{"id":"39d5cd90-db64-4641-900c-c78713ea5b3c","arxiv_id":"2507.23328","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":4,"one_line_summary":"First feedback-efficiency estimates for narrow absorption line outflows give kinetic luminosities that clear the AGN feedback threshold, provided the absorbers are genuinely quasar-associated.","lead":"Astronomers estimated how much kinetic energy narrow absorption line outflows in eight bright quasars carry, and found the energy can rival or exceed the broader, faster outflows usually studied. If these thin gas streams are truly ejected by the quasars, the most common outflow type could be a major driver of galaxy feedback.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The feedback claim depends on the 11 NALs being intrinsic, but §3 concedes partial coverage is the only test and that compact intervening clouds can mimic it; if even a few systems are foreground, R, Ṁ, Ė_k, and ε_k lose physical meaning.","rationale":"The paper is an honest first application of the excited/ground-state density diagnostic to NALs, and it openly discloses the decisive weakness. The strongest claim—efficiencies comparable to or larger than BALs—is only meaningful if the absorbers are quasar-associated. On that point the evidence is a single diagnostic (partial coverage) with known degenerate explanations (compact intervening clouds, unresolved saturation), and §3 explicitly says there is no independent corroboration. I therefore agree with the reader’s weakest-assumption identification. A concrete multi-epoch experiment can put the assumption to a direct test: intrinsic NALs should vary coherently in the doublet members or respond to continuum changes over rest-frame years, while stable, non-varying profiles would support the intervening-cloud alternative. I also checked the numerics: recomputing ε_k from Table 1 L_Edd and Table 3 Ė_k does not reproduce several printed entries (e.g., Q1548+0917 z=2.6082 gives 4.6 rather than 7.8; HS1700+6416 gives 17.7 rather than 22), so the extreme quantitative claims are additionally fragile. Neither issue forces rejection—the paper labels its values as limits and asks for follow-up—but both block acceptance. The conditional verdict stands.","tokens_in":18643,"tokens_out":21961,"duration_ms":249486,"concrete_test":"Re-observe the eight quasars with high-resolution UV spectroscopy (Keck/HIRES or VLT/UVES) in two epochs separated by at least 2–3 yr in the quasar rest frame, and measure C IV λλ1548,1551 and Si IV λλ1394,1403 doublet profiles, residual fluxes, and covering factors. Coordinated variability in both doublet members, or a response to a change in the quasar’s ionizing continuum, would corroborate intrinsic origin; a null result over a decade-long baseline would strengthen the alternative that the apparent C_f<1 arises from unresolved saturated components or foreground compact clouds. This test directly targets the assumption on which Eqs. (2)–(6) depend.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that NAL outflows can reach feedback efficiencies above 0.5% L_Edd and rival BALs—rests on the unverified classification of the 11 systems as intrinsic. The derivation chain is: partial coverage (C_f<1) implies quasar origin; Q(H) in Eq. (2) is then the target quasar’s ionizing flux; R from Eq. (2) is a distance from that quasar; Eqs. (4)–(5) give a mass outflow rate and kinetic luminosity; Eq. (6) gives a feedback efficiency. If an absorber is intervening, every quantity after the first step is physically meaningless, and a lower limit on Ė_k for unrelated foreground gas is not a conservative bound. The paper selects on partial coverage alone, and §3 states the problem explicitly: caveat 1 notes that compact, low-ionization clouds of a few AU in (sub-)DLA systems can show partial coverage with no quasar connection, and caveat 2 concedes “we cannot corroborate that they are intrinsic by an additional and separate test.” The paper excludes only systems with log N_HI>19, but Table 2 contains several systems with log N_HI≈17–18 for which compact clouds remain plausible; caveat 3 adds that only one velocity component shows C_f<1 while Lyα is black, a structure also expected for a small dense core embedded in a larger medium. Thus the abstract’s own statement that “the results depend critically on this selection method” is the main vulnerability: if even a few of the 11 are intervening, the 8/11 super-threshold headline and the comparison to BALs in Figure 4 collapse.