{"id":"f5bcc0e3-fbed-4d30-93ea-70be450eabbb","arxiv_id":"2607.06302","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":9,"one_line_summary":"Soft X-ray emission from SN 1987A's equatorial ring becomes broader and extends inward after ~2012, while hard X-rays remain compact, indicating growing contribution from reverse-shock-heated interior material.","lead":"This paper uses 26 years of Chandra X-ray observations to show that the soft X-ray glow from supernova 1987A's ring has started spreading inward over the last decade, while the hard X-rays stay put. A smart generalist might read it because it tracks, in real time, how a supernova transitions from blasting its surroundings to being reshaped by its own inward-collapsing shock waves.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"Soft-band torus broadening may be partly driven by using the hard-band centroid to center the soft-band image; all cross-checks share this procedure, so a growing soft–hard centroid offset would mimic the claimed energy-dependent width evolution.","rationale":"The paper presents a careful, well-scoped analysis with multiple cross-checks, and the central claim of energy-dependent radial evolution is physically motivated and consistent with simulations and multiwavelength data. However, the specific procedure of determining the center from the hard band and applying it to the soft band creates a structural vulnerability: any time-evolving soft–hard centroid offset would produce exactly the observed signature (soft-band broadening appearing after ~2012), and no existing cross-check can detect it because all share the same centering procedure. This is not a fatal flaw — the effect may be small or absent — but it is the single most load-bearing untested assumption, and the concrete test proposed above is straightforward to execute with existing data. The reader correctly identified the internal alignment as the weakest assumption; I am refining the concern from 'merged-image registration' to the more specific issue of 'hard-band center applied to soft-band image within each observation.' The Fe K results are appropriately framed as confirmation and are not affected by this concern. Given that the central imaging claim rests on an untested systematic that could account for a meaningful fraction of the signal, a CONDITIONAL verdict is warranted pending the self-centering test. If the test shows negligible centroid offsets, the verdict should return to ACCEPT.","tokens_in":22836,"tokens_out":3352,"duration_ms":191721,"concrete_test":"Re-fit the torus model to the 0.5–1.5 keV merged images for each 4-year epoch with (x0, y0) as free parameters rather than fixed from the hard-band fit. Compare the best-fit soft-band centers to the hard-band centers epoch by epoch. If the soft–hard centroid offset grows to ≳0.05 arcsec after 2012 and the soft-band width decreases toward the hard-band value when self-centered, the energy-dependent broadening claim is substantially weakened. If the offset remains negligible (<0.02 arcsec) or the width difference persists unchanged, the claim is secure.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The center position for each observation is determined exclusively from the 1.5–7.0 keV band (Section 2.3: 'We use images in the 1.5–7.0 keV band, constructed with fluximage, where the ring structure is relatively well defined') and then applied unchanged to the 0.5–1.5 keV images (Section 2.4: 'The same procedure is applied to other energy bands using the center position determined from the 1.5–7.0 keV band'). If the soft-band emission centroid migrates relative to the hard-band centroid over time — which is physically plausible given that the two bands trace increasingly different components (ER vs. interior/reverse-shock emission) and that the east–west asymmetry reverses around 9500 days, roughly when the width divergence begins — the soft-band torus fit would be systematically broadened because the model's center is offset from the true soft-band centroid. The width difference in Figure 3b is ~0.05–0.10 arcsec in the latest epochs; a soft–hard centroid offset of comparable size (within the scatter seen in Table 3 center coordinates, which ranges over ~0.1–0.3 arcsec across observations) could account for a substantial fraction of this signal. Critically, all three analysis pathways — merged torus fitting (Section 3.1), individual-observation fitting (Appendix C), and radial-profile FWHM (Appendix D) — use the same hard-band-derived center positions. A shared systematic bias would therefore not be caught by any of these cross-checks. The fact that early epochs show similar widths in both bands argues against a purely static instrumental effect, but does not rule out a time-growing centroid offset that tracks the physical evolution of