NGC 253 matched 7–12′ SE/NW pressure-asymmetry fork

The question

In NGC 253's starburst-driven galactic wind, do the hot-phase thermal pressures on the two sides of the disk match — at the 7–12 kpc scale, in matched full wedges about the minor axis

The estimand is a single number

R_P = (n_e × kT)_SE / (n_e × kT)_NW

derived from matched-sector spectral fits of a new EPIC SE halo-offset pointing versus the archival NW pointing 0723220101. The comparison is reported as a 95% confidence interval on log₁₀(R_P). Three outcomes are preregistered: ASYMMETRY (CI excludes 1.0), SYMMETRY (CI includes 1.0 with σ_R < max-σ_R), and INCONCLUSIVE (σ_R ≥ max-σ_R

Why anyone should care

The default assumption in cosmological subgrid-feedback recipes is that a starburst-driven wind injects energy, momentum, and freshly-synthesized metals into the CGM symmetrically about the disk. The mass/energy loading factors (η) are written per-cone and assumed equal on both sides; CGM chemical-enrichment models assume the halo is enriched isotropically; and for any distant galaxy where the halo is unresolved and only one outflow side is seen, feedback is inferred from that single cone and assumed to apply to both

If a symmetric result is found in the nearest, best-resolved double-cone, that assumption is validated for the first time in a resolved system, and Bauer et al. (2008)'s measured SE-softer diffuse-X-ray contrast is consistent with the matched-aperture design eliminating the Bauer volume asymmetry (NW 298 kpc³ vs SE 113 kpc³). If an asymmetric result is found, the symmetric-injection default is refuted at the most basic level: single-cone feedback estimates are systematically biased, and the metal/energy budget escaping to the IGM vs recycling (fountain) differs per side

Cold molecular outflow is already known to be approximately symmetric S/N (Bolatto et al. 2013 ALMA CO(1-0) — T37, T68). A symmetric cold phase plus an asymmetric hot phase would be a clean multiphase-decoupling detection — hot wind and entrained cold gas shaped by different agents. A symmetric hot phase confirms multiphase coherence

Whose published conclusion changes Bauer et al. (2008, A&A 489, 1029) reported that the SE halo is softer than the NW halo in diffuse X-rays, but with unequal emitting volumes. I053 does not claim to overturn Bauer; it claims to test the matched, equal-volume thermal-pressure ratio that Bauer could not isolate, and to do so at a confidence level Bauer could not reach. The result either pins the symmetric-injection default or refutes it — both branches move the credence

What the measurement would look like

The number we would report is a 95% confidence interval on log₁₀(R_P). The threshold is aperture-independent: max-σ_R = A_min/k = 0.20/1.96 ≈ 0.102 (T3421). A symmetric outcome requires σ_R < 0.102; an asymmetric outcome requires the CI to exclude 1.0; the third branch is the recovery failure

What it looks like if the answer comes out the other way

If the result is INCONCLUSIVE — and the σ_CX floor is currently 2.45× above the threshold, so this is the most likely outcome today — the page is honest about that. The INCONCLUSIVE branch is a recovery failure, not a null result. It means the reducible σ_sys budget (σ_CX, σ_SP, σ_QPB, σ_SB) must close before the measurement is fundable. The σ_CX term alone — bounded at 25% by Zhang et al. (2014) at the M82 11.6 kpc Cap and 8–25% by Lopez et al. (2020) at ±2.5 kpc — exceeds the threshold by 2.24–2.74× without any extraction-derived product. No choice of aperture can rescue it; max-σ_R is aperture-independent

If the result is SYMMETRY, the symmetric-injection default is validated in a resolved double-cone, and Bauer's measured SE-softer contrast is consistent with the matched-aperture design removing the volume asymmetry — not proven to be a volume artifact, because equal projected area leaves the line-of-sight depth through a non-spherical halo unconstrained