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper selects 11 narrow absorption line (NAL) systems in 8 luminous quasars from Misawa et al. (2007a), based on partial coverage, the presence of low-ionization C II or Si II lines, and coverage of Lyα. Using upper limits on n_e from undetected excited-state lines of C II and Si II, the authors derive lower limits on the absorber distance R, the total hydrogen column N_H (via a Cloudy model at log U = -2.6), and then, following Borguet et al. (2012), lower limits on the mass outflow rate, kinetic luminosity, and feedback efficiency ε_k = Ė_k/L_Edd. Eight of the 11 systems have ε_k > 0.005, with two extreme systems reaching >7.8 and >22, leading the authors to conclude that NAL outflows may be a significant AGN feedback channel. The paper is explicit that the results depend critically on the unverified identification of the absorbers as intrinsic.","tokens_in":18804,"tokens_out":9214,"duration_ms":101576,"significance":"If the central claim holds, the paper would establish that NAL outflows, the most common absorption-line outflow class, can have feedback efficiencies rivaling or exceeding BALs, with important implications for AGN feedback. The work is methodologically useful: it applies the excited-state/resonance-line diagnostic to NALs for the first time, the arithmetic of Eqs. (4)-(6) reproduces the tabulated values given the stated assumptions, and the authors are commendably transparent about the selection caveats. However, the conclusion is conditional on two fragile premises: the intrinsic nature of the absorbers and the single adopted ionization parameter. The small sample and the non-independence of multiple systems from the same quasars further limit the statistical weight of the '8 of 11' headline. The paper is better framed as a method demonstration with conditional limits than as an established measurement of NAL feedback efficiency.","major_comments":[{"comment":"The headline result that eight of eleven systems exceed ε_k > 0.005 depends entirely on the eleven absorbers being intrinsic, but the only selection test is partial coverage, and §3 explicitly concedes (caveats 1 and 2) that compact low-ionization clouds in (sub-)DLA systems can show partial coverage without any quasar connection and that no independent corroborating test is available. Because Eq. (2) uses the target quasar's Q(H) and the geometric R from that equation enters Eqs. (4)–(6), any intervening contaminant makes Ṁ, Ė_k, and ε_k unphysical; a lower limit on Ė_k for foreground gas is not a conservative bound on AGN feedback. The authors should estimate the expected contamination fraction from the literature or at least show how many of the eleven systems would need to be intervening to eliminate the super-0.5% conclusion, and they should carry that contingency through the abstract and §6 conclusions.","section":"§3; Tables 2–3"},{"comment":"The ionization parameter log U = -2.6 is not independently constrained: it is chosen because it reproduces the sample-averaged N(CII)/N(CIV) ≈ 0.3, and it is then used both to convert N_HI to N_H with Cloudy and to compute R from Eq. (2). Since Ė_k and ε_k scale directly with N_H and R, the derived efficiencies are contingent on this single-phase, single-U assumption rather than measured. The authors should present a sensitivity test over a plausible range of log U (e.g., -3.5 to -1.5) and discuss the single-phase assumption, quantifying how many of the eight super-threshold systems survive.","section":"§4, after Eq. (2)"},{"comment":"Two of the headline systems (Q1548+0917 at z_abs = 2.6082 and HS1700+6416 at z_abs = 2.4330) have ε_k lower limits of 7.8 and 22, implying Ė_k of 4.6 and 17.7 times L_Edd respectively (and even larger multiples of L_bol). This is energetically challenging for radiatively driven quasar outflows and suggests that the assumed N_H (from log U = -2.6), f_c = 0.5, or R lower limits may be mutually inconsistent. The paper should address this tension explicitly and identify which assumption drives the extreme values before using these systems to argue that NALs rival BALs.","section":"§5, Table 3"},{"comment":"The comparison with BAL, SIV, and EUV500 samples mixes lower limits computed with the assumptions of this paper (f_c = 0.5, log U = -2.6, single-phase) with upper limits and measurements from other work that use different SEDs, covering fractions, and fitting methods. The statement that NALs have 'exceptionally large efficiency compared to the other outflow classes' could be an artifact of these methodological differences; a common-metric comparison or a matched-treatment reanalysis is needed to support it.","section":"§5, Figure 4"},{"comment":"The global covering fraction is set to f_c = 0.5 because intrinsic NALs are found in at least 50% of quasars, but the detection rate of NAL absorbers is not the same as the solid-angle covering fraction of a single outflow; since Ṁ, Ė_k, and ε_k are all proportional to f_c, the choice directly sets the normalization of the feedback efficiencies. The authors should vary f_c over a plausible range (e.g., 0.1–1.0) and report which of the eight systems remain above ε_k = 0.005.","section":"§5, below Eq. (5)"}],"minor_comments":[{"comment":"The