the remnant. The paper does not test whether the soft and hard band centroids diverge over time.","agreement_with_reader":"partial"},"referee_report":{"model":"glm-5.2","summary":"This paper presents a systematic Chandra imaging and spatially resolved spectroscopic study of SN 1987A spanning 1999–2025, combining ACIS and HETG zeroth-order data in 4-year bins. The authors employ EDSER subpixel reconstruction, MARX-simulated PSFs, and Richardson–Lucy deconvolution to characterize the projected torus morphology in soft (0.5–1.5 keV) and hard (1.5–7.0 keV) bands. The central new result is that the soft-band torus width becomes systematically broader than the hard-band width after the early 2010s, and when radius and width are considered together, the soft-band emission extends inward to r_ell ~ 0.5 arcsec in the 2020s. The authors interpret this as evidence for an increasing contribution from reverse-shock-heated and/or high-latitude material. A complementary spectral analysis tracks the Fe K line flux evolution and its east–west asymmetry. The analysis is methodologically careful, with Monte Carlo error estimation (300 realizations), cross-checks against individual observations (Appendix C), and a model-independent radial-profile analysis (Appendix D).","tokens_in":23186,"tokens_out":1547,"duration_ms":315886,"significance":"The paper provides a valuable, uniformly analyzed long-term Chandra dataset that quantifies the emergence of energy-dependent radial structure in SN 1987A as it transitions beyond the ER-dominated phase. The finding that the soft-band emission extends inward in the 2020s is physically well-motivated and consistent with HD/MHD predictions (Orlando et al. 2015–2025) and JWST reverse-shock observations (Larsson et al. 2023). The Fe K spatial analysis, while not a new discovery, provides an independent Chandra-based confirmation of the east–west asymmetry and extends its possible detection to earlier epochs. The multiwavelength comparison in Section 3.3 provides useful context. The analysis pipeline (RL deconvolution with MARX PSFs, Monte Carlo uncertainties, multiple cross-checks) is thorough and reproducible in principle.","major_comments":[{"comment":"Section 2.3–2.4: The center position for each observation is determined exclusively from the 1.5–7.0 keV band and then applied unchanged to the 0.5–1.5 keV images. If the soft-band emission centroid migrates relative to the hard-band centroid over time — which is physically plausible given that the two bands trace increasingly different components and the east–west asymmetry reverses around 9500 days (Section 4) — the soft-band torus fit would be systematically broadened because the model center is offset from the true soft-band centroid. The width difference in Figure 3b is ~0.05–0.10 arcsec in the latest epochs; a soft–hard centroid offset of comparable size (within the scatter seen in Table 3 center coordinates, which ranges over ~0.1–0.3 arcsec across observations) could account for a substantial fraction of this signal. Critically, all three analysis pathways — merged torus fitting,","section":null},{"comment":"individual-observation fitting (Appendix C), and radial-profile FWHM (Appendix D) — use the same hard-band-derived center positions, so a shared systematic bias would not be caught by any of these cross-checks. The paper should either (a) demonstrate quantitatively that a soft–hard centroid offset of the plausible magnitude cannot reproduce the observed width difference (e.g., by injecting simulated offsets and re-fitting), or (b) repeat the soft-band torus fitting with centers determined independently from the soft-band images and show that the width evolution persists. Without one of these tests, the central claim of energy-dependent width evolution remains vulnerable to this systematic.","section":null}],"minor_comments":[{"comment":"Section 2.1: The pileup fraction is stated as '≲10%' for most observations, but several bright epochs between 2002 and 2004 reached higher values. Please quantify the peak pileup fraction in those epochs and briefly discuss whether it could bias the 2000–2003 or 2004–2007 torus measurements.","section":null},{"comment":"Table 1: The 2008–2011 NONE observations (~20 ks) are excluded because contemporaneous HETG data provide deeper coverage. Please state explicitly whether any HETG data from this interval exist or whether there is a temporal gap, as this affects the continuity of the time series.","section":null},{"comment":"Figure 3: Error bars are not visible or not described in the caption. Please confirm that Monte Carlo uncertainties (Appendix B) are plotted and state this in the caption.","section":null},{"comment":"Section 3.2, Table 2: Several early-epoch fits are marked 'N/C' (not constrained). Please define this abbreviation in the