{"id":"73f56289-259a-4757-892c-9c918fb68775","arxiv_id":"2412.13250","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Deeper MUSE spectroscopy of MACS J0138-2155 releases 107 redshifts, two newly discovered lensed Lyman-alpha sources, and a Faber-Jackson calibration for cluster members that anchors strong lensing models for H0.","lead":"New MUSE observations of the lensing cluster MACS J0138-2155 produce a catalogue of 107 galaxy redshifts, 13 lensed multiple images, and stellar velocity dispersions for 16 galaxies. The authors calibrate the Faber-Jackson relation for cluster members, a key input for mass models that will turn the time delays of two lensed supernovae into an H0 measurement.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unresolved velocity-dispersion discrepancies and unpublished simulation calibration leave the Faber-Jackson slope vulnerable to systematic bias.","rationale":"The paper is a strong observational contribution: the 107-object redshift catalogue and the identification of 13 lensed images are well supported by the shown spectra and QF system. The kinematic measurements are the least secure link in the chain that leads to the FJ calibration and hence to the lens-model priors used for H0. The paper honestly discloses the two large discrepancies with Flowers et al. and the reliance on an in-prep. simulation calibration, but disclosure does not remove the risk. The proposed test directly quantifies how sensitive the headline FJ slope is to the disputed measurement and checks which pipeline is biased. If the slope is insensitive, the central claim survives; if it shifts, the calibration should be revised before the priors are used at full weight. This does not change the reader's CONDITIONAL verdict; it sharpens the condition.","tokens_in":21049,"tokens_out":4243,"duration_ms":39833,"concrete_test":"Refit the FJ relation for the 13 cluster members twice: once with ID 930 at 97 ± 5 km/s and once with ID 930 at 167 ± 15 km/s, using the published F160W magnitudes. If the slope changes by more than the quoted ±0.05, the disputed kinematic measurement materially affects the central calibration. Also re-extract and re-fit ID 930 from the released MUSE cube using the Flowers et al. single-template setup and a full spectral library with pPXF; the pipeline that reproduces the higher value is the biased one, and the paper's sigma_v for this member would need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the Faber-Jackson relation with alpha = 0.25 ± 0.05 is reliable rests on the stellar velocity dispersions in Table 1 being accurate and their uncertainties correctly estimated. Two entries, ID MUSE 930 (97 ± 5 km/s) and the background galaxy ID MUSE 873 (238 ± 2 km/s), disagree with the independent Flowers et al. (2024) values (167 ± 15 and 291 ± 3 km/s) by far more than the quoted error bars. The paper attributes this to the shallower data used by Flowers et al., but the opposite bias—in the present pPXF measurements—is not ruled out. The error calibration itself rests on 16,000 simulated spectra described only in the unpublished Granata et al. (in prep.), so the uncertainty estimates cannot currently be independently checked. If even one cluster member's sigma_v is biased as strongly as ID 930, the 13-point FJ fit can shift, and the slope and scatter used as lens-model priors in Ertl et al. (2025) and Suyu et al. (in prep.) inherit the bias, potentially affecting the H0 inference from SN Encore/Requiem time delays.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents VLT/MUSE spectroscopy of the galaxy cluster MACS J0138−2155, totalling 3.7 hours, and uses it to build a spectroscopic catalogue of 107 objects with reliable redshifts, including 50 cluster members and 13 multiply lensed images from four background sources (among them four images of the host galaxy of SNe Requiem and Encore). It measures line-of-sight stellar velocity dispersions for 14 cluster members and two background galaxies from pPXF fits with a carefully selected stellar template library, and calibrates the Faber-Jackson relation L ∝ σ^{1/α} for 13 early-type cluster members using F160W magnitudes from a companion paper, finding α = 0.25 ± 0.05 and intrinsic scatter 25 km/s. The calibrated relation is compared with six other strong-lensing clusters and found to be consistent with no significant redshift evolution. The kinematic catalogue and scaling relation are intended as priors for strong lensing models used for H0 measurement from the