, unc-11036 4 . ± .
44 2 . ± .
31 1275 . ± . (cid:38) .
70 broad, cmplx-23025 1 . ± .
14 0 . ± .
15 587 . ± . . ± .
05 broad-30887 0 . ± .
05 0 . ± .
07 936 . ± . . ± .
00 mBAL?, broad, unc
Mini-BAL Properties & Statistics
Overall we find 105 high-velocity C iv mini-BALs in44 quasars out of 638 total quasars in our sample. Ta-ble 3 provides a more detailed break down of some ofthe sample statistics. For example, 44 quasars (38 non-BALQSOs) have ≥ iv mini-BAL while 25 quasars(21 non-BALQSOs) have ≥ iv mini-BALs.Figure 4 shows the numbers of mini-BALs hav-ing different values of FWHM(1548), velocity shift,REW(1548), and covering fractions. The top panelin Figure 4 shows that most of the mini-BALs in ourstudy are relatively narrow. In particular, the me-dian FWHM(1548) is 300 km s − and ∼
25% of themini-BALs have FWHM(1548) (cid:46)
200 km s − . Sim-ilarly, most of the mini-BALs are weak, with me-dian REW(1548) ∼ > − adopted by Rodr´ıguez Hidalgo (2008) for theirSDSS study to our sample would miss 75% of the mini-BALs in our catalog.The distribution of velocity shifts shown in Figure 4rises steeply toward small outflow velocities. Note thatthe distribution is truncated at v ∼ − − by ourselection constraint (Section 2.2). Overall, the velocityrange of the mini-BALs is similar to BALs, including3 mini-BALs ( ∼
3% of the sample) at high velocities
Catalog of C iv Mini-BALs < − ,
000 km s − . These rare high-velocity cases arehighlighted by orange boxes in Figure 3. Conversely, ∼
35% of the mini-BALs in our sample are at low velocitiesfrom − − − .The bottom panel in Figure 4 shows separately the dis-tributions of well-measured covering fractions for whichwe have direct values (from saturated lines and/or re-solve doublets) or firm lower limits (Section 3). Thesecovering fraction distributions indicate that most andpossibly all C iv mini-BALs in our catalog have partialcovering. In particular, all of the well-measured C val-ues are <
1. Therefore, considering only well-measuredsystems, 100% have C < ∼
75% have C (cid:46) .
5, and ∼
7% have definite C < .
1. The weak mini-BALs withwell-measured C (cid:46) . v (cid:46) − − we consider. In the HIRESsample, we find only two quasars, J100841+362319 andJ102325+514251, that were targeted (by our team foranother study) to have complexes of AALs that could in-clude outflow lines (Chen et al. 2019; Simon & Hamann2010b; Simon et al. 2012). All of the other quasars ap-pear to have been targeted for studies of i) interveningabsorption lines (e.g., DLAs and LLSs) that should beunbiased for mini-BALs in the spectra, or ii) associatedabsorption lines (D’Odorico et al. 2004) at low velocitiesthat are not part of our catalog.The raw fraction of quasars with at least one mini-BAL in our study, 44/638 ≈ ∼
9% of quasars have at least one mini-BAL meeting the definitional requirements of our study(v (cid:46) − − and 70 (cid:46) FWHM(1548) (cid:46) − , Section 2.2). Excluding BAL quasars from the par-ent sample makes only a small difference, such that ∼ Correlation Tests
FWHM (km/s)
Velocity Shift (1000 km/s)
REW ( Å ) Covering Fraction C with True Values C with Lower Limits N u m b e r o f m i n i - B A L s Figure 4. C iv mini-BAL parameter distributions, from toppanel to bottom: FWHM(1548), velocity shift, REW(1548),and covering fraction. Rare cases with large uncertainties(due to poorly constrained fits, Section 3) are excluded. Thedashed vertical lines show the median values (300 km s − , − . × km s − , 0.51 ˚A, and 0.34, respectively, from topto bottom panel) in the distributions. Chen et al.
Table 3.
Number of quasars with C iv mini-BALs and numberof C iv mini-BALs in our catalog. Results for all quasars andnon-BAL quasars in our dataset are listed separately. Categories ≥ ≥ ≥ ≥ Figure 5 shows various C iv mini-BAL parametersplotted against the velocity shift and covering fractionto test for trends. There is generally large scatter.The only significant trend is for larger FWHM(1548)at larger velocity shifts (top panel). This relationshiphas a probability of occurring by chance of P = 0 . ∼ − − − .There is also tentative trend for mini-BALs at largervelocity shifts to have smaller covering fractions (thirdpanel from top). The probability of this correlationoccurring by chance P = 6%. The broader lineswith smaller covering fractions and nominally shallowertroughs result in the lines having no significant trend inREW(1548) with velocity shift (second panel from thetop). 4.3. Mini-BALs in Outflow Complexes
Figure 6 shows an expanded view of all C iv mini-BALs that belong to outflow absorption-line complexesas defined in Section 4.0, e.g., with 3 or more compo-nents that appear physically related less than 3000 kms − apart (see also Tables 1 and 2). These absorption-line complexes are interesting because they trace spa-tially complex outflow structures near the quasars.Roughly 45% of the quasars with mini-BALs in ourstudy have an outflow complex by this definition, and ∼
53% of all mini-BALs in our catalog appear in sucha complex. The outflow complexes in J100841+362319,J102325+514251 and J212329-005052 were studied indetail by (Chen et al. 2019; Simon & Hamann 2010a;Hamann et al. 2011). Note that some of the lines in thesecomplexes are narrower and/or at smaller velocity shiftsthan we consider for the mini-BALs in our study. Theirvelocity shifts overall span a wide range from ∼ − , ∼ − while the FWHMs in the complex -30 -25 -20 -15 -10 -5Velocity Shift (10 km/s)050010001500200025003000 F W H M ( k m / s ) slope=-0.028±0.006-30 -25 -20 -15 -10 -5Velocity Shift (10 km/s)0123456 R E W ( Å ) slope=(-19±14) ×10 -30 -25 -20 -15 -10 -5Velocity Shift (10 km/s)0.00.20.40.60.81.0 C o v e r i n g F r a c t i o n slope=(96±50) ×10 F W H M ( k m / s ) slope=99±89 Figure 5.