table notes (it currently appears only in the footnote marker).","section":null},{"comment":"Section 3.2: The Gaussian width is fixed at sigma = 0.04 keV based on XRISM results. Since XRISM has much higher spectral resolution than ACIS, please briefly justify why this narrow-line assumption is appropriate for the ACIS data and whether the Fe K flux is sensitive to this choice.","section":null},{"comment":"Figure 4 caption: 'Sourth' should be 'South' in panel (e) label.","section":null},{"comment":"Appendix D: The FWHM is described as measured from RL-deconvolved profiles, but the method for determining the half-maximum level (especially in the presence of the central depression or asymmetric wings) is not specified. Please add a brief description.","section":null},{"comment":"Section 3.3: The multiwavelength torus radii are compared, but the X-ray values come from a single ObsID (25514) while the radio/optical/NIR values come from different instruments and modeling approaches. A sentence acknowledging that the absolute values are not directly comparable across wavelengths would strengthen this section.","section":null},{"comment":"Table 3: The center coordinate scatter for HETG observations in 2018 (e.g., ObsIDs 21037, 21038, 21042) shows RA offsets of ~0.3–0.5 arcsec from the mean. Please comment on whether these outliers affect the merged image quality for the 2015–2018 bin.","section":null}],"recommendation":"major_revision","confidential_remarks":"The alignment concern raised in review is, in my assessment, the key issue for this paper. The central claim (energy-dependent width evolution) is load-bearing for the paper's contribution, and the alignment strategy creates a plausible systematic that is not addressed by the existing cross-checks because they all share the same centering procedure. This is fixable — the authors can either re-run the soft-band fits with independently determined centers or perform a simulation-based sensitivity test — but it requires new analysis work. I would not reject on this basis because the fix is within scope and the physical interpretation is sound and well-motivated. The Fe K results are secondary and appear solid."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful and constructive report. The referee raises one major concern: the center position used for torus fitting is determined exclusively from the hard band (1.5–7.0 keV) and applied unchanged to the soft band (0.5–1.5 keV), which could systematically broaden the soft-band torus if a soft–hard centroid offset exists. We agree this is a valid and important concern. We will address it by repeating the soft-band torus fitting with centers determined independently from the soft-band images and by performing a simulation-based test injecting known centroid offsets. We expect the width evolution to persist, but if it does not, we will revise the central claim accordingly.","responses":[{"response":"We agree that this is a legitimate and important concern. The referee is correct that all three analysis pathways (merged torus fitting, individual-observation fitting, and radial-profile FWHM) share the same hard-band-derived center positions, so they cannot independently catch a systematic bias from a soft–hard centroid offset. We will address this in the revised manuscript by implementing both tests the referee suggests. First, we will repeat the soft-band torus fitting using centers determined independently from the 0.5–1.5 keV images for each epoch. We note that at early epochs (before ~2012), the soft and hard morphologies are similar, so the centroid offset should be minimal; the critical test is whether the width difference persists at late epochs when the soft-band center is allowed to differ. Second, we will perform a simulation-based test: we will inject known centroid offsets (ranging from 0.05 to 0.3 arcsec, covering the scatter seen in Table 3) into synthetic soft-band images, re-fit the torus model with the hard-band center, and quantify the resulting artificial broadening. This will allow us to determine whether a plausible offset can account for the observed ~0.05–0.10 arcsec width difference. We expect the width evolution to persist with independently determined soft-band centers, because the radial-profile analysis in Appendix D already shows that the soft-band profiles have more extended inner wings (not just a symmetric broadening that a pure centroid offset would produce), but we will present the quantitative results transparently and revise the central claim if the tests indicate the signal is not robust. We will add a new subsection or appendix describing these tests and their outcomes.","revision_made":"yes","referee_comment":"Section 2.3–2.4: The center position for each observation is determined exclusively from the 1.5–7.0 keV band and then applied unchanged to the 0.5–1.5 keV images. If the soft-band emission centroid migrates relative to the hard-band centroid over time, the soft-band torus fit would be systematically broadened. All three analysis pathways use the same hard-band-derived centers, so a shared systematic bias would not be caught by cross-checks. The paper should either (a) demonstrate quantitatively that a soft–hard centroid offset cannot reproduce the observed width difference, or (b) repeat the soft-band torus fitting with independently determined soft-band centers and show the width evolution persists."