# RMO104 — saved evidence, clear bounds and the next observation

## What is complete

This stage consolidates the saved R98–R103 results. Their JSON records were
read and their claims checked for consistency. No numerical solver, shock
audit or observation reduction was rerun. The existing R72 four-family model
demonstration remains a model demonstration, not four solar identifications.

The useful result is a conditional chain: partial inputs can exclude fast;
wider bounds can admit both a complete fast state and a complete slow state;
an independent magnetic constraint can exclude fast again. The required bound
depends on which component and field direction are known. None of this alone
identifies the solar front shown in the viewer.

## Six saved checks

| Check | Result | Inputs / precision | Limit | Next useful measurement |
|---|---|---|---|---|
| R98 | One published slow-shock polar reproduced. | 177 sampled slow states; seven printed-curve points differ by at most 0.556°, within the declared ±1.5° reading tolerance. | This checks one published model case. Graph-reading tolerance is not an observational error. | For a solar application: matched local plasma states and geometry. |
| R99 | Partial inputs can exclude fast. | Exact compression r₀ ≈ 2.167774 exceeds the necessary fast upper bound 2 at normal acoustic Mach M₀ = √3, γ = 5/3. | Removing the normal-flow or thermal constraint admits checked fast alternatives. Excluding fast does not identify slow. | Normal plasma flow and a justified thermal constraint; image speed alone is insufficient. |
| R100 | Error size and allowed combinations matter. | Compression r₀ ±0.05 and Mach M₀ ±5% exclude fast. Mach ±10% loses that scalar certificate; an imposed joint strip with the same wider marginal ranges retains it. | The strip is a declared model assumption. Individual ± bounds do not specify dependence. | The actual joint measurement constraints, including shared calibration or geometry. |
| R101 | The wider partial inputs really admit both families. | A complete fast state at r = 2.15 and M/M₀ = 1.09 lies inside the wider box, alongside the saved slow reference. | Different full states fit the same partial bounds. Identical full states have not received two diagnoses. | An additional independent constraint capable of separating the remaining states. |
| R102 | Normal Alfvén-speed bounds restore fast exclusion. | About cAn,₀ = w₀, ±20% passes; the strict symmetric half-width limit for this criterion is about 24.4%. | Compression ±0.05 and Mach ±10% are included. This concerns the normal component, with hard model bounds. | An independent bound on the normal Alfvén speed, or field strength, density and direction. |
| R103 | Field direction controls the usable speed error. | About total cA,₀ = w₀/cos30°, total speed ±20% with angle 30° ±5° excludes fast; angle ±10° does not certify it. | The strict angular half-width limit is about 5.08859°. With angle ±10°, this criterion requires total-speed error below about ±14.55%. Same compression/Mach bounds; fixed front normal. | A field-to-normal angle constraint and, for an uncertain normal, its linked effect on normal flow. |

## How to state the errors

Use the centre and ± amount/percentage, units, hard-bound or statistical
meaning, and the allowed joint combinations. Here r₀ ≈ 2.167774 and M₀ = √3;
±0.05 in compression is absolute, while ±10% in Mach is relative. Mach refers
to the normal shock-frame plasma flow divided by the upstream sound speed.
The normal-speed centre is w₀; the R103 total-speed centre is w₀/cos30°.
R100–R103 bounds are declared model bounds, not inferred 1σ, 2σ or 3σ errors.
The strict thresholds certify this exclusion criterion, not a universal
instrument requirement or a demonstrated feasible boundary of all shocks.

Unknown parameters may be represented by explicitly sourced assumptions and
defensible ranges. They must be labelled assumed, kept separate from measured
values, and varied where the conclusion depends on them. Shared density,
field or geometric inputs must be linked consistently. A chosen correlation
must not be introduced merely to obtain a preferred family. Conditional solar
model support is a valid outcome, with its assumptions visible.

## Solar cases already in the project

This table updates saved readiness; it is not a new analysis of fourteen
events. E08 remains a completed control and the Kumar/IRIS hold remains in
force. E09 is not the 2017 SUVI event. E11 includes the later Ma results even
though an older catalogue field predates them.

