{"id":"e2dda515-809e-4d4b-a422-9a39003dfa63","arxiv_id":"2502.00927","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Cosmic rays injected at the large-scale jet-driven shock front rather than near the black hole produce more extended pressure support and more effectively quench cooling flows in a 10^14 solar mass halo simulation.","lead":"This paper simulates a massive galaxy cluster to test where cosmic rays produced by an AGN jet should be injected to best stop cooling flows and star formation. It finds that injecting cosmic rays at the large-scale jet shock is more effective than injecting them at the black hole, and that jet precession with a roughly 100 million year period helps.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Shock-injection trigger is not validated against actual shock sites, so the central BH-vs-shock comparison may be an artifact of when the spawn particles decelerate rather than a real feedback effect.","rationale":"The reader's weakest_assumption identifies exactly the same shock-injection trigger (Sec. 2.3) and the energy non-conservation of the injection (Sec. 5.3) as the main sources of concern. I agree with that identification and with the CONDITIONAL verdict. My stress-test does not identify a new independent flaw that would move the verdict to REJECT: the paper is explicit that the injection model is a toy (Sec. 5.3), the qualitative direction of the result (shock-front injection extends CR profiles and mitigates BH-vicinity density suppression) is robust to the details of the toy model, and the comparison across precession geometries is internally consistent for a fixed prescription. The most load-bearing single concern remains the untested mapping from the 1/4-velocity trigger to real shock sites, because the kinematic trigger is not the same as a shock detection and the energy non-conservation is acknowledged rather than quantified. That is exactly why CONDITIONAL is the right verdict, with a concrete test that would resolve the concern rather than a reason to reject the paper. I would keep the verdict UNCHANGED relative to the reader, because the reader's conditional acceptance already encodes the right caution: the central claim is supported but not proven until the shock-trigger assumption is validated against true shock diagnostics.","tokens_in":24616,"tokens_out":1627,"duration_ms":14628,"concrete_test":"Re-run at least one live-accretion comparison pair (CRshock0.1-pr100Myr and CRBH0.1-pr100Myr) with on-the-fly or post hoc shock detection (e.g., velocity-jump/temperature-jump or Mach-number criterion across cell faces) and deposit the same fixed eCR,jet only at cells identified as shocked, with the total injected CR energy budget enforced by conserving total energy. If the extended CR profile and the higher sustained jet flux persist, the central claim stands; if the CR deposition site or energy budget shifts materially, the BH-vs-shock comparison is an artifact of the velocity-trigger prescription.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Sec. 4.5, Conclusions) is that the locus of CR injection is a key control: shock-front injection yields higher sustained jet flux, more extended CR profiles, and more quenching than BH-vicinity injection. The load-bearing assumption is the shock-injection prescription of Sec. 2.3: CRs are deposited when a spawned jet particle's velocity first drops to 1/4 of launch speed, asserted to mean it 'almost surely encountered a strong shock.' This is a kinematic criterion (velocity decrement), not a shock identifier. A spawned particle can decelerate by adiabatic expansion, turbulent drag, or numerical dissipation in a cocoon without crossing a strong shock; conversely, a real shock may be passed without reducing the particle to exactly v/4. Such a trigger would bias CR deposition to a specific kinetic-energy loss contour of the jet plasma, so the BH-vs-shock comparison in Sec. 4.1, Fig. 6, Fig. 8 could be re-expressing a difference in where that velocity-loss contour lies rather than where physical CR production at shocks occurs. Sec. 5.3 explicitly concedes the model is a toy, ignores shock properties, and does not strictly conserve energy during CR injection; the energy added at the injection site could overstate the CR pressure support that the paper attributes to shock acceleration. Without an independent check that the 1/4-velocity contour tracks actual shock sites, the headline result is conditional.