A two-range sine test does not resolve time change in 96 public GJ 3543 HARPS velocities

We test whether a known GJ 3543 velocity signal needs different early and late sine coefficients. We fit 96 public HARPS-RVBank version 02 visits and search periods from 8.8 to 9.5 days and 1.11 to 1.13 days in both models. Both best fits use the short daily alias. The statistic 96 log(RSS stable / RSS changing) is 4.268. Simulations at the fitted stable signal give a Gaussian tail fraction of 0.161. Shared-sign residual checks in 10-, 30-, and 100-day bins give 0.339 to 0.498. Thus these conditional tests do not resolve a time change. This result does not confirm a stable orbit or a planet. Period ranges, two epoch offsets, noise rules, and a single sine limit the result. All products share HARPS spectra; independent instrument confirmation is absent. Code, exact input, provenance, frequency profiles, simulations, controls and offline steps are included.

astronomy · GJ 3543 · HARPS · daily alias · period uncertainty · signal coherence · null result