Transit-referenced TESS fits place two of eight DS Tuc events nearer B for all 24 tested models

Two of eight fixed DS Tuc brightness events stay nearer star B than star A in all 24 tested TESS models. We fit public TESS pixel data from seven sectors. The eight event intervals come from a prior SCI study. Joint transit images give a reference for star A, the known planet host. Gaia DR3 gives the direction and separation of star B. The main fit places five events nearer A and three nearer B. Two events stay nearer B in all 24 tested models for the local baseline and image fit. Their main distances from B are 0.85 and 2.27 arcsec. One has a large offset across the binary direction. Six models for the first event give a negative signal for one reference transit. All 18 models with positive signals in every reference still place it nearer B. All 122 accepted model signals added to nearby real backgrounds recover the correct nearest star under the adopted pixel response function. These controls share that function with the fit. Individual transit references can differ by several arcseconds. Model shifts, reference scatter and unknown color errors prevent a firm physical source assignment. The result is a set of conditional source positions and failure checks for eight fixed events. The package includes exact selected pixels, source hashes, code, all model results, noise draws, controls and figures.

DS Tuc · TESS · stellar flares · binary stars · difference images · transit reference · source positions · reproducible astronomy