| Event | Role | Saved result / status | Remaining question |
|---|---|---|---|
| E01 · 19 May 2007 | Long · imaging/cadence comparison | Saved literature control. | Match one front, sector, interval and cadence before comparing physical interpretations. |
| E02 · 27 July 2010 | Chen & Wu · leading and trailing structures | Saved literature comparison. | Track the two structures separately; slower image motion does not establish a slow shock. |
| E03 · Liu 2012 paper | Wave trains | Saved card; event date not yet extracted. | Identify the event and one crest with its time series and uncertainty. |
| E04 · 15 February 2011 | Vanninathan · DEM compression and heating | Published thermal constraints; no new RMO observation diagnosis. | Pair density and temperature estimates in the same region/time, preserving background and line-of-sight assumptions. |
| E05 · 16 February 2011 | Veronig/Vršnak · EIS and AIA | Image speed and plasma Doppler speed distinguished. | Confirm slit crossing, exposure, control interval and density significance. |
| E06 · 7 June 2011 | CASHeW/Kozarev | Published geometry and PFSS provenance saved. | Match front edge and local state estimates at the same place/time. |
| E07 · 12 December 2013 | CASHeW/Kozarev · C4.6 | Published imaging case saved. | Constrain local plasma states; flare class does not label the wave family. |
| E08 · 13 June 1998 | Harra · outer front / filament control | CONTROL_ONLY; completed audit retained. | Keep the outer front and filament spectra separate. Do not reopen the completed audit. |
| E09 · 22 June 2015 | Ye · flare-loop comparison | Date corrected in the saved project; author slow-shock interpretation distinguished. | Test conductive/isothermal versus adiabatic assumptions before transferring the interpretation. |
| E10 · 27 January 2011, 08:45 UT | Muhr/Veronig/Kienreich/Vršnak · B6.6 | Exact catalogue row and speed convention saved. | Add normal, compression and local states; do not mix other events that day. |
| E11 · 13 June 2010 | Ma · EUV/radio/thermal comparison | R76 input audit and later R84–R94 conditional fast-branch checks already saved. | Match diagnostics in place/time, errors, plasma flow and geometry. The old catalogue status is superseded by those later results. |
| Additional · 13 February 2009 | T. Podladchikova et al. 2019 · stereo geometry | Saved same-patch speed comparison: about 260 → 207 km/s. | 3D normal, flow and field require more than a tabulated crest height. |
| Additional · 9 May 2014 | Kumar · mode conversion | HOLD on the earlier IRIS crossing task. | Keep mode conversion distinct from identification of a local shock. |
| Additional · 10 September 2017 | GOES-16/SUVI · separate new example | 51-frame supplied viewer; public primary event sources located. | Exact archive matching, timestamps and calibration are still needed for quantitative tracking. Optional if acquisition delays progress. |

Readiness sources in the project: RMO/RMO_event_check_coverage_R94.md,
results/literature_input/catalogue_current.json, the saved scientific assessments. These are restored saved statuses; the
primary articles for all catalogue entries were not reread at R104.

The 2017 event sources located during preparation are
[Veronig et al. 2018](https://arxiv.org/html/1810.09320v1) and
[Seaton & Darnel 2018](https://arxiv.org/html/1712.06003v3).
The [NOAA archive listing](https://data.ngdc.noaa.gov/platforms/solar-space-observing-satellites/goes/goes16/l1b/suvi-l1b-fe195/2017/09/)
includes 10 September. This establishes an acquisition route, not the exact
identity, calibration or observing cadence of the supplied JPEG sequence.
No raw observation files were downloaded. The viewer retains frame 23,
Original: GOES/SUVI and all existing processing.

## One next bounded step, not run

One proposed next observational test: use E04 (15 February 2011) to compare paired published DEM density and temperature changes with reversible adiabatic compression, T₂/T₁ = (ρ₂/ρ₁)^(2/3), for γ = 5/3. First identify the same region and interval and the meaning of its errors. Finish with compatibility, incompatibility, or an explicit limit from missing paired values. This is a thermal-consistency test, not a fast/slow identification. It uses a new observational constraint rather than repeating the Ma reconstruction. It has not been run. SUVI tracking can follow separately and is not required to finish this step.

The article conclusion affected is whether the measured thermal response
requires departures from the stated reversible compression model. A compatible
response would not prove a shock; an incompatible response would first require
checking emission weighting and the adopted density/temperature estimates.
Published time evolution ranges must not be relabelled measurement errors.

Source to inspect for that step: [Vanninathan et al. 2015](https://arxiv.org/abs/1509.05269),
already present as E04. The first task is source-level pairing, not a bulk
observatory download. This proposed step is not automatically launched here.

## Interface reconstruction and limits

 R104 is rebuilt directly from
the exact R102a HTML. The accepted title/logo arrangement, the entire title row,
all existing scripts/styles, the embedded viewer and earlier scientific sections
are preserved. The saved R103 result is copied as a separate section. This
synthesis is another closed section; it does not lengthen the header summary.

The title and footer identify R104 / interface 0.4.42. A separate single-file
HTML and a ZIP containing that exact HTML are supplied. The screenshot failure
of the previously delivered R103 has no established root cause: the exact file
opened in that screenshot has not been supplied. Source preservation and ZIP
integrity checks are not native-browser acceptance. R104 is a reconstruction
from the user-working baseline, not a claim that the original fault has been
reproduced. No browser-policy workaround was attempted.