What the framework does NOT do
- Prescribe specific emissions reductions. The framework provides consistency-anchored measurement and projection under one diagnostic apparatus. It does not turn into a policy recommendation.
- Replace climate models. CMIP6 ensembles are the shadows the framework reads. The framework adds a unified diagnostic on top of model output; it does not produce the model trajectories.
- Make policy recommendations. The arc's findings inform policy debate; they do not resolve it. The σ_cross paradox is a risk-assessment finding, not a policy directive.
- Resolve scientific debates. Where the framework's cross-shadow consensus disagrees with single-instance findings (e.g. RAPID AMOC weakening 2008–2012 not surviving framework BIC test), the framework reports its reading honestly — it does not arbitrate.
What the framework DOES do
The framework's principal contribution to climate is itself a Law-V finding: a unified cross-observable diagnostic in domain-internal units, paired with the explicit identification of where the diagnostic does and does not apply.
- Where Theorem 3 binds (annual cadence, narrow log-range), the framework declines to make a Layer-A claim and falls back to Layer-B direction-and-rate-sign consensus that retains real informational content.
- Where Theorem 8 forced-rate degeneracy fires (constant rate ⇒ log|ρ| does not scale with log|Φ|), the framework returns β ≈ 0 and reports forced-rate scope through \(\mathscr{A}\).
- Where Theorem 10 anti-shadow detection fires (\(\mathfrak{A} \gg \tau_{T3}\)), the framework reports physical-shadow-mismatch — instances #25 (ozone), #28 (ENSO Niño regions), #30 (tropical cyclone wind vs pressure). The candidate set shadows different cascades; that is honest physics, not poor methodology.
Synthesis-level scope-reporter
By Theorem 12 of foundations.md, every framework finding admits a scope-reporter \(\mathscr{A}\) = (tested, excluded, \(\tau_{T3}\), gauge \(G\)). The synthesis-level \(\mathscr{A}\) for the climate arc:
- Energy-budget scenario-fan: only ~24% complete (6/25 model-SSP pairs) due to the heavy ocean-heat-content integration. Partial results show closure ratios of 0.397 to 0.549 — broadly consistent with instance #18's observational 78%.
- SSP1-1.9 and SSP3-7.0 missing for AMOC msftmz and Amazon NPP in the cached Pangeo catalogue — honest catalogue gap, not experimental choice.
- Theorem 3 binds forward: extending the projection horizon does not enlarge log(\(\Phi_{\max}/\Phi_{\min}\)) for sea-ice and AMOC unless Φ approaches zero. β extraction precision is intrinsically poor at narrow log-range, and SSP fans do not fix this.
- Sea-ice and AMOC have NO scenarios that clear "no PELT regime change pre-2050" — the SSP forcing transition itself constitutes a PELT-detected regime change. The criterion was too strict for forced-physics shadows.
- Brake-p does not apply to forced-rate cascades (constant rate ⇒ log|ρ| does not scale with log|Φ|). All statistical-fan β values are near zero by construction. The framework's predictive utility is in cross-shadow σ_cross and PELT regime-detection, not in single-scenario β.
The open critique
The framework is subject to all the standard critiques of cross-model ensemble analyses: the CMIP6 ensemble is itself a sample with its own biases; the SSPs are a discretisation of a continuous space of possible forcing trajectories; the historical record is one shadow per observable. The framework does not argue these critiques away. It provides a diagnostic that respects these limitations — by reporting σ_cross honestly, by acknowledging Theorem 3 binding, by distinguishing Layer-A from Layer-B from Layer-C calls, and by publishing its own \(\mathscr{A}\) scope-reporter for every finding.