15 climate questions. One diagnostic. 15 answers.
The framework has been applied to 15 specific climate cascades — across CMIP6 ensembles, satellite records, and ice-core paleoclimate. Each question gets the same operator and the same verdict vocabulary. Below: the headline finding for each cluster.
Tipping elements · 4 instances · CMIP6 + NSIDC + RAPID
Arctic sea-ice loss is robust. AMOC weakening is partially robust. Amazon NPP is inconclusive.
Of the three IPCC-prominent tipping observables in the CMIP6 SSP5-8.5 ensemble: 21/21 models agree on Arctic sea-ice direction; 19/19 agree on AMOC direction but only 21% agree year-by-year; 5/25 models project Amazon NPP rising rather than falling — genuine disagreement, not measurement noise. The framework also shows the literature's 2008–2012 RAPID AMOC weakening claim does not survive a BIC change-point test.
Sea-level rise · 1 instance · 6 observational shadows + AR6 fan
Aggregate sea-level rise looks reassuring under aggressive mitigation. The Antarctic component does not.
Under SSP1-1.9 the aggregate stays sub-rate (β = −1.12) — but the dominant component, the Antarctic Ice Sheet, is super-rate (β = +3.76) under the same pathway. Component decomposition reveals what aggregation hides. AIS post-tipping rate scales as Φ³·⁵ — the fastest acceleration anywhere in the climate catalogue.
CO₂ & energy budget · 2 instances · Mauna Loa + Vostok + EPICA + CERES
Anthropogenic CO₂ is a fundamentally different cascade-class from natural CO₂.
Separated at 5.7σ. Modern Mauna Loa CO₂ rate exponent β ≈ +4.04. Paleoclimate Vostok and EPICA β ≈ +3.00 (cross-cadence σ = 0.024 — natural CO₂ is scale-invariant from millennia to 100,000 years). The framework reads anthropogenic CO₂ as belonging to a distinct cascade class, not just a faster natural process. The energy-budget instance independently reproduces Trenberth–Fasullo's "missing energy" finding (78% closure) from first principles, in domain-internal units.
Paleoclimate · 3 instances · 3 ice cores + 5 proxy records
Some past regime changes were cascade reorganisations. Others were just frequency shifts.
The Mid-Pleistocene Transition (~900,000 years ago) reorganised dominant Milankovitch periodicity from 41 kyr to 100 kyr — but the brake-rate scaling did not change. The framework distinguishes resonance reorganisation from cascade-class change. Younger Dryas and Heinrich events reveal the high-latitude AMOC-coupled cascade behaves as one Heinrich-class across H1–H6, while tropical/marine proxies carry event-specific information — supporting the heterogeneous-trigger picture.
Single systems · 5 instances · Ozone, permafrost, ENSO, cyclones, tipping prediction
The Montreal Protocol was the principal lever. Permafrost has tipped. ENSO is not a brake-cascade.
Counterfactual analysis: without the 1987 protocol, Antarctic ozone hits zero by 2010. The original intervention did the heavy lifting; further compliance produces only marginal recovery. Permafrost cross-model first-break warming threshold = +0.36 ± 0.12 K — already exceeded. ENSO returns R² < 0.01 across all indices: oscillatory, not brake-cascade — the framework correctly reports its own non-applicability. Tropical-cyclone wind-vs-pressure shadows shadow different cascades in 5/6 basins (Theorem-10 anti-shadow detection).
All 15 cluster into the synthesis
Each instance feeds the cross-observable scenario-fan synthesis at scenarios and the best-pathway analysis at mitigation. Source documents: comprehensive-record.md for per-instance methodology; scenario_fan_synthesis.md for the synthesis.