Testing the Chronos Change-Clock Framework: Public Clock Transformation, Method Robustness, and Leap-Event Discrimination

This paper presents a coordinated series of public-data tests of the Chronos change-clock framework. The central premise is that different clocks can describe the same progression while assigning distinct mathematical coordinates, making clock provenance and explicit transformation scientifically important. The study examines atomic and civil time, Earth rotation, Earth orbital progression, and pulsar-based timing. It evaluates pairwise clock transformations, three-clock closure, nanosecond reference-clock response, agreement between frequentist and Bayesian pulsar reconstructions, and historical leap-second event discrimination. The atomic, rotational, and orbital analysis used 1,952 shared daily events with chronological calibration, validation, and untouched test periods. The pulsar-clock analysis recovered a predetermined reference-clock transformation on held-out epochs with subnanosecond closure error. A separate state-plus-change coordinate distinguished leap-second periods from no-event periods with a held-out ROC AUC of 0.9881 and an exact permutation probability of 0.000445. The unchanged coordinate also correctly classified three later IERS no-leap decisions. Failed tests, gauge limitations, small-event-count restrictions, and interpretation boundaries are retained in the paper. The results support the operational usefulness of clock provenance, declared transformations, state-plus-change coordinates, chronological testing, and closure residuals. They do not independently establish every physical claim of Chronos Theory. The accompanying reproducibility archive contains the analysis reports, source-code packages, numerical audits, public-data provenance, figures, checksums, and the frozen prospective protocol for IERS Bulletin C73.