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses FIRE-2 MHD simulations of a ~10^14 Msun halo to compare AGN jet feedback models in which cosmic rays (CRs) are injected either in the black hole vicinity or at large-scale jet-driven shock fronts, using both constant-flux jets and live black hole accretion, with variations in precession, opening angle, jet velocity, and CR energy fraction. The central claim is that injection at the jet-driven shock front produces more extended CR distributions, preserves a higher sustained jet energy flux, lengthens the black hole accretion duty cycle, and suppresses cooling flows and star formation more effectively than injection near the black hole. The paper also identifies an 'optimal' precession period of order tens of Myr that places shocks near the cooling radius, and it estimates gamma-ray fluxes, finding all runs below the Fermi-LAT Coma limit.","tokens_in":24957,"tokens_out":7811,"duration_ms":72403,"significance":"If the shock-front injection prescription is physically faithful, the result is significant: it identifies the CR production locus, not just the total CR energy budget, as a control on AGN feedback self-regulation, and it proposes a numerically practical method for including large-scale CR acceleration in galaxy simulations. The paper is honest about its limitations in Sec. 5.3 and does not overclaim success: even the most successful run only intermittently reaches SFR < 5 Msun/yr, and stricter quenching definitions yield even shorter periods. Strengths include the systematic parameter coverage, live gravitational-torque black hole accretion rather than fixed energy injection, anisotropic CR transport with streaming and diffusion, and a central comparison that is not defined in terms of fitted parameters. The main weaknesses are that the shock trigger is an unvalidated kinematic proxy and that the energy budget at CR injection is not conserved in a quantified way; these issues directly affect the headline comparison.","major_comments":[{"comment":"The shock-front CR injection model rests on the assertion that a spawned jet particle whose velocity decelerates to 1/4 of its launch velocity 'has almost surely encountered a strong shock.' This is a kinematic trigger, not a shock identification: deceleration to that level can occur through adiabatic expansion, turbulent drag, or numerical dissipation without a strong shock, while a real shock crossing need not reduce the particle speed to exactly one quarter. The BH-vs-shock comparison in Sec. 4.1 and Figs. 6, 8, and 9 could therefore trace where the v = v_launch/4 contour lies rather than where physical CR acceleration at shocks occurs. Because this assumption is load-bearing for the central claim, I request an independent check: for at least a subset of runs, identify actual shock surfaces in post-processing (e.g., via velocity jumps, entropy jumps, or Mach-number criteria) and compare them with the injection sites, or vary the trigger threshold to demonstrate that the qualitative contrast is robust. The paper's own Sec. 5.3 limitation statement correctly flags this as a toy model, but the central claim is currently conditional on this trigger tracking real shocks.","section":"Sec. 5.3"},{"comment":"The statement that 'we did not strictly conserve energy when injecting cosmic ray energy at the shock front' is a concern for the mechanism claimed in Secs. 3.2 and 4.1, because the quenching effect is attributed to CR pressure gradients. If CR energy is added at the trigger without a compensating subtraction from the gas kinetic or thermal energy, or from the jet energy budget defined in Eq. (5), the CR pressure support could be artificially enhanced. The paper says the injected CRs contribute less than 0.3% of the energy budget, but no derivation, table, or figure supports this number, and it appears in tension with the nominal CR energy fractions of 0.1-0.3 quoted in Tables 2 and 3. Please clarify the implementation (what energy, if anything, is removed at injection) and provide a quantitative energy-accounting check for a representative shock-injection run.","section":"Sec. 5.3"},{"comment":"The identification of an 'optimal range of jet precession periods (~ tens of Myr)' goes beyond the simulated values. The constant-flux and live-accretion runs include precession periods of 10 Myr and 100 Myr, plus an effectively very short period for the isotropic wind, but no intermediate period such as 30-50 Myr. The finding that 100 Myr places shocks near 30 kpc while 10 Myr gives faster early suppression does not uniquely determine a 'tens of Myr' optimum; it is an interpolation between two points. Please add an intermediate-period run or soften the claim to 'periods between 10 and 100 Myr' with an explicit caveat that the location of the optimum is not resolved by the present grid.","section":"Sec. 4.5 and Abstract"},{"comment":"The gamma-ray 'prediction' is partially circular. The fiducial diffusivity kappa_CR = 1e29 cm^2/s was adopted because, as stated in Sec. 2, previous work matching observed gamma-ray luminosities requires this value. Equation (8) and the accumulated gamma-ray fluxes in Fig. 12 use this same kappa_CR, so the statement that all runs fall within the Fermi-LAT Coma limit is a consistency check with a parameter already calibrated to gamma-ray