Fits of a fixed profile show a decrease in KELT-9 b half-depth transit width in six TESS sectors from 2019 to 2024

We test whether the observed transit width of KELT-9 b changes in six public TESS sectors from 2019 to 2024. We fit 90 transits in uniform 120-second SPOC light curves. Each event has a free center, depth, time scale, and local baseline. Width means the full interval at half of the maximum loss of light in the fitted, exposure-averaged profile. The Sector 82 minus Sector 14 width change is -3.35 minutes, or -1.61%. Whole-event resampling gives a 95% interval of -3.70 to -2.99 minutes. A weighted line gives -0.643 minute per year, with a conditional 95% interval of -0.698 to -0.587. SAP flux, baseline and window changes, a fixed asymmetric reference, and separate polynomial wing fits all give negative slopes. An artificial signal with constant width gives a recovered mean endpoint change of +0.05 minute. These checks use the same observations. The intervals do not include changes shared by a whole sector or year. The result extends the photometric width record in a system with known nodal precession. It is not a physical contact duration or a unique measure of orbital geometry. The package contains raw data, hashes, code, results, figures, and a verified offline repeat.

KELT-9 b · TESS · transit width · half-depth width · nodal precession · gravity darkening · public astronomical data · reproducible photometry