Cleaning WASP-33 b transits from the host star photometric variability: analysis of TESS data from two sectors

Based on TESS observations of a $\delta$ Scuti variable WASP-33 obtained in 2019 and 2022, we thoroughly investigate the power spectrum of this target photometric flux and construct a statistically exhaustive model for its variability. This model contains $30$ robustly justified harmonics detected in the both TESS sectors simultaneously, $13$ less robust harmonics detected in a single TESS sector, a red noise and a quasiperiodic noise terms. This allowed us to greatly improve the accuracy of the exoplanet WASP-33 b transit timings, reducing the TTV residuals r.m.s. drastically, by a factor of $3.5$, from $63$ s to $18$ s. Finally, our analysis does not confirm existence of a detectable orbital phase variation claimed previously based on WASP-33 TESS photometry of 2019.