data, not an independent prediction. Please state this explicitly in Sec. 5.2 and, if possible, show the sensitivity of the gamma-ray fluxes to kappa_CR (for example, a factor of 3 variation) so the reader can judge whether the conclusion is robust.","section":"Sec. 5.2, Eq. (8), Fig. 12"}],"minor_comments":[{"comment":"'sbugrid' is a typo for 'subgrid'; please correct.","section":"Sec. 2.3"},{"comment":"The heading 'Eenergy flux required for cosmic rays to quench star formation' contains a typo ('Eenergy').","section":"Sec. 5.1"},{"comment":"In both table captions, 'processing period' should be 'precession period'; the Table 2 caption also repeats 'completely completely'.","section":"Tables 2 and 3"},{"comment":"The legend entry 'all CR at BH' is not defined in the captions; please identify the run (live-v3e3-CRBH1) and note that its CR energy fraction is order unity.","section":"Figs. 6 and 8"},{"comment":"The parenthetical values in the magnetic-energy columns (mass-weighted averages) are easy to misread; a footnote or explicit header would improve clarity.","section":"Tables 2 and 3"},{"comment":"The notation switches between n_H and n_gas in the same equation; please define n_H explicitly and use a consistent symbol.","section":"Sec. 5.2, Eq. (8)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is honest about its limitations, and the simulation campaign is substantial. My main reservation is not the existence of caveats but that the central BH-vs-shock comparison rests on an unvalidated kinematic trigger, and the energy non-conservation at injection is not quantified. These issues are fixable within the manuscript's scope via post-hoc shock identification, an energy-accounting audit, and either softening or resolving the optimal-precession claim. I do not see grounds for rejection, but I would not accept the paper without these checks."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the systematic comparison, in live-accretion FIRE simulations, of cosmic-ray injection at the black hole versus at the large-scale jet-driven shock. The result—that shock-front injection preserves accretion duty cycles, spreads CRs farther out, and suppresses the cooling flow more effectively than BH-vicinity injection—is a real advance over the previous constant-flux treatments. The paper also does something I wish more simulation papers did: it states its own limitations plainly, including that the shock-injection model is a toy, does not conserve energy at injection, and ignores shocks not tied to the jet (Sec. 5.3). The authors are not overselling this as a finished subgrid model.\n\nThe main soft spot is exactly where the stress-test points. The injection trigger is purely kinematic: deposit CRs when a spawned jet particle decelerates to 1/4 of its launch speed. The paper justifies this with the strong-shock jump condition, but nothing in the simulations verifies that those deceleration events actually coincide with strong shocks. Adiabatic expansion or numerical dissipation in the cocoon could trigger the same contour. That said, the qualitative contrast between the two injection sites does not obviously hinge on the precise location of that contour—the BH-vicinity case deposits energy before the jet escapes, while the shock case deposits it after substantial propagation. The non-conservation of energy at injection is shown to be small (<0.3% of the budget), so it is unlikely to explain the pressure support attributed to CRs. Still, the referee should ask for some demonstration that the 1/4-velocity contour tracks shocks—for instance, a comparison with a simple shock-finding criterion in a subset of snapshots.\n\nThe other concerns are minor. The 'optimal precession period of tens of Myr' in the abstract is a bit stronger than the simulated values of 10 and 100 Myr; the body more carefully says ≲100 Myr. The gamma-ray comparison is partly circular because the same diffusivity used in the simulation was previously calibrated to Fermi-LAT limits; the authors call it a consistency check rather than a prediction, which is fair, but it should be labeled as such in the abstract. The definition of quenching as SFR <5 Msun/yr is loose, and the authors admit stricter definitions shorten the quenched periods; that caveat belongs in the main text, not just Sec. 5.3.