Mini-BAL parameter relationships. Fromtop panel to bottom: FWHM(1548) vs. velocity shift;REW(1548) vs. velocity shift; covering fractions C vs. ve-locity shift; and FWHM(1548) vs. covering fraction. As inFigure 5, rare cases with large uncertainties (estimated tobe larger than their formal values returned from the fits) areexcluded. The red lines show linear least squares fits to theplotted points. In the two bottom panels, the fits apply onlyto C with derived specific values. The linear slopes andassociated 1 σ errors are given in each panel. The shadedregions indicate the 3 σ uncertainties in the linear fits. Catalog of C iv Mini-BALs ∼
70 to ∼ − , i.e., thefull range of FWHMs considered in our study.4.4. P v and Other Ions For every C iv mini-BAL detected in our survey,we search for other absorption lines at the same ve-locity shift, such as Ly α λ iv λ v λ vi λ v λ ii λ ii λ v absorption is particularly important asan indicator of large outflow column densities (Sections1 and 5). Thus we expand our search for P v absorp-tion to include strong outflow systems at lower velocityshifts than our survey limit v < − − . Figure 7shows all of the outflow systems where P v absorptionis clearly detected. One of the quasars with P v absorp-tion in Figure 7 is a BAL, J221531-174408. The othersare mini-BALs or some type of broad outflow AALs.To our knowledge, there is only one other published re-port of P v absorption in an individual non-BAL quasar(by Hamann et al. 2019b). However, the study of me-dian quasar spectra by Hamann et al. (2019a) showsthat BAL and mini-BAL systems typically have P v ab-sorption at a fraction of the strength of C iv . The actualnumber of quasars with P v absorption in our mini-BALsample is likely to be larger what is shown in Figure 7because secure detections of P v lines can be thwartedin our study by 1) unrelated absorption lines in theLy α forest, 2) observed wavelength coverages that usu-ally do not include P v and 3) the spectra often havinglow signal-to-noise ratios at short observed wavelengthswhere the P v lines are found.All of the mini-BAL systems with securely detectedP v lines have saturated absorption in C iv , Si iv andO vi (when available) based on ∼ v lines should be accompa-nied by large optical depths in these other lines, specif-ically τ (cid:38) iv . The measured P v absorptiontroughs in Figure 7 also tend to be (much) narrower thanC iv and other high-ionization lines like O vi . These re-sults are in good agreement with other measurementsof P v in individual BAL quasars (Chamberlain et al.2015; Capellupo et al. 2017; Moravec et al. 2017, andrefs therein) and with the median composite spectra oflarge BAL and mini-BAL quasar samples presented in(Hamann et al. 2019a).Figures 8 and 9 show all of the remaining mini-BALsystems in our catalog with lines other than C iv clearlydetected. All of these multi-ion systems, includingthose with P v detections in Figure 7, exhibit a general trend for deeper absorption troughs and wider profilesin strong transitions of higher ions such as C iv , N v ,and O vi compared lower ions like Si iv and weaker linesdue to low abundance like P v . The mini-BAL systemswith resolved doublets in Figures 7 and 8 indicate thatall of the lines are saturated with moderate depths dueto partial covering. This implies that the progressionin line depths and velocity widths with ionization andline optical depth is directly tied to a trend in the cov-ering fractions, namely, that the strong lines of higherions form in spatially larger regions while weaker tran-sitions and lines of lower ions tend to form in smallerpockets of higher column density gas. These resultsare consistent with other studies of individual outflowquasars and composite BAL and mini-BAL spectra (e.g.,citations in previous paragraph). We discuss the impli-cations of these results for the physical conditions andspatial structure of quasar outflows in Section 5.4.5. Line-locked mini-BALs
Line-locked absorption-line systems have distinct com-ponents at velocity separations equal to prominent dou-blets such as the C iv , N v , or Si iv . Lines formed indifferent outflow clumps can become locked at thesedoublet separations due shadowing effects in radiatively-driven outflows. Line-locked systems can therefore be asignature of outflows driven by radiative forces in thevicinity of a quasar (Milne 1926; Scargle 1973; Braun& Milgrom 1989). It is also possible that lines appear-ing at the doublet separation in observed spectra resultchance alignments of physically unrelated absorption-line clouds. However, observations of multiple line-locksin the same spectrum argue strongly for the reality ofphysical line locks in at least some cases (e.g., Ganguly& Brotherton 2008; Hamann et al. 2011). The largestatistical study by Bowler et al. (2014) showed thatroughly two-thirds of SDSS quasars with multiple C iv NALs at speeds up to ∼ ,
