Three fixed-period HD 10180 curves increase error in 96 later HARPS exposures after six larger signals are fitted

We test three weak radial velocity curves in HD 10180 with later public HARPS data. We learn their amplitudes and phases from 240 early exposures at periods of 1.17766, 9.655, and 67.55 days. We test the fixed curves on 96 later exposures from July 2010 to August 2017. We fit six larger periodic signals and instrument offsets separately in each set. The primary noise model adds 1.5 m/s of independent noise to the quoted errors. The three fixed curves together increase later weighted squared error by 27.03. Each of the six larger curves improves later error in a separate control. We then include early coefficient uncertainty and average over the published period bounds with early-data weights. The predictive log density ratio for all three weak curves is −5.80 relative to no weak curve. The 67-day component changes from −2.12 at its central period to +0.84 with the period average. An early-only period fit near 66.67 days also improves its later score. Thus period uncertainty changes the 67-day result. The smaller test set after July 2012 also changes the sign for the 9.655-day curve. These are conditional curve tests under specified linear signal and noise models. They do not establish or exclude planets. We preserve the source data, requests, row split, code, results, period samples, figures, and reproduction steps.

HD 10180 · HARPS · radial velocity · exoplanet candidates · time split · period uncertainty · public data · reproducible astronomy