two lensed supernovae.","tokens_in":21227,"tokens_out":7638,"duration_ms":68753,"significance":"If the measured velocity dispersions and the resulting Faber-Jackson calibration are reliable, this paper provides a valuable data product for cluster lensing cosmology: a complete spectroscopic member sample, new multiply imaged background sources, and a direct kinematic prior on the dPIE velocity dispersion parameters that break a known degeneracy in cluster lens models. The data reduction is careful and the redshift catalogue is a concrete deliverable. The main weakness is that the two load-bearing ingredients for the FJ claim, the velocity dispersion error calibration and the FJ fit, rest on materials that are either unpublished or disputed by an independent measurement, so the paper's central reliability claim is not yet fully supported.","major_comments":[{"comment":"The velocity dispersion for ID MUSE 930 (97 ± 5 km/s) differs by ~70 km/s from the independent measurement of Flowers et al. (2024) (167 ± 15 km/s), and ID MUSE 873 differs by ~50 km/s (238 ± 2 vs 291 ± 3 km/s). Because ID 930 is one of the 13 points used to fit the Faber-Jackson relation, the authors should demonstrate that the fitted parameters (α, σ_v,ref, Δσ_v) are robust to this possible systematic. A straightforward sensitivity test would be to refit the relation without ID 930 (or using the Flowers et al. value) and report the resulting parameters; without such a test, the claim that the calibrated slope is reliable is not yet established.","section":"Section 4.1, Table 1"},{"comment":"The uncertainty on each σ_v value is derived from a calibration based on 16,000 simulated spectra that is described only in the unpublished Granata et al. (in prep.) paper. Since these error bars enter directly into the Faber-Jackson likelihood and the quoted intrinsic scatter, the paper should either provide the essential elements of the simulation (input galaxy models, noise recipes, the fitted relative-uncertainty versus S/N relation) or include them in an appendix, so that the error estimates are reproducible and can be assessed by the reader.","section":"Section 4.1"},{"comment":"The Faber-Jackson fit treats the F160W magnitudes from Ertl et al. (2025) as fixed (not explicitly including their uncertainties). If magnitude errors are non-negligible, the fitted slope α could be biased. Please clarify whether the likelihood adopted from Bergamini et al. (2019) includes errors on both variables, and if not, provide a quantitative statement of the impact of magnitude uncertainties on α and σ_v,ref.","section":"Section 4.2, Eq. (1) and Table 2"}],"minor_comments":[{"comment":"In the Summary, the paper states 'we report 4 lensed images of the host of SN Encore and SN Requiem, and 11 images from three other lensed background sources', which sums to 15, but the catalogue in Section 3.2 and Table A.1 lists 13 multiply lensed images in total, with 9 images from the three other sources. Please correct the count.","section":"Section 5"},{"comment":"There are typographical issues, e.g., 'reduced e ffective exposure time' and 'James WebbSpace Telescope'; please fix the spacing.","section":"Section 2"},{"comment":"The sentence 'we probed the relative uncertainty on the value of sigma as a function of S/N from the 1σ scatter of the recovered σv values about the input velocity dispersion of the 16,000 simulated spectra' could be reworded for clarity, e.g., 'we measured the relative uncertainty as a function of S/N from the 1σ scatter of the recovered σv values about the input values.'","section":"Section 4.1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript relies on several companion papers in preparation (Ertl et al. 2025, Acebron et al. 2025, Suyu et al. in prep., Granata et al. in prep.) for the magnitudes, lens models, and error calibration. This is common in this community, but the FJ slope claim depends critically on Granata et al. (in prep.) not being available. I recommend asking the authors to include the sensitivity test for ID 930 and a short appendix on the simulation calibration before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a solid, workmanlike contribution that moves the SN Encore/Requiem H0 program forward. The new MUSE data (2.9 h of ToO) yield a deep redshift catalogue for MACS0138: 107 objects, 50 cluster members, 13 lensed images from four background sources including two newly discovered Lyman-alpha emitters. The kinematic catalogue (14 cluster members, 2 background galaxies) is the first for this cluster and directly attacks the sigma_v-r_t degeneracy that plagues cluster lens models. The Faber-Jackson calibration, alpha = 0.25 +/- 0.05 with 25 km/s scatter, is consistent with six other clusters in the same redshift range - a good cross-check.