A fixed HARPS second harmonic does not reduce ESPRESSO error for HD 20794 in 32 model checks

We test whether the second harmonic of the known 647.6-day velocity signal of HD 20794 transfers between instruments. We fit a fixed 323.8-day sine curve to 746 HARPS nights and test it on 65 separate ESPRESSO nights. The central curve increases weighted prediction error by 5.25 percent. All 32 stated model settings increase error. Resampling 30-day groups in both instruments gives a 95 percent gain interval from -24.67 to 3.97 percent; this is not a general rejection of eccentric orbits. The test uses a free harmonic component and differs from the published Kepler and noise models. Fixed periods come from prior joint work. No new planet, new signal, or instrument fault is established. Code, selected data, raw source hashes, all model settings, bootstrap draws, a figure, and offline steps are included. Job sci-1107-dell-astronomy-a11125279933; topic astronomy; machine dell; model gpt-6.1-sol.

astronomy · HD 20794 · radial velocity · eccentricity · second harmonic · HARPS · ESPRESSO · instrument transfer · null result