Local linear fits to 168 TESS orbits give a 0.119 percent mean loss in the 46-minute K2-22 b window

We tested mean dimming and intermittent deep transits of the known dust planet K2-22 b. We used public two-minute TESS light curves from sectors 45, 46 and 72 in 2021 and 2023. Local linear fits and fixed coverage cuts retain 168 orbits. A 46-minute exposure-integrated rectangular template gives a mean loss of 0.119 percent of target flux. Its conditional basic bootstrap 95 percent interval is 0.067 to 0.170 percent with eight-orbit noise blocks. Aperture flux and quadratic baselines give similar means. A half-orbit control is near zero. An exploratory paired contrast is positive with a wider interval. These results support weak mean dimming at the expected phase under the tested model. They do not identify its physical source within the aperture. The one-sided conditional 99 percent upper limit on the mean is 0.179 percent. For fixed 1 percent deep events, the conditional rate limit is 21.3 percent for independent orbits and 54.6 percent for activity in 32-orbit blocks. These rate bounds assume additive centered bootstrap noise, a common rate, and the stated processed signal. They have no demonstrated general coverage. The contribution includes original FITS inputs, measurements for each orbit, sampling audits, code, figures and reproduction checks. It does not add new TESS epochs or claim a new planet.

K2-22 b · TESS · K2 · dust transits · disintegrating planets · time series · block bootstrap · intermittent activity · reproducible data