notation 'log(Ṁ/M⊙ s⁻¹) > 79–(3.1×10⁵)' (and the similar expression in §6) is malformed and should be replaced with the actual range of log values (approximately 1.9–5.5 in M⊙ yr⁻¹) or a properly formatted range.","section":"Abstract and §6"},{"comment":"The sentence 'Nonetheless, we do not that, (a) photoionization models ...' contains a typo and should read 'we do note that'.","section":"§3, caveat paragraph"},{"comment":"The bullet 'ε_k ≳ 0–22' is imprecise; it should read something like 'ε_k > 0.005 to >22' to match Table 3.","section":"§6, summary bullet"},{"comment":"The text says equation (3) uses critical densities from Tayal (2008a,b), while Figure 2 is described as calculated with the CHIANTI 10.0 database; please clarify which calculation is used for the reported density limits.","section":"§4 and Figure 2"},{"comment":"Multiple systems from the same quasar (three from Q1548+0917 and two from HS1946+7658) are treated as independent in the '8 of 11' counting; the paper should either report quasar-level statistics or explicitly discuss the clustering.","section":"Tables 2–3"},{"comment":"The affiliation 'Institute for Gavitation and the Cosmos' should be 'Institute for Gravitation and the Cosmos'.","section":"Affiliations"}],"recommendation":"major_revision","confidential_remarks":"The paper is transparent about its caveats, and the arithmetic chain is reproducible, but the central astrophysical claim—that NAL outflows can rival BALs as a feedback channel—is more conditional than the title and abstract suggest. The stress-test concern about intervening contaminants is well placed: the paper's own §3 concedes that partial coverage is the only test and that compact intervening clouds can mimic it. I would ask the authors to either add a quantitative contamination estimate and a sensitivity analysis for log U and f_c, or substantially reframe the paper as a method demonstration with conditional upper/lower limits. The small, non-independent sample also argues against a strong population-level conclusion."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is the first attempt to put feedback numbers on intrinsic NAL outflows, and the arithmetic is reproducible. But the headline claim—that NAL outflows can reach efficiencies above 0.5% L_Edd and rival BALs—rests on an unverified selection and a chain of assumptions that produces some physically impossible output values. The authors themselves flag most of this, which makes the paper honest but not convincing.\n\nWhat's genuinely new: applying the Borguet et al. (2012) excited/ground-state density method to NALs and deriving lower limits on M_dot, E_dot_k, and epsilon_k for 11 systems. That hasn't been done before, and the paper is transparent about the method's limitations. The comparison to BAL and other outflow classes in Fig. 4 is a useful addition, even if the plotted points carry large systematic uncertainties.\n\nThe soft spots, in order of severity. First, the intrinsicness of the absorbers. The selection relies on partial coverage alone, and the paper concedes (§3) that compact low-ionization clouds in intervening DLAs can produce the same signature and that no independent test corroborates intrinsic origin. If even a few of the 11 are foreground, R, M_dot, E_dot_k, and epsilon_k are not conservative bounds—they are numbers about unrelated gas. Second, the ionization parameter: log U = -2.6 is chosen because it reproduces the sample's average CII/CIV ratio, then used to convert N_HI to N_H and to derive R. That's a circularity. Third, the extreme outputs: two systems have epsilon_k > L_Edd, and one has R > 3.7 Mpc at 98 km/s, requiring a travel time of hundreds of Gyr. These values should have triggered a sanity check rather than being reported as headline results. There are also minor typographical issues, e.g., the abstract's unit for M_dot.\n\nNone of this is fatal to the basic method. The paper is a legitimate first step and the authors clearly know where the weaknesses are. But the main conclusion needs to be reframed as conditional on the intrinsicness of the sample and on a more physical treatment of the extreme systems.