The dissipation-avalanche module (dissipation_avalanche.py)

Tests the prediction that the dissipation events of developed turbulence are marginal-stability avalanches (ledger row IV/size law). On the validated ns3d.py field it computes the full-resolution dissipation ε(xv)=2ν S_ijS_ij from spectral strain, thresholds it at h⟨ε⟩, and labels the connected high-dissipation structures with periodic boundaries (6-connectivity plus a union–find merge across the three periodic face pairs).

)} Tests the prediction that the dissipation events of developed turbulence are marginal-stability avalanches (ledger row IV/size law). On the validated ns3d.py field it computes the full-resolution dissipation ε(xv)=2ν S_ijS_ij from spectral strain, thresholds it at h⟨ε⟩, and labels the connected high-dissipation structures with periodic boundaries (6-connectivity plus a union–find merge across the three periodic face pairs).

)} Tests the prediction that the dissipation events of developed turbulence are marginal-stability avalanches (ledger row IV/size law). On the validated ns3d.py field it computes the full-resolution dissipation from spectral strain, thresholds it at , and labels the connected high-dissipation structures with periodic boundaries (6-connectivity plus a union–find merge across the three periodic face pairs). The pooled volume distribution is fitted by a discrete maximum-likelihood estimator with a KS-selected lower cutoff, and cross-checked by logarithmic-bin regression; the structure fractal dimension is measured from gyration-radius–volume scaling on non-wrapping components, and a log-likelihood ratio tests the power law against a fitted exponential. The pre-registered Lin–Wyart band is . Across resolution the measured exponent is () and (), with –, power-law form (–) confirmed and the exponential rejected at every threshold. A checkpointed driver (dev64.py, embed96.py, analyze.py, consolidate_p4.py) reproduces the two-resolution trend; captured output ships with the module.