{"id":"bce41c89-8917-4c37-951c-83f129621a86","arxiv_id":"2501.18090","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Differential emission measure analysis of SN 1987A from 2007 to 2021 shows the cool ring component fading while a hot ejecta component grows, consistent with MHD simulations.","lead":"A new analysis of 14 years of XMM-Newton X-ray spectra of the supernova remnant SN 1987A maps the hot gas's temperature structure over time. It finds the original ring emission fading and a new high-temperature component from freshly shocked ejecta brightening after about 2014.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (3)'s power-law net–kT relation is the load-bearing assumption; it is untested at the epochs where the claimed 3–5 keV secondary peak appears, so that peak could be an artifact of the ionization model.","rationale":"The reader's weakest-assumption analysis identifies Eq. (3) as the key vulnerability, and I agree: the power-law net–kT relation is the single most load-bearing step because every temperature bin's ionization state, and hence the emissivity used to convert counts into EM, is set by that relation. The paper's own caveat in Section 4.3 is an explicit admission that this assumption is questionable, and the MHD simulations show a non-power-law, component-dependent net(T) with large scatter. The secondary peak at 3–5 keV in 2020–2021 is the main new feature that drives the ejecta-brightening interpretation, and it appears exactly in the temperature range where the simulated net(T) deviates most strongly from a power law. The existing validation tests are reassuring but not decisive: the synthetic MHD test in Figure 1d is for 2014 and verifies only that the main peak and a tail are recovered, not that a secondary peak is real; the vpshock-based DEM in Appendix C uses the same power-law net–T assumption, so it is not an independent check of that assumption; and the Fe K centroid decline is only moderately significant (about 2–3 sigma, Appendix D). I therefore do not see a reason to change the reader's CONDITIONAL verdict: the observational trends are credible and the modeling is thorough, but the specific identification of a secondary peak and its physical attribution to newly shocked ejecta should be tested against an alternative net(T) prescription before the claim is accepted as robust. My concrete test directly targets that gap by refitting the two decisive epochs with the MHD-derived net(T) profile and with extreme beta values; persistence across these would substantially increase confidence, while disappearance would indicate that the peak is a modeling artifact.","tokens_in":37406,"tokens_out":4650,"duration_ms":55626,"concrete_test":"Refit the 2020 Nov and 2021 Dec RGS+EPIC-pn spectra with the same c7pvnei model but replace Eq. (3) with the EM-weighted net(T) relation taken from the B18.3 MHD simulations (middle panels of Figure 4), and also with beta fixed at -1 and +1. If the 3–5 keV secondary peak and the post-2014 major-peak EM decline persist under all three net(T) prescriptions, the conclusion is robust; if the secondary peak disappears or shifts, Eq. (3) is the load-bearing assumption and the claim should be softened to 'a high-temperature tail is present' pending a more flexible ionization treatment.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the ring is fading and ejecta are brightening rests on two DEM features: the post-2014 decline of the major peak and the growth of a high-temperature tail into a 3–5 keV secondary peak in 2020–2021. The DEM reconstruction assigns emission measure to temperature bins through 40 vnei components whose ionization parameters are tied by Eq. (3), tau_i = tau_1keV (T_i/1keV)^beta. In NEI plasmas, temperature and net are strongly degenerate: an underionized high-temperature plasma can mimic cooler CIE emission, so a wrong net(T) can redistribute EM in T and even create peaks. The paper's own Section 4.3 states that the power-law assumption 'might be over-simplified' and that the results 'must be treated with caution,' and the MHD simulations in Figure 4 show that the true EM-weighted net(T) is not a single power law (plateau at 0.3–1 keV, decrease at 1–5 keV, rise at >5 keV). The validation in Figure 1d uses the 2014 MHD synthetic spectrum, not the 2020/2021 epochs where the secondary peak emerges; it therefore does not demonstrate that Eq. (3) preserves the very feature being claimed. If the power-law relation forces high-temperature bins to have too high an ionization parameter, the fit can create a spurious 3–5 keV concentration of EM, and the 'ejecta brightening' conclusion would be weakened. This is a correctness risk, not merely a model-choice preference, because the secondary peak is the main new observable result.