An early 22.91-day depth model does not improve later predictions for KIC 12557548 in the main Kepler test

We test whether a fixed rotation clock fitted to early Kepler depths predicts later depths of the known dusty system KIC 12557548. At the main phase half width of 0.05, a 22.91-day sinusoid fitted on quarters 1 to 8 increases squared prediction error in quarters 9 to 17 by 1.12 percent in aperture flux and 1.07 percent in corrected flux. Later projections onto the early models are 0.321 and 0.335, with quarter resampling intervals that include zero. A period grid selected on early data also gives negative later error gains. A wider event window gives small positive fixed-period gains. Thus, the main test does not show a robust prediction from one unchanged early clock. It does not exclude a changing relation to stellar rotation. We retain all 24 tests and off-transit controls. We verified the signed source bundle and all 17 raw-file hashes, regenerated the event table, and checked the regression with an independent calculation. There is no new object claim or independent telescope confirmation.

Kepler · KIC 12557548 · dust · stellar rotation · prediction · transit variability