000 km/s have at least oneline-lock pair. Those results indicate that physical line-locking due to radiative forces is both real and commonin quasar outflows.In our study, five pairs of mini-BALs in four quasarsappear line-locked at the C iv doublet separation (498km s − ) to an accuracy <
20% of the FWHM of thenarrowest fitted line in the pair. This amounts to 4/25 = ∼
16% of quasars with multiple C iv mini-BALs havingat least one line-locked C iv pair. These five line-lockpairs are marked by in blue brackets in Figure 6. We alsofind apparent line-locks in the doublet lines of other ions,O vi , N v and Si iv . These line-lock cases are marked byblue brackets in Figures 7 and 8. The velocity offsetsin these cases match the doublet separations very well,8 Chen et al. low-v AALs
J015327-431137-15000 -14000 -130000.00.51.01.5 J103921-271916-24000 -21000 -18000 -150000.00.51.01.5 J104032-27274910500 10000 9500 9000 85000.51.0 J104733+052454-12000 -9000 -6000 -3000 00.00.51.01.5 low-v AALs
J115122+020426 5000 4000 3000 2000 1000 0 10000.500.751.001.25 low-v AALs
J144653+0113566000 4000 2000 00.00.51.01.5 low-v AALs
J151352+085555 8000 6000 4000 20000.00.51.01.5 low-v AALs
J211654-433234-13000 -11000 -9000 -70000.00.51.01.5 J212329-005052
Aug 2008Aug 2006 -20000 -18000 -160000.00.51.01.5 J214159-4413258000 6000 4000 20000.00.51.01.5 low-v AALsBAL
J221531-1744086000 4000 2000 00.00.51.01.5 low-v AALs
J100841+3623196000 5000 4000 30000.00.51.01.5 low-v AALs
J102325+514251 10000 8000 6000 4000 2000 00.00.51.01.5 low-v AALs
J115940-0032038000 7000 6000 5000 40000.00.51.01.5 J131215+423900-5000 -4000 -3000 -20000.00.51.01.5 low-v AALs
J141719+41323712000 11000 10000 9000 80000.00.51.01.5 complex NALs
J160455+381214
Apr 2005May 2009 low-v AALs
J160843+071508
Apr 2007Jul 2008
Oct 2004Aug 2006Jul 2008 -12000 -10000 -8000 -60000.00.51.01.5 J231324+003444
Velocity shifts (km/s) N o r m a li z e d F l u x Figure 6.
Normalized spectra of all C iv mini-BALs that appear in multi-component outflow complexes (having 3 or moreC iv absorption doublets that appear related to each other based on similar profiles and velocity shifts (cid:46) − apart,sometimes including low-velocity lines not in our catalog, see also ‘cmplx’ in Tables 1 and 2). The vertical green solid lines markthe velocity threshold at − − for consideration in our mini-BAL catalog. The blue dotted lines mark the centroids ofmini-BAL profiles in C iv λ iv doublets included in our mini-BAL catalog. Bluebrackets indicate mini-BAL pairs that appear to be line-locked (Section 4.5). C iv absorption lines not in our catalog (AALs,BALs, or low-velocity systems) are marked by green brackets above the spectra. The velocities on the horizontal axis pertain tothe short-wavelength lines in the doublets. Spectra from different observing epochs (when available) are over-plotted in differentcolors. Catalog of C iv Mini-BALs . C I V / J + . S i I V / . N V / - - - - - - V e l o c i t y s h i f t s ( k m / s ) . P V / C I V / J + S i I V / N V / - - - - - V e l o c i t y s h i f t s ( k m / s ) P V / C I V / J + S i I V / N V / O V I / - - - - - - V e l o c i t y s h i f t s ( k m / s ) P V / C I V / J - S i I V / - - - - - - - - V e l o c i t y s h i f t s ( k m / s ) P V / C I V / J - S i I V / N V / O V I / - - - - - V e l o c i t y s h i f t s ( k m / s ) P V / Normalized Flux
Figure 7.
Normalized multiple ion absorption profiles in five quasar spectra showing P v λ Chen et al. . C I V / J - . S i I V / - - - - - - - - - - V e l o c i t y s h i f t s ( k m / s ) . N V / C I V / J - N V / - - . - - . - - . V e l o c i t y s h i f t s ( k m / s ) O V I / C I V / J - S i I V / N V / - - - - - - V e l o c i t y s h i f t s ( k m / s ) O V I / C I V / J + S i I V / - - - - - V e l o c i t y s h i f t s ( k m / s ) O V I / C I V / J - S i I V / N V / - - - - - - V e l o c i t y s h i f t s ( k m / s ) O V I / Normalized Flux
Figure 8.
Normalized multi-ion absorption profiles on a velocity scale relative to the quasar redshift. See Figure 7 for additionalnotes.
Catalog of C iv Mini-BALs . C I V / J + . S i I V / . N V / - - - - - - V e l o c i t y s h i f t s ( k m / s ) . O V I / C I V / J + S i I V / N V / - - - - - - V e l o c i t y s h i f t s ( k m / s ) O V I / C I V / J + S i I V / - - - - - V e l o c i t y s h i f t s ( k m / s ) N V / C I V / J + A p r ( K e c k ) J u l ( K e c k ) S i I V / - - - - - V e l o c i t y s h i f t s ( k m / s ) N V / C I V / J - S i I V / - - - - - - - V e l o c i t y s h i f t s ( k m / s ) N V / Normalized Flux
Figure 8. continued. Chen et al. within (cid:46)
10% of the line FWHMs. Including all of thesesystems in the tally above, we find that 7/25 = 28% ofquasars with multiple mini-BALs have at least one line-lock pair. 4.6.
