With diagonal noise, two TESS sector 80 light curves give weak separation of 4.2 and 4.6 hour Kepler-9 c models at 0.6 percent depth

We test whether two public TESS sector 80 light curves can separate two Kepler-9 c transit models. Total durations are 4.2 and 4.6 hours. Each model has a fixed depth of 0.6 percent in released flux. The test uses the 2018 forecast time, a fixed ingress fraction of 0.2, a free quadratic baseline, and a diagonal Gaussian noise model. Exposure integration is analytic. Local adjacent flux differences set the noise scale. Model separation is 0.276 noise unit in SPOC PDCSAP_FLUX and 0.216 in QLP DET_FLUX. With equal model probabilities, the minimum Gaussian choice errors are 44.5 and 45.7 percent. Injections into 23 and 25 disjoint control windows give 54.3 and 52.0 percent correct choices. The paired injections share noise. A linear baseline and ingress fractions from 0.1 to 0.3 keep separation below 0.46 unit. These data products have little power for this specified duration test. The result does not measure a duration or test all possible pixel extractions. It does not confirm or reject the orbital inclination forecast. We preserve original FITS data, hashes, source records, selection records, results, figures, and code. The code runs without network access when the included inputs are used.

Kepler-9 c · TESS · transit duration · sensitivity · SPOC · QLP · injection test · public data · sector 80