\n\nBottom line: worth sending to referees, but expect major revision. I would not cite it yet.","headline":"First NAL outflow efficiency estimates, but the feedback claim rests on unverified intrinsicness and extreme outputs that strain physical plausibility.","tokens_in":19579,"tokens_out":4047,"would_cite":false,"duration_ms":40366,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Narrow absorption-line outflows in quasars can carry as much kinetic energy as broad absorption-line winds, and eight of the 11 systems studied exceed the 0.5% Eddington feedback threshold, with two exceeding it by more than an order of…","keywords":["quasar absorption lines","narrow absorption lines","AGN feedback","quasar outflows","partial coverage","kinetic luminosity","excited-state absorption lines","Eddington ratio"],"falsifier":"Detect the excited-state lines C II* 1336 or Si II* 1265/1533 at the absorption redshift of any of these systems; a detection would replace an electron-density upper limit with a measured density, shorten the derived distance $R$, and lower the quoted feedback efficiencies. Conversely, showing that any selected system is an intervening (sub-)DLA cloud, for instance by detecting a foreground galaxy along the line of sight at $z_{\\rm abs}$ with no association to the quasar, would remove that system's support for the feedback claim.","tokens_in":18219,"feed_emoji":"🌪️","tokens_out":9055,"duration_ms":93559,"temperature":0.7,"pith_summary":"This paper asks whether narrow absorption lines (NALs, gas signatures narrower than 500 km/s) in quasar spectra, the most common absorption-line outflow class, carry enough kinetic energy to count as AGN feedback. Using 11 NAL systems in eight optically luminous quasars selected for \"partial coverage,\" a signature thought to mark gas tied to the quasar, the authors place upper limits on electron density from the absence of excited-state C II and Si II lines and convert those into lower limits on distance, mass outflow rate, kinetic luminosity, and feedback efficiency. Eight of the 11 systems exceed the 0.5% Eddington-luminosity feedback threshold, and two reach lower limits of $\\varepsilon_k > 7.8$ and $> 22$. The authors' own conclusion is conditional: if the NALs are truly intrinsic, their feedback efficiency is comparable to or larger than that of broad absorption line outflows, but the paper also states that the connection of these distant absorbers to the quasar central engine remains open.","feed_headline":"Narrow quasar outflows rival broad winds in feedback power","feed_subtitle":"Eight of 11 narrow absorption-line systems exceed the wind-energy threshold thought to shape galaxies.","key_machinery":"The load-bearing object is the excited-to-ground-state column density ratio of C II and Si II, which acts as an electron-density probe: because no excited-state lines are detected, each system yields only an upper limit on $n_e$ through the collisional-excitation relation of the paper's equation (3). That upper limit enters the ionization-parameter definition $U = Q(H)/(4\\pi R^2 n_H c)$ with an assumed $\\log U = -2.6$, turning the density limit into a lower limit on the radial distance $R$ (about 100 kpc to 4 Mpc). Distance, total hydrogen column density, and measured ejection velocity then feed the standard outflow formulas of equations (4) and (5), $\\dot M = 4\\pi R f_c \\mu m_p N_H v_{\\rm ej}$ and $\\dot E_k = \\frac{1}{2}\\dot M v_{\\rm ej}^2$, with a global covering fraction $f_c = 0.5$, and the result is divided by the Eddington luminosity to get $\\varepsilon_k$. All derived quantities are lower limits, because the distances are lower limits and the outflow may be instantaneous rather than continuous.","core_discovery":"The central claim is that intrinsic narrow absorption line outflows can be energetically dominant feedback agents, not a minor byproduct of quasar winds. From single-epoch spectra of 11 systems, the paper derives lower limits on the kinetic luminosity, $\\log(\\dot E_{\\rm k}/{\\rm erg~s}^{-1}) > 42.9$ to $49.8$, and feedback efficiency $\\varepsilon_k = \\dot E_{\\rm k}/L_{\\rm Edd}$ that are comparable to or larger than those measured for BAL, high-ionization S IV, and EUV500 outflows. The physical picture that emerges is of low-density gas ($n_e < 0.2$ to $18~{\\rm cm}^{-3}$) located hundreds of kiloparsecs from the nucleus, meaning the energy is deposited in the circumgalactic medium rather than in the immediate vicinity of the black hole. The authors state plainly that this conclusion holds only if the selected NALs are genuinely intrinsic, and that the large distances raise an open question about how these absorbers connect to the quasar-driven outflow.","pith_inferences":["If partial-coverage NALs really sit at hundreds of kiloparsecs, a direct test of the feedback interpretation is to search for the transverse proximity effect: gas near these quasars' sightlines should be over-ionized compared with the general intergalactic medium at the same redshift.","The extreme $\\varepsilon_k > 7.8$ and $\\varepsilon_k > 22$ limits cannot be steady-state efficiencies; taken literally they demand that either the assumed global covering fraction $f_c = 0.5$ is too large, the single ionization parameter is wrong for those systems, or the Eddington luminosities are underestimated.","A detection of C II* or Si II* excited lines in any of these systems, for example with higher signal-to-noise ultraviolet spectra, would convert an upper limit into a density measurement, shrink $R$, and lower the feedback