\n\nThe paper is also honest. It openly discusses the two velocity dispersions that disagree with Flowers et al. (2024): ID 930 (97 vs 167 km/s) and ID 873 (238 vs 291 km/s), and offers a reasonable explanation based on Flowers' shallower data. That does not, however, rule out bias in the present pPXF measurements. And the error calibration from 16,000 simulated spectra is only described in Granata et al. (in prep.), so the quoted uncertainties aren't independently checkable. If even one cluster member is biased at the level of ID 930, the FJ slope and scatter shift, and those priors propagate into the lens models used for H0. This is the weak spot, and it's the same one the stress-test note identifies.\n\nNone of this is fatal. The catalogue and the calibration are likely correct; the discrepancies are caution flags, not a smoking gun. I'd send this to a serious referee, and I'd want the authors to either release the simulation calibration or fold a summary into the paper, and to do a head-to-head re-analysis of the two outliers. Once the companion lens-model papers (Ertl et al. 2025, Suyu et al. in prep.) land, this will be a standard reference for the cluster. I'd cite it, and I'd bring it to the reading group as a case study in how kinematics and lens models interact.","headline":"A solid, honest data paper that moves the SN Encore/Requiem H0 program forward; the two unresolved sigma_v discrepancies and unpublished error calibration are real but not fatal.","tokens_in":21869,"tokens_out":3801,"would_cite":true,"duration_ms":35365,"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":"This paper establishes the spectroscopic and kinematic groundwork that turns the two strongly lensed supernovae in the cluster MACS J0138–2155 into a Hubble-constant measurement.","keywords":["galaxy clusters: individual (MACS J0138.0−2155)","gravitational lensing: strong","galaxies: distances and redshifts","galaxies: kinematics and dynamics","galaxies: elliptical and lenticular, cD","supernova time delays","Hubble constant","Faber-Jackson relation"],"falsifier":"Take the 14 cluster members with measured velocity dispersions and re-observe them at higher spectral resolution and S/N with an independent instrument and fitting code. If the recovered dispersions differ from the published values by more than the quoted uncertainties (as two of the values already differ from an independent shallower catalogue), then the Faber–Jackson calibration and the lens-model priors built on it are biased. A cheaper intermediate check is to re-run pPXF on the same spectra with the single-template approach used by that shallower catalogue and see whether the two discrepant galaxies move toward its values.","tokens_in":20833,"feed_emoji":"🔭","tokens_out":11563,"duration_ms":101485,"temperature":0.7,"pith_summary":"Gravitational lensing by the galaxy cluster MACS J0138–2155 produces multiple images of background sources, including four images of the host galaxy of the two supernovae Requiem and Encore; the time delays among those images depend on the Hubble constant, but only if the cluster's total mass distribution is modelled accurately. This paper supplies the observational ingredients that make such a model possible. From 3.7 hours of MUSE integral-field spectroscopy it builds a catalogue of 107 reliable redshifts, identifying 50 cluster members and 13 lensed images from four background sources, and it measures line-of-sight stellar velocity dispersions for 14 bright cluster members. Combining those kinematics with HST F160W magnitudes, it calibrates the Faber–Jackson relation, $L \\propto \\sigma^{1/\\alpha}$, finding $\\alpha = 0.25^{+0.05}_{-0.05}$ with an intrinsic scatter of $25^{+6}_{-4}$ km/s. That relation sets the velocity-dispersion parameter of every member galaxy in the strong-lensing mass model, breaking the degeneracy between member velocity dispersion and truncation radius. The result is a cluster mass model from which the Requiem–Encore time delays can be converted into an $H_0$ measurement.","feed_headline":"MUSE spectra pin the mass law behind the Requiem–Encore H0","feed_subtitle":"A measured mass–luminosity slope anchors the cluster model needed to turn Requiem–Encore delays into