Variable mini-BALs
A small subset of the quasars in our study were ob-served more than once with VLT-UVES or Keck-HIRES.We check for line variability between these observationsby visual inspecting the normalized spectra plotted ontop of each other (e.g., as in Figures 6 to 8). We specif-ically search for flux mismatches in absorption troughsaccompanied by good matches in the continuum ad-jacent to the troughs. The main uncertainty in theseassessments is the continuum placement (not signal-to-noise ratios in the spectra). We record only obviousoccurrences of line variability with ‘var’ in the Notes inTables 1 and 2. We then fit the lines separately in eachobserving epoch for these definite variability cases.We identify four quasars in our sample with definitemini-BAL variability, all with changes in more that onemini-BAL. They are J160455+381214 shown in Figure 9,J160843+071508 in Figure 8, and J175603+574848and J212329-005052 in Figure 6. The variability inJ160455+381214 was studied previously by Misawaet al. (2005); Bachev et al. (2005); Misawa et al. (2007b);Carini et al. (2007); Misawa et al. (2014); Horiuchi et al.(2016) using data from other telescopes, mainly Sub-aru. The variability in J160843+071508 was studied byMacLeod et al. (2012); Chen et al. (2015) using spectrafrom the SDSS, and the variability in J212329-005052was studied in detail by Hamann et al. (2011) usingthe same data as our current study. There is no pre-vious study to our knowledge of the line variability inJ175603+574848.We briefly summarize the measured mini-BAL vari-abilities as follows. The fitted C iv mini-BALs inJ160455+381214 changed by ∼
20% in REW caused bya ∼
20% change in covering fraction with no obviouschanges ( (cid:46) vi and with greater mag-nitude in N v (Figure 9). The C iv mini-BAL includedin our catalog for J160843+071508 changed by ∼ ∼
11% in covering fraction, ∼
7% in FWHM,with no obvious change in velocity shifts in 1.25 yr (0.32yr in the quasar rest frame). This quasar also showsweak variability in N v and Si iv absorptions (Figure 8).The lone C iv mini-BAL we record for J175603+574848(in a complex blend) changed by ∼
24% in REW, ∼ ∼
13% in FWHMs, and no changesin velocity shifts within 3.76 yr (1.21 yr in the quasar -12 -11 -10 -9 -8 -70.000.250.500.751.001.251.50 CIV 1548/51complex NALs
Apr 2005 (Keck)May 2009 (Keck) -12 -11 -10 -9 -8 -70.000.250.500.751.001.251.50 NV 1239/43-12 -11 -10 -9 -8 -70.000.250.500.751.001.251.50 OVI 1032/38
Velocity shifts (1000 km/s) N o r m a li z e d F l u x Figure 9.
Normalized multi-ion absorption profiles inJ160455+381214 on a velocity scale relative to the quasarredshift. Variability for multiple ion absorptions in two ob-servations. There are obvious changes in the absorption-linestrength between the two epochs. See Figure 7 for additionalnotes. rest frame). The five C iv mini-BALs in J212329-005052show the changes by ∼ − ∼
56% in the REWs, ∼ vi absorption lines. SUMMARY & DISCUSSIONWe present a unique new catalog of high-velocity C iv mini-BALs identified by visual inspection of 638 quasarsin the VLT-UVES and Keck-HIRES archives (Tables 1and 2). One important feature of our mini-BAL cata-log is that the high-resolution and high-signal-to-noisespectra in these archives are much more sensitive toweak and narrow mini-BALs than all previous outflowline surveys using medium-resolution spectra such as theSDSS. We identify mini-BALs based on smooth roundedBAL-like absorption profiles with velocity widths in therange 70 (cid:46) FWHM(1548) (cid:46) − . The ap-proximate upper bound on FWHM avoids BALs whilethe lower bound limits contamination from unrelatedintervening absorption lines (see Section 2.2). We alsolimit our mini-BAL catalog to velocity shifts (cid:46) − Catalog of C iv Mini-BALs − from the quasar frame to avoid associate ab-sorption lines that can reside complex blends or havesquare profiles that are not BAL-like and unlikely toform in BAL-like outflows. We then fit every C iv mini-BAL meeting these criteria to obtain basic line proper-ties like REW(1548) (which ranges from 0.03 to 5.09 ˚A),FWHM(1548) (from 70 to 2693 km/s), and the line-of-sight covering fraction, C (from ∼ ∼ iv mini-BAL meeting the selec-tion criteria described above, after correcting for lim-ited spectra coverage and noisy spectral regions butnot accounting for undercounting due to blends (Sec-tion 4.1). This fraction is similar to the value quoted byRodr´ıguez Hidalgo (2008); Hamann et al. (2012) whofound that C iv mini-BALs appear in ∼
11% of brightSDSS quasars. However, there is little overlap in thetypes of mini-BALs counted in these two studies becauseof the very different sensitivities, spectral resolutions,and definitional properties of the mini-BALs considered.For example, the SDSS spectral resolution is ∼
150 kms − compared to ∼ ∼ − in our study. TheSDSS-based study by Rodr´ıguez Hidalgo (2008) consid-ered any blueshifted velocity < − but also set avery high minimum width threshold at FWHM(1548) (cid:38)
700 km s − (to avoid contamination by unresolvedblends). In contrast, we count mini-BALs up to 10 timesnarrower but only if they are at relatively large velocityshifts v < − − . The steeply-rising numbersof mini-BALs at low velocities in our study (Figure 4)indicates that we exclude substantial numbers of mini-BALs at smaller velocity shifts. Similarly, the FWHMdistribution in our study (Figure 4) indicates that theSDSS studies substantially undercount mini-BALs byconsidering only broad ones. Therefore, the true to-tal fraction of quasars with any type of mini-BAL withFWHM(1548) >
