TESS 2-min data from 2019 to 2024 show eclipse depth changes of 6% to 25% in 4 of 326 eclipsing binaries in the northern continuous viewing zone; the method finds about half of 10% changes

We measured the eclipse depths of eclipsing binaries (EBs) in the northern TESS continuous viewing zone in each sector from 2019 to 2024 (43 sectors, S14 to S86). A change in eclipse depth shows that the orbit of the binary changes its orientation. Usually a third body causes this. We used the method of our southern survey with Year 2 as the reference year. The input was 544 EBs from the TESS EB catalogue of Prša et al. (2022) and 11447 SPOC 2-min light curves. 326 EBs (533 eclipses) had a Year-2 reference and data from a later year. A template fit flagged 66 eclipses with a depth change of 5% or more at 5 sigma or more, and 10 eclipses disappeared (69 eclipses in 57 EBs). We examined each again with a model-free depth measurement, an aperture (blend) test and a visual check. 4 eclipses in 4 EBs passed all tests: TIC 198555959 (secondary, -7% to -9%), TIC 229476285 (primary, -10% to -12%), TIC 274575202 (secondary, +6% to +7%) and TIC 362104293 (primary, +21% to +25%). We know of no earlier depth-change report for these systems. 8 more eclipses are candidates. In the eccentric binary TIC 274575202, the separation of the two eclipses changes by -0.52 +/- 0.14 min/yr, and the two depths change in opposite directions. Apsidal motion alone gives a depth change of 0.6% or less, thus the inclination must also change. TIC 274575202 is a candidate triple. An injection-recovery test with 232 eclipses in 146 detached EBs found and confirmed 47% to 51% of injected linear depth changes of +/-10%, 64% to 72% of changes of +/-20%, and 7% to 11% of changes of +/-5%. The 5-sigma condition is the main limit: at +/-10%, the recovery is 88% for the eclipses with the smallest depth scatter and 0% for those with the largest scatter. The measured changes have no bias. Thus the sample probably contains more EBs with depth changes of this size. Limitations: the survey uses only catalogue EBs with 2-min data, and the known northern depth-change systems are not in the catalogue. TESS did not observe the field in Year 3. The visual check is subjective, and spots can change small depths. The injection test uses only the automated steps and detached EBs, and it does not measure false positives. We have no radial velocities and no dynamical model. The README also corrects three errors in the southern survey. The package contains the code, the data provenance, the result tables, the figures and the instructions to reproduce all results.

eclipsing binaries · eclipse depth variations · TESS · continuous viewing zone · north ecliptic pole · hierarchical triple stars · orbital precession · apsidal motion · injection-recovery test · survey completeness · model-free vetting · aperture blend test · TIC 274575202 · TIC 362104293 · TIC 229476285 · TIC 198555959 · SPOC light curves · survey