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a differential emission measure (DEM) analysis of XMM-Newton RGS and EPIC-pn observations of SN 1987A taken between 2007 and 2021. The authors model the thermal X-ray spectra with a Chebyshev-parameterized DEM composed of 40 vnei temperature bins, with the ionization parameter of each bin tied to temperature through a power law (Eq. 3). They find a persistent major DEM peak at ~0.5–1 keV whose emission measure declines after about 2014, and a high-temperature tail that grows in EM and appears to form a secondary peak at ~3–5 keV in 2020–2021. The DEM results are compared with MHD simulations, leading the authors to interpret the major peak as the equatorial ring (now fading) and the secondary peak as newly shocked ejecta (now brightening). A complementary analysis of the Fe K line centroid shows a possible recent decrease, interpreted as additional evidence for newly shocked ejecta.","tokens_in":37842,"tokens_out":4654,"duration_ms":51606,"significance":"If the central claims hold, the paper provides a valuable continuous temperature-structure characterization of a benchmark supernova remnant, extending the epoch coverage to 2021 and offering a direct comparison with MHD simulations. The work is significant because it moves beyond discrete-temperature fits and identifies a temporal transition from ring-dominated to ejecta-dominated X-ray emission, a prediction that can be tested with future XRISM and other missions. The authors have also included several commendable strengths: validation of the DEM reconstruction on synthetic one-, two-, and three-temperature spectra and on a 2014 MHD synthetic spectrum; MCMC-based uncertainty estimation with full corner plots and best-fit tables in appendices; and a cross-check with an alternative vpshock-based DEM model. The main concern is that the load-bearing assumption of a power-law net–kTe relation (Eq. 3) is acknowledged by the authors themselves to be oversimplified, and the specific new feature (the 3–5 keV secondary peak) appears in an epoch not covered by the model validation.","major_comments":[{"comment":"The power-law net–kTe relation is load-bearing for the secondary-peak claim, but the paper's own Section 4.3 states that this assumption is 'over-simplified' and that the MHD simulations show an EM-weighted net(kTe) that is not a power law, with a plateau at 0.3–1 keV, a decrease at 1–5 keV, and an increase at >5 keV. The claimed secondary peak at 3–5 keV lies exactly in the temperature range where the simulated net decreases, whereas the fitted power-law index beta is generally >= 0, forcing an increasing net with temperature. This could artificially concentrate emission measure at high temperatures. The validation in Figure 1d uses the 2014 synthetic spectrum, not the 2020–2021 epochs where the secondary peak emerges, so it does not demonstrate that Eq. (3) preserves the very feature being claimed. Please add a synthetic-spectrum test based on the 2020 and 2021 MHD epochs, and/or a systematic exploration of alternative net(T) prescriptions (e.g., the simulated EM-weighted relation), to demonstrate that the secondary peak is not an artifact of the assumed ionization model. This is the main correctness risk of the paper.","section":"Section 3.1, Eq. (3)"},{"comment":"The assumption that metal abundances are identical across all temperature bins, while acknowledged as a simplification, may bias the DEM distribution if the high-temperature component is indeed dominated by shocked ejecta with abundances different from those of the ring. The paper's own comparison with Maitra et al. (2022), who allowed different abundances for different components and found significantly higher abundances in the hot component, indicates that this assumption may not be physically adequate. Since the physical interpretation of the secondary peak as ejecta depends on the separation of components in temperature, please quantify the sensitivity of the DEM profile, especially in the 3–5 keV region, to allowing abundance differences between low- and high-temperature bins (or at least between the ring and ejecta components). Without such a test, the identification of the secondary peak with ejecta is less secure.","section":"Section 3.1, abundances"},{"comment":"The component identification (major peak = ring, secondary peak = ejecta) relies on comparing the observed DEM with the same group's MHD simulations (Orlando et al. 2020; Greco et al. 2022). While this is a reasonable interpretive framework, it is not an independent validation, and the paper itself notes discrepancies (observed peak temperatures are higher and absolute DEMs are lower than simulated). The observed trends of major-peak decline and tail brightening could in principle be produced by a single evolving plasma component (e.g., the forward shock encountering lower-density regions) without requiring a distinct ejecta component. Please make the model dependence of the component labels more explicit, and consider a quantitative test, such as fitting the 2020/2021 spectra with a model that includes a ring component plus an independent ejecta component with separate abundances and ionization states, to strengthen the claim that a newly shocked ejecta component is required by the data.","section":"Section 4.1, Figure 4"}],"minor_comments":[{"comment":"The sentence 'Our DEM results consistent well with