XARTATOMS · M104 热晕 X 射线分析手册XARTATOMS · M104 hot-halo X-ray analysis handbook

面向大一物理学生的半实时工作台A semi-live workbench for a first-year physics undergraduate
目录（全部页面）Contents (all pages)

论文草稿：M104 热晕结果（ApJ 风格，待评审）

Paper draft: M104 hot-halo results (ApJ style, not reviewed)

cognition-os/07_REVISION_LOG.md（R107–R131 与 Synthesis 段）；04_DECISION_LOG.md（D146–D157）；REDTEAM_QUEUE.md（RT-102…RT-113）；.agent/jobs/2026-10-20-p54-paper-results-draft/out/sol/results.tex（sol 交付的 AASTeX 结果片段）；.agent/jobs/2026-10-20-p54-paper-results-draft/out/sol/summary.txt；.agent/jobs/2026-10-20-p47-morphology-synthesis/REPORT.sol.md；.agent/jobs/2026-10-20-p34-b031-rebuild/REPORT.sol.md；.agent/jobs/2026-10-20-p37-2t-bias-bound/REPORT.sol.md；.agent/jobs/2026-10-20-p43-hardband-2t-constraint/REPORT.sol.md；.agent/jobs/2026-10-20-p44-2t-resolving-instruments/REPORT.sol.md；.agent/jobs/2026-10-20-p50-projection-vs-2t/REPORT.sol.md；.agent/jobs/2026-10-20-p52-soft-instrument-forecast/REPORT.sol.md；.agent/jobs/2026-10-20-p53-sz-xray-joint-2t/REPORT.sol.md；.agent/jobs/2026-10-20-p56-lit-stellar-cvab/REPORT.hermes.md；cognition-os/references/literature_m104_hot_halo_2026-10-16.md

状态：论文草稿——draft v2 已被红队否决，v3 进行中。本页是 P54 交付的 AASTeX 结果片段的读者地图，不是已评审的论文。红队 RT-117（R149，2026-10-20 16:48 EDT）对 P54b 的 draft_v2 给出 6 项 MATERIAL 级反对，判定"不可接受"：(1) 草稿完全遗漏 R141/R142 的降级，而它的两条头条句恰是被降级的那两句；(2) "双锥下限 ≥1.486"为假——生产者自己的网格里有一个 90° 双锥在 0.995；(3) "七情景"压力包络是数字转置——真包络是 [−4.759, −4.564]；(4) tab:cones 把 B3 5–8′ 标成 Provisional，而生产者、草稿自己的脚注和 tab:gas 都说 Questioned；(5) "旋转束缚气体进入两个 68% 区间"缺 R142 的警示（旋转族是一个横跨整个数据区间的自由旋钮）；(6) 恒星注入"未被排除"的判定挑了唯一一个不排除它的估计器——另一个估计器以 2.9σ 排除它。P54c（draft v3）已派出修订。每个数字都带"来源"；标有 Provisional / Questioned 的结论不得作为已确立结果引用。P54 的完整草稿见下方链接。

Status: paper draft — draft v2 has been REJECTED by the red team; v3 is in progress. This page is a reader's map of the P54-delivered AASTeX results section, not a peer-reviewed paper. Red team RT-117 (R149, 2026-10-20 16:48 EDT) returned 6 MATERIAL objections to P54b's draft_v2 and ruled it "not acceptable": (1) the draft omits the R141/R142 downgrades entirely, and its two headline sentences are exactly the ones those records downgraded; (2) the bicone floor "≥ 1.486" is false — the producer's own grid contains a 90° bicone at 0.995; (3) the "seven-scenario" stress envelope is a digit transposition — the true envelope is [−4.759, −4.564]; (4) tab:cones labels B3 5–8′ "Provisional" where the producer, the draft's own footnote and tab:gas all say "Questioned"; (5) "rotating bound gas enters both 68 % intervals" lacks R142's caveat (the rotation family is a free knob spanning the whole data interval); (6) the stellar-injection "not excluded" verdict cherry-picks the one estimator that does not exclude it — the other estimator excludes it at 2.9σ. P54c (draft v3) is dispatched to revise. Every number carries a source; conclusions labelled Provisional / Questioned must not be quoted as established results. The full P54 draft is linked below.

