A 91-day Ross 128 sampling block gives weak choice between nearby fixed periods in Gaussian tests

We test the ability of real Ross 128 observing times to separate fixed circular-wave periods. With a 1.7 m/s wave and independent Gaussian errors, 138 HARPS times over 91 days choose a generating period of 9.8658 days over 9.89 days in 56.0% of trials, and over 9.8556 days in 52.6%. The full 305 times give at least 99.99% correct choices for these pairs under the main model. Shared night noise and group slopes retain the short-block limit. Projection geometry gives exact phase bounds on expected excess error. No observed velocities are fitted. The result is a conditional design limit, not a measured period or planet claim. Original code, aggregate results, a figure, exact source hashes and retrieval steps are included. External database records are excluded.

astronomy · Ross 128 · GJ447 · HARPS · sampling · period alternatives · Gaussian simulation · design sensitivity