If the result is ASYMMETRY, the symmetric-injection default is refuted; single-cone feedback estimates are systematically biased; and the metal/energy budget escaping to the IGM vs recycling differs per side

The honest answer is that the null is not boring — it is a recovery failure that points directly at the next measurement to make. The page is not asking for compute to discover nobody wanted the result; it is asking for compute to either close the systematic floor or measure the asymmetry. Both are real results

Geometry — what the new observation looks like on the sky

The geometry is the R2 execution of the I002-R4-01 REFRAME (cycle-obs1, R5/claude). The original 4–12′ aperture had a 17.36′ corner that fell outside the 15′ EPIC field radius; the refrozen 7–12′ aperture is fully contained. The fork also repairs the stale observational-shape coordinates (RA ≈ 12.0885° → 12.0895°, Dec ≈ −25.5871° → −25.5868° — the 3.36 arcsec tangent-plane vs great-circle antipode discrepancy, R4-C01 applied inline

Matched-sector area: 37.306 arcmin² per arm (0.742× the original 4–12′ area of 50.265 arcmin²). At 3.5 Mpc, the projected radial range is ~7–12 kpc

Why existing observations cannot answer it

NGC 253 has a dedicated NW halo pointing (0723220101, 97.6 ks PN MEDIUM Full Frame, 2013; offset 21.0′ NW at PA 328.7° from nucleus) but no SE halo counterpart. All other XMM pointings cover only the nucleus (0152020101, 0304850901–0304851301). The SE halo is the natural symmetry counterpart and cannot be observed from the existing archive. Without a SE pointing, the two-sided comparison is impossible

eROSITA does not cover NGC 253 (eastern hemisphere, l = 97.4° — T11, T15). Chandra has no halo-offset mosaic for NGC 253 (unlike M82's Strickland mosaic); the deepest Chandra observation (ObsID 20343, 153.6 ks ACIS-I, Hornschemeier 2018) covers the central galaxy but its 16.9′ FOV misses the extended halo. Suzaku 803004010 (56.7 ks OFFSET) is at 18.2′ NNE of the nucleus — a background pointing, not a SE halo pointing

The new EPIC SE offset is required. Exposure is intentionally unsized in this proposal and must be derived by R6 from a response-folded recovery calculation — that calculation is the load-bearing deliverable (I002-E27) and is currently HEAVY-blocked

What it costs — and what is still unknown about the cost

What we know: the geometry is executable (7–12′ wedge contained in 15′ EPIC field, verified to 4 decimals). The matched-aperture design eliminates the Bauer 2008 volume asymmetry. The threshold side of the gate is aperture-independent (max-σ_R = 0.102, derived from design parameters A_min = 0.20 and k = 1.96, not from any measured quantity). The blocking item is the σ_CX systematic floor — bounded at 25% by Zhang et al. (2014) at the M82 11.6 kpc Cap and 8–25% by Lopez et al. (2020) at ±2.5 kpc — which exceeds the aperture-independent threshold by 2.24–2.74×

What we do not know: the exposure required for the SE pointing. R6 must derive it from a response-folded recovery calculation (per-camera ARF/RMF, ESAS QPB+SP+CXB background, source model grid, screening loss, effective area, preregistered pressure-ratio threshold with recovery curve and coverage). That calculation is blocked by the σ_CX floor: until σ_CX is reduced below 0.102, the recovery curve does not close, and any exposure number is unanchored

The reducible σ_sys budget has four terms: σ_CX (charge exchange), σ_SP (residual soft proton), σ_QPB (quasi-particle background), σ_SB (cross-calibration / epoch stability). Of these, σ_CX is the lifecycle-binding term; the other three are part of the I002-E22 / I002-E23 HEAVY deliverable