摘要（大一学生可读版）

Abstract (plain-language version)

我们研究蝎子星系（M104）的热气体晕，用 XMM-Newton 的三台相机（MOS1、MOS2、pn）和 Chandra 的光谱，测量气体温度随半径的变化。核心结果：在 2–3.5′ 处，晕沿星系盘方向比垂直于盘方向亮，B0 能段的合并小轴/大轴亮度比是 0.582±0.066，比“圆”的 1 低 4.75–6.31σ，与 K 波段星光的形状（0.484）一致；扣掉恒星贡献后，热气体本身仍压扁到 0.567 [0.497,0.642]。中心 2′ 因点源太密，两种“无洞”面亮度图法在全部 9 个能段的独立注入检验都不合格，全部隔离。温度方面，外围（8–10′）减内区（2–4′）的温差在两种前景约定下分别为 +0.08 keV（天空 A）和 +0.18 keV（天空 B），95% 区间都包含 0，所以既不能证实也不能排除外升；两种方法（谱拟合与 Fe-L 硬度）其实是同一批数据的两遍投影，不是两次独立探测。更深层的问题是“简并”：我们证明，对 M104 来说，藏着一团更热的气体和外面真的更热，在现有 X 射线谱上给出几乎一样的信号。现有仪器（EPIC、RGS、Resolve、Xtend、ACIS-S）、投影模型、甚至热 SZ 都无法把两者分开。CoPI 建议把“温度升还是不升”记为一个正式的简并（负面）结果，等待 PI 决定。

We study the X-ray spectra of the hot gas halo of M104 (the Sombrero Galaxy) with XMM-Newton's three cameras (MOS1, MOS2, pn) and Chandra, measuring how the gas temperature changes with radius. The core result: at 2–3.5′ the halo is brighter along the stellar disk than perpendicular to it — the combined B0-band minor/major brightness ratio is 0.582±0.066, 4.75–6.31σ below round (1), consistent with the K-band starlight shape (0.484); after removing the stellar contribution the hot gas itself is still flattened at 0.567 [0.497,0.642]. The central 2′ is quarantined: neither hole-free surface-brightness method qualifies there (independent injection tests fail in all 9 bands), so no central value is plotted. On temperature, the outer-minus-inner contrast kT(8–10′) − kT(2–4′) is +0.08 keV (sky A) or +0.18 keV (sky B), and both 95 % intervals contain 0, so we can neither confirm nor exclude an outward rise; the two methods (spectral fitting and Fe-L hardness) are one projection applied twice to the same data, not two independent detections. The deeper issue is degeneracy: we show that for M104, a hidden hotter phase and a genuine outer rise give almost identical X-ray spectra. Current instruments (EPIC, RGS, Resolve, Xtend, ACIS-S), explicit projection models, and even thermal SZ cannot separate them. The CoPI proposes recording Q-B as a formal degeneracy (negative) result, pending the PI's decision.