70 km s − at any blueshifted veloci-ties is probably close to the sum of the two fractionsfrom these studies, e.g., 15% to 20%. Furthermore, weneed to note that there is an important selection effectin our sample. Most of the parent sample of spectra forour study were taken for intervening absorption lines.Therefore, it is very probable that many objects show-ing intrinsic absorptions were deliberately not chosen forEchelle spectroscopy, as the intrinsic absorbers compli-cate the studies of intervening absorption. This selectioneffect will cause another possible underestimation of thefrequency of mini-BAL and BALs.All of the mini-BAL absorbers identified by C iv in ourcatalog are highly ionized based on the absence of ac-companying low-ionization lines such as Si ii or C ii andthe strong presence of high-ionization lines such as N v and O vi (when within our spectral coverage). P v ab-sorption is detected unambiguously in 2 mini-BAL sys-tems in our catalog, plus three other quasars with broadoutflow lines not in the catalog because their velocityshifts are too low (Section 4.4 and Figure 7). TheseP v detections indicate that the absorption lines of moreabundant ions like Si iv , C iv , N v , and O vi are highlysaturated and that the total outflow column densitiesare very large (at least at the velocities where P v isdetected). Photoionization models presented in otherstudies show, specifically, that P v detections indicateC iv optical depths of several hundred to ∼ N H > cm − (for solarabundances, with specific column densities depending onthe observed line strengths and widths, Hamann 1998;Leighly et al. 2009, 2011; Capellupo et al. 2014; Cham-berlain et al. 2015; Moravec et al. 2017; Hamann et al.2019a,b). An intense radiation field with large ionizationparameters is needed to produce these large column den-sities of ionized gas for Pv absorption. Hamann et al.(2019a) combined this result with minimum gas den-sities inferred from excited-state absorption lines (e.g.,in median spectra of BAL and mini-BAL quasars) toconclude that the outflows reside typically within a fewtens of pc of the central quasars (see also Hamann et al.2019b).The majority of C iv mini-BALs with resolved dou-blets in our study, and all of the cases with flat-bottomedtroughs or accompanying P v detections, indicate thatthe lines are optically thick with partial covering of thebackground light source (see also Moravec et al. 2017, forsimilar mini-BAL results). Other cases with line blend-ing yield only lower limits on the covering fractions andoptical depths (see Tables 1 and 2). Altogether, morethan half of the C iv mini-BALs in our survey overallare confirmed to be optically thick with partial cover-ing. This requires typical absorbing structures that arenot much larger than the projected area of the quasaremission regions. Given that our catalog considers onlylarge velocity shifts ( v < − − ) from the C iv broad emission lines, most or all of the mini-BAL partialcovering applies to the quasar continuum source. This isthe UV-emitting accretion disk < iv (J222006-280323, Figure 3) and 0.03 in Si iv (J015327-431137,Figure 8), imply that the outflow absorbing structurescan be (cid:46) vi andN v are within our spectral coverage, they tend havedeeper and broader absorption troughs compared to4 Chen et al. C iv . Similarly, the C iv lines tend to be deeper andbroader than lower-ionization and/or low-abundancelines such as Si iv and P v when they are within thewavelength coverage (Figures 7 and 8). We infer fromthis that 1) all of the lines are typically optically thick,and 2) the progression in the line depths and velocitywidths is related to their optical depth-dependent cov-ering fractions in spatially inhomogeneous absorbing re-gions.Spatially inhomogeneous (e.g., clumpy) absorberspresent a range of column densities and line opti-cal depths across the projected area of the continuumsource. Schematic illustrations of this type of absorb-ing geometry can be found in Hamann et al. (2001) andHamann & Sabra (2004). This can naturally lead to dif-ferent derived covering fractions in different lines if theyare optically thick over different projected areas. Weaklines with narrower profiles like the P v and Si iv mini-BALs form primarily in smaller pockets of high-columndensity (and perhaps lower-ionization) gas, while thebroader and deeper absorption troughs of higher ionslike C iv and O vi form in lower column density regionsthat cover larger spatial areas and a wider range of ve-locities (see also Arav et al. 2005, 2008; Moravec et al.2017; Leighly et al. 2018; Hamann et al. 2019a,b; Choiet al. 2020, for more discussion).One interesting consequence of this clumpy absorbinggeometry is that, without confinement by an externalpressure, structures < < K corresponding to thermal speed (for hy-drogen) of v th ∼
13 km s − , the dissipation times are (cid:46)
150 yr to (cid:46)
750 yr for the maximum cloud sizes givenabove. However, the mini-BAL profile widths suggestthat the actual internal velocities are much larger. Themedian mini-BAL width in our study, FWHM ≈ − (Figure 4) indicates that the actual cloud dis-sipation times are only (cid:46) (cid:46)
30 yr. These shortsurvival times suggest that either 1) the clouds are verynear the quasars where the flow times (at the observedoutflows speeds, nominally ∼ − ) are shorterthan the survival times, or 2) they created in situ po-tentially at much larger distances (see Chen et al. 2018,2019, for more discussion).The subset of multi-component C iv absorption linecomplexes are an interesting subset of mini-BALs in ourcatalog because they identify multiple distinct yet physi-cally related absorbing structures in the outflows. Theirvelocity shifts in our sample range from ∼ − ,