}],"tokens_in":22813,"tokens_out":669,"duration_ms":146104,"standing_objections":[]},"desk_editor":{"model":"glm-5.2","letter":"The main thing to know: this paper reports the first quantitative, energy-resolved measurement of the soft-band torus becoming systematically broader than the hard band after ~2012, with the soft emission extending inward to r_ell ~ 0.5 arcsec in the 2020s. That is a new result — not present in Racusin et al. 2009, Frank et al. 2016, or Ravi et al. 2024 — and it directly connects to what Orlando et al.'s HD/MHD simulations predict for this evolutionary phase. The Fe K work is confirmation and extension, not discovery, and is framed that way honestly. The analysis is methodologically careful: RL deconvolution with MARX-simulated PSFs, 300-realization Monte Carlo error estimation, cross-checks on individual observations (Appendix C), and an independent radial-profile FWHM analysis (Appendix D). The torus model is simple and uniform across epochs, which is the right call for consistency. The multiwavelength 2022 snapshot is a nice contextual addition. The stress-test concern about internal alignment is the right thing to worry about: the center for each observation is determined from the 1.5–7 keV band and applied unchanged to the soft band, and all three cross-check pathways share this procedure. If the soft and hard centroids diverge over time — physically plausible given the east-west asymmetry reversal around 9500 days — a growing offset could artificially broaden the soft-band torus fit. The width difference in Figure 3b is ~0.05–0.10 arcsec at late epochs, and the center-coordinate scatter in Table 3 ranges over ~0.1–0.3 arcsec, so the concern is quantitatively reasonable. However, I do not think it sinks the result. The early-epoch agreement between bands argues against a static instrumental effect, and the fact that the soft band broadens while the hard band stays narrow is the opposite of what a simple centroid drift would produce if the hard band were the reference — a shared offset would broaden both. The paper could and should directly test whether soft and hard centroids diverge over time; that is a straightforward check and its absence is the main gap. But the physical interpretation — increasing interior/reverse-shock contribution in the soft band — is well-motivated and consistent with JWST and simulation predictions. This is for researchers working on SN 1987A evolution, SNR shock-CSM interaction, and anyone benchmarking HD/MHD models against observations. It deserves a serious referee. The referee should ask for the centroid-divergence test, but the core result is solid enough to warrant that engagement rather than summary rejection.","headline":"Energy-dependent torus broadening in SN 1987A is a real new result; the alignment concern is worth raising but probably does not sink the signal","tokens_in":24003,"tokens_out":633,"would_cite":true,"duration_ms":175429,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["98.38.Mz","95.85.Nv","97.60.Bw"],"model":"glm-5.2","headline":"Supernova 1987A's X-ray ring grows energy-dependent after 2012","keywords":[],"falsifier":"If the energy-dependent broadening of the soft-band torus width disappears or reverses when a different center-determination or alignment method is used, the result would be called into question.","tokens_in":22995,"feed_emoji":"🔭","tokens_out":974,"duration_ms":242904,"temperature":0.7,"pith_summary":"This paper analyzes 26 years of Chandra X-ray observations of Supernova 1987A, from 1999 to 2025, to track how the remnant's ring-shaped X-ray emission has evolved. The authors model the emission as a projected torus and measure its radius and width in two energy bands: soft (0.5–1.5 keV) and hard (1.5–7.0 keV). They find that before roughly 2012, the soft and hard X-ray rings had similar widths. After that, the soft-band ring became systematically broader than the hard-band ring. When radius and width are considered together, the soft-band emission extends inward, with its inner boundary reaching down to about 0.5 arcseconds by the 2020s — comparable to the ring radius seen around the year 2000. The outer boundary stays similar between bands. The authors