预计章节、现状与来源

Planned sections, current status, and sources

章节Section 现状Status 内容要点Content 来源Source
1. 观测与数据归约 1. Observations and data reduction Provisional 暂定Provisional XMM-Newton MOS1+MOS2+pn，四个点指 0900170101/0501/0601/0701（约 390 ks 等效曝光）加归档 0136950201（Li+2026）；v18 管线实际拟合源 0.375–9.0 keV、背景 0.375–6.23 keV（R023，非脚本注释中常被引用的 0.4–3.2 keV）；源幂律 SrcPo 冻结在 0，0.7 keV 前景槽位 MWhothalo 在全部 150 个保存拟合中为零且冻结（R019/D62）。Chandra 9532/9533 是 ACIS-I（D61 更正早先“内环 S3 主导”的说法），四个环只覆盖名义环带的 35–62 %（P19/D99）。 XMM-Newton MOS1+MOS2+pn, four pointings 0900170101/0501/0601/0701 (~390 ks effective) plus archival 0136950201 (Li+2026); the v18 pipeline actually fits source 0.375–9.0 keV and background 0.375–6.23 keV (R023, not the often-quoted 0.4–3.2 keV); the source power law SrcPo is frozen at 0 and the 0.7 keV foreground slot MWhothalo is zero and frozen in all 150 saved fits (R019/D62). Chandra 9532/9533 are ACIS-I (D61 corrects the earlier 'inner ring S3-dominated' claim), and the four rings cover only 35–62 % of the nominal annuli (P19/D99). cognition-os/07_REVISION_LOG.md R023、R019、R059/D99、D61/D62；04_DECISION_LOG.md D55、D65；cognition-os/references/literature_m104_hot_halo_2026-10-16.md（Li+2026） cognition-os/07_REVISION_LOG.md R023, R019, R059/D99, D61/D62; 04_DECISION_LOG.md D55, D65; cognition-os/references/literature_m104_hot_halo_2026-10-16.md (Li+2026)
2. 背景与前景 2. Background and foreground Questioned 存疑Questioned 天空 A（继承，MW 晕 kT 0.1905）与天空 B（在外围 8–10′ 自身拟合，kT 0.2044）两种约定把 ΔkT 从 +0.094 移到 +0.198 keV，摆动 +0.104 keV，是已计入系统误差的 2.4 倍（R085/D128）。天空 B 是事后的、在同一个 8–10′ 谱上拟合，循环（R042/D82）。v18 中 sky_background_scale 从未被拟合（RT-49 OBJ-1/6）。外部约束：eROSITA DR1 第 1495 号 bin MW 晕 kT 0.1646 +0.0070/−0.0089 keV（距 M104 1.68°，已重放）、HaloSat 北部 CGM 0.179 +0.005/−0.007 keV（Bluem+2022）——两者都比项目值更冷，把晕 kT 冻在外部值后对比度变成 ≈0 或负（RT-84/R088/D130；P31 重放 R091）。RASS 数据集在 v18 中从未接入模型，不约束任何东西（RT-27/R032/D72）。 Sky A (inherited, MW-halo kT 0.1905) vs sky B (fitted on the outer 8–10′ annulus itself, kT 0.2044) move ΔkT from +0.094 to +0.198 keV, a +0.104 keV swing that is 2.4× the booked systematic (R085/D128); sky B is post hoc, fitted on the same 8–10′ spectrum, i.e. circular (R042/D82). v18's sky_background_scale was never fitted (RT-49 OBJ-1/6). External constraints: eROSITA DR1 bin-1495 MW-halo kT 0.1646 +0.0070/−0.0089 keV (1.68° from M104, replayed) and HaloSat northern CGM 0.179 +0.005/−0.007 keV (Bluem+2022) — both cooler than the project values, and freezing the halo kT at them drives the contrast to ≈0 or negative (RT-84/R088/D130; P31 replay R091). The RASS dataset in v18 is never attached to a model and constrains nothing (RT-27/R032/D72). cognition-os/07_REVISION_LOG.md R085、R088、R091、R092/D134、R130、Synthesis；04_DECISION_LOG.md D130、D133、D134、D157 cognition-os/07_REVISION_LOG.md R085, R088, R091, R092/D134, R130, Synthesis; 04_DECISION_LOG.md D130, D133, D134, D157