000 km s − to ∼ − and their FWHMs range from ∼ ∼ − . The lines at higher velocities in thesecomplexes tend to be broader and weaker, with smallercovering fractions, than the lines in the same complexat lower velocities (see also Simon & Hamann 2010b;Chen et al. 2019). The complexes with such tiny sizesand short survival times indicate they are some typeof highly-structured outflows very close to the centralquasars.Finally, we note that in large surveys ∼ ∼ ∼
35% of the BAL quasarsthat remain have mini-BALs, and in roughly half ofthose appear in the wings of the BAL toughs, suggestingthat the BAL and mini-BAL outflows are physically re-lated. The partial covering situation that is common tothese features (see also Hamann et al. 2019a) suggeststhat the narrower and generally weaker mini-BALs formin outflow clumps or filaments that cover smaller spatialareas (to produce smaller covering fractions) and smallerranges in velocity (to produce narrower profiles). BALsmight form be spatially larger clumps or in collections ofmany mini-BAL-like clumps that blend together to pro-duce broader deeper troughs in observed spectra (Hallet al. 2007; Hamann et al. 2008, 2013).Similarly, the narrow mini-BALs in our study, withFWHMs as small as ∼
74 km s − , have widths overlap-ping with the narrow absorption lines (NALs) catalogedin other surveys (e.g. Wild et al. 2008; Nestor et al.2008). There is, again, no clear classification bound-ary between mini-BALs and outflow NALs. Moreover,some of the mini-BALs in our study reside in multi-component absorption-line complexes that include evennarrower lines than our survey threshold FWHM (cid:38) − . These results indicate that at least some mini-BALs are physically related to the outflow NALs. Morework is needed to understand the physical nature andrelationships of all of these different outflow absorption-line features. ACKNOWLEDGMENTS Catalog of C iv Mini-BALs
Discovery Projects grantsDP0877998, DP110100866, and DP130100568.REFERENCES
Anderson, S. F., Weymann, R. J., Foltz, C. B., & Chaffee,Jr., F. H. 1987, AJ, 94, 278Arav, N., Brotherton, M. S., Becker, R. H., et al. 2001,ApJ, 546, 140Arav, N., Kaastra, J., Kriss, G. A., et al. 2005, ApJ, 620,665Arav, N., Korista, K. T., & de Kool, M. 2002, ApJ, 566, 699Arav, N., Moe, M., Costantini, E., et al. 2008, ApJ, 681, 954Bachev, R., Strigachev, A., & Semkov, E. 2005, MNRAS,358, 774, doi: 10.1111/j.1365-2966.2005.08708.xBarlow, T. A., Hamann, F., & Sargent, W. L. W. 1997, inAstronomical Society of the Pacific Conference Series,Vol. 128, Mass Ejection from Active Galactic Nuclei, ed.N. Arav, I. Shlosman, & R. J. Weymann, 13Berg, T. A. M., Ellison, S. L., Prochaska, J. X., Venn,K. A., & Dessauges-Zavadsky, M. 2015, MNRAS, 452,4326, doi: 10.1093/mnras/stv1577Borguet, B. C. J., Arav, N., Edmonds, D., Chamberlain,C., & Benn, C. 2013, ApJ, 762, 49.https://arxiv.org/abs/1211.6250Bowler, R. A. A., Hewett, P. C., Allen, J. T., & Ferland,G. J. 2014, MNRAS, 445, 359,doi: 10.1093/mnras/stu1730Braun, E., & Milgrom, M. 1989, ApJ, 342, 100Capellupo, D. M., Hamann, F., & Barlow, T. A. 2014,MNRAS, 444, 1893, doi: 10.1093/mnras/stu1502Capellupo, D. M., Hamann, F., Herbst, H., et al. 2017,MNRAS, 469, 323, doi: 10.1093/mnras/stx870Carini, M. T., Noble, J. C., Taylor, R., & Culler, R. 2007,AJ, 133, 303, doi: 10.1086/509774Chamberlain, C., Arav, N., & Benn, C. 2015, MNRAS, 450,1085, doi: 10.1093/mnras/stv572Chartas, G., Charlton, J., Eracleous, M., et al. 2009, NewAstronomy Review, 53, 128.https://arxiv.org/abs/0907.3464Chen, C., Hamann, F., Simon, L., & Barlow, T. 2018,MNRAS, 481, 3865, doi: 10.1093/mnras/sty2534Chen, C., Hamann, F., Simon, L., & Ma, B. 2019, ApJ,887, 78, doi: 10.3847/1538-4357/ab53d8 Chen, Z.-F., Gu, Q.-S., Chen, Y.-M., & Cao, Y. 2015,MNRAS, 450, 3904, doi: 10.1093/mnras/stv813Choi, H., Leighly, K. M., Terndrup, D. M., Gallagher,S. C., & Richards, G. T. 2020, ApJ, 891, 53,doi: 10.3847/1538-4357/ab6f72Cicone, C., Maiolino, R., Sturm, E., et al. 2014, A&A, 562,A21, doi: 10.1051/0004-6361/201322464Dawson, K. S., Schlegel, D. J., Ahn, C. P., et al. 2013, AJ,145, 10, doi: 10.1088/0004-6256/145/1/10de Kool, M., Arav, N., Becker, R. H., et al. 2001, ApJ, 548,609, doi: 10.1086/318996de Kool, M., Korista, K. T., & Arav, N. 2002, ApJ, 580, 54Debuhr, J., Quataert, E., & Ma, C.-P. 2012, MNRAS, 420,2221, doi: 10.1111/j.1365-2966.2011.20187.xDi Matteo, T., Springel, V., & Hernquist, L. 2005, Nature,433, 604D’Odorico, V., Cristiani, S., Romano, D., Granato, G. L.,& Danese, L. 2004, MNRAS, 351, 976Dunn, J. P., Bautista, M., Arav, N., et al. 2010, ApJ, 709,611. https://arxiv.org/abs/0911.3896Eisenstein, D. J., Weinberg, D. H., Agol, E., et al. 2011,AJ, 142, 72. https://arxiv.org/abs/1101.1529Elvis, M. 2012, in Astronomical Society of the PacificConference Series, Vol. 460, AGN Winds in Charleston,ed. G. Chartas, F. Hamann, & K. M. Leighly, 186.https://arxiv.org/abs/1201.3520Filiz Ak, N., Brandt, W. N., Hall, P. B., et al. 2013, ApJ,777, 168, doi: 10.1088/0004-637X/777/2/168—. 2014, ApJ, 791, 88, doi: 10.1088/0004-637X/791/2/88Gabel, J. R., Kraemer, S. B., Crenshaw, D. M., et al. 2005,ApJ, 631, 741Ganguly, R., Bond, N. A., Charlton, J. C., et al. 2001, ApJ,549, 133Ganguly, R., & Brotherton, M. S. 2008, ApJ, 672, 102Ganguly, R., Eracleous, M., Charlton, J. C., & Churchill,C. W. 1999, AJ, 117, 2594Gibson, R. R., Jiang, L., Brandt, W. N., et al. 2009, ApJ,692, 758, doi: 10.1088/0004-637X/692/1/758Guo, Z., & Martini, P. 2019, ApJ, 879, 72,doi: 10.3847/1538-4357/ab2590 Chen et al.
