Local TESS fits give later times than the 2024 forecasts for four 2025 TOI-1338 b transits

We test four 2025 transit forecasts for TOI-1338 b from Wang and Liu (2024). We fit public TESS two-minute light curves from sectors 89, 93, 96, and 98. We compare the numeric BJD values. The forecast table does not specify its time scale or midtime rule. Each local fit has a free time, width, depth, and linear baseline. The primary model fixes the impact parameter and limb darkening and includes binary-star acceleration. The fitted closest-approach times for sectors 89, 93, and 98 are 35, 57, and 56 minutes later than the tabulated forecasts. Their statistical error estimates are 6, 11, and 6 minutes. Changes to flux type, baseline, and transit shape preserve those offsets. The best Sector 96 fit is about 75 minutes late. That event has two local egress solutions and needs more caution. The closest-approach times of three older events agree with published direct-fit times within two minutes. Synthetic tests exposed a failure of one local starting point. A grid and several starts removed that failure in the tested sample. The package preserves original FITS files, input hashes, code, timing tables, resampling draws, and figures. These measurements can constrain a future joint orbit fit. They do not identify an additional planet. The forecast errors have no published covariance. We therefore give no combined rejection probability.

TOI-1338 b · BEBOP-1 · TESS · circumbinary planet · transit timing · transit duration · forecast test · public data