2023 Barnard's Star date means give a positive 4.1244-day band term at the 2021–2022 phase under independent Gaussian errors

We test whether a known blue minus red velocity feature in Barnard's Star repeats between observing seasons. We use 112 published MAROON-X velocity pairs to form 103 daily groups in BJD. All four published periods stay fixed. A fit to 2021 and 2022 sets the direction of the 4.1244-day sine and cosine pair. Its signed projection in 23 date means from 2023 is 0.449 m/s. A separate 95% Student t interval is 0.139 to 0.760 m/s. This interval assumes independent Gaussian date errors and relative variances from the quoted channel errors. The model uses separate run lines and one unknown variance scale in 2023. Constant and quadratic run trends give 0.452 and 0.463 m/s. The interval includes zero under two fixed damped cosine noise models with a 10-day decay time. Thus the evidence for a repeated band term depends on the error model. The test is retrospective. The period and feature were already selected with these observations. This is not a new independent detection, and the physical cause is not identified. We also give four model-based simultaneous 95% upper bounds on band amplitude. For b, c, d, and e, these are 0.477, 0.690, 0.450, and 0.412 m/s under the primary model. The package preserves source files, code, provenance, results, figures, and reproduction instructions.

Barnard's Star · MAROON-X · radial velocity · wavelength dependence · time split · conditional inference