simulations' should read 'Our DEM results are consistent with simulations.'","section":"Abstract"},{"comment":"Given the explicit statement that the power-law net–kTe assumption 'might be over-simplified' and that 'the results must be treated with caution,' the abstract and conclusions should carry a corresponding caveat about the model-dependent nature of the secondary peak and the ejecta-brightening claim.","section":"Section 4.3"},{"comment":"The statistical significance of the Fe K centroid decrease is modest (f-test p-value ~0.05, corresponding to roughly 2σ; the extrapolation test is also caveated by large scatter in the earlier epochs). The text should describe this as a 'possible' or 'tentative' decrease rather than stating it as a firm detection, and it should be made clear that this is not an independent confirmation of the DEM result.","section":"Appendix D"},{"comment":"The claim that the DEM model provides 'significantly better fit' than discrete-temperature models should be qualified, since the reported Delta C values range from -4 to -109; for epochs with Delta C close to -4, the improvement is not statistically significant. Please state the average improvement and note the epoch-to-epoch variation.","section":"Section 3.3"},{"comment":"The validation on the 2014 MHD synthetic spectrum is useful, but the text should explicitly note that this validation does not cover the 2020–2021 epochs where the new secondary peak appears; this limitation is central to the robustness of the main conclusion.","section":"Section 3.2, Figure 1d"},{"comment":"The 2019 observation has a much shorter exposure (32.4 ks for EPIC-pn, PGTI 11.1 ks) than the other epochs; the DEM uncertainties for that epoch appear larger in Figure 2. Please discuss whether the 2019 data point could be affected by the lower signal-to-noise ratio, and whether it influences the inferred onset of the tail rise.","section":"Table 1 and Figure 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid observational study of an important target, and the DEM methodology is generally well executed. The main reservation is that the headline new result—the emergence of a secondary DEM peak at 3–5 keV—rests on an ionization-parameter prescription that the authors themselves describe as oversimplified, and that is not validated at the epochs where the feature appears. The requested additional tests (synthetic 2020/2021 spectra, alternative net(T) prescriptions, abundance sensitivity) are feasible within the manuscript's scope and would substantially strengthen the central claim. The reliance on the authors' own simulations for the component identification is not circular in a strict sense, but the interpretive labels should be presented with more caution, especially in the abstract."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First the punchline: this is a useful, honest paper with a real new data product, the 14-epoch DEM evolution of SN 1987A from 2007 to 2021, and the central qualitative trend (the 0.5-1 keV component peaks around 2011-2014 and then declines, while a hot tail keeps growing) is probably robust. The strongest new claim, the emergence of a 3-5 keV secondary peak in 2020-2021 that they read as newly shocked ejecta, is much less secure. It rests on the power-law net–kT relation in Eq. 3, which the paper itself calls over-simplified, and the validation is thinnest exactly in that regime.\n\nWhat the paper does well: the DEM framework is applied carefully. They fit RGS and EPIC-pn simultaneously over 14 epochs, give MCMC uncertainties, and test the model on synthetic 1/2/3-temperature spectra and on a 2014 MHD synthetic spectrum. They also cross-check with a vpshock-based DEM and get similar qualitative behavior, which is reassuring. The Fe K centroid decline, while only moderately significant (2-3 sigma), is an independent, physically sensible second handle. And they are candid about the limitations of Eq. 3. The comparison with previous DEM work (Zhekov, Alp) is fair and clear.\n\nSoft spots, in proportion. The stress test is right: Eq. 3 is load-bearing for the secondary peak. The MHD EM-weighted net–kT curve in Figure 4 is not a power law, and the synthetic validation in Figure 1d is for 2014, not for the 2020/2021 epochs where the secondary peak appears. If the power-law forces the high-temperature bins to have too high an ionization parameter, the fit can concentrate EM in a spurious 3-5 keV bump. I do not think this undermines the main-peak decline, because that part is constrained by the well-resolved RGS lines below 2 keV, where the net–T relation is flatter and less problematic. But the ejecta-brightening conclusion, insofar as it hangs on the secondary peak, is model-dependent. The component identification also leans on simulations from the same group; that is not a fatal circularity since the DEM is an independent product, but the labels are not observationally locked down.