\n\nWho is this for? The AGN feedback simulation community, particularly people working on CR transport and cool-core clusters. It is a solid parameter study with honest reporting, and it deserves a serious referee. My recommendation: send it out, with the request that the authors either validate the injection trigger against a shock-finding method or clearly reframe the central claim as conditional on the toy prescription. The qualitative conclusion is likely to survive that test, but the paper will be stronger if the load-bearing assumption is actually checked.","headline":"A useful, honest simulation study showing that where AGN cosmic rays are injected (shock front vs black hole vicinity) changes whether feedback self-regulates; the central result is probably robust but rests on an acknowledged toy prescription that needs sharper validation.","tokens_in":892,"tokens_out":994,"would_cite":true,"duration_ms":28247,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Cosmic rays generated at the jet's large-scale shock front, rather than at the black hole, are what let an AGN jet suppress cooling flows and quench star formation in a massive halo.","keywords":["AGN feedback","cosmic rays","jet-driven shocks","cooling flows","galaxy quenching","magnetohydrodynamic simulations","jet precession","circumgalactic medium"],"falsifier":"Run one of the live-accretion jet models with an explicit on-the-fly shock finder that deposits cosmic rays wherever strong shocks are actually detected; if the differences between black-hole injection and shock-front injection in accretion duty cycle and star formation shrink or vanish, the proposed mechanism is not the operative one. Alternatively, measure the gamma-ray radial profile of a cool-core cluster hosting a precessing jet: the model predicts a less centrally concentrated gamma-ray flux for jets with roughly 10 to 100 Myr precession than for black-hole-injected cosmic rays.","tokens_in":24429,"feed_emoji":"🌌","tokens_out":7000,"duration_ms":70033,"temperature":0.7,"pith_summary":"Massive galaxies and clusters stay red and dead only if something stops their hot gas from cooling and forming stars. This paper argues that the location of cosmic-ray production in AGN jets is that something. Using magnetohydrodynamic simulations of a $10^{14}\\,M_\\odot$ halo with live black-hole accretion, it shows that cosmic rays made at the large-scale jet-driven shock front, rather than near the black hole, leave the jet with a higher energy flux, spread the cosmic rays to larger radii, and suppress cooling flows and star formation more effectively. The paper also identifies a sweet spot for jet precession, periods of roughly tens of megayears, that puts the shock front at the cooling radius where it does the most work.","feed_headline":"Cosmic rays born at jet shocks quench galaxies best","feed_subtitle":"Simulations of a massive cool-core halo show shock-front cosmic rays beat black-hole injection for suppressing star formation.","key_machinery":"The central mechanism is a particle-spawning jet with a deceleration trigger for cosmic-ray injection. New gas particles are launched from the black hole; when one decelerates to one quarter of its launch velocity, which for a strong shock in $\ngamma=5/3$ gas is the post-shock velocity, the code deposits a fixed cosmic-ray energy per unit mass at that location, modeling Fermi acceleration at the large-scale jet cocoon shock. The argument then runs through the cosmic-ray pressure gradient: shock-injected cosmic rays build a pressure gradient that offsets gravity at the cooling radius, while black-hole-injected cosmic rays push on dense gas near the hole, suppress accretion, and starve the jet.","core_discovery":"This paper argues that where cosmic rays are put into an AGN jet matters as much as how much energy they carry. In simulations of a cool-core $10^{14}\\,M_\\odot$ halo, injecting cosmic rays near the black hole inflates pressure in the accretion region, throttles black-hole accretion, and leaves the jet too weak to reach large radii; this produces episodic accretion and little impact on the cooling flow. Injecting the same cosmic-ray energy at the large-scale jet-driven shock front instead preserves a higher overall jet energy flux, disperses cosmic rays to larger radii, and maintains a cosmic-ray pressure gradient at roughly $10$\\,--\\,$30$ kpc that balances gravity near the cooling radius. The result is more effective, longer-lived suppression of the cooling flow and of star formation. The paper further finds that a precession period of roughly tens of megayears places the shock front at the inner circumgalactic medium, which is where the cooling flow is strongest, and that the combination of shock-front injection, a $\nlesssim$100 Myr precession period, and a cosmic-ray energy fraction near 0.3 of the jet energy produces the strongest quenching seen in its live-accretion runs.","pith_inferences":["If the injection locus is the controlling variable, sub-grid AGN feedback implementations that deposit cosmic rays only at the black hole will systematically underestimate cosmic-ray pressure support at the cooling radius and overestimate the episodicity of black-hole accretion.","The same logic gives a direct observational discriminant: cool-core clusters hosting precessing jets with tens-of-megayear periods should show flatter gamma-ray radial profiles than those with steady jets, a signature that deeper gamma-ray observations could test.","Because