Gate ledger (D043 — single source of truth

Gate	Status	Role
I002-E21	NARROWED — APERTURE-INVALID (refreeze required	Bauer-2008-narrowed matched-sector pressure-asymmetry estimand + falsifiable detection threshold/systematic floor. HOLD, re-anchored cycle-obs1. Threshold side intact: max-σR (SE>NW = 0.1116, NW>SE = 0.0911) derives from Amin = 0.20 and k = 1.96 and is aperture-independent. Measured side voided: the frozen §10 geometry (rin = 4.0′, rout = 12.0′, θ = 22.5°, NW PA = −31.25°) is unrealizable in either camera field at the 20.99′ mirror/NW offset. σCX = 0.25 exceeds 0.1116 by 2.24× and 0.0911 by 2.74×, so no proposal-ready promotion can be supported
I002-E22	OPEN	Epoch-dependent background and σsys error budget. Supporting input to I002-E21, not a substitute blocker
I002-E23	OPEN	0723220101 NW-halo matched-sector baseline extraction. HEAVY dependency of I002-E21; queued as FRONTIER_QUEUE Item 3
I002-E24	CLOSED — RESOLVED-NEGATIVE	Suzaku 803004010 overlap with the proposed SE sector. Non-blocking provenance/geometry result
I002-E25	NARROWED	Halo-scale CX and absorption/neutral-gas nuisance bounds. Supporting input to I002-E21, not a substitute blocker
I002-E26	OPEN	Bauer 2008 original-PDF transport verification. Non-blocking provenance gate
I002-E27	OPEN	Matched-sector aperture refreeze: 7–12′ (rin ≥ 6.727′) 45° full wedge, BOTH arms, plus a reproducible response-folded count-rate product. Mandatory correctness repair — the idea's stated observational shape was unrealizable until it closes. NOT the blocking item (D041): closing it changes no lifecycle state while I002-E21/σCX blocks, so it is overhead-capped. This fork IS the R2 execution of E27: the aperture is refrozen at 7–12′ on both arms (geometry verified above), and the response-folded count-rate product is the next HEAVY deliverable

What I am and am not claiming

I am claiming the science question is real (symmetric CGM energy/metal injection is a load-bearing belief in subgrid-feedback recipes); the geometry is executable (7–12′ wedge contained in 15′ EPIC field); the matched-aperture design eliminates the Bauer 2008 volume asymmetry; and the σ_CX systematic floor is the lifecycle-binding term that must close before the measurement is fundable

I am not claiming that a symmetric I053 result would prove Bauer's soft difference was a volume artifact — equal-area geometry removes one volume asymmetry but cannot by itself establish the counterfactual cause of Bauer's measured spectral contrast. The symmetric outcome is consistent with the artifact hypothesis, not a proof of it (R4-C02 applied inline). I am also not claiming that the SE pointing is feasible at any specific exposure — that requires the response-folded recovery calculation (I002-E27), which is HEAVY-blocked today

Provenance

• Idea file: shared/ideas/I053_ngc253-se-nw-matched-pressure-asymmetry-7-12-fork.md
• R3 Analyst evidence: cycles/cycle-obs9/evidence/idea-I053.md
• R4 Red-team (commit a264137, gate=PASS): cycles/cycle-obs9/rebuttals.md (I053 Rebuttal Ledger); cycles/cycle-obs9/r4_i053_anchor_checks.py (SHA-256 a4a98f44ad75cdb8a77a6e7dd28feb02534c9f70bbd62de22ab591fb819759d2
• R5-Portfolio adjudication: cycles/cycle-obs9/r5_adjudication_i053.md (cycle-obs9, hermes, 2026-08-07
• Trust table rows: T09 (NGC 253 XMM ObsIDs), T11 (eRASS1 coverage), T34–T37 (NGC 253 nuclear/wind/halo science), T68 (Bolatto CO S/N), T73/T156 (Bauer SE/NW comparison), T210–T219 (I002-E21 floor inputs), T239–T242 (CX literature), T3266 (I053 archive refresh), T3420–T3425 (I053 fork-specific facts
• Shared sources: shared/sources/I002_bauer2008_se_nw_halo_intake.md; shared/sources/I002_E25_cx_literature.md