Reproduction kit for "A parameter-free closed-form prediction for the structure-function exponents of three-dimensional turbulence"

Closed-form, parameter-free prediction for the structure-function exponents zeta(p) of fully-developed, statistically isotropic, incompressible three-dimensional Navier-Stokes turbulence at high Reynolds number. The prediction follows from a binary multiplicative cascade construction whose two weights are uniquely fixed by total-measure conservation and a self-similarity ratio constraint, with no fitted parameters at any stage.

This deposit contains the complete analysis pipeline used in the accompanying manuscript (under review at Europhysics Letters, May 2026):

- pipeline/ — closed-form predictions, structure-function and multifractal estimators (MFDFA, partition function), bootstrap helpers, JHTDB clients (givernylocal backend with legacy SOAP fallback).- runners/ — live JHTDB pull, bulked-ensemble orchestrator, and a multi-convention multifractal-spectrum diagnostic suite (finite-N bias, q-range sweep, partition function, positive-q analysis).- manuscript/ — the manuscript source (Markdown), compiled PDF, and figure generators.- data/ — cached intermediate JSON results from a representative live run on the JHTDB isotropic1024coarse dataset (64 lines × 4096 points across 4 decorrelated time slices, 262,144 access tokens).

The reproduction guide (REPRODUCE.md) walks through regenerating every figure and table in the manuscript from raw data; data/bulked_results.json alone is sufficient to regenerate Figure 1, Table I, and Table II without re-querying JHTDB.

Empirical headline: the closed-form prediction has bootstrap-mean absolute deviation <|Delta zeta|> = 0.033 [0.027, 0.040] from JHTDB Extended-Self-Similarity exponents across p in [0.5, 8], smaller by a factor of ~1.5 than She-Léveque (0.050), KO62 (0.054), and K41 (0.074), with bootstrap pairwise differences excluding zero at p < 0.01.

The raw JHTDB velocity data analysed here is publicly available at the Johns Hopkins Turbulence Database (Li et al. 2008, doi:10.1080/14685240802376389).