H0.","key_machinery":"The load-bearing mechanism is the Faber–Jackson relation, $L \\propto \\sigma^{1/\\alpha}$, calibrated by a Markov-chain Monte Carlo fit from 13 early-type cluster members whose line-of-sight velocity dispersions $\\sigma_v$ come from pPXF full-spectrum fits of MUSE spectra, and whose luminosities come from HST F160W magnitudes. A dual pseudo-isothermal elliptical mass distribution (a dPIE, a truncated mass profile with a velocity-dispersion parameter and a half-mass radius) is the standard model component used for cluster members; for vanishing core radius its velocity-dispersion parameter is well approximated by the central stellar velocity dispersion. The calibrated relation sets that parameter from observed magnitude for every member, so the strong-lensing model no longer has to explore the degenerate velocity-dispersion/truncation-radius plane freely. The companion strong-lensing models adopt exactly this prior, which is how the kinematics enter the time-delay prediction.","core_discovery":"The paper's central claim is that the new MUSE data yield a pure, complete spectroscopic sample of the cluster and a reliably calibrated Faber–Jackson relation for its early-type members, and that these are what make the upcoming $H_0$ measurement from the two lensed supernovae trustworthy. The catalogue contains 107 secure or likely redshifts, with 50 cluster members in the narrow range $0.324<z<0.349$ and 13 lensed images from four background sources spanning $0.767 \\le z \\le 3.420$, including the four images of the SN host at $z=1.95$; two of the background sources are new Lyman-$\\alpha$ emitters found by scanning the cube. For the kinematics, pPXF full-spectrum fits on surface-brightness-weighted apertures give velocity dispersions for 14 cluster members and two background galaxies. Fitting $L \\propto \\sigma^{1/\\alpha}$ to 13 early-type members yields $\\alpha = 0.25^{+0.05}_{-0.05}$, $\\sigma_{v,\\mathrm{ref}} = 206^{+14}_{-13}$ km/s at the reference magnitude, and an intrinsic scatter of $25^{+6}_{-4}$ km/s. The slope is consistent with Faber–Jackson calibrations in six other lensing clusters at $0.31\\le z \\le 0.59$, so the paper concludes there is no statistically significant redshift evolution of the relation in that range. Because the dPIE velocity-dispersion parameter of a member is well approximated by its central stellar velocity dispersion, this calibrated relation gives the lens model an observational prior for every member's mass, which is what breaks the velocity-dispersion/truncation-radius degeneracy that would otherwise bias the predicted time delays.","pith_inferences":["If the Faber–Jackson prior is as informative as claimed, the resulting $H_0$ from Requiem–Encore should be systematically more precise than it would be with the member masses left free; this can be tested by comparing the $H_0$ posterior with and without the kinematic prior in the same lens model.","The reported tens-of-km/s disagreements for two objects with an independent shallower catalogue suggest the systematic floor of the kinematic measurements is not yet fully settled, so a targeted re-observation of those two galaxies at higher S/N with a different template set would be a decisive check.","The S/N $\\ge 10$ threshold and the quoted uncertainties rest on 16,000 simulated spectra described in a companion paper that has not yet appeared; publishing those simulations would let other groups reproduce the error budget and extend the method to shallower data."],"forward_implications":["The 107-object redshift catalogue pins down which galaxies are cluster members and which are lensed background objects, so the strong-lensing model of MACS0138 no longer has to fit the member list and image redshifts as free unknowns.","With the Faber–Jackson prior assigning a velocity dispersion to every member from its F160W magnitude, the lens model can predict the time delays between the four images of the SN host with a controlled member-mass uncertainty, making the subsequent $H_0$ measurement from Requiem and Encore possible.","The two newly discovered Lyman-$\\alpha$ sources at $z=3.152$ and $z=3.420$ add multiple-image constraints at high redshift, strengthening the model against the mass-sheet degeneracy.","The consistency of the calibrated slope with six other clusters at $0.31\\le z\\le 0.59$ implies the same mass-luminosity scaling can be used as a prior for other cluster lenses in that redshift range without independent kinematics for every