Hall, P. B., Anderson, S. F., Strauss, M. A., et al. 2002,ApJS, 141, 267Hall, P. B., Anosov, K., White, R. L., et al. 2011, MNRAS,411, 2653. https://arxiv.org/abs/1010.3728Hall, P. B., Sadavoy, S. I., Hutsemekers, D., Everett, J. E.,& Rafiee, A. 2007, ApJ, 665, 174Hamann, F. 1998, ApJ, 500, 798Hamann, F., Barlow, T., Cohen, R. D., Junkkarinen, V., &Burbidge, E. M. 1997a, in Astronomical Society of thePacific Conference Series, Vol. 128, Mass Ejection fromActive Galactic Nuclei, ed. N. Arav, I. Shlosman, & R. J.Weymann, 19–+Hamann, F., Barlow, T. A., Junkkarinen, V., & Burbidge,E. M. 1997b, ApJ, 478, 80Hamann, F., Chartas, G., McGraw, S., et al. 2013,MNRAS, 435, 133, doi: 10.1093/mnras/stt1231Hamann, F., Herbst, H., Paris, I., & Capellupo, D. 2019a,MNRAS, 483, 1808, doi: 10.1093/mnras/sty2900Hamann, F., Kanekar, N., Prochaska, J. X., et al. 2011,MNRAS, 410, 1957. https://arxiv.org/abs/1008.3728Hamann, F., Kaplan, K. F., Hidalgo, P. R., Prochaska,J. X., & Herbert-Fort, S. 2008, MNRAS, 391, L39.https://arxiv.org/abs/0808.3998Hamann, F., & Sabra, B. 2004, in Astronomical Society ofthe Pacific Conference Series, Vol. 311, AGN Physicswith the Sloan Digital Sky Survey, ed. G. T. Richards &P. B. Hall, 203–+Hamann, F., Sabra, B., Junkkarinen, V., Cohen, R., &Shields, G. 2003, ArXiv Astrophysics e-printsHamann, F., Simon, L., Hidalgo, P. R., & Capellupo, D.2012, in Astronomical Society of the Pacific ConferenceSeries, Vol. 460, AGN Winds in Charleston, ed.G. Chartas, F. Hamann, & K. M. Leighly, 47.https://arxiv.org/abs/1204.3791Hamann, F., Tripp, T. M., Rupke, D., & Veilleux, S. 2019b,MNRAS, 487, 5041, doi: 10.1093/mnras/stz1408Hamann, F. W., Barlow, T. A., Chaffee, F. C., Foltz, C. B.,& Weymann, R. J. 2001, ApJ, 550, 142Hewett, P. C., & Foltz, C. B. 2003, AJ, 125, 1784Hopkins, P. F., & Elvis, M. 2010, MNRAS, 401, 7.https://arxiv.org/abs/0904.0649Hopkins, P. F., Hernquist, L., Cox, T. J., & Kereˇs, D. 2008,ApJS, 175, 356. https://arxiv.org/abs/0706.1243Horiuchi, T., Misawa, T., Morokuma, T., et al. 2016, PASJ,68, 48, doi: 10.1093/pasj/psw044Kauffmann, G., & Haehnelt, M. 2000, MNRAS, 311, 576King, A. 2003, ApJL, 596, L27, doi: 10.1086/379143Knigge, C., Scaringi, S., Goad, M. R., & Cottis, C. E. 2008,MNRAS, 386, 1426 Leighly, K. M., Dietrich, M., & Barber, S. 2011, ApJ, 728,94Leighly, K. M., Hamann, F., Casebeer, D. A., & Grupe, D.2009, ApJ, 701, 176. https://arxiv.org/abs/0906.2149Leighly, K. M., Terndrup, D. M., Gallagher, S. C.,Richards, G. T., & Dietrich, M. 2018, ApJ, 866, 7,doi: 10.3847/1538-4357/aadee6MacLeod, C. L., Ivezi´c, ˇZ., Sesar, B., et al. 2012, ApJ, 753,106, doi: 10.1088/0004-637X/753/2/106Matthews, J. H., Knigge, C., Long, K. S., et al. 2016,MNRAS, 458, 293, doi: 10.1093/mnras/stw323Milne, E. A. 1926, MNRAS, 86, 459,doi: 10.1093/mnras/86.7.459Misawa, T., Charlton, J. C., & Eracleous, M. 2014, ApJ,792, 77, doi: 10.1088/0004-637X/792/1/77Misawa, T., Charlton, J. C., Eracleous, M., et al. 2007a,ApJS, 171, 1Misawa, T., Eracleous, M., Charlton, J. C., & Kashikawa,N. 2007b, ApJ, 660, 152Misawa, T., Eracleous, M., Charlton, J. C., & Tajitsu, A.2005, ApJ, 629, 115Misawa, T., Kawabata, K. S., Eracleous, M., Charlton,J. C., & Kashikawa, N. 2010, ApJ, 719, 1890.https://arxiv.org/abs/1007.0448Moravec, E. A., Hamann, F., Capellupo, D. M., et al. 2017,MNRAS, 468, 4539, doi: 10.1093/mnras/stx775Murphy, M. T., Kacprzak, G. G., Savorgnan, G. A. D., &Carswell, R. F. 2019, MNRAS, 482, 3458,doi: 10.1093/mnras/sty2834Nestor, D., Hamann, F., & Rodriguez Hidalgo, P. 2008,MNRAS, 386, 2055O’Meara, J. M., Lehner, N., Howk, J. C., et al. 2015, AJ,150, 111, doi: 10.1088/0004-6256/150/4/111Ostriker, J. P., Choi, E., Ciotti, L., Novak, G. S., & Proga,D. 2010, ApJ, 722, 642,doi: 10.1088/0004-637X/722/1/642Pˆaris, I., Petitjean, P., Ross, N. P., et al. 2017, A&A, 597,A79, doi: 10.1051/0004-6361/201527999Prochaska, J. X., & Hennawi, J. F. 2009, ApJ, 690, 1558.https://arxiv.org/abs/0806.0862Prochaska, J. X., Hennawi, J. F., & Herbert-Fort, S. 2008,ApJ, 675, 1002, doi: 10.1086/526508Prochaska, J. X., O’Meara, J. M., Fumagalli, M., Bernstein,R. A., & Burles, S. M. 2015, ApJS, 221, 2,doi: 10.1088/0067-0049/221/1/2Reichard, T. A., Richards, G. T., Hall, P. B., et al. 2003,AJ, 126, 2594Rodr´ıguez Hidalgo, P. 2008, PhD thesis, University ofFlorida, Gainesville