\n\nWho this is for: anyone working on SN 1987A or early SNR evolution, and anyone doing DEM analysis on NEI plasmas. It deserves serious peer review. My recommendation: send it to referees, and require the authors to test how much the secondary peak depends on Eq. 3, either by using a more flexible net–kT parametrization or by repeating the synthetic validation on the 2020 simulation. If they cannot, the claim should be softened to a model-dependent suggestion, not a firm detection.","headline":"Useful 14-epoch DEM evolution of SN 1987A, with a central trend that is likely real; the flashy new 3-5 keV ejecta peak is still hostage to an oversimplified net–kT power law.","tokens_in":38366,"tokens_out":2317,"would_cite":true,"duration_ms":27403,"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":"By tracking SN 1987A's full X-ray temperature distribution from 2007 to 2021, this paper shows the equatorial ring is fading while shock-heated ejecta are brightening into a 3-5 keV component.","keywords":["supernova remnants","SN 1987A","X-ray spectroscopy","differential emission measure","shock-heated ejecta","equatorial ring","non-equilibrium ionization","XMM-Newton"],"falsifier":"Re-fit the same spectra while letting the ionization parameter vary freely at each temperature bin, or fix it to the temperature-dependent values from the MHD simulations used in the paper, and check whether the 3–5 keV secondary peak and the post-2014 decline of the 0.5–1 keV peak persist; if they vanish, they are artifacts of the assumed power-law ionization relation.","tokens_in":37239,"feed_emoji":"💥","tokens_out":9301,"duration_ms":82380,"temperature":0.7,"pith_summary":"This paper argues that the X-ray gas in SN 1987A is not well described by the two or three discrete temperature components used in earlier models, and that its true structure is a continuous temperature distribution whose shape has been changing since about 2014. Fitting a differential emission measure (DEM) model to XMM-Newton RGS and EPIC-pn spectra taken over 14 years, the authors find a major emission peak near 0.5–1 keV whose total emission measure peaked around 2011–2014 and has since declined, alongside a high-temperature tail extending past 5 keV that kept brightening and by 2020–2021 grew into a secondary peak at 3–5 keV. Comparing these DEMs with three-dimensional magnetohydrodynamic simulations, they identify the fading peak with the shocked equatorial ring and the rising secondary peak with newly reverse-shock-heated ejecta. A drop in the centroid energy of the Fe K line over the last few epochs adds independent support for a growing, less-ionized ejecta component. The result matters because it ties observations directly to numerical simulations and defines the stage at which the remnant's X-ray emission shifts from ring-dominated to ejecta-dominated.","feed_headline":"X-ray map shows SN 1987A ring fading, ejecta brightening","feed_subtitle":"A 14-year temperature-mapping study reveals a 3-5 keV ejecta component growing while the equatorial ring cools.","key_machinery":"The load-bearing object is the differential emission measure (DEM) model c7pvnei, an extension of the XSPEC c6pvmkl model to non-equilibrium-ionization plasmas. It represents the plasma with 40 temperature bins logarithmically spaced between 0.1 and 10 keV; each bin contributes a single vnei (non-equilibrium ionization) spectrum with temperature T_i and emission measure given by a Chebyshev polynomial of order 7 in log-temperature. To keep the fit tractable, the model assumes the ionization parameter at each temperature follows a single power law, τ(T) = τ_1keV (T/1 keV)^β, plus a single set of metal abundances across all bins. The result is a quasi-continuous temperature distribution that can be compared directly, point by point, with the continuous emission-measure distribution extracted from MHD simulations, which is precisely what the paper does to identify the ring and ejecta components.","core_discovery":"The paper's central discovery is that the X-ray-emitting plasma in SN 1987A has been undergoing a slow regime change. Using a Chebyshev-parameterized DEM model (c7pvnei) with 40 temperature bins from 0.1 to 10 keV, the authors reconstruct the continuous emission measure distribution at 14 epochs. In all epochs the distribution contains a major peak at ~0.5–1 keV and a tail to ≳5 keV, but the evolution is especially clear: the major peak's emission measure rises to a maximum around 2011–2014 and then declines, while the tail's emission measure increases continuously and in 2020–2021 appears as a secondary peak at ~3–5 keV. The observed DEM profiles agree with MHD simulation predictions, allowing the authors to attribute the major peak to the shocked equatorial ring and the secondary peak to reverse-shock-heated ejecta. The average metal abundances derived from the DEM are systematically higher than those from discrete-temperature fits, and the Fe K line centroid has dropped from ~6.65–6.7 keV to ≶6.6 keV, both of which point to a growing contribution from metal-poorer, less-ionized outer ejecta.","pith_inferences":["Extending the same DEM analysis to the pre-2007 Chandra LETG/HETG epochs would test whether the major-peak emission measure indeed