For 45 contact binaries with strong TESS 2018-2026 period curvature, earlier ASAS-SN and SuperWASP timings show about 40 percent of the predicted offset; 403 systems show no V1309 Sco-like decay

Question: do the orbital-period changes that TESS measures in contact binaries continue as a constant dP/dt in earlier, independent data, and does any system show a fast, steady decrease like the merger progenitor V1309 Sco? Approach: we measured per-sector eclipse-timing shifts with a Fourier light-curve template for 749 contact binaries of the Prsa et al. (2022) TESS catalog (P = 0.2-1.0 d, 5,001 SPOC 2-min light curves, 2018-2026). We fit linear and quadratic ephemerides (403 vetted binaries). Then we measured season timings with the same templates in ASAS-SN (2012-2025) and SuperWASP (2004-2008) data and tested the TESS ephemerides backwards in time. Results: 118 of 403 binaries have a TESS quadratic term at 5 sigma or more (68 increases, 50 decreases; median |dP/dt| 0.068 s/yr). For strongly curved binaries, the observed earlier offset is a median 0.42 of the TESS-quadratic prediction about 4 years before TESS (45 binaries) and 0.34 about 12 years before (22 binaries); a constant dP/dt gives 1. The decade-mean dP/dt between the survey epoch and TESS is a median 0.35 of the TESS-window value. Apart from one rejected spot or blend case, no binary has a significant decrease faster than 0.44 s/yr, more than 100 times slower than V1309 Sco before its merger. Conclusion: most TESS-window period changes of contact binaries are parts of longer, non-secular variations. Limitations: timings are defined only modulo P/2, an empirical -274 s ASAS-SN time offset and a 98 s systematic term were applied, SuperWASP placement across 2008-2014 uses a trapezoid rule, and no cyclic models were fitted.

contact binaries · W UMa stars · eclipse timing · O-C diagram · orbital period change · TESS · ASAS-SN · SuperWASP · stellar mergers · V1309 Sco · luminous red novae · out-of-sample test