Catalog of C iv Mini-BALs Rodr´ıguez Hidalgo, P., Eracleous, M., Charlton, J., et al.2013, ApJ, 775, 14, doi: 10.1088/0004-637X/775/1/14Rodr´ıguez Hidalgo, P., Hamann, F., Eracleous, M., et al.2012, in Astronomical Society of the Pacific ConferenceSeries, Vol. 460, AGN Winds in Charleston, ed.G. Chartas, F. Hamann, & K. M. Leighly, 93Rogerson, J. A., Hall, P. B., Rodr´ıguez Hidalgo, P., et al.2016, MNRAS, 457, 405, doi: 10.1093/mnras/stv3010Rupke, D. S. N., G¨ultekin, K., & Veilleux, S. 2017, ApJ,850, 40, doi: 10.3847/1538-4357/aa94d1Rupke, D. S. N., & Veilleux, S. 2013, ApJ, 768, 75,doi: 10.1088/0004-637X/768/1/75Scannapieco, E., & Oh, S. P. 2004, ApJ, 608, 62Scargle, J. D. 1973, ApJ, 179, 705, doi: 10.1086/151910Schneider, D. P., Richards, G. T., Hall, P. B., et al. 2010,AJ, 139, 2360, doi: 10.1088/0004-6256/139/6/2360Silk, J., & Rees, M. J. 1998, A&A, 331, L1 Simon, L. 2011, High Redshift Quasar Abundances andEnvironments: Connecting Black Hole and Host GalaxyEvolution (University of Florida).https://books.google.com/books?id=l1JlnQAACAAJSimon, L. E., & Hamann, F. 2010a, MNRAS, 407, 1826,doi: 10.1111/j.1365-2966.2010.17022.x—. 2010b, MNRAS, 409, 269,doi: 10.1111/j.1365-2966.2010.17306.xSimon, L. E., Hamann, F., & Pettini, M. 2012, inAstronomical Society of the Pacific Conference Series,Vol. 460, AGN Winds in Charleston, ed. G. Chartas,F. Hamann, & K. M. Leighly, 52Trump, J. R., Hall, P. B., Reichard, T. A., et al. 2006,ApJS, 165, 1Weinberger, R., Springel, V., Hernquist, L., et al. 2017,MNRAS, 465, 3291, doi: 10.1093/mnras/stw2944Weymann, R. J., Morris, S. L., Foltz, C. B., & Hewett,P. C. 1991, ApJ, 373, 23Wild, V., Kauffmann, G., White, S., et al. 2008, MNRAS,388, 227. https://arxiv.org/abs/0802.4100 Chen et al. F l u x original spectrumcontinuumfitting line Wavelength ( Å ) Figure 10.
We draw a continuum over the mini-BAL systemat -18,003 km/s in J024221+004912, and the fitting line isshown in red.
APPENDIX A. NOTES ON INDIVIDUAL QUASARSThis Appendix provides additional notes on mini-BALsystems in our catalog that have ambiguities in the mini-BAL identifications, complex line blends, or uncertain-ties in the line profile fits that are not already explainedin the Notes in Tables 1 and 2.J024221+004912: For the broad system with velocityshift -18,003 km/s (Figure 10), it could be combinedby a broad and shallow profile and a narrower and steepprofile. But the continuum is uncertain due to the broadfeature. Therefore, we only fit the steep and narrowerprofile.J103921-271916: The C iv mini-BAL at -42,626 km/sis blended with Si iv (Figure 11). We estimate the upperand lower limits.J104032-272749: we confirm the broad system at -21,767 km/s is a C iv mini-BAL, because we find otherlines N v λ iv λ vi λ ∼ − ∼ − v at the same redshift. It is likely a blend.J231324+003444: The system at -30,887 km/s showsa broad, non-smooth, and asymmetric profile. It couldbe a blend (Figure 3).The mini-BAL system at -11,259 km/s in J104033-272308, the systems at -13,416 km/s and -27,118 km/s -15 -14.5 -14 -13.5 -13 -12.5 -12 -11.50.00.20.40.60.81.01.2 CIV 1548/51-15 -14.5 -14 -13.5 -13 -12.5 -12 -11.50.00.20.40.60.81.01.2 CIVSiIV 1394/1403 Velocity shifts (1000 km/s) N o r m a li z e d F l u x Figure 11.