3. 形态学：内晕沿盘压扁 3. Morphology: the inner halo flattened along the disk Provisional 暂定Provisional P47b（RT-112 质疑吸收后）B0 2–3.5′ 合并小轴/大轴 0.582±0.066：对圆 1 剖面/线性 −4.75/−6.31，对 K 光 0.484 +1.49/+1.56；背景单算 ≈1，减背景只放大不制造。P57：尘埃抬高 ×1.16–1.17，内禀 ≈0.50。P58：扣掉恒星（小轴 8–16 %、大轴 9–13 %）后条件气体 B0 0.567 [0.497,0.642]（B3 0.627 [0.524,0.753]、B4 0.668 [0.575,0.777]）。P59：Chandra 0.585±0.042，Chandra−XMM +0.003±0.079/+0.013±0.127（仪器无关）。P60：非旋转 HSE 0.85–0.96 与双锥风 ≥1.49 在 95% 外排除；旋转束缚气体与恒星质量损失注入可行。3.5–5′ 仍低于 1；5–8′ 不合格（pn 分母为负、Fieller 无界）。中心 2′ 隔离（P35 第 2 轮 9 个能段无一通过、0/9 选方法通过全部门；P38 0/18；P48）。 P47b (after RT-112's objections) B0 2–3.5′ combined minor/major 0.582±0.066: vs round 1 profile/linear −4.75/−6.31, vs K-light 0.484 +1.49/+1.56; the background-only ratio is ≈1, so subtraction only amplifies. P57: dust raises it by ×1.16–1.17, intrinsic ≈0.50. P58: after removing stars (8–16 % minor, 9–13 % major) the conditional gas B0 is 0.567 [0.497,0.642] (B3 0.627 [0.524,0.753], B4 0.668 [0.575,0.777]). P59: Chandra 0.585±0.042, Chandra−XMM +0.003±0.079/+0.013±0.127 (instrument-independent). P60: non-rotating HSE 0.85–0.96 and a bicone wind ≥1.49 excluded outside 95 %; rotating bound gas and stellar-mass-loss injection viable. At 3.5–5′ still below 1; 5–8′ not qualified (pn denominator negative, Fieller unbounded). Central 2′ quarantined (P35 round 2: no band's centre passes in 9 bands, 0/9 selected methods pass all gates; P38 0/18; P48). cognition-os/07_REVISION_LOG.md R134、R136、R137、R138、R107、R115、R131、R086/D128；.agent/jobs/2026-10-20-p47b-morphology-corrected/REPORT.sol.md；.agent/jobs/2026-10-20-p57-dust-lane-absorption/REPORT.sol.md；.agent/jobs/2026-10-20-p58-gas-only-morphology/REPORT.sol.md；.agent/jobs/2026-10-20-p59-chandra-inner-morphology/REPORT.sol.md；.agent/jobs/2026-10-20-p60-flattening-models/REPORT.sol.md；REDTEAM_QUEUE.md RT-112、RT-115 cognition-os/07_REVISION_LOG.md R134, R136, R137, R138, R107, R115, R131, R086/D128; .agent/jobs/2026-10-20-p47b-morphology-corrected/REPORT.sol.md; .agent/jobs/2026-10-20-p57-dust-lane-absorption/REPORT.sol.md; .agent/jobs/2026-10-20-p58-gas-only-morphology/REPORT.sol.md; .agent/jobs/2026-10-20-p59-chandra-inner-morphology/REPORT.sol.md; .agent/jobs/2026-10-20-p60-flattening-models/REPORT.sol.md; REDTEAM_QUEUE.md RT-112, RT-115
4. 温度对比与两相简并 4. Temperature contrast and the two-phase degeneracy Questioned 存疑Questioned ΔkT = kT(8–10′) − kT(2–4′)，目标 0。P34 自身线重建（B031 别名修正后，六行保持符号与 0-in-68 % 判定）：天空 A +0.078312 [−0.113342,+0.237629]；天空 B derived +0.179595 [+0.064830,+0.296842]（循环）；天空 B regen +0.309587（未收敛，质疑）；晕 0.179 −0.007 [−0.184,+0.145]；晕 0.1646 −0.105 [−0.286,+0.028]；晕自由（0.220）+0.180608 [−0.145,+0.487]。符号未确立（R091/R088）。两相：P43 的 56 组冷+热混合，0/56 被现有可观测量在 3σ 排除，单温 ΔkT 偏差 −1.56…+1.67 keV，两边都能偏（P37/R113）。D146：PI 裁定现有数据在 +0.2 keV 处结案，只有 |ΔkT| ≳ 0.3 keV 可探测。 ΔkT = kT(8–10′) − kT(2–4′), target 0. P34 own-line rebuild (after the B031 line-alias fix; all six rows keep sign and 0-in-68 % verdict): sky A +0.078312 [−0.113342,+0.237629]; sky B derived +0.179595 [+0.064830,+0.296842] (circular); sky B regen +0.309587 (did not converge, Questioned); halo 0.179 −0.007 [−0.184,+0.145]; halo 0.1646 −0.105 [−0.286,+0.028]; halo free (0.220) +0.180608 [−0.145,+0.487]. The sign is not established (R091/R088). Two-phase: P43's 56 cold+hot mixtures — 0/56 excluded at 3σ by any existing observable, single-temperature ΔkT bias −1.56…+1.67 keV, both signs possible (P37/R113). D146: the PI rules the existing data closed at +0.2 keV; only |ΔkT| ≳ 0.3 keV is detectable. cognition-os/07_REVISION_LOG.md R091、R088、R085、R119、R113、R104、R066/D107、D146；.agent/jobs/2026-10-20-p34-b031-rebuild/REPORT.sol.md 表；.agent/jobs/2026-10-20-p37-2t-bias-bound/REPORT.sol.md；.agent/jobs/2026-10-20-p43-hardband-2t-constraint/REPORT.sol.md cognition-os/07_REVISION_LOG.md R091, R088, R085, R119, R113, R104, R066/D107, D146; .agent/jobs/2026-10-20-p34-b031-rebuild/REPORT.sol.md table; .agent/jobs/2026-10-20-p37-2t-bias-bound/REPORT.sol.md; .agent/jobs/2026-10-20-p43-hardband-2t-constraint/REPORT.sol.md
