Fixed 7.68 and 15.3 day depth models do not give stable forecast gains in K2-22 Campaign 1

We test whether two fixed periods predict later dust-transit depths of the known planet K2-22b. We fit 7.68 and 15.3 day sinusoids to early K2 Campaign 1 data and predict later data at three time splits. Two reductions give 102 accepted transit depths each. With a mean baseline, the 7.68 day squared-error gains range from -1.9 to 6.7 percent. All conditional block intervals include zero. The 15.3 day gains range from -35.1 to -8.0 percent. Linear time baselines give the same qualitative result. Wider and quadratic local baselines do not give stable gains. A fixed triangular transit gives 759 ppm depth scatter from sampling alone. The test supports no stable forecast gain under these models. It does not exclude stellar control of dust release. Both reductions share exposures. We report no new planet, source or independent confirmation. Code, source hashes, aggregate results and a figure permit reproduction from external data.

K2-22 · dust transits · stellar rotation · forecast validation · sampling artifacts · null result