{
 "claim_retracted": "peak pressure decays monotonically across all 26 outputs (362 -> 146 GPa)",
 "errors": [
  "'146 GPa' was a misread of 3.69e11 dyn/cm^2 (= 36.9 GPa), not 1.46e12",
  "the sequence is NOT monotonic across all 26: it rises 2.52x at the final output"
 ],
 "corrected": {
  "monotonic_span": "plt2d_00113 through plt2d_01248 (24 consecutive outputs), 361.8 -> 14.6 GPa, strictly decreasing",
  "final_output_GPa": 36.9,
  "final_rise_factor": 2.52,
  "anomaly_diagnosis": "ONE cell of 1,572,864 (0.0001% of the domain) at r=10811, z=-9814 cm in rarefied rock (rho=1.20, i.e. 0.44 rho_0) on the crater wall. Only 1 cell exceeds 20 GPa. The physical shock front at that time is at z ~ -37,700 cm, a different structure entirely. Dense-rock (rho>2.0) percentiles are P99=11.6, P99.9=13.2, P99.99=13.6 GPa -- continuing the decay trend from 14.6 GPa. So the release wave IS behaving correctly; the global max is an isolated single-cell spike in the cavitation-regularized expanded branch, not a physical peak.",
  "implication": "Global max over all cells is the wrong statistic for late-time decay -- it is single-cell-noise dominated once the shock weakens. Phase 2 must fit the decay on radially binned peak pressure restricted to rho>2.0, and should report a high percentile rather than the raw max at late times."
 }
}