peaked around 2011–2014, or whether the apparent turnover is a model artifact.","The 3–5 keV secondary peak should be visible in broadband hard X-ray imaging; targeted NuSTAR or future high-resolution missions may be able to image the ejecta component separately from the ring.","Relaxing the power-law ionization assumption by fixing τ(T) to the MHD-simulated values for each component would turn the DEM fit into a stricter test of the simulations and would reveal how much of the secondary peak is required by the data rather than by the assumed relation.","If the Fe K centroid decline is confirmed at higher significance with future microcalorimeter missions, the measured centroid can be inverted to estimate the mass of newly shocked ejecta, providing a quantitative link to the DEM's secondary peak."],"forward_implications":["The equatorial ring of SN 1987A is now past its X-ray peak; its soft X-ray flux and surface brightness should continue to decline as the forward shock moves into lower-density material.","The newly shocked ejecta will become the dominant thermal X-ray source over the next several years, offering a time-resolved view of reverse-shock heating in a young remnant.","Because the DEM-fit abundances are higher than those from two- or three-temperature fits, earlier abundance estimates for SN 1987A are likely biased low, and similar biases may affect other SNR abundance measurements.","The Fe K centroid decline predicts that the average iron ionization state will keep dropping until the reverse shock reaches the metal-rich core, at which point line centroids and abundances should rise again.","The DEM-vs-simulation comparison provides a template for using continuous temperature distributions instead of discrete components when connecting SNR X-ray spectra to hydrodynamic models."],"supporting_citations":[{"why":"Supplies the earlier XMM-Newton monitoring, two-temperature fits, and Fe K centroid baseline that this work extends to 2020–2021.","marker":"Sun et al. 2021"},{"why":"Provides the B18.3 3D MHD simulation whose continuous EM and net distributions are the direct comparison target.","marker":"Orlando et al. 2020"},{"why":"Produces the synthetic X-ray spectra from those simulations at 2012, 2014, and 2020 that are compared with observed spectra and with the Fe K centroid prediction.","marker":"Greco et al. 2022"},{"why":"Introduced the power-law ionization-parameter assumption and the first DEM-style fit to SN 1987A with Chandra LETG/HETG.","marker":"Zhekov et al. 2006"},{"why":"Extended the DEM analysis to 2007, giving a bimodal temperature distribution that this paper refines with higher-quality data.","marker":"Zhekov et al. 2009"},{"why":"The RGS+NuSTAR continuous-temperature analysis that found three peaks and disagreed with the earlier bimodal results; the present work reconciles this by finding a single broad peak plus a growing tail.","marker":"Alp et al. 2021"},{"why":"The original c6pvmkl DEM model upon which the c7pvnei model is based.","marker":"Lemen et al. 1989"}],"fun_headline_variants":["SN 1987A's ring fades as ejecta takes the X-ray stage","X-ray glow shift reveals SN 1987A ejecta outshining ring","SN 1987A: ring dims, ejecta brightens in 14-year X-ray view","SN 1987A's ring fades while ejecta flares in X-rays","New X-ray spike hints SN 1987A's ejecta overtake ring"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The weakest link is the assumption that the ionization age of every plasma component is a single power-law function of temperature, which the paper itself calls 'over-simplified'; if the real relation differs by component, the reconstructed DEM peaks—including the claimed secondary ejecta peak—could shift or be spurious.","fun_headline_variants_meta":{"raw":{"variants":["SN 1987A's ring fades as ejecta takes the X-ray stage","X-ray glow shift reveals SN 1987A ejecta outshining ring","SN 1987A: ring dims, ejecta brightens in 14-year X-ray view","SN 1987A's ring fades while ejecta flares in X-rays","New X-ray spike hints SN 1987A's ejecta overtake ring"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000787,"raw_usage":{"total_tokens":3546,"prompt_tokens":1094,"completion_tokens":2452,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":710,"completion_tokens_details":{"reasoning_tokens":2340}},"tokens_in":710,"tokens_out":2452,"duration_ms":17189,"temperature":1.0,"reasoning_tokens":2340,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T00:43:26.499098+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-fit the same spectra while letting the ionization parameter vary freely at each temperature bin, or fix it to the temperature-dependent values from the MHD simulations used in the paper, and check whether the 3–5 keV secondary peak and the post-2014 decline of the 0.5–1 keV peak persist; if they vanish, they are artifacts of the assumed power-law ionization relation.","supporting_citations":[{"cited_title":"R., Mewe , R., Schrijver , C","cited_arxiv_id":null,"evidence_quote":"The original c6pvmkl DEM model upon which the c7pvnei model is based."}],"review_version":1}