5. 哪条路都分不开简并 5. No route separates the degeneracy Questioned 存疑Questioned 仪器：P52，10000 ks 时 EPIC 排除 56 组中的 15–20 组；Xtend、ACIS-S（现行响应+实测本底，0/56）、RGS 1′/2′（0/56）；Resolve 阀门打开（假设）3/56；LEM 类 2 eV 概念机 1 Ms 时 15/56（危险的 32 组中排除 9 组）；25/56 对一切仪器与单温谱简并（RT-110/R126 撤回 P44 的“0/56 排除”——Resolve 响应阀门关闭、ACIS-S O VII 面积 0.0065 cm²）。投影模型：P50，两相真值 8/9200 选 2T（余量 ≥2 的 0 次），梯度真值 104/400 选其家族（余量 ≥2 的 0 次）→ Kondo+2026 路线在 M104 信噪比下失效。热 SZ：P53，SO 基线最大期望 z=0.091（L=100 kpc），0/56 达到 3σ；ACT y 图在 M104 处被遮蔽，SPT-3G 不覆盖，Planck 噪声不可得；放开视线深度后单温平板可精确复现混合的 y。 Instruments: P52, at 10000 ks EPIC excludes 15–20 of 56 mixtures; Xtend, ACIS-S (current response + measured stowed background) 0/56; RGS 1′/2′ 0/56; Resolve gate-valve-open (hypothetical) 3/56; a LEM-like 2 eV concept at 1 Ms 15/56 (9 of 32 dangerous); 25/56 degenerate with a single-temperature spectrum for every instrument (RT-110/R126 withdraw P44's '0/56 excluded' — Resolve's response is gate-valve-closed, ACIS-S's O VII area 0.0065 cm²). Projection model: P50, two-phase truths picked as 2T in 8/9200 mocks (0 with margin ≥2), gradient truths recover their family in 104/400 (0 with margin ≥2) → the Kondo+2026 route fails at M104's S/N. Thermal SZ: P53, SO baseline max expected z = 0.091 (L = 100 kpc), 0/56 reach 3σ; ACT y maps masked at M104, SPT-3G does not cover it, Planck noise unavailable; with depth free a single-temperature slab reproduces the mixture y exactly. cognition-os/07_REVISION_LOG.md R126、R128、R129、R130、Synthesis；.agent/jobs/2026-10-20-p44-2t-resolving-instruments/REPORT.sol.md；.agent/jobs/2026-10-20-p50-projection-vs-2t/REPORT.sol.md；.agent/jobs/2026-10-20-p52-soft-instrument-forecast/REPORT.sol.md；.agent/jobs/2026-10-20-p53-sz-xray-joint-2t/REPORT.sol.md cognition-os/07_REVISION_LOG.md R126, R128, R129, R130, Synthesis; .agent/jobs/2026-10-20-p44-2t-resolving-instruments/REPORT.sol.md; .agent/jobs/2026-10-20-p50-projection-vs-2t/REPORT.sol.md; .agent/jobs/2026-10-20-p52-soft-instrument-forecast/REPORT.sol.md; .agent/jobs/2026-10-20-p53-sz-xray-joint-2t/REPORT.sol.md
6. 总结与下一步 6. Summary and next steps 待定 PendingPending CoPI 提案 D157（待定，等 PI 决定）：(A) 把 Q-B 写成正式的简并（负面）结果——CoPI 推荐；(B) 留给未来的软 X 射线微量热计（LEM 类，1 Ms 排除危险的 32 组中的 9 组）；(C) 多晕 SZ 叠加（随机误差按 N^−1/2 改善，但共同下限仍在）。新数据设计（P36/P36b/R118）没有任何 XMM 设计被证明能在 +0.2 keV 真值下到 2σ；前景模型的选择本身是未被跨越的系统项。 CoPI proposal D157 (pending, awaiting the PI's decision): (A) write Q-B up as a formal degeneracy (negative) result — the CoPI's recommendation; (B) keep it open for a future soft-X-ray microcalorimeter (LEM-like: at 1 Ms, 9 of the 32 dangerous mixtures excluded); (C) stack many haloes for SZ (random error improves as N^−1/2 but a common floor remains). New-data design (P36/P36b/R118): no XMM design is shown to reach 2σ at a true +0.2 keV; the sky-model choice itself is the unspanned systematic. cognition-os/04_DECISION_LOG.md D157；cognition-os/07_REVISION_LOG.md Synthesis、R118、R127 cognition-os/04_DECISION_LOG.md D157; cognition-os/07_REVISION_LOG.md Synthesis, R118, R127