the velocity-deceleration trigger is only a proxy for shock acceleration, rerunning the live-accretion cases with an explicit shock finder and continuous cosmic-ray injection would test whether the optimal precession period shifts with the actual shock location.","The proposed optimal precession timescale of roughly tens of megayears suggests that jet direction changes on that timescale may be a physical requirement for efficient AGN self-regulation, not just a numerical choice."],"forward_implications":["At fixed jet power, shock-front cosmic-ray injection keeps the black-hole accretion duty cycle long, roughly 0.5 to 1 Gyr, rather than episodic on a 100 to 200 Myr timescale, so the jet carries more total energy to the halo.","A jet precession period of roughly 10 to 100 Myr places the shock front at a few tens of kiloparsecs, near the cooling radius; non-precessing jets push cosmic rays into a narrow beam beyond 100 kpc and barely suppress the cooling flow.","Raising the cosmic-ray fraction to about 0.3 of the jet energy, with shock-front injection and a 100 Myr precession period, gives the strongest quenching, with star formation rates below about $5\\,M_\\odot\\,\\mathrm{yr}^{-1}$ sustained with duty cycles of at least 0.5 Gyr.","Predicted gamma-ray fluxes stay below the Fermi-LAT upper limit of about $1.8\\times10^{42}\\,\\mathrm{erg\\,s^{-1}}$, and shock-front injection lowers the central gamma-ray flux because the cosmic-ray distribution is less concentrated.","Including cosmic rays from supernovae adds to the cosmic-ray pressure and works together with AGN cosmic rays to suppress cooling flows more thoroughly."],"supporting_citations":[{"why":"Supplies the isolated cool-core halo initial conditions and establishes that non-AGN feedback alone is insufficient to resolve the cooling-flow problem.","marker":"(Su et al. 2019)"},{"why":"Sets the requirement that successful AGN jet models need energy fluxes comparable to the free-fall flux at the cooling radius, which anchors the jet energies used here.","marker":"(Su et al. 2020)"},{"why":"Previous study showing cosmic-ray-dominated AGN jets are effective quenchers and giving the cosmic-ray flux scale needed to quench, which this paper extends to shock-front injection.","marker":"(Su et al. 2021)"},{"why":"Provides the gravitational-torque black-hole accretion model used to couple jet fluxes to live black-hole growth.","marker":"(Hopkins & Quataert 2010)"},{"why":"Supplies the particle-spawning method used to launch the jet with controlled mass, velocity, temperature, and magnetic-field properties.","marker":"(Torrey et al. 2020)"},{"why":"Implements cosmic-ray transport with streaming, diffusion, and losses and calibrates the fiducial diffusivity of $10^{29}\\,\\mathrm{cm^2\\,s^{-1}}$.","marker":"(Chan et al. 2019)"},{"why":"Provides the hadronic-loss and gamma-ray emissivity calculation used to compare predicted gamma-ray fluxes with observations.","marker":"(Guo & Oh 2008)"}],"fun_headline_variants":["Where AGN cosmic rays are born decides galaxy quenching","Shock-front cosmic rays beat black-hole injection","Cosmic ray injection site tips AGN feedback balance","Jet shock cosmic rays quench better than black-hole rays"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result rests on treating a jet particle's slowdown to one quarter of its launch speed as a reliable stand-in for crossing a strong shock, and on depositing a fixed cosmic-ray energy per unit mass at that moment without strictly conserving total energy.","fun_headline_variants_meta":{"raw":{"variants":["Where AGN cosmic rays are born decides galaxy quenching","Shock-front cosmic rays beat black-hole injection","Cosmic ray injection site tips AGN feedback balance","Jet shock cosmic rays quench better than black-hole rays"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00016,"raw_usage":{"total_tokens":1323,"prompt_tokens":1125,"completion_tokens":198,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":741,"completion_tokens_details":{"reasoning_tokens":135}},"tokens_in":741,"tokens_out":198,"duration_ms":2661,"temperature":1.0,"reasoning_tokens":135,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T17:12:09.305704+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run one of the live-accretion jet models with an explicit on-the-fly shock finder that deposits cosmic rays wherever strong shocks are actually detected; if the differences between black-hole injection and shock-front injection in accretion duty cycle and star formation shrink or vanish, the proposed mechanism is not the operative one. Alternatively, measure the gamma-ray radial profile of a cool-core cluster hosting a precessing jet: the model predicts a less centrally concentrated gamma-ray flux for jets with roughly 10 to 100 Myr precession than for black-hole-injected cosmic rays.","supporting_citations":[],"review_version":1}