interpret this as evidence that emission from reverse-shock-heated material inside the ring, and possibly from higher-latitude structures, is becoming an increasingly important contributor to the remnant's X-ray output. The paper also confirms and extends prior detections of Fe K line emission (around 6.7 keV), showing it was already present in 2007–2009 Chandra data and is consistently stronger on the eastern side of the ring.","feed_headline":"Supernova 1987A's X-ray ring splits: soft band grows broader after 2012","feed_subtitle":"26 years of Chandra data reveal the remnant's soft X-rays extending inward, signaling a transition from ring-dominated to interior-emission","key_machinery":"The analysis uses Chandra ACIS and HETG observations binned into 4-year intervals, subpixel event repositioning (EDSER) at quarter-pixel scale, Richardson–Lucy deconvolution with MARX-simulated point-spread functions, and an elliptical Gaussian torus model fitted to the deconvolved images to extract ring radius and width as a function of energy band and epoch. Spatially resolved spectroscopy divides the remnant into four quadrants (East, West, North, South) and fits an absorbed power-law plus Gaussian model in the 5–8 keV band to characterize Fe K line emission.","core_discovery":"The central finding is that the X-ray ring of SN 1987A has developed an energy-dependent radial structure: after the early 2010s, soft X-rays (0.5–1.5 keV) trace a broader distribution that extends inward to smaller radii than hard X-rays (1.5–7.0 keV). This inward extension of soft emission, with the inner boundary reaching roughly 0.5 arcseconds by the 2020s, signals a growing contribution from interior or high-latitude plasma — likely reverse-shock-heated ejecta — marking the remnant's transition away from a purely equatorial-ring-dominated phase.","pith_inferences":[],"forward_implications":["If the inward soft-band extension traces reverse-shock-heated ejecta, continued monitoring should show the soft X-ray morphology becoming increasingly distinct from the hard-band ring as more ejecta is shock-heated.","The eastern enhancement of Fe K emission, if confirmed to predate 2018, may point to an intrinsic asymmetry in the explosion or in the circumstellar environment that predates the current epoch.","The transition from equatorial-ring-dominated to interior-emission-dominated X-ray morphology provides an observational benchmark for testing 3D hydrodynamic and magnetohydrodynamic simulations of supernova remnant evolution at 30–40 years post-explosion.","Future X-ray observatories with higher angular resolution could resolve whether the inward soft emission is structured (e.g., bubble-like, as suggested by JWST) or smoothly distributed, distinguishing between reverse-shock and high-latitude origin scenarios."],"fun_headline_variants":["SN 1987A soft X-rays broaden inward post-2012 as interior emission grows","Chandra tracks SN 1987A ring: soft X-ray band extends inward after early 2010s","SN 1987A X-ray ring develops energy-dependent radial structure over 26 years","Soft X-rays in SN 1987A remnant trace broader radius than hard band after 2012","Late-time SN 1987A shows soft X-rays extending inward beyond the hard band"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The internal alignment strategy uses the remnant itself as a reference for registering multi-epoch images rather than external point sources, which assumes that the centroid of the X-ray emission does not shift between epochs in a way that would bias the merged image. If the emission centroid migrates — for example, due to the east-west brightness asymmetry reversal reported around 9500 days — the alignment could introduce systematic radial structure differences between bands","fun_headline_variants_meta":{"raw":{"variants":["SN 1987A soft X-rays broaden inward post-2012 as interior emission grows","Chandra tracks SN 1987A ring: soft X-ray band extends inward after early 2010s","SN 1987A X-ray ring develops energy-dependent radial structure over 26 years","Soft X-rays in SN 1987A remnant trace broader radius than hard band after 2012","Late-time SN 1987A shows soft X-rays extending inward beyond the hard band"]},"model":"glm-5.2","effort":"high","cost_usd":0.0,"raw_usage":{"total_tokens":746,"prompt_tokens":626,"completion_tokens":120,"prompt_tokens_details":null},"tokens_in":626,"tokens_out":120,"duration_ms":43597,"temperature":1.0,"reasoning_tokens":null,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-08T10:15:40.900243+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"If the energy-dependent broadening of the soft-band torus width disappears or reverses when a different center-determination or alignment method is used, the result would be called into question.","supporting_citations":[],"review_version":1}