efficiencies, providing a concrete way to sharpen or overturn the claim."],"forward_implications":["NAL outflows, which are present in roughly half of quasars, must be included in AGN feedback budgets; ignoring them would undercount the kinetic energy available to heat or expel gas.","Eight of the 11 systems have lower limits on $\\varepsilon_k$ above 0.5% of $L_{\\rm Edd}$, so if the selection is right these outflows alone meet the threshold thought to affect host-galaxy star formation.","The implied distances (hundreds of kpc to roughly 4 Mpc) mean most of the outflow energy is delivered to the circumgalactic medium, not to the inner kiloparsecs, which changes where and when feedback acts.","Because the method needs only a single epoch of spectra, it can measure feedback efficiency for stable NAL systems that time-variability studies cannot probe.","Two systems have formal lower limits $\\varepsilon_k > 7.8$ and $> 22$, implying kinetic luminosities above the Eddington luminosity when the lower limits are treated as actual values."],"supporting_citations":[{"why":"Supplies the parent NAL survey, the partial-coverage classification into reliability classes, and the 11 systems analyzed here.","marker":"Misawa et al. 2007a"},{"why":"Provides the mass-outflow-rate and kinetic-luminosity formulas in equations (4) and (5), and the lower-limit argument for instantaneous outflows.","marker":"Borguet et al. 2012"},{"why":"Provides the photoionization calibration that fixes log U = -2.6 to match the average C II/C IV column-density ratio.","marker":"Hamann 1997a"},{"why":"Supplies the conventional segmented power-law SED used to convert L_bol into Q(H) and the distance formula behind equation (2).","marker":"Narayanan et al. 2004"},{"why":"Sets the 0.5% Eddington-luminosity threshold used to judge whether the outflows cause significant AGN feedback.","marker":"Hopkins & Elvis 2010"},{"why":"Demonstrates the excited/resonance-line method for measuring electron densities that this paper follows.","marker":"Arav et al. 2013"},{"why":"Provides the Cloudy photoionization model used to convert neutral hydrogen column density into total hydrogen column density at log U = -2.6.","marker":"Ferland et al. 2017"},{"why":"Supplies the critical density of C II used in equation (3) to convert the excited-to-ground-state ratio into an electron-density upper limit.","marker":"Tayal 2008a"},{"why":"Supplies the critical density of Si II used in equation (3) for the same conversion.","marker":"Tayal 2008b"}],"fun_headline_variants":["NAL outflows rival broad winds in AGN feedback efficiency","Intrinsic NAL outflows can rival broad absorption winds","Narrow-line quasar outflows match broad winds in feedback energy","NAL wind feedback efficiency reaches BAL levels at hundreds of kpc","Quasar NAL outflows deposit feedback energy at hundreds of kpc"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes both that the 11 selected absorption systems are truly gas ejected by the quasar rather than small dense clouds in unrelated foreground galaxies, and that one adopted ionization parameter (log U = -2.6) describes all of them; if either premise fails, the derived distances, outflow rates, and efficiencies no longer describe quasar-driven winds.","fun_headline_variants_meta":{"raw":{"variants":["NAL outflows rival broad winds in AGN feedback efficiency","Intrinsic NAL outflows can rival broad absorption winds","Narrow-line quasar outflows match broad winds in feedback energy","NAL wind feedback efficiency reaches BAL levels at hundreds of kpc","Quasar NAL outflows deposit feedback energy at hundreds of kpc"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000973,"raw_usage":{"total_tokens":4230,"prompt_tokens":1134,"completion_tokens":3096,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":750,"completion_tokens_details":{"reasoning_tokens":3009}},"tokens_in":750,"tokens_out":3096,"duration_ms":23834,"temperature":1.0,"reasoning_tokens":3009,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T10:55:12.449878+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Detect the excited-state lines C II* 1336 or Si II* 1265/1533 at the absorption redshift of any of these systems; a detection would replace an electron-density upper limit with a measured density, shorten the derived distance $R$, and lower the quoted feedback efficiencies. Conversely, showing that any selected system is an intervening (sub-)DLA cloud, for instance by detecting a foreground galaxy along the line of sight at $z_{\\rm abs}$ with no association to the quasar, would remove that system's support for the feedback claim.","supporting_citations":[{"cited_title":"doi:10.1086/380781","cited_arxiv_id":null,"evidence_quote":"Supplies the conventional segmented power-law SED used to convert L_bol into Q(H) and the distance formula behind equation (2)."}],"review_version":1}