member.","The quoted intrinsic scatter of 25 km/s around the relation gives a direct estimate of how much member mass varies at fixed luminosity, which can be propagated into the final $H_0$ uncertainty budget."],"supporting_citations":[{"why":"Defines the luminosity–velocity-dispersion power law that the paper calibrates for the cluster members.","marker":"Faber & Jackson 1976"},{"why":"Introduces the pPXF penalized pixel-fitting method used to measure the stellar velocity dispersions.","marker":"Cappellari & Emsellem 2004"},{"why":"Supplies the MCMC fitting procedure for the Faber–Jackson relation and the argument that the dPIE velocity-dispersion parameter equals the central stellar dispersion, breaking the velocity-dispersion/truncation-radius degeneracy.","marker":"Bergamini et al. 2019"},{"why":"Provides the redshift-measurement and template-matching method followed to build the spectroscopic catalogue.","marker":"Caminha et al. 2023"},{"why":"Previous application of the kinematic extraction and weighting procedure on which the velocity-dispersion measurements are based.","marker":"Granata et al. 2023"},{"why":"Reports the discovery of SN Requiem and the earlier [OII]-emitter multiple images in this cluster.","marker":"Rodney et al. 2021"},{"why":"Announces SN Encore and the four JWST-detected images of the shared host galaxy that the MUSE data confirm at z=1.95.","marker":"Pierel et al. 2024b"},{"why":"Companion strong-lensing models that incorporate the member catalogue, multiple-image constraints, and Faber–Jackson prior to predict the SN time delays.","marker":"Ertl et al. 2025"}],"fun_headline_variants":["MUSE spectra calibrate the mass law for the Requiem–Encore H0","Faber-Jackson slope of 0.25 from MUSE anchors the lensing H0","MUSE at VLT provides 107 redshifts for the two-SN H0 lensing","Faber-Jackson calibration from MUSE sets prior for SN H0","107 MUSE redshifts prepare the cluster for a reliable H0"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the velocity dispersions recovered from pPXF fits to MUSE spectra are unbiased estimates of the dPIE velocity-dispersion parameter that the lens model assigns to each cluster member; if those kinematics are biased, the Faber–Jackson relation and every strong-lensing model that uses it inherit that bias.","fun_headline_variants_meta":{"raw":{"variants":["MUSE spectra calibrate the mass law for the Requiem–Encore H0","Faber-Jackson slope of 0.25 from MUSE anchors the lensing H0","MUSE at VLT provides 107 redshifts for the two-SN H0 lensing","Faber-Jackson calibration from MUSE sets prior for SN H0","107 MUSE redshifts prepare the cluster for a reliable H0"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000921,"raw_usage":{"total_tokens":4129,"prompt_tokens":1305,"completion_tokens":2824,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":921,"completion_tokens_details":{"reasoning_tokens":2728}},"tokens_in":921,"tokens_out":2824,"duration_ms":19925,"temperature":1.0,"reasoning_tokens":2728,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:19:22.261616+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the 14 cluster members with measured velocity dispersions and re-observe them at higher spectral resolution and S/N with an independent instrument and fitting code. If the recovered dispersions differ from the published values by more than the quoted uncertainties (as two of the values already differ from an independent shallower catalogue), then the Faber–Jackson calibration and the lens-model priors built on it are biased. A cheaper intermediate check is to re-run pPXF on the same spectra with the single-template approach used by that shallower catalogue and see whether the two discrepant galaxies move toward its values.","supporting_citations":[{"cited_title":"2019, , 631, A130","cited_arxiv_id":null,"evidence_quote":"Supplies the MCMC fitting procedure for the Faber–Jackson relation and the argument that the dPIE velocity-dispersion parameter equals the central stellar dispersion, breaking the velocity-dispersion/truncation-radius degeneracy."},{"cited_title":"B., Grillo , C., Rosati , P., et al","cited_arxiv_id":null,"evidence_quote":"Provides the redshift-measurement and template-matching method followed to build the spectroscopic catalogue."},{"cited_title":"2023, , 679, A124","cited_arxiv_id":null,"evidence_quote":"Previous application of the kinematic extraction and weighting procedure on which the velocity-dispersion measurements are based."}],"review_version":1}