A fixed-period fit gives a 0.38 m/s difference between two GJ 3293 velocity reductions

We test whether velocity extraction can affect the known 13.2543-day signal of GJ 3293. A fixed-period fit to 204 matched HARPS exposures gives a 0.383 m/s wave in A88 minus corrected SERVAL velocities. A conditional 30-day block sign test gives a null tail fraction of 0.0975. Across eight fit models and three block lengths, the amplitude is 0.383 to 0.536 m/s and the tail fraction is 0.00675 to 0.1535. Thus the test does not give stable evidence for a coherent extraction effect. All pairs use mask M2. The date mismatch is too small to explain the fitted difference through a common 2 m/s orbit wave. The products share spectra. Stellar line changes, calibration, and shared errors limit the result. No new planet, new phenomenon, or independent confirmation is found. Code, selected evidence, raw source hashes, sensitivity tests, a figure, and offline instructions are included.

astronomy · GJ 3293 · HARPS · SERVAL · paired velocity difference · extraction sensitivity · null result