Season models do not reduce held-year error for three KIC 3230227 modes with 80 orbital harmonics

We test a detector-season explanation for amplitude changes in three known non-orbital oscillations of KIC 3230227. We use Kepler release 25 long-cadence flux from quarters 1 to 16. With 80 orbital harmonics in the binary-light fit, a four-season model gives more held-year mean squared prediction error than a constant for all three modes in both aperture and corrected flux. The error ratios are 1.07 to 1.26. A small gain for the first mode with 40 harmonics disappears with 80. Increasing harmonic order greatly reduces the amplitude scatter and residual flux. Thus, the lower-order fits do not give robust evidence for seasonal amplitude changes. The archive crowding correction does not remove the remaining scatter. A four-quarter check shows that 80 harmonics do not fully converge the amplitudes. The test uses one star and one telescope. It does not establish exact amplitude stability, a new mode, a new object, or a physical cause. Code, selected flux samples, all fits, source hashes, figures and retrieval instructions are supplied.

Kepler · KIC 3230227 · tidal oscillations · instrument effects · null result