This upload contains a manuscript and supporting analysis materials for a study of an unexpected timing pattern found in global precision-clock data. Atomic clocks are usually treated as highly local instruments: each lab clock has its own noise, and each satellite or ground station clock has its own processing errors. This study tests that assumption by looking for shared patterns across two largely independent timing systems: the BIPM international atomic-clock network and GNSS/GPS station and satellite clock products. The main result is a reproducible pattern with a period of about 10.5 to 16 days. It appears in BIPM Circular T clock residuals across many national laboratories, and it also appears independently in GNSS station-clock and satellite-clock products from two analysis centers. The analysis uses null tests, cross-checks, pair-difference controls, calendar-effect removal, and product-regressor audits to test whether the pattern can be explained by simple causes such as lunar/fortnightly terms, weekly calendar effects, satellite orbit products, Earth-orientation products, or mapped GNSS/GPSPPP transfer infrastructure. The mechanism is not identified. The remaining possibilities include a hidden common feature of timing/geodesy processing, a shared-input or transfer-function effect, or a physical signal filtered through both systems.
Michael Geil (Wed,) studied this question.