草稿版本链：.agent/jobs/2026-10-20-p54-paper-results-draft/out/sol/results.tex（v1，sol 交付的"结果"片段：两个结果小节、六个表格、两个仅标题的图块，所有数字带同行来源注释，Provisional 语句用 \todo{} 标记；仅供调度员审阅，尚未纳入任何手稿）→ .agent/jobs/2026-10-20-p54b-paper-revision/out/hermes/draft_v2.md（v2，已被 RT-117 否决，见上方状态框）→ .agent/jobs/2026-10-20-p54c-paper-revision/（v3，修订中：逐条回答 RT-117 的 OBJ-1..7）。红队报告：.agent/jobs/2026-10-20-redteam-rt117-p54b-draft/REPORT.hermes.md（R149）。

Draft version chain: .agent/jobs/2026-10-20-p54-paper-results-draft/out/sol/results.tex (v1, sol's delivered "results" fragment: two results subsections, six tables, two caption-only figure blocks, every number carrying a same-line source comment, Provisional statements marked \todo{}; for dispatcher review only, not yet incorporated into any manuscript) → .agent/jobs/2026-10-20-p54b-paper-revision/out/hermes/draft_v2.md (v2, rejected by RT-117 — see the status box above) → .agent/jobs/2026-10-20-p54c-paper-revision/ (v3, in progress: answering RT-117's OBJ-1..7 one by one). Red-team report: .agent/jobs/2026-10-20-redteam-rt117-p54b-draft/REPORT.hermes.md (R149